Proceedings - Johannes Gutenberg-Universität Mainz

Proceedings Volume 2, 2011 

Earthquake Geology and Archaeology: 

Science, Society and Critical Facilities 

Editors 

C. Grützner, R. Pérez-López, T. Fernández Steeger, I. Papanikolaou 

K. Reicherter, P.G. Silva, and A. Vött 

PROCEEDINGS 

2nd INQUA-IGCP 567 International Workshop on Active 

Tectonics, Earthquake Geology, Archaeology and Engineering 

19-24 September 2011 

Corinth (Greece) 

ISBN: 978-960-466-093-3

Proceedings 

Earthquake Geology and Archaeology: 

Science, Society and Critical facilities 

2 nd INQUA-IGCP 567 International Workshop on Active Tectonics, Earthquake Geology, Archaeology and Engineering 

Editors 

C. Grützner, R. Pérez-López, T. Fernández-Steeger, I. Papanikolaou 

K. Reicherter, P.G. Silva and A. Vött 

This Volume of Proceedings has been produced for the 2 nd INQUA-IGCP 567 International Workshop 

on Active Tectonics, Earthquake Geology, Archaeology and Engineering held in Corinth (Greece), 

19-24 September 2011. The event has been organized jointly by the INQUA-TERPRO Focus Area on 

Paleoseismology and Active Tectonics and the IGCP-567: Earthquake Archaeology. 

This scientific meeting has been supported by the INQUA-TERPRO #0418 Project (2008-2011), the IGCP 567 Project, the Earthquake Planning 

and Protection Organization of Greece (EPPO – ) and the Periphery of the Peloponnese

Printed by 

The Natural Hazards Laboratory, 

National and Kapodistrian University of Athens 

 

Edited by 

INQUA-TERPRO Focus Area on 

Paleoseismology and Active Tectonics 

& IGCP-567 Earthquake Archaeology 

INQUA-IGCP 567 Proceedings, Vol.2 

© 2011, the authors 

I.S.B.N. 978-960-466-093-3 

PRINTED IN GREECE 

Vol. 2: Earthquake Geology and Archaeology: Science, Society and Critical facilities (C. Grützner, R. Pérez-López, T. Fernández-Steeger,, I. Papanikolaou, K. 

Reicherter, P.G. Silva & A. Vött, Eds.). ISBN 978-960-466-093-3. Printed in Greece, 2011. 2 nd INQUA-IGCP 567 International Workshop, Corinth (Greece). 

Vol. 1: Archaeoseismology and Palaeoseismology in the Alpine-Himalayan collisional Zone (R. Pérez-López, C. Grützner, J. Lario, K. Reicherter & P.G. Silva, Eds.). 

ISBN 978-84-7484-217-3. Printed in Spain, 2009. 1 st INQUA-IGCP 567 International Workshop. Baelo Claudia, Cádiz (Spain).

Preface 

After the very successful 1 st Workshop on Earthquake Archaeology and Paleoseismology held in the ancient roman site of Baelo 

Claudia (Spain, 2009), the INQUA Focus Group on Paleoseismology and Active Tectonics decided to elaborate a bi-annual 

calendar to support this joint initiative with the IGCP-567 “Earthquake Archaeology”. This second joint meeting moved to the 

eastern Mediterranean, a tectonically active setting within the Africa-Eurasia collision zone and located in the origins of the 

pioneer’s works on archaeoseismology. However, for the coming year 2012, at least a part of us will move also to the New World, 

where the 3 rd INQUA-IGCP 567 international workshop will take place in Morelia, Mexico in November 2012. It is planned to 

proceed with the meeting, so we are thinking of Aachen, Germany, to be the host in 2013, possibly together with Louvain, Belgium. 

The aim of this joint meeting is to stimulate the already emerging comparative discussion among Earthquake Environmental Effects 

(EEE) and Earthquake Archaeoseismological Effects (EAE) in order to elaborate comprehensive classifications for future 

cataloguing and parametrization of ancient earthquakes and palaeoearthquakes. One of the final goals our collaborative workshops 

is the integration of archaeoseismological data in Macroseismic Scales such as the Environmental Seismic Intensity Scale ESI- 

2007 developed within the frame of the International Union for Quaternary Research (INQUA). In this second workshop we offer 

again a multidisciplinary and cross-disciplinary approach and program, since there is an urgent necessity to share the knowledge 

and objectives among geologists, seismologists, geodesists, archaeologists and civil engineers in order to improve seismic hazard 

assessments and analyses in a near future. Also, we intend to sharpen geoscientists and their research more in the direction of 

critical facilities, which are of world-wide public and political interest after the dramatic catastroph in Fukushima, Japan. 

The last two years provided significant dreadful earthquake scenarios, which were in most of the cases oversized in relation to the 

data provided by the historical and instrumental seismicity. The Haiti Mw 7.0 (Haiti, Jan 2010), Malua Mw 8.8 (Chile, May 2010), 

Christchurch Mw 6.3 (New Zealand, Feb 2011), Tohoku Mw 9.0 (Japan, Mar 2011) and Lorca Mw 5.1 (Spain, May 2011) events 

illustrates that both extreme subduction earthquakes or moderate events can generate severe damage in relation to relevant 

secondary coseismic effects or Earthquake Environmental Effects (EEE). Most of these recent events have clearly demonstrated 

that the vibratory ground shaking is not the unique, or even most significant, source of direct damage, and it is by no means the 

only parameter that should be considered in seismic hazard assessments. The lessons offered by the aforementioned events 

corroborate once again the relevance of liquefaction, tsunamis, rockfalls, landslides, ground subsidence, uplift or failure as a major 

source of hazard. But this also underpins the need of re-evaluating the significance of macroseismic intensity as an empirical 

measurement of earthquake size. In fact, as highlighted in the last volume produced by the INQUA Focus Area (Serva et al., 2011), 

intensity is a parameter able to describe a complete earthquake scenario, based on direct field observation and suitable to be 

preserved in the geological, geomorphological and archaeological records. 

With this aim the INQUA TERPRO #0418 Project (2008-2011) has implemented a world-wide online EEE Catalogue based on 

Google Earth in order to promote the use of the ESI-2007 Scale for seismic hazard purposes 

www.eeecatalog.sinanet.apat.it/terremoti/index.php. On the other hand the IGCP-567 is promoting an interesting shared approach 

of EEE data and EAE data for the same purpose. Examples of this variety of original research coming from this collaborative 

approach are the Geological Society of London Special Volume 316 (2009) Paleoseismology: Historical and Prehistorical 

records of Earthquake Ground Effects for Seismic Hazard Assessment (K. Reicherter, A.M. Michetti & P.G. Silva, Eds.), the 

Geological Society of America Special Papers 471, Ancient Earthquakes (2010) (M. Sintubin, I.S. Stewart, T. Niemi & E. Altunel, 

Eds.) and the Special Volume of Quaternary International (2011) Earthquake Archaeology and Paleoseismology (P.G. Silva, M. 

Sintubin & K. Reicherter, Eds.). In the same way, this abstract volume contains more than 80 contributions from researchers of 

more than 27 different countries and illustrates the upgrading shared knowledge on palaeo-, ancient, historical and instrumental 

earthquakes and images an impressive growth of our community. Our workshop was co-ordinated through the newly established 

website www.paleoseismicity.org, where earthquake info and blogs are openly shared. 

Finally, we wish all participants a fruitful conference and workshop in the vicinity of the ancient sites of the Classical Greece around 

the Corinth Gulf, where earthquake science, wonderful landscapes, ancient cultures, amazing sunny days, fantastic “Greek 

cooking”, nice beaches, daily cool beers, wine tasting events and late night gin tonics mixed with hot discussions are waiting for all 

of us. A special “efharisto poli” goes to Christoph Grützner and Raul Pérez-López for their invaluable work with the organisation 

and the abstract handling. 

The Organizers of the 2 nd INQUA-IGCP 567 Workshop 

Dr. Ioannis Papanikolaou Prof. Klaus Reicherter Prof. Andreas Vött Prof. Pablo G. Silva 

Agricultural University of Athens (GRE) RWTH Aachen University (GER) University of Mainz (GER) University of Salamanca (ESP) 

Aon-Benfield Hazard Research Centre, 

University College London, (UCL) (UK)

2 nd INQUA-IGCP-567 International Workshop on Active Tectonics, Earthquake Geology, Archaeology and Engineering, Corinth, Greece (2011) 

EARTHQUAKE 

ARCHAEOLOGY 

INQUA PALEOSEISMOLOGY 

AND ACTIVE TECTONICS 

FIRST PALEOSEISMIC EVIDENCES IN ECUADOR: THE PALLATANGA FAULT RECORD 

Baize, Stéphane (1, Laurence Audin (2), Thierry Winter (3), Alexandra Alvarado (4), Luis Pilatasig (5), Mercedes Taipe (4), Paul 

Kauffmann (1), Pedro Reyes (4) 

(1) Institut de Radioprotection et de Sûreté Nucléaire, BP 17, 92262 Fontenay-aux-Roses, France. stephane.baize@irsn.fr 

(2) Institut de Recherche pour le Développement, ISTERRE, Grenoble, France. 

(3) Bureau de Recherche Géologiques et Minières, Service Risques Naturels, Orléans, France. 

(4) Escuela Politécnica Nacional, Instituto de Geofísica, Quito, Ecuador. 

(5) Ministerio de Recursos Naturales No Renovables, Quito, Ecuador. 

Abstract (First paleoseismic evidences in Ecuador: the Pallatanga fault record): The Pallatanga fault (PF) is a prominent 

strike-slip fault of Central Ecuador. This structure is suspected to have hosted large earthquakes, including the 1797 Riobamba 

event (M~7.5). The scope of the study is to evaluate the paleoseismic history of the fault, together with enhancing the 

seismotectonic model of this part of the Andes and improving the seismic hazard assessment. From 3 trenches, we could infer that 

the PF experienced several strong events (M7.2 to 7.7) in the last 8500 years. According to a new mapping campaign, we also 

could evidence that the fault propagates north to the Riobamba outskirts, suggesting that faulting occurred nearby this big city. 

Key words: Paleoearthquakes, Holocene, Pallatanga fault, Ecuador 

INTRODUCTION 

The Pallatanga fault (PF) is a NNE-SSW segment of 

the Dolorès-Guayaquil Megashear. This megastructure 

is the large deformation zone 

accommodating the dextral displacement between 

the Northern Andean Block and the South America 

Plate with a rate of 6 to 8 mm/a. The PF is a 50 kmlong 

fault, for which a previous morphotectonic study 

validated its Holocene right-lateral motion (Winter et 

al., 1993). This fault is probably the source of one of 

the largest crustal earthquake in South-America, 

occurred in Riobamba in 1797 (M~7.5: Beauval et al., 

2010) (figure 1). The “Riobamba Antigua” city was 

destroyed (25,000 casualties) and then replaced at 

its current place. It may also have generated a 

Mw6.1 event in 1911. 

The scope of the study is to (1) assess the 

occurrence of potential large paleoearthquakes and 

quantify their number, magnitudes and recurrence 

times, (2) improve the seismotectonic model of the 

area, and (3) enhance the seismic hazard 

assessment of this inhabited region (Riobamba city, 

200,000 inhabitants). 

PREVIOUS MORPHOTECTONIC STUDY OF PF 

A detailed topographic survey in the Pangor Rio 

valley (Winter et al., 1993) evidenced the fault plane 

characteristics (N40°E, 75°W) and allowed estimating 

the “near-field” displacement to be ~40m for the 

dextral component and ~8m for the reverse one. 

Winter et al. (1993) deduced a cumulated striae 

dipping slightly to the south (10°). They also 

proposed an average slip rate of 2.9 to 4.6 mm/yr, 

according to regional geomorphic correlation. 

RESULTS OF THE TRENCH SURVEY 

We focused on paleoseismological analyses at 

Rumipamba and on a detailed mapping of the active 

fault trace continuation towards Riobamba city. 

Trenches 

Figure 1: Location of the study area, with the main faults 

and historical seismicity. Figure extract from Beauval et al. 

(2010). Blue disks: historical events, with magnitude 

proportional to radius (ex. 1797: M=7.6; 1911: M=6.2; 1958: 

M=5). Specific faults: Pallatanga fault in blue and bounded 

by red arrows, Guamote-Huigra fault in green, Llanganates 

fault in cyan, Pucara fault in magenta. Pink lines: provinces 

boundaries. Names of provinces in pink; names of towns in 

black. 

At Rumipamba, the fault trace runs parallel to the 

mountain front and cuts perpendicularly the erosional 

features going downhill (figure 2). The mountain 

slope sedimentation is exclusively dominated by ash 

falls, with rare colluvium and alluvium. Unfortunately, 

3


EARTHQUAKE 

ARCHAEOLOGY 



we could not find any channel deposits in test 

trenches and it was thus impossible to find linear 

features that could account for lateral offsetting. In 

further developments, we thus used the 

“morphological slip vector” of Winter et al. (1993). 

In the three excavated trenches, ash falls are 

strongly pedogenetized in black organic soils 

(andisols), over thicknesses of 2 to 3.5 meters. 

last event in 1797 should then have a higher 

magnitude than mentioned in table 1. Moreover, the 

event #2, which is also recorded in trench #1 and 

well constrained by 14 C datings, is associated with 2 

successive and discordant colluvial wedges. If we 

assume this is not partly due to non-tectonic 

processes (climate degradation for ex.), then arises 

the issue of possible under-representation of 

earthquake number in trench #2 (and correlative 

overestimation of magnitude). 

Trenches 

#1 

#2 & 3 

Fault trace 

EQ 

Date 

approx. 

(cal y.) 

Vertical 

throw 

(cm) 

Total 

throw 

(cm) 

Mw 

SRL 

(km) 

5 1797 25 145 7.2 60 

4 ~1,000 60 345 7.5 100 

3 ~0 90 520 7.7 140 

2 ~ -1,000 90 520 7.7 140 

1 ~ -4,500 70 400 7.6 120 

Figure 2: South-eastward view of the mountain front, cut by 

the Rumipamba segment of the PF. Trenches were 

excavated in the the same area. 

Available 14 C datings of these soils (23 samples) 

range from 6650 cal BC to 1650 cal AD, scattered 

into 7 classes of ages (-6650, -5500, -3500, -2800, - 

1800, 800, 1650). They lie above the volcanic 

basement (Mesozoic-Tertiary). 

The trenches’ survey pointed out several interbeds 

with variable oxidized material contents -including 

basement fragments and root remnants- in the 

vicinity of the fault strands. Our basic assumption is 

that these “clastic layers” are stratigraphic markers 

of earthquakes (colluvial wedges) because they are 

associated with the erosion of a basement scarp 

appeared during surface faulting. 

Among the 3 trenches, the trench #2 is especially 

interesting in order to reconstruct the paleoseismic 

history of the fault. There, the stratigraphic series is 

more complete and stratigraphic units can be 

correlated on each side of the fault (figure 3). In 

trenches #1 and #3, the fault strands are obvious but 

correlations are difficult. In trench #2, the fault splits 

in 3 strands propagating from the basement fault 

gouge up to the Holocene deposits and the modern 

soil. The fault zone downthrows the eastern wall by 

about 3 m. 

By performing a retrodeformation of the trench, we 

could infer 5 strong events with individual vertical 

throws from 0.25 m up to 0.90 m (table 1). Assuming 

a slip vector dipping of 10° (Winter et al., 1993), the 

total throws are between 1.45 to 5.20 m, which 

provides magnitudes from 7.2 to 7.7 according to 

Wells & Coppersmith (1994) relations. To generate 

such earthquakes, the same empirical relations 

suggest that the fault rupture can be as long as 60 to 

140 km (which is larger than the mapped PF). 

Some uncertainties remain after trench survey. For 

example, the retrodeformation of trench #2 can lead 

to only 4 paleoearthquakes because arguments for 

separating events #4 and #5 (Table 1) are weak. The 

Table 1: synthetic table of the earthquake history along the 

Rumipamba segment of the Pallatanga fault. Vertical throw 

is the observed data; total throw is calculated from vertical 

throw and morphological slip vector; Mw is calculated from 

total throw with W&C (1994) relation (average 

displacement); SRL is estimated from total throw 

CONTRIBUTION OF MAPPING 

From its southern tip (Western Cordillera front) to the 

trench site (figure 1), the PF is mapped and reaches 

a length of about 60 km. Given the previous 

conclusion of large fault ruptures (SRL>60 km), it has 

been decided to investigate the “unknown” northern 

continuation of the Rumipamba segment where 

morphological features suggested its existence and 

where the rupture may have propagated during large 

past events. The interest is that section is also that it 

can help in improving the seismotectonic model of 

the region and in clarifying the seismic hazard 

assessment of this inhabited area. 

Before field mapping, fault traces were tracked down 

with analysis of remote-sensing images (SPOT5). 

Field control and mapping allow validating most of 

this preliminary work thanks to morphologic and 

geologic observations. The Rumipamba segment 

splits into 2 fault zones when entering the Inter- 

Andean valley (figure 4). The western branch 

(Rumipamba-Huacona and La Merced segments) 

seems to rotate into a N-S direction, with pure lateral 

kinematics, whereas the eastern ones changes to a 

NE-SW strike, with a more important vertical 

component (Cajabamba, Gatazo and Lican- 

Riobamba segments). This last segment cuts the 

Riobamba basin and seems to offset the surface 

deposits of the Upper Pleistocene (40 ka, Bernard et 

al., 2008) and later morphological features. The 

scarp is almost 100 m high and, locally, we could find 

a dextral offset of 50 m of a valley slope. To the 

north-east, some morphological features (mountain 

front spurs, stream profiles) and outcrops (normal 

faulting and tilting of recent sediments) suggest that 

active deformation propagates from the Riobamba 

basin north-eastwards along the Cordillera Reale 

front (Penipe area), along the Chambo segment. 

4


EARTHQUAKE 

ARCHAEOLOGY 



Figure 3: Mosaic picture and sketch interpretation of trench #2. Ages are given in cal BC/AD. 

Figure 4: Sketch map of the northern extension of the PF within the Riobamba basin, and example of the morphological imprint of 

the fault. DS: segment with prevailing dextral component; N: with normal component (symbols on the downthrown block side); R: 

with reverse component; V: segment with undifferenciated vertical offset; VS: segment with vertical and dextral strike-slip 

component (symbols on the upthrown block side, for the 3 last). 

5


EARTHQUAKE 

ARCHAEOLOGY 



These preliminary conclusions need to be completed 

by further mapping, especially in the Chambo valley 

and along the western rim of the Inter-Andean valley 

(north of San Juan). 

All along the prospected area, several places have 

been quoted to be potential sites for further 

paleoseismic investigations. 

DISCUSSION 

For the first time in Ecuador, paleoseismic 

investigations prove the occurrence of 4 or 5 strong 

events which shook the Riobamba region in the past. 

We probably found the trace in sediments of the 

Riobamba event (1797) and then confirm that it has 

been generated by the Pallatanga fault. Magnitudes, 

inferred from geological signal and applying empirical 

relations, range between 7.2 and 7.7, which is in 

agreement with the independent estimation of the 

Riobamba earthquake by Beauval et al. (2010). 

Reccurrence times of these events seem variable 

(Table 1: 800 to 3,500 years) with a mean value 

around 1,000-1,500 years. 

Nonetheless, the trench survey and its interpretation 

are soiled by uncertainties and arises the issue of 

under or over-representation of the event number, 

which is a crucial point in seismic hazard assessment 

(Mc Calpin, 2009). In addition, the 3 trenches 

revealed variable tectono-sedimentary signals. These 

nuances weaken the above conclusions dealing with 

magnitudes of paleoearthquakes and hazard 

assessment for the PF, and we clearly need 

additional investigations to strengthen the results. In 

this first-step work, we assumed for instance during 

magnitude assessment that the observed coseismic 

offsets are “average” values of the event slip and this 

has clearly to be validated. The mapping clearly 

shows that the trenches were not performed at the 

northern tip of the fault and new trenching sites must 

be found out to increase the slip vs rupture length 

dataset. 

Dealing with the fault slip rate of the fault, our results 

based on radiometric datings outstandingly validate 

the rate published by Winter et al. (1993). Roughly, 

trenches show vertical displacement of about 3 

meters of recent soils -i.e. 20 meters following the 

hypothesis of a 10° slip vector- which have been 

produced during the last 5,000 years. This leads to a 

mean slip rate of ~4 mm/a. With respect to the total 

displacement rate (~6-8 mm/a) of the North-Andean 

Block, it suggests that other segments accommodate 

a significant part of the relative motion of North 

Andean Block wrt South America. Up to now, no 

other faults are potential candidates. 

What we know from our mapping campaign is that 

deformation probably propagates both to the north 

(along the western edge of the Inter-Andean valley) 

and to the north-east (towards the eastern rim of the 

Cordillera Reale). This last segment could be the 

structural link between the Pallatanga fault (to the 

south) and the Pucara/Llanganates faults (to the 

north), one of these being probably the source of the 

Pelileo destructive earthquake (1949) (Beauval et al., 

2010). This propagation of deformation to the north 

also drastically increases the seismic hazard for 

Riobamba city, because rupture segments probably 

run all along the outskirts of this vulnerable big city. 

Acknowledgements: The authors give special thanks to 

the Indian communities of Rumipamba and neighboring 

villages, for their warm welcome and “technical” contribution 

to logistics and trench cleaning. Thanks also to the various 

authorities of Ecuador who make this research possible 

(prefecture and the civil safety services). We also benefit of 

the help and fruitful discussions with people from the 

Escuela Politecnica Nacional de Quito. 

References 

Beauval, C., H. Yepes H., W. H. Bakun, J. Egred, A. 

Alvarado & J.-C. Singaucho, (2010). Locations and 

magnitudes of historical earthquakes in the Sierra of 

Ecuador (1587–1996). Geophys. J. Int., doi: 

10.1111/j.1365-246X.2010.04569.x. 

Bernard, B., B. van Wyk de Vries, D. Barba, H. Leyrit, C. 

Robin, S. Alcaraz & P. Samaniego, (2008). The 

Chimborazo sector collapse and debris avalanche: 

Deposit characteristics as evidence of emplacement 

mechanisms. Journal of Volcanology and Geothermal 

Research 176, 36–43. 

Mc Calpin, J. P., (2009). Paleoseismology. 

INTERNATIONAL GEOPHYSICS SERIES, volume 95, 

Academic Press, 802 pages. 

Wells, D. L. & K. J. Coppersmith, (1994). New empirical 

relationships among magnitude, rupture length, rupture 

width, rupture area and surface displacement. Bull. 

Seism. Soc. America 84, 974-1002. 

Winter, T., J.P. Avouac & A. Lavenu, (1993). Late 

Quaternary kinematics of the Pallatanga strike-slip fault 

(Central Ecuador) from topographic measurements of 

displaced morphological features. Geophys. J. Int. 115, 

905-920. 

6


EARTHQUAKE 

ARCHAEOLOGY 



SOME NOTES ON EARTHQUAKE AND FAULT RELATIONSHIPS FOR DIP-SLIP EVENTS 

Salvatore Barba (1) and Debora Finocchio (1, 2) 

(1) Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, Roma, Italy. salvatore.barba@ingv.it 

(4) Università di Urbino, Italy 

Abstract (Some notes on earthquake and fault relationships for dip-slip events): We developed a conceptual model for dipslip 

earthquakes to predict the coseismic surface throw and the dislocation as a function of depth along-dip (1-D averaged). By 

performing a Monte Carlo experiment in Matlab to test the conceptual model against earthquake data for years 1990-2006, we 

found that, for reverse faulting, the surface throw is better reproduced with the slip distributions which show a maximum slip closer 

to the surface. Our model predicts a surface throw for reverse faulting earthquakes that is larger than that contained in the Wells 

and Coppersmith (1994) relationships. The use of downward or upward directivity in the rupture changes dramatically the 

assessment and the perception of the earthquake hazard, as shown by the catastrophic 2011 Japan earthquake. 

Key words: Normal fault, thrust fault, coseismic rupture, Monte Carlo model. 

INTRODUCTION 

It is commonly accepted in the literature that 

deviatoric stress increases with depth (see, e.g., Das 

and Scholz, 1983; Henry and Das, 2001). The 

coseismic dislocation is predicted to initiate at the 

base of the seismogenic layer and to propagate 

upward. However, two classes of observations, 

coseismic throw and depth directivity, are hardly 

explained with such hypotheses in the case of dipslip 

earthquakes. Reverse-faulting earthquakes 

occurred in the past 20 years exhibited a much larger 

throw at the surface, with respect to normal-faulting 

earthquakes of similar magnitudes. Also, in certain 

cases of reverse-faulting earthquakes, seismological 

observations show a downward directivity of the 

rupture (Carminati et al., 2004). 

the USGS catalogue, and the depth distributions from 

regional earthquake catalogue. Based on these 

distributions, we generated a very large synthetic 

earthquake catalogue. For each of these synthetic 

earthquakes we set the average slip, the length and 

the width of the rupture based on a-priori 

relationships. Then, we predicted the surface throw 

based on two classes of slip distributions at depth: 

one class based on four a-priori distributions (our 

METHOD AND DATA 

In this work we developed a conceptual model for 

dip-slip earthquakes to predict the coseismic surface 

throw and the dislocation as a function of depth 

along-dip (1-D averaged). To this purpose, we 

determined relationships among the surface throw 

and the earthquake parameters (slip, dip, depth, and 

moment magnitude Mw) for two emblematic normal 

and reverse faults. Finally, assuming that the (1-D 

averaged) slip-distribution at depth may be related to 

the deviatoric stress, we discussed the difference 

between our model and Das e Scholz (1983) model. 

This work is thus developed in two parts: the 

generation of synthetic relationships between 

coseismic throw and magnitude, and the comparison 

of model predictions with a 20-years long earthquake 

dataset (Table 1). We made a Monte Carlo 

experiment in Matlab to test the conceptual model 

against data. 

The synthetic relationships are based on standard 

seismic catalogues. We determined the frequency 

distributions for dip, Mw and seismic moment from 

7 

Fig. 1: Cumulative percentage of earthquakes that 

ruptured the surface for reverse (TF) and normal faulting 

(NF). Lines represent the model prediction, the symbols 

represent the data. 

choice), and one class based on literature data, 

which have been grouped in three slip distributions at 

depth. These four (or three) distributions differ as it 

follows: (1) the maximum slip is in the center of the 

rupture, (2) it is closer to the bottom or (3) to the top 

edge of the rupture; (4) the slip is uniform. For each 

magnitude bin we computed the average and the


EARTHQUAKE 

ARCHAEOLOGY 



standard deviation of the predicted surface throw. 

Last, we compared the predicted maximum and 

average surface throw with the observed throw for 

earthquakes occurred in years 1990-2008. As a 

cross check, we computed the percentage of 

synthetic earthquakes that produce surface throw, as 

a function of magnitude (Fig. 1). 

RESULTS 

We found differences between reverse and normal 

faulting earthquakes. For reverse faulting, the surface 

throw is better reproduced with the slip distributions 

which show a maximum slip closer to the surface, or 

with the uniform slip (Fig. 2). For normal faulting, the 

best fit occurs when the maximum slip is closer to the 

bottom edge of the rupture. 

reverse faulting earthquakes that is larger than that 

contained in the Wells and Coppersmith (1994) 

relationships. The opposite occurs for normal faulting 

earthquakes. In the dataset of Wells and 

Coppersmith (1994), the fault dimensions are derived 

sometimes from seismological data, sometimes from 

geodetic data. As we used only earthquakes 

occurred in recent times, our dataset is based on 

seismological data and is therefore more 

homogeneous. Although we have fewer data, our 

result appears robust. In the hypothesis that the slip 

distribution can be used as a proxy for deviatoric 

stress, the maximum of the deviatoric stress is 

expected to be close to the upper edge of the fault 

rupture at least for earthquakes with low dip angle 

(=5.9). 

N Code Date Location Mw Depth (Km) Type Avg Throw (m) 

1 EV-022 1990/11/06 Iran 6.6 11.1 TF 2.6 

2 EV-073 1992/08/19 Turkey 7.2 17 TF 3 

3 EV-101 1993/05/17 California 6.1 7 NF 0.02 

4 EV-117 1993/09/29 India 6.2 14.1 TF 1.93 

5 EV-187 1995/05/13 Greece 6.5 14 NF 0.05 

6 EV-192 1995/06/15 Greece 6.5 14 NF 0.02 

7 EV-206 1995/10/01 Turkey 6.4 9 NF 0.2 

8 EV-282 1997/09/26 Italy 6 6 NF 0.03 

9 EV-330 1999/09/07 Greece 6 10 NF 0.08 

8


EARTHQUAKE 

ARCHAEOLOGY 



N Code Date Location Mw Depth (Km) Type Avg Throw (m) 

10 EV-331 1999/09/20 Taiwan 7.6 9 TF 5 

11 EV-405 2002/02/03 Turkey 6.5 5 NF 0.15 

12 EV-414 2002/06/22 Iran 6.5 11 TF 1 

13 EV-431 2002/11/03 Alaska 7.2 4.2 TF 4 

14 EV-518 2005/02/22 Iran 6.4 7 TF 0.8 

15 EV-532 2005/10/08 Pakistan 7.6 19.1 TF 3.7 

9


EARTHQUAKE 

ARCHAEOLOGY 



HOLOCENE COASTAL NOTCHES IN THE MEDITERRANEAN: PALAEOSEISMIC OR 

PALAEOCLIMATIC INDICATORS? 

Boulton, S. J. (1) and Stewart, I. S. (1) 

(1) School of Geography, Earth and Environmental Sciences, University of Plymouth, Plymouth, Devon, PL4 8AA, U.K. 

Sarah.boulton@plymouth.ac.uk 

Abstract (Holocene Coastal Notches in the Mediterranean: Palaeoseismic or Palaeoclimatic indicators?): Bioerosion and 

bioconstruction along rocky coastlines can lead to the development of coastal notches that are preserved when uplifted or 

submerged above or below the swash zone and thus can be used to quantify relative vertical coastal motions in tectonically active 

areas. There are two main models for the genesis of notch profiles. The first is tectonically driven, with notches forming during 

relative still-stands of sea-level; notches are subsequently raised (or lowered) relative to sea-level due to local seismic events. The 

second model favours a climatic origin for notch formation, where stable periods of Holocene climate allow enhanced erosion. 

Here, we explore these two models using a database of Eastern Mediterranean notches. We conclude that the spatial and 

temporal distribution of the notches favours a dominantly tectonic control on formation. 

Key words: notches, tectonics, climate, Holocene. 

INTRODUCTION 

Along rocky coastlines bioerosion and 

bioconstruction can lead to the development of 

coastal notches that can be preserved when uplifted 

or submerged above the swash zone (Stewart and 

Morhange, 2009). As these geomorphic features 

form at sea level, palaeoshorelines could be used to 

quantify relative coastal uplift and subsidence in 

tectonically active areas when the sea level history is 

known (e.g., Pirazzoli & Thommeret, 1977; Pirazzoli 

et al., 1981, 1982, 1989, 1991, 1994a, 1994b; Stiros 

et al., 1992; Stewart et al., 1997; Stiros et al., 2000; 

Antonioli et al., 2006; Shaw et al., 2008). 

Despite this large body of work there is some 

disagreement on the process of notch formation. 

The most commonly assumed theory is that notches 

form when the rate of sea level change matches the 

uplift (or subsidence) rate along a coastline resulting 

in a relative sea level still stand. Notches are then 

raised higher than sea level when a seismic event 

results in rapid coastal uplift preserving the feature 

from further erosion. However, in many other areas it 

is not clear whether notches reflect short-term 

seismic events or the long-term uplift rate averaged 

over many seismic cycles (Stewart & Vita Finzi, 

1996). Rarely, can individual notches be matched to 

documented earthquakes (Pirazzoli, 1994a). 

Recently, Cooper et al. (2007) have linked notch 

formation to Holocene periods of climatic stability. 

Specifically, when the rate of sea level rise is lower 

than the tectonic uplift rate but when climatic 

conditions favour high productivity, increased levels 

of bioerosion develop a notch. This enhanced 

erosion ceases during periods of rapid climate 

change (RCC), allowing notches to become 

emergent due to continuing coastal uplift. Cooper et 

al. (2007) discount any correlation between notches 

and individual seismic events, stating that individual 

earthquakes are too small to raise a notch clear 

above sea level and too numerous to explain the 

formation of the four notch intervals that they 

observed on the Perachora Peninsula (Greece). This 

is a dramatic reinterpretation of coastal notches 

which, if correct, means these phenomena cannot be 

used for palaeoseismic indicators and associated 

seismic hazard assessments. So what drives notch 

development: tectonic instability (earthquakes) or 

climatic stability (enhanced bioerosion)? 

If tectonic uplift is the dominant control in the 

Mediterranean, then notches will develop only in 

coasts where slow regional emergence is augmented 

by abrupt uplift events (earthquakes) and the age of 

notches will relate broadly to local palaeoseismic 

episodes. If climate is the dominant control on notch 

formation, then notches can also form on subsiding 

coasts and will date only to periods of climatic 

stability (no RCC) that are consistent across the 

region. Mayewski et al., (2004) propose five periods 

of stable climate dating to 8000-6000, 5400-4200, 

3800-3300, 2500-1200 and 1000-600 years BP. That 

implies that no more than five notches ought to be 

formed in any one place, each corresponding to an 

intervening periods of no RCC during the last 10,000 

yrs (Mayewski et al., 2004). Yet in some parts of the 

Mediterranean coast more than five notch levels are 

recorded; along the coast of Crete, nine to ten well 

preserved superimposed shorelines can be observed 

(Pirazzoli et al., 1982) and eight are present on 

Rhodes (Pirazzoli et al., 1989). 

10


EARTHQUAKE 

ARCHAEOLOGY 



a. 

Height (m) 

Frequency 

b. 

Sample age (recalibrated years B.P) 

Figure 1a) Histogram showing frequency of samples against age from notches around the Eastern Mediterranean in 100 year 

groups; b) Graph of radiocarbon age against height of all notch data for the Eastern Mediterranean region. Grey bars on both 

graphs indicate the periods of proposed Holocene rapid climate change. 

Such field observations seem inconsistent with the 

climate model. However, to appraise any relationship 

with climate in a more robust manner requires a 

systematic analysis of the Holocene palaeo-shoreline 

data for the Mediterranean region as a whole. To do 

that, we have compiled a dataset of palaeo-shoreline 

data that span several decades of published 

research in the region. The database itself comprises 

428 dated samples derived from 40 separate studies 

that were undertaken at locations from across the 

whole of the Eastern Mediterranean. These data are 

used to examine age and height relationships and 

test correlations with periods of RCC. 

The results of the database are summarised in figure 

1. Generally, with the exception of a few outliers, the 

initial phase of notch development in the 

Mediterranean occurs in the period 5000 – 6500 

years BP (Fig. 1). This period correlates with the mid- 

Holocene slowdown of global sea-level rise. The 

eustatic slowdown at this time ensures that along 

many tectonically uplifting coasts, rates of 

emergence and sea level rise are in unison and so 

notch development takes place. It is noteworthy that 

under the climate model this period is one of rapid 

climate change and so a time window in which notch 

formation ought to be muted. 

Indeed, overall, figure 1 shows that there appears to 

be no correlation between stable climatic periods and 

notch occurrence, with numerous notches dating to 

periods of RCC (grey bars). This strongly suggests 

that climate is not the controlling factor for notch 

formation. Instead, it appears that the complex age 

and height pattern of notches around the region more 

likely reflects local tectonic histories of slow 

interseismic crustal deformation overprinted by 

abrupt seismic movements of the coastline. 

Discriminating ambiguous palaeoseismic information 

from this complex shoreline record, however, 

remains a difficult task. 

Acknowledgements: We would like to thank Richard 

Martin for his help in constructing the notch database. 

11


EARTHQUAKE 

ARCHAEOLOGY 



References 

Antonioli, F., Ferranti, L., Lambeck, K., Kershaw, S., 

Verrubbi, V., Dai Pra, G., (2006). Late Pleistocene to 

Holocene record of changing uplift rates in southern 

Calabria and northeastern Sicily (southern Italy, Central 

Mediterranean Sea). Tectonophysics, 422, 23-40. 

Cooper, F.J., Roberts, G.P., Underwood, C.J., (2007). A 

comparison of 10 3 -10 5 climate stability and the formation 

of coastal notches. Geophysics research letters, 34, 

L14310, doi:10.1029/2007GL030673 

Mayewski, P.A., Rohling, E., Stager, C., Karlén, K., Maasch, 

K., Meeker, L.D., Meyerson, E., Gasse, F., van Kreveld, 

S., Holmgren, K., Lee-Thorp, J., Rosqvist, G., Rack, F., 

Staubwasser, M., and Schneider, R., (2004), Holocene 

climate variability, Quaternary Research 62, 243-255. 

Pirazzoli, P.A., Thommeret, J., (1977). Datation 

radiométrique d’une ligne de rivage à +2.5 m près de 

Aghia Roumeli, Crète, Grèce. Le Comte Rendu 

d’Academie Science, Paris, Series D, 97, 1255-1257. 

Pirazzoli, P.A., Thommeret, J., Thommeret, Y., Laborel, J., 

Montaggioni, L.F. (1981). Les rivages emerges 

d’Antikythira (Cerigotto): correlations avec le Crete 

occidentale et implications cinematiques et geodynamics. 

Niveax Marins et Tectonique Quaternaires dans l’Aire 

Mediterraneene, Paris, UNRS and Universitie Paris I, pp. 

49-65. 

Pirazzoli, P.A., Thommeret, J., Thommeret, Y., Laborel, J., 

Montaggioni, L.F. (1982). Crustal block movements from 

Holocene shorelines: Crete and Antikythira (Greece). 

Tectonophysics, 86, 27-43. 

Pirazzoli, P.A. 1986, Marine notches In: van de Plassche, 

O. (Ed), Sea-level Research: A Manual for the Collection 

and Evaluation of Data, Geo Books, Norwich pp. 361– 

400. 

Pirazzoli, P.A., Montaggioni, L.F., Saliege, J.F., Segonzac, 

G., Thommeret, Y., Vergnaud-Grazzini, C., (1989). 

Crustal block movements from Holocene shorelines: 

Rhodes Island (Greece). Tectonophysics, 170, 89-114. 

Pirazzoli, P.A., Laborel, J., Saliege, J.F., Erol, O., Kayan, I., 

Person, A. (1991). Holocene raised shorelines on the 

Hatay coasts (Turkey): Palaeoecological and tectonic 

implications. Marine Geology, 96, 295-311. 

Pirazzoli, P.A., Stiros, S.C., Laborel, J., Laborel-Deguen, F., 

Arnold, M., Papageorgiou, S., Morhange, C. (1994a). 

Late-Holocene shoreline changes related to 

palaeoseismic events in the Ionian Islands, Greece. The 

Holocene, 4, 397-405. 

Pirazzoli, P.A., Stiros, S.C., Arnold, M., Laborel, J., Laborel- 

Deguen, F., Papageorgiou, S., (1994b). Episodic uplift 

deduced from Holocene shorelines in the Perachora 

Peninsula, Corinth area, Greece. Tectonophysics, 229, 

201-209. 

Shaw, B., Ambraseys, N.N., England, P.C., Floyd, M.A., 

Gorman, G.J., Higman, T.F.G., Jackson, J.A., Nocquet, 

J,-M., Pain, C.C., Piggott, M.D., (2008). Eastern 

Mediterranean tectonics and tsunami hazard inferred from 

the AD 365 earthquake. Nature Geoscience, 1, 268-276. 

Stewart, I.S., Vita-Finzi, C., (1996). Coastal uplift on active 

normal faults: The Eliki Fault, Greece. Geophysical 

Research Letters, 23, 1853-1856. 

Stewart, I.S., Morhange, C. (2009). Coastal geomorphology 

and sea-level change. In J.C. Woodward (Ed.). The physical 

geography of the Mediterranean (pp. 383–413). Oxford: Oxford 

University Press. 

Stewart, I.S., Cundy, A., Kershaw, S., Firth, C., (1997). Holocene 

coastal uplift in the Taormina area, northeastern Sicily: 

Implications for the southern prolongation of the Calabrian 

seismogenic belt. Journal of Geodynamics, 24, 37-50. 

Stiros, S.C., Arnold, M., Pirazzoli, P.A., Laborel, J., Laborel,, 

Papageorgiou, S., (1992). Historical coseismic uplift on Euboea 

Island, Greece. Earth and Planetary Science Letters, 108, 109- 

117. 

Stiros, S.C., Laborel, J., Laborel-Deguen, F., Papageorgiou, S., 

Evin, J., Pirazzoli, P.A., (2000). Seismic coastal uplift in a 

region of subsidence: Holocene raised shorelines of Samos 

Island, Aegean Sea, Greece. Marine Geology, 170, 41-58. 

12


EARTHQUAKE 

ARCHAEOLOGY 



ANALYSING THE LANDSLIDE SUSCEPTIBILITY WITH STATISTICAL METHODS IN 

MAILY-SAY, KYRGYZSTAN 

Braun, Anika (1), Tomas M. Fernandez-Steeger (1), Hans-Balder Havenith (2), Almaz Torgoev(2), Romy Schlögel (3) 

(1) Department of Engineering Geology and Hydrogeology, RWTH Aachen University, Lochnerstraße 4-20, 52064 Aachen, 

GERMANY. Email: braun@lih.rwth-aachen.de 

(2) Georisks and Environment, Department of Geology, University of Liège, B20 Sart Tilmann, 4000 Liège, BELGIUM. 

(3) School and Observatory of Earth Sciences, Institut de Physique du Globe de Strasbourg, 5 rue Descartes, F-67084 

Strasbourg Cedex 

Abstract (Analysing the landslide susceptibility with statistical methods in Maily-Say, Kyrgyzstan): A landslide 

susceptibility analysis was carried out for Maily-Say, a former uranium mining town in Kyrgyzstan. Numerous landslides threaten 

inhabitants, infrastructure and uranium tailings. Besides typical factors responsible in this region, like seismicity, geology, 

geomorphology and climatic conditions, land use plays an important role in Maily-Say. In order to predict landslide susceptibility 

and to clarify the interplay of different factors two statistical methods were implemented, a bivariate statistical method and a data 

mining approach based on a multi-temporal digital landslide inventory. Generally, with both methods areas could be mapped that 

show a high potential for future landslides. Furthermore, the correlation of landslides with the landform and loess deposits provide 

some information on the seismic effects on slope stability. 

Key words: landslide susceptibility, bivariate statistical analysis, data mining 

INTRODUCTION 

In this study a landslide susceptibility analysis in 

Maily-Say, Kyrgyzstan, was undertaken. In the 

vicinity of the former uranium mining and milling town 

more than 200 landslides were present in 2007. 

Besides frequent damage to houses and 

infrastructure, landslides already caused several 

fatalities. Landslides damming the main river during 

spring runoff lead to flooding (Havenith et al., 2006b). 

Numerous radioactive, partially instable uranium 

tailings and waste dumps are threatened by 

landslides. Since the main river leads to the 

Ferghana Valley, a densely populated and 

agricultural region covered mainly by Uzbekistan, the 

destabilisation of nuclear waste tailings in Maily-Say 

bears the potential of a major environmental 

catastrophe (Blacksmith Institute, 2006). 

SETTING 

Maily-Say is located in the western foothills of the 

seismically active Tien Shan high mountain belt, on 

the northern rim of the Ferghana Basin. The Tien 

Shan is an old orogenic belt from Variscan times 

which was reactivated during the collision of India 

and Eurasia 55 Ma ago and started to rise 10 Ma ago 

(Molnar & Tapponier, 1975; Bullen et al., 2001), see 

figure 1. The peaks exceed heights of 7000 m. The 

geology in Maily-Say is related to the transitional 

position between high mountains and a basin 

dominated by partially soft Jurassic, Cretaceous and 

Paleogene sedimentary rocks, see figure 4. A 

relatively weathering resistant formation is the 

Cretaceous limestone that was also mined for 

The term landslide susceptibility implies the spatial 

probability of occurrence of slope failures (Aleotti & 

Chowdury, 1999). By analysing geological and 

geomorphological situations that have lead to slope 

failures in the past, it is possible to predict the 

landslide susceptibility to a certain degree (Varnes, 

1984). In this study two different statistical methods 

were used for predicting the landslide susceptibility in 

Maily-Say, a bivariate statistical approach and data 

mining. Besides pointing out endangered areas, the 

study aimed at analysing the main factors causing 

slope failures in Maily-Say and their temporal 

development with the help of a multi-temporal 

landslide inventory. 

The idea behind this approach is to extract a 

maximum of information from a simple dataset 

(geology, digital elevation model, landslide inventory) 

to provide a first localisation of endangered areas 

without having been to the field. 

Fig. 1: Schematic tectonic map of Southern Central Asia, 

from Bossu et al. (1996) with the outline of figure 2 (a). 

13


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 2: Digital elevation model of Kyrgyzstan (SRTM data from Reuter et al. (2007)) with mapped landslide sites and Ms 5 

earthquakes, based on data of the Kyrgyz and world seismic catalogue. 

uranium. The landscape is characterised by a quiet 

rough relief with heights reaching from 700 m to 

4000 m. The climate is predominantly dry-continental 

with snowfall in winter and high run off in spring. 

Due to the tectonic, geological, geomorphological 

and hydrometerological conditions this region in 

general is highly prone to landslides (Roessner et al., 

2005), see also figure 2). In Maily-Say also the 

uranium mining activities from 1946 to 1968 are 

supposed to have an important impact on the 

destabilisation of slopes. Actually, before 1946 there 

were only few landslides present (Havenith et al., 

2006b). Then, landslide activity increased due to 

mining activities (Torgoev et al., 2002). Until 1962 the 

number of landslides reached 157. The direct link 

between mining activities and landslides may be 

explained by rock weakening because of the 

extraction works, collapse of underground mining 

galleries and rising groundwater levels in the 

abandoned galleries (Havenith et al., 2006b). But 

also indirect processes like changed land use due to 

population growth and increasing traffic may play a 

role. After the mining activities stopped in 1968, the 

increase of landslide activity was going on and even 

accelerating in the 90s. Large landslides, like the 

Koytash, the Tektonik and the Isolith landslides, 

involving up to 5 Million m 3 of rocks and soils formed 

during this period (Minetti, 2004). It is not clear, 

whether this are still long-term effects of the mining 

activities or if other factors, like seismic events or 

climatic trends contribute to this development. For 

instance a massive collapse of the Tektonik landslide 

occurred 7 weeks after a Ms 6,2 earthquake in 1992, 

at 20 km in the south of Maily-Say (Havenith et al., 

2006b). Unfortunately documentation e.g. on 

earthquake damages are scarce, so it is not possible 

to draw clear connections between earthquake 

events and single landslides. 

14 

METHODS, RESULTS AND DISCUSSION 

A multi-temporal landslide and scarp inventory of the 

years 1962, 1984, 1996, 2002 and 2007 as well as 

geological and geomorphological data for the Maily- 

Say Valley were available and prepared in ArcGIS 

(ESRI). The landslide and scarp inventories were 

developed on the basis of existing digital inventories, 

aerial photographs from the 50s and 60s, satellite 

images from 2002 and 2007 and field observations. 

Bivariate statistical analysis 

For the bivariate statistical analysis landslide and 

scarp maps were compared to different factor maps 

on a pixel base using a GIS. According to the method 

of van Westen (1997), a weight index (W i ) was 

calculated expressing the probability of landslide 

occurrence for each raster cell according to the 

distinct parameter. Parameters investigated were e.g. 

slope angle, altitude, curvature, distance to rivers as 

erosion base, geology, distance to faults and loess 

deposits. Loess deposits in the Maily-Say valley 

mainly remained on plateaus which are due to the 

Fig. 3: Histogram showing the W i for scarps vs. loess 

boundary.


EARTHQUAKE 

ARCHAEOLOGY 



lacking slope angle not very vulnerable to landslides. 

Often there are steeper slopes at the verge of these 

plateaus and the loess deposits. This is where loess 

landslides are likely to occur. Therefore, a 30 m 

buffer around the boundaries of loess was also 

analysed. The weight indices were compared for 

scarps and landslide bodies separately and between 

the different years. A landslide susceptibility map was 

created following the InfoVal method (Saha et al., 

2005) based on the 1962 landslide inventory. 

In the following paragraph some interesting results 

are summarized. A seismic triggering is suggested by 

the correlation of the landslide scarps with convex 

landforms and with the boundaries of loess deposits. 

Basically, concave landforms are supposed to be 

more susceptible to landslides because of runoff 

convergence and lower depth to water table. On the 

other hand, in convex landforms the seismic waves 

are amplified when reflected at the surface. Havenith 

et al. (2006a) observed this phenomenon in the 

Suusamyr region, also within the Tien Shan. Hence, 

the correlation of landslide scarps with concave 

landforms indicates a seismic triggering. The 

correlation of landslide scarps with the boundaries of 

loess (see figure 3 and 4) is increasing over time and 

stagnating after 1996. Earthquakes can directly 

trigger loess landslides, e.g. by loess liquefaction as 

observed in China and Tajikistan (Wang et al., 2004) 

or indirectly by forming fractures that allow rapid 

infiltration and resulting collapse during the next 

rainfall event (Havenith & Bourdeau, 2010). The fact 

that this observed trend is rising abruptly between 

1984 and 1996 and then stagnating may indicate a 

connection to the 1992 Ms 6,2 earthquake event near 

Maily-Say. Not only a direct triggering but also a 

weakening effect of the aforementioned processes 

and a temporal delayed failure of slopes is possible. 

In addition to that, evidence for a climatic change 

was figured out. A shift of landslide correlation with 

slope aspect from S in 1962 and 1984 to NW in 1996 

indicates the contribution of a wetter climate to 

increasing slope failure. 

The method caused problems e.g. in implementing 

the geology (figure 4). Here landslides correlate 

strongly with the Sarybia formation, a Jurassic 

sandstone formation. This is because only a very 

small extend of this formation is included into the 

analysis extend. This small extend includes a very 

large landslide, which leads to an overestimation. 

The landslide susceptibility map created with this 

method based on the 1962 landslide inventory was 

compared to landslide inventories of the years after 

1962. The comparison showed agreement between 

regions mapped as highly susceptible and landslides 

that formed after 1962. 

Data Mining 

With data mining methods it is possible to detect 

landslides in a dataset using classification algorithms. 

This method aims at simulating the reasoning 

process, e.g., the one of a geologist as in this case 

(Fernandez-Steeger, 2002). An artificial neural 

network (ANN) and a Bayesian network were 

developed for analysing the multi-temporal landslide 

inventory. Since this method can handle large input 

Fig. 4: Geological map of the working area showing distribution of loess deposits and outlines of landslides in 2007. Geology after 

de Marneffe (2010). 

15


EARTHQUAKE 

ARCHAEOLOGY 



datasets, an abundance of geological, morphological 

and hydrological factors were used as input data. 

The dataset was prepared in a ArcGIS on a pixel 

base, while the main modelling was executed with 

the PASW Modeler 14 (SPSS Inc.), a user interface 

for common data mining algorithms as well as data 

pre- and post-processing The results of the 

classification were transferred back to the GIS for 

evaluation. While the ANN was adjusted quite 

optimal to the landslides and provided good 

classifications, the Bayesian networks developed the 

ability to identify also zones for possible future 

landslides. At this point of the work the contribution of 

each factor to the result is still not very transparent. 

Further development is needed here. Whereas ANNs 

have already successfully been implemented for 

landslide susceptibility analysis (Fernandez-Steeger, 

2002; Lee et al., 2003) there are no studies based on 

Bayesian networks. A comparison between the 

results of the InfoVal method and a Bayesian network 

is shown in figure 5 

Fig. 5: Comparison of the results of the statistical and the 

data mining approach, a) InfoVal method, b) Bayesian 

networks. Both based on landslide data from 1962. 

CONCLUSIONS 

Two statistical methods were applied to analyse 

landslide susceptibility in Maily-Say, Kyrgyzstan. 

Both methods provide the possibility of a remote 

landslide analysis based on relatively simple 

datasets. This is an advantage for remote places like 

Kyrgyzstan, where landslide sites are not easily 

accessible or experts are simply lacking. Especially 

the bivariate statistical method is a simple method 

that provides a useful first idea of landslide 

susceptibility. The data mining approach is also 

promising for the task of predicting landslide 

susceptibility, but the whole process of data 

preparation, modelling and validation requires more 

time and mathematical skills. 

References 

Aleotti, P., Chowdhury, R. (1999). Landslide hazard 

assessment: summary review and new perspectives. 

Bulletin of Engineering Geology and the Environment, 58, 

21–44. 

Blacksmith Institute (2006). The world’s worst polluted 

places-The Top Ten. 

Bossu, R., Grasso, J. R., Plotnikova, L. M., Nurtaev, B., 

Fréchet, J., Moisy, M. (1996). Complexity of 

Intracontinental Seismic Faultings: The Gazli, 

Uzbekistan, Sequence. Bulletin of the Seismological 

Society of America, 86, 959–971. 

Bullen, M.E., Burbank, D.W., Garver, J.I., Abdrakhmatov, 

K.Y. (2001). Late Cenozoic tectonic evolution of the 

northwestern Tien Shan: New age estimates for the 

initiation of mountain building. Geological Society of 

America Bulletin, 113, 1544–1559. 

De Marneffe, C. (2010): Cartographie et modelisation 3D de 

la geologie de la vallee de Mailuu-Suu, Tien Shan, 

Master's Thesis, University of Liège, unpublished, 93p. 

Fernandez-Steeger, T. M. (2002). Erkennung von 

Hangrutschungssystemen mit Neuronalen Netzen als 

Grundlage für Georisikoanalysen. Ph.D. Thesis, 

Karlsruhe University. 

Havenith, H. B., Bourdeau, C. (2010). Earthquake-induced 

Landslide Hazards in Mountain Regions: A Review of 

Case Histories from Central Asia. Geologica Belgica, 13, 

135–150. 

Havenith, H. B., Strom, A., Caceres, F., Pirard, E. (2006a). 

Analysis of landslide susceptibility in the Suusamyr 

region, Tien Shan: statistical and geotechnical approach. 

Landslides, 3, 39–50. 

Havenith, H. B., Torgoev, I., Meleshko, A., Alioshin, Y., 

Torgoev, A., Danneels, G. (2006b). Landslides in the 

Mailuu-Suu Valley, Kyrgyzstan-Hazards and Impacts. 

Landslides, 3, 137–147. 

Lee, S., Ryu, J.-H., Lee, M.-J., Won, J.-S. (2003). Use of an 

articial neural network for analysis of the susceptibility to 

landslides at Boun, Korea. Environmental Geology, 44, 

820–833. 

Minetti, L. (2004). Kyrgyz Republic proposed naturall 

disaster mitigation project, 1st IDA Preparation Mission. 

Ministry of Environment and Emergency (MEE), Bishkek, 

40p, inedit. 

Molnar, P., Tapponier, P. (1975). Cenozoic Tectonics of 

Asia: Effects of a continental collision. Science, 189, 

419–426. 

Reuter, H. I., Nelson, A., Jarvis, A. (2007). An evaluation of 

void filling interpolation methods for SRTM data. 

International Journal of Geographic Information Science. 

21:9, 983-1008. 

Roessner, S., Wetzel, H. U., Kaufmann, H., Sarnagoev, A. 

(2005). Potential of Satellite Remote Sensing and GIS for 

Landslide Hazard Assessment in Southern Kyrgyzstan 

(Central Asia). Natural Hazards, 35, 395–416. 

Saha, A. K., Gupta, R. P., Sarkar, I., Arora, M. K., 

Csaplovics, E. (2005). An approach for GIS-based 

statistical landslide susceptibility zonation-with a case 

study in the Himalayas. Landslides, 2, 61–69. 

Torgoev, I. A., Alioshin, Y. G., Havenith, H. B. (2002). 

Impact of uranium mining and processing on the 

environment of mountainous areas of Kyrgyzstan. In: 

Merkel, Planer- Friedrich and Wolkersdorfer (eds), 

Uranium in the aquatic environment, pp. 93–98, Springer, 

Berlin Heidelberg New York. 

Wang, L., Wang, Y., Wang, J., Li, L., Yuan, Z. (2004). The 

Liquefaction Potential of Loess in China and its 

Prevention. 13th World Conference On Earthquake 

Engineering, Vancouver, B.C., Canada, 13p. 

Vandenhove, H. Q. H., Clerc, J.-J., Lejeune, J.-M., Sweeck, 

L., Sillen, X., Mallants, D., Zeevaert, T. (2003). Final 

report in frame of EC-TACIS Project N SCRE1/N 38 

remediation of uranium mining and milling tailing in 

Mailuu-Suu district Kyrgyzstan, Mol (Belgium). 

Van Westen, C. J. (1997). Statistical landslide hazard 

analysis. In: Application guide, ILWIS 2.1 for Windows. 

ITC, Enschede, The Netherlands, 73–84. 

Varnes, D. J. (1984). Landslide hazard zonation: a review of 

principles and practise. UNESCO, Paris, 63 p. 

16


EARTHQUAKE 

ARCHAEOLOGY 



TECTONIC INTERPRETATION OF THE 2008 WENCHUAN EARTHQUAKE: WHY IT ONLY 

PROPAGATED IN ONE DIRECTION - THE FUTURE? 

Burchfiel, B. C. (1), Royden, L. H. (1) 

(1) Department of Earth, Atmospheric and Planetary Sciences, 54-1010, Massachusetts Institute of Technology, Cambridge, MA 

02139, USA. EMAIL: bcbruch@mit.edu 

Abstract (Tectonic interpretation of the 2008 Wenchuan earthquake): The 2008 Wenchuan earthquake, (M = 7.9) occurred on 

a listric thrust fault that at the hypocenter (15-20 km) dipped ~30-40° NW and steepened upward to near vertical at the surface 

(Zhang et al., 2008) Maximum slip was 9 m vertical and 6 m right-slip. The fault propagated from the epicenter to the northeast 

for ~ 200 km with increasing right-slip eastward and broke across at least two fault segment boundaries. Although the earthquake 

occurs along the steep topographic slope of the Longmen Shan at eastern margin of the Tibet plateau, it was unexpected because 

data suggests that the recurrence interval on the fault zone was 2500 to 4000 years. 

Key words: Earthquake, Wenchuan, China, 2008 

INTRODUCTION 

The earthquake fault, Pengguan-Beichuan fault, 

broke in two segments, the main segment was along 

the east side of a major uplift of Precambrian 

basement rocks, the Pengguan massif, in the 

Longmen Shan at the eastern margin of the Tibet 

plateau, southwestern Sichaun, China, and 

propagated northeast into Paleozoic rocks. The 

Pengguan uplift is a northeast plunging anticline with 

its sedimentary cover still exposed at the north end. 

The east side of the anticline is the Pengguan- 

Beichuan fault. Because of the plunge to the north 

the east vergent thrust sheets are exposed also to 

the north and are cut by the Pengguan-Beichuan 

fault (figure 1). These geological relations were 

discussed by Burchfiel et al., (1995). 

The question of why the fault only propagated to the 

northeast and its long recurrence interval can be 

explained by the geological framework in the area 

that was known in 1995. The west side of Pengguan 

uplift and the Pengguan-Beichuan fault are cut off at 

the south end by the Wenchuan-Maowen fault, a 

steep fault that has active right-slip, but was not 

reactivated during the earthquake, that also cuts off 

the sedimentary cover and Mesozoic thrust sheets so 

that the southern Penguan Precambrian has no 

sedimentary cover and it is not clear how much it has 

been uplifted. How these two faults interact is critical, 

but is at present unknown. The Wenchuan-Beichuan 

fault has a low-grade mylonitic fabric that shows the 

west side down (a normal fault), that is consistent 

with the geology where the rocks west of the 

Penguan massif are metamorphosed Paleozoic that 

belong to the Mesozoic thrust sheets. 

DISCUSSION 

Why the earthquake did not propagate also to the 

southwest and what are the prospects of a large 

earthquake to the southwest can be explained by the 

geological framework of the area. To the south, the 

Precambrian is uplifted again in the Baoxing massif 

that also has the characteristics of an east vergent 

fold with a steep thrust along its eastern side (Fig. 1). 

The Wenchuan-Maowen fault appears to continue 

south and cuts off the west side of the Baoxing uplift 

and Paleozoic and even Triassic strata lie west of the 

fault against Precambrian rocks to the east. 

Regionally east of the Chengdu plain (the Quaternary 

area east of Pengguan uplift) is an single curving 

fold, the Longchaun anticline) that intersects the 

Longmen Shan obliquely in the north but trends away 

from the Longmen Shan to the south and curves 

south east of Chengdu and continue farther south 

east of Emei shan (Figure 1). Within the southern 

Chengdu plane several north-plunging anticlines 

appear and grow in amplitude southward so that the 

elevation to the south rises and rapidly becomes the 

eastern part of the Tibetan Plateau. These folds 

involve Precambrian rocks in the south. Thus what 

appears to be a horizontal decollement with the 

sedimentary section beneath the Chengdu plane in 

the north that connects the Longchaun anticline with 

the Pengguan-Beichuan fault, in the south the 

decollement must drop into the Precambrian rocks 

where the topography becomes part of the Tibetan 

Plateau. 

The geometry of active deformation suggests 

stresses focused mainly at the Pengguan uplift in the 

north, become distributed southward from the 2008 

epicenter across a broad of the area suggesting that 

the stress is being relieved in the southern area 

where is more distributed and being relieved by a 

broader and more active area of seismic activity. 

However, along the topographic eastern margin of 

plateau, mountains increase in elevation so that at 

the Siguniang Shan (4 girls) they reach more than 

6200 m and at Gonga Shan farther south they reach 

7550 m. The mountain front east of the 4 girls is very 

17


EARTHQUAKE 

ARCHAEOLOGY 



steep, suggesting again rapid uplift like along the 

Pengguan massif. This is similar to the conditions at 

the area of the Wenchuan earthquake, and needs to 

be tectonically and geomorphologically researched 

for recurrence interval in this area that may also be 

longer that recorded history. looked at in terms of the 

even though the stress appears to be more 

distributed to the east in this region. Like all geology 

relations there are ambiquities: the steepness of the 

topographic margin of the plateau in this area is a 

clue as to how structurally active the plateau margin 

is but the broader distribution of active structures and 

earthquakes suggests stresses are being distributed. 

Farther east there is second steep mountain front on 

the east side of Emei Shan with an active fault at its 

base. Another alarming relationship. 

The change in structure from north to south east of 

the Longmen Shan can be interpreted to be related 

to the geometery of the Xianshuihe fault, an arcuate 

NW-tending fault major active right-slip that 

separates two major crustal provinces in SW China: 

a province to the south that is part of a crustal 

fragment that rotates clockwise at 10-12 mm/yr 

around the Eastern Himalayan syntaxis, from the 

Longmen Shan province that moves NE ~parallel to 

the eastern margin the Tibetan Plateau with only a 

very slow convergence rate (1-2 mm/yr) with respect 

to South China to the east as shown by GPS data. In 

fact, the convergence rate is so slow that the eastern 

part of the Tibetan Plateau moves east nearly at the 

same rate as South China as first shown by King et 

al. (1997). The slow convergence rate across the 

topographic margin of Tibet Plateau explains the 

large recurrence interval of the Wenchuan 

earthquake, but does not easily explain the steep 

topographic margin and the rise in elevation of the 

front to the south. 

West of Kanding Xianshuihe fault curves more 

sharply that either to the NW or SE forming a 

restraining bend indicates there is a component of 

compression across the fault that forms the high 

topographic area east of the fault and where 

basement rocks are involved in the structure. At the 

bend the mountains reach 7550m where the large 

Cenozoic pluton that underlies the Gonga Shan. If so 

the it is also where to the east of the fault the Tibetan 

plateau was elevated and the folds that come from 

the southern part of the Sichuan basin involve 

Precambrian rocks. There are cross structures here; 

WNW trending thrust faults that involve Precambrian 

cut across the N-S trending folds from the southern 

Sichuan basin that also involve Precambrian. The 

decollement here is in the basement beneath the 

folds and probably beneath the huge NW-trending 

Danba antiform of Precambrian east of Kanding. 

This raises the question of how deep is the 

decollement and how does it interact with the 

Xianshuihe fault; the decollement is deep and the 

Xianshuihe fault may be a crustal feature. 

Fig. 1: Generalized map of the Longmen Shan and 

Southern Sichuan basin area showing structures of late 

Cenozoic to active age. Faults are in red decorated with 

barbs for thrust faults, arrows for strike-slip faults and 

double ticks for normal faults. Blue are the folds of late 

Cenozoic age, except in the southeast where they are of 

Middle Cretaceous age (dark blue). These structures 

overprint the early Mesozoic Longmen Shan thrust faults 

that moved the Songpan Ganze Unit (pale red) eastward 

above the Yangtze Unit (white).The area of the 

southwestern Sichuan basin shown in yellow, the Chengdu 

plain, has a thin cover of Pleistocene sediments derived 

from the Longmen Shan and ponded behind the active 

Longquan anticline (LQA). Large black dot at south end of 

Pengguan massif (PM) is location of the 2008 Wenchuan 

earthquake epicenter. 4G=Four Girls peaks, AF=Anninghe 

fault, BF=Beichuan fault, BSM=Baoxing massif, DA=Danba 

antiform, EA=Emei Shan anticline, GS=Gonga Shan, 

HF=Huya fault, PM=Pengguan massif, SF=Shimian fault, 

WMF=Wenchuan-Maowen fault, XF=Xianshuihe fault, 

XSP=Xushan platform, XPA=Xiong Po anticline. 

Acknowledgements: This work was supported by NSF 

Grant 000357. 

References 

Burchfiel, B. C., Z. Chen., Y. Liu, & L. H. Royden, (1995). 

Tectonics of the Longmen Shan and adjacent regions. 

International Geological Review, 37(8), 661-736. 

King, R. W., F. Shen, B. C. Burchfiel, Z. Chen, Y. Liu, L. H. 

Royden, E. Wang, X. Zhang, & J. Zhao, (1997). 

Geodetic measurement of crustal motion in southwest 

China. Geology, 25 (2), 179-182. 

Zhang, P-Z., X-Z. Wen, Z-K. Shen, & J-H. Chen, (2008). 

Oblique, High-Angle, Listric-Reverse Faulting and 

Associated Development of Strain: The Wenchuan 

Earthquake of May 12, 2008, Sichuan. Annual reviews 

of Earth and Planetary Sciences, 38, 353-382. 

18


EARTHQUAKE 

ARCHAEOLOGY 



CHARACTERIZATION OF LATE PLEISTOCENE-HOLOCENE EARTHQUAKE-INDUCED 

“HOMOGENITES” IN THE SEA OF MARMARA THROUGH MAGNETIC FABRIC. 

IMPLICATION FOR CO-SEISMIC OFFSETS DETECTION AND MEASUREMENTS 

Campos Corina (1)(2), Beck Christian (1), Crouzet Christian (1), Carrillo Eduardo (3) 

(1) Institut des Sciences de la Terre ISTerre, UMR CNRS 5275, Grenoble University, 73376 Le Bourget du Lac Cedex, 

France. Email : Corina.Guzman@etu.univ-savoie.fr 

(2) Departamento de Ciencias de la Tierra, Universidad Simón Bolívar, Sartenejas, Baruta, Venezuela. 

Email: campossc@usb.ve. 

(3) Instituto de Ciencias de la Tierra, Universidad Central de Venezuela, Av. Los Ilustres, Caracas, Venezuela. 

Email: eduardo.carrillo@ciens.ucv.ve. 

Abstract: The Marmara Sea is located in the eastern part of the Mediterranean region, in an area with strong seismic instability 

associated with the activity of North Anatolian Fault. In order to analyze the impact of seismicity on the sedimentation in the 

Marmara Sea we have studied three giant piston cores (27 to 37.7m long). They represent the last 20 kyr of sedimentation. The 

upper section (marine stage) is predominantly composed by fine grained terrigenous material (clay-silty) and in less percentage by 

the silty-sandy laminated intervals. The lower section (non marine) is composed by abundant fine grained terrigenous material, 

numerous turbidites sequences and levels with deformation structures (possible seismites). Some turbidites show an abrupt 

contact separating the coarse grain basal part (bed load) of the fine grain upper part (suspended load), this later are defined as 

“homogenites” 

Key words: Sea of Marmara, Earthquakes, Turbidites, Homogeneites. 

Introduction 

The Marmara Sea (Northwestern Turkey) is a pullapart 

basin developed along the North Anatolian 

Fault (Hancock and Barka, 1981). East-West 

elongated (200 km) it is composed of several aligned 

sub-basins (Tekirda, Orta, Kumburgaz and Çinarcic 

basin) (Fig.1). The North Anatolian Fault (N.A.F) is a 

1200-km-long dextral strike-slip fault (engör et al., 

2004) and is considered as a major active boundary 

between Anatolia and Eurasia plates. (Armijo el al., 

1999; McClusky et al., 2000; Flerit et al., 2003). 

The northern branch of the N.A.F. crosses the 

different deep sub-basins, where giant pistons cores 

(from 27 to 37,7 m) have been retrieved, with location 

based on high resolution seismic reflection imaging. 

This survey aimed to detect and characterize coseismic 

sedimentary episodes - especially 

“homogenites” – for a 20 000 yr-long period. Different 

previous studies have demonstrated the interest of 

sedimentary record (isolated marine basins and 

lakes) as archives of seismic activity (Hempton and 

Dewey, 1983; Calvo et al., 1998; Chapron et al., 

1999; among many others). Movements of water 

masses (seiche effect, reflected turbidites) and mass 

wasting (triggered by major earthquakes) are inferred 

to combine themselves for specific depositions which 

locally compensate vertical offset of seafloor, as in 

the Sea of Marmara central basin (Beck et al, 2007; 

Stegmann et al., 2007; Strasser et al., 2006). Thus, 

the present work is dedicated to the Central (Orta) 

and eastern (Çinarcic) basins to detect, characterize, 

and correlate, these sedimentary “events”, as a 

contribution to pecise paleoseismic data (associated 

fault offset, chronology). 

Fig.1 Simplified geodynamic setting of the Sea of 

Marmara for the present day with coring locations (after 

Armijo et al., 1999; Beck et al., 2007). GPS velocity 

vector with fixed Eurasia Plate, from McClusky et al., 

2000. The Anatolia Plate is bounded by major strike-slip 

faults systems (North and East) and subduction (South). 

The objective of this work is identify and characterize 

the impact of seismicity over the sedimentation and 

distinguish them from “normal” sedimentary 

processes as hemipelagic-type fallout and flooding, 

19


EARTHQUAKE 

ARCHAEOLOGY 



subsequently we hope to determine the earthquake 

recurrence and contribute to improve hazard 

estimates. We have studied 3 giant piston cores 

collected during the MARMACORE Cruise in August- 

Septembre 2001. The cores were taken in the 

Çinarcic basin (MD01-2425), and in the Orta basin 

(MD01-2429, MD01-2431), at depths between 1230 

and 1170m. The sedimentary record in theses cores 

represent the sedimentation of the last 20 kyr. 

METHODS 

One of the challenges is the distinction between finegrained, 

slow and continuous, hemipelagic 

sedimentation, from quite instantaneous resuspension 

and re-deposition. For this we combined 

different tools in order to characterize the textures: 

grain size (laser diffraction grain size analyzer, 

Malvern TM ), anisotropy of magnetic susceptibility 

(Kappabridges MFK1-FA AGICO) and X-ray imagery 

(D.G.O.’s SCOPIX). Compositions were controlled 

through magnetic susceptibility (BARTINGTON TM 

MS2 contact sensor), and microscopic analysis. The 

chronology was established from 

14 C (AMS 

measurements) derived from wood, plants and 

sapropelic muds. We calibrated the ages with Oxcal 

Program v4.1. The ages found represent the 

Holocene and part of the late Pleistocene. 

RESULTS 

The analyzed sections are composed of fine grained 

terrigenous material (clay-silty) intercalated with siltysandy 

laminated intervals, turbidites sequences and 

liquefaction features as ball and pillow. The upper 

marine part is predominantly compose by the claysilty 

slightly calcareous and in less percentage by the 

silty-sandy laminated intervals, theses intervals 

consist of milimetric’s lenticular and parallel planar 

beddings. 

In the lower non marine part the fine grained 

terrigenous material is abundant but we find 

numerous turbidites sequences with thicknesses 

ranging from centimeters to decimeters. This 

turbidites can show erosive bases, normal gradation 

and ripples, others can show an abrupt contact 

separating the coarse grain basal part (bed load) of 

the fine grain upper part (suspended load)(Fig.2), this 

later are defined as “homogenites”. In this non 

marine section we can also find in less proportion the 

slumps and the levels with deformation structures 

(possible seismites). 

Fig.2 Textural and compositional characterization of three homogenites present in the non marine section (core MD01-2425). The 

mean size, magnetic susceptibility and lineation magnetic don’t show differences between the hemipelagic normal sedimentation 

and the homogeneous coseismic sedimentation. The foliation magnetic (AMS) is higher in the homogenites than the hemipelagic 

deposits, this high foliation contrast has to be explained by different arrays of phyllosilicates, associate a specific settling 

conditions related to water mass oscillation. 

20


EARTHQUAKE 

ARCHAEOLOGY 



The Magnetic Susceptibility signature in the marine 

part is in general lower than the non marine part 

(average in the marine part is10x10 -5 SI and 30x10 -5 

SI in the non marine part). In general this signature is 

strong in the silty-sandy laminated intervals and in 

the turbidites sequences. In the turbidites sequences 

this signature has a behavior similar to the one 

observed in the profiles of grain-size, being higher in 

the sand layers than the fine grain, and showing 

constants values in the homogenites. On the other 

hand, the foliation determined from the Anisotropy of 

Magnetic Susceptibility is much higher in the levels 

the fine grain size belonging to the homogenites 

(average of 1.12) than any other level of thin grain 

size (average 1.06) (Fig.2). As both types of levels 

(homogenites and hemipelagic deposits) have similar 

grain-sizes and not very different composition, this 

high foliation contrast has to be explained by different 

arrays of phyllosilicates, which, at their turn, cannot 

be explained by different compactions. Thus, we infer 

specific settling conditions related to water mass 

oscillation. 

In the study area others analyses are in processes 

such as the carbonates contents, clay minerals, 

analysis of terrigenous fraction (mineralogy), organic 

matter, etc. 

References 

Armijo, R., Meyer, B., Hubert, A., Barka, A. (1999). 

Westward propagation of the North Anatolian Fault into 

the northern Aegean: timing and kinematics. Geology 27 

(3), 267–270. 

Beck C., Mercier de Lepinay B., Schneider J.L., Cremer M., 

Cagatay N., Wendenbaum E., Boutareaud S., Menot G., 

Schmidt S., Weber O., Eris K., Armijo R., Meyer B., 

Pondard N., Gutscher M.A., Turon J.L., Labeyrie L., 

Cortijo E., Gallet Y., Bouquerel H., Gorur N., Gervais A., 

Castera M.H., Londeix L., de Resseguier A., Jaouen A. 

(2007) Late Quaternary co-seismic sedimentation in the 

Sea of Marmara's deep basins, Sedimentary Geology, 

199, 65-89. 

Calvo, J.P., Rodriguez-Pascua, M., Martin-Velazquez, S., 

Jimenez, S., De Vicente, G. (1998). Microdeformation of 

lacustrine laminate sequences from Late Miocene 

formations of SE Spain: an interpretation of loop bedding. 

Sedimentology 45, 279–292. 

Chapron, E., Beck, C., Pourchet, M., Deconinck, J.-F. 

(1999). 1822 AD earthquake-triggered homogenite in 

Lake Le Bourget (NW Alps). Terra Nova 11, 86–92. 

Flerit, F., Armijo, R., King, G.C.P., Meyer, B., Barka, E. 

(2003). Slip partitioning in the Sea of Marmara pull-apart 

determined from GPS velocity vectors. Geophysical 

Journal International 154, 1–7. 

Hancock, P.L., Barka, A.A. (1981). Opposed shear senses 

inferred from neotectonic mesofractures systems in the 

North Anatolian fault zone. J. Struct. Geol. 3, 383–392. 

Hempton, M.R., Dewey, J.F. (1983). Earthquake-induced 

deformational structures in young lacustrine sediments, 

East Anatolian Fault, southwest Turkey. Tectonophysics 

98, 7–14. 

McClusky, S., Bassalanian, S., Barka, A., Demir, C., 

Ergintav, S., Georgiev, I., Gurkan, O., Hamburger, M., 

Hurst, K., Hans-Gert, H.-G., Karstens, K., Kekelidze, G., 

King, R., Kotzev, V., Lenk, O., Mahmoud, S., Mishin, A., 

Nadariya, M., Ouzounis, A., Paradissis, D., Peter, Y., 

Prilepin, M., Relinger, R., Sanli, I., Seeger, H., Tealeb, A., 

Toksöz, M.N., Veis, G.(2000). Global positioning system 

constraints on plate kinematics and dynamics in the 

eastern Mediterranean and Caucasus. Journal of 

Geophysical Research 105, 5695–5719. 

Sengör, A.M.C., Tuysuz, O., Imren, C., Sakinc, M., 

Eyidogan, H., Gorur, N., Le Pichon, X., Claude Rangin, 

C. (2004). The North Anatolian Fault. A new look. Ann. 

Rev. Earth Planet. Sci. 33, 1-75. 

Stegmann, S., Strasser, M., Anselmetti, F.S., Kopf, A. 

(2007). Geotechnical in situ characterisation of 

subaquatic slopes: The role of pore pressure transients 

versus frictional strength in landslide initiation. 

Geophysical Research Letters, 34, L07607 

Strasser, M, Anselmetti, F.S., Fäh, D., Giardini, D., 

Schnellmann, M. (2006). Magnitudes and source areas of 

large prehistoric northern Alpine earthquakes revealed by 

slope failures in lakes. Geology, 34 (12) 

21


EARTHQUAKE 

ARCHAEOLOGY 



PALEOSEISMOLOGICAL EVIDENCE FOR HISTORICAL SURFACE RUPTURE EVENTS IN 

S. MIGUEL ISLAND (AZORES) 

Carmo, Rita (1, José Madeira (2), Ana Hipólito (1), Teresa Ferreira (1) 

(1) Centro de Vulcanologia e Avaliação de Riscos Geológicos, Universidade dos Açores, Complexo Científico, 3º Piso – Ala Sul, 

Rua Mãe de Deus, 9500-321 Ponta Delgada, Açores, Portugal. Email: Rita.L.Carmo@azores.gov.pt 

(2) Dpto. Geologia, LATTEX/IDL, Faculdade de Ciências da Universidade de Lisboa, Edifício C6, Campo Grande, 1749-016 

Lisboa, Portugal. Email: jmadeira@fc.ul.pt 

Abstract (Paleoseismological evidence for historical surface rupture events in S. Miguel Island (Azores)): 

The Azores archipelago is located in the triple junction between the Eurasian, Nubian and North American 

lithospheric plates. The Achada das Furnas plateau, located in the central part of S. Miguel Island, between Fogo 

and Furnas volcanoes, is dominated by several basaltic cinder cones that define several WNW-ESE and E-W 

alignments. Two E-W trending scarps were identified by aerial photo analysis. Trenches were open across the 

scarps to confirm their tectonic nature exposing two active normal faults (the Altiprado Faults). At least four 

paleoearthquakes were deduced, three of which in historical times. Radiocarbon ages are in agreement with this 

interpretation. 

Key words: Azores, active faults, paleoseismology 

GEOLOGICAL SETTING 

The Azores archipelago is located at the Eurasia 

(Eu), Nubia (Nu) and North America (NA) triple 

junction (Fig. 1). The Mid Atlantic Ridge separates 

the NA from Eu and Nu plates, while the Azores- 

Gibraltar Fault Zone (Terceira Rift and Gloria Fault) is 

the boundary between Eu and Nu plates. The 

archipelago comprises nine islands distributed by 

three groups: the western islands lie on NA plate 

while the central and eastern groups are located 

along the western segment of the Azores-Gibraltar 

Fault Zone (Fig. 1). As a result of its complex tectonic 

setting, the Azores archipelago is subject to frequent 

seismic and volcanic activity. 

S. Miguel Island was settled in 1439-1443. It has 

three active explosive central volcanoes with summit 

calderas linked by zones of fissural volcanism. The 

main tectonic structures trend NW-SE to WNW-ESE, 

NNW-SSE and NE-SW. 

The Achada das Furnas plateau, located in the 

central area of the island, between Fogo and Furnas 

volcanoes, is dominated by several basaltic cinder 

cones defining WNW-ESE and E-W alignments (Fig. 

2). Aerial photo analysis identified the presence of 

two E-W trending scarps produced by the Altiprado 

Faults (AF1 and AF2, Fig. 2). 

Fig. 1: Azores tectonic setting. NA – North American 

plate; Eu – Eurasian plate; Nu – Nubian plate; MAR – 

Mid Atlantic Ridge; TR – Terceira Rift (s.l.); EAFZ – 

East Azores Fracture Zone; GF – Gloria Fault. World 

topography and bathymetry from ESRI (2008). 

Fig. 2: Main tectonic and volcanic structures of 

Achada das Furnas plateau. 

22


EARTHQUAKE 

ARCHAEOLOGY 



THE ALTIPRADO FAULTS 

Altiprado Fault 1 

The Altiprado Fault 1 (AF1) trace is marked by an 

835 m-long, 3 m-high, south-facing scarp, trending 

N87ºE (Fig. 3). To the east and to the west its trace 

becomes uncertain. 

A ~18 m-long trench exposed the fault, trending 

N87ºE and dipping 65ºS, displacing a stratigraphic 

succession that comprises 6 pumice fall deposits 

(units) separated by paleosols produced by eruptions 

of Fogo and Furnas volcanoes (Fig. 4): 1 - olivebrown 

paleosol containing basalt lapilli fragments, 

corresponding to Fogo A deposit (4520±90 years BP; 

Wallenstein, 1999); 2 – pumice fall deposit composed 

of alternating yellowish lapilli and light olive brown 

ash beds, probably related to Fogo C deposit; 3 - 

strong brown pumice ash fall deposit with some lapilli 

at the base (3a), and a yellowish pumice fall deposit 

in the central part (3b); probably Fogo D deposit; 4 - 

grey pumice ash deposit, corresponding to Furnas C 

deposit (1900 years BP; Guest et al., 1994), topped 

by a very dark grey soil (1440-1500 cal AD); 5 - 

stratified pumice fall deposit of alternating beds of 

fine to medium greyish white lapilli and ash 

containing sanidine crystals (from 1563 AD historical 

eruption in Fogo volcano), topped by a very dark grey 

paleosol rich in coal fragments (1660-1700 cal AD); 6 

- remobilised deposit from the underlying unit. 

The AF1 affects all stratigraphic units and the 

existing scarp is an uneroded free-face almost devoid 

of soil that corresponds to the projection of the fault 

to the surface (Fig. 4). Several WNW-ESE to E-W 

trending fractures and open cracks, sometimes filled 

with material from overlying units, and a colluvium 

composed of material of units 3 and 4 were also 

exposed in the trench (Fig. 4). 

Units 2 to 4 (soil 1440-1500 cal AD) are displaced by 

1.0m and units 5 (soil 1660-1700 cal AD) and 6 by 

0.38m, indicating two surface rupturing 

paleoearthquakes of Mw 6.7 and 6.4 (using the Wells 

& Coppersmith’s, 1994, M/MD correlation), 

respectively, accounting for an accumulated dip-slip 

of 1.38m. 

Two earthquakes in 571-511 years indicate a 

recurrence interval that ranges from 286 to 256 

years, yielding a slip rate of 2-3 mm/year. 

Altiprado Fault 2 

The Altiprado Fault 2 trace is marked by an almost 

imperceptible 1690 m-long and ~40 cm-high south 

facing scarp, trending N87ºE (Fig. 3). To the east its 

location becomes uncertain. 

Fig. 3: Vertical aerial photograph of Altiprado region, Achada das Furnas plateau, showing the geomorphic expression of 

Altiprado Faults (Aerial photo from Direcção Geral de Planeamento Urbanístico, 1974). 

Fig. 4: Map of the east wall of the Altiprado Fault 1 trench. 

23


EARTHQUAKE 

ARCHAEOLOGY 



A 29 m-long trench exposed two subvertical faults 

(AF2-1 and AF2-2) that display frequent changes in 

dip sense (N75-89ºE, generally dipping 62-88ºS) and 

an E-W trending paleo-channel (filled with unit 6) 

(Fig. 5). The exposed stratigraphic sequence is the 

same observed in AF1 trench, with exception of unit 

1 that is absent (Fig. 5). There are also open 

fractures trending ENE-WSW to E-W, sometimes 

filled with material from overlying units, and a 

colluvium, containing material of unit 3, deposited 

against the AF2-1 fault (Fig. 5). 

The faults produced differential vertical separations 

on units 2 and 3 (younger than ~4500 years BP), and 

on units 4 (~1900 years BP; soil 1440-1500 cal AD) 

and 5 (1563 AD): AF2-1 - accumulated dip-slip of 

33cm (26+7cm); AF2-2 - accumulated dip-slip of 

15cm (11+4cm). Assuming that the ruptures in AF2-1 

and AF2-2 were produced by the same earthquakes, 

displacement values of 0.37m (0.26+0.11cm) and 

0.11m (0.07+0.04cm), correspond to earthquakes of 

Mw of 6.4 and 6.0, respectively (using the Wells & 

Coppersmith’s, 1994, M/MD correlation). The 

accumulated displacement is 0.48m (0.37m+0.11m). 

The existence of an E-W trending paleo-channel 

suggest that it may have been developed at the base 

of a previous fault scarp, once the regional drainage 

system in this area is oriented N-S. 

Altiprado Faults evolution 

The Altiprado Faults evolution was deduced from 

geometric analysis of the trenches. As they are 

geographically close and affect the same 

stratigraphic succession, this allows correlating their 

evolution (Fig. 6): 

a) Deposition of units 1 (4520±90 years BP) to 3; 

b) Surface rupture at AF2-1 and AF2-2 with normal 

separations of 26cm and 11cm, respectively (Mw 6.4 

Fig. 5: Map of the west wall of the Altiprado Fault 2 trench. 

earthquake); 

c) Erosion truncating unit 3, with the formation of 

gullies in AF1 fault zone, and fault scarp retreat at 

AF2-1 with the formation of a colluvial wedge (C) with 

material from unit 3; 

d) Deposition of unit 4 (~1900 years BP) and soil 

development (1440-1500 cal AD); 

e) Surface rupture (Mw 6.7) at AF1 with normal 

separation of 1.0m; 

f) Erosion with minor fault scarp retreat with formation 

of a colluvial wedge (C) in AF1; 

g) Deposition of unit 5 (1563 AD – Fogo Volcano 

historical eruption) with sin-eruptive ruptures (Mw 

6.0?) in AF2-1 and AF2-2, without geomorphic 

expression, of 7cm and 4cm, respectively; 

h) Erosion truncating the top of unit 5 and formation 

of a new paleo-channel in AF2 fault zone; 

development of a soil (1660-1700 cal AD); 

i) Erosion truncating unit 5 and deposition of unit 6; 

j) Surface rupture (Mw 6.4) at AF1 with normal 

separation of 0.38m; 

k) Erosion truncating units 5 and 6; development of 

present top soil. 

DISCUSSION 

Analysing the geological history, the first surface 

rupture earthquake is associated to AF2, originating 

an accumulated dip-slip of 37cm in AF2-1 and AF2-2. 

It occurred in pre-historical times, after the deposition 

of unit 3 (Fogo D deposit,


EARTHQUAKE 

ARCHAEOLOGY 



earthquakes in the island and the youthful aspect of 

the scarp, this event is probably related to October 

22 nd , 1522 historical earthquake. This was one of the 

most destructive events that occurred in Azores, 

causing 5000 deaths. The epicentre was located 

inland, a few km southwards from AF1 (Fig. 7) and 

triggered several landslides and severe damage, 

mainly in the central eastern part of the island. 

Fig. 8: Epicentre distribution of 2005 seismic swarm 

(May 10 th 

to December; data from CVARG/CIVISA, 

2005). 

Fig. 7: Isoseismal map for October 22 nd , 1522 

earthquake (Silveira et al., 2003). 

The third rupture is associated with AF2. As it is not 

clear if the last displacement (11cm in AF2-1 and 

AF2-2) is affecting only the lower part of unit 5 (1563 

AD) or affects the entire unit, this could be associated 

with the intense seismic activity that accompanied 

the 1563 historical eruption of Fogo volcano. The 

earthquakes were felt in a wide region and caused 

severe damage in Ribeira Grande and Ribeira Seca 

villages. Otherwise it could have occurred after this 

volcanic event. 

The fourth earthquake is related with AF1 and 

occurred after the deposition of unit 6 (>1660-1700 

cal AD) with normal separation of 0.38m. 

If the last three earthquakes occurred in historical 

period, the recurrence interval for Altiprado Faults 

zone is 163 years. 

There is no evidence of a high magnitude earthquake 

(Mw 6.4) with inland epicentre in historical records, 

as well in the instrumental seismicity data that could 

justify the most recent displacement. However, there 

is a long time record of seismic swarms in this area, 

with earthquakes of low to moderate magnitude (e.g. 

1967, 1989 and 2005). 

The most recent seismic swarm started on May 10 th , 

2005 and continued till the end of that year, 

registering more than 46000 earthquakes (Fig. 8). 

The strongest events occurred on September 20 th 

and 21 th and had ML 4.1 and 4.3, respectively, 

causing several landslides and ground cracking. 

GPS monitoring data allowed Trota (2008) to verify 

that there was also ground deformation associated 

with the seismic activity. Considering the seismicity 

pattern, the low seismic energy released and the 

GPS data, this activity has been related to a 

magmatic intrusion. The previous seismic swarms 

were not geodetic monitored and the seismic network 

was limited, so there are no evidences of previous 

episodes of crustal deformation and the seismic 

parameters may have significant errors associated. 

Nevertheless, one possibility is that the 0.38m 

displacement observed in AF1 trench could be an 

accumulated value as a consequence of several 

moderate earthquakes during the periods of seismic 

activity increments in this area that might have been 

associated with crustal deformation. 

Acknowledgements: Rita Carmo is supported by a PhD 

Grant from Fundo Regional da Ciência e Tecnologia. 

Fieldwork was financed by Centro de Informação e 

Vigilância Sismo-Vulcânica dos Açores (CIVISA). 

References 

Guest, J.E., Duncan, A.M., Cole, P.D., Gaspar, J.L., 

Queiroz, G., Wallenstein, N., Ferreira, T. (1994) – 

Preliminary report on the volcanic geology of Furnas 

volcano, São Miguel, Azores. ESF Furnas Laboratory 

Volcano, Eruptive History and Volcanic Open File Report 

1, 24 p. 

Silveira, D., Gaspar, J.L., Ferreira, T., Queiroz, G (2003) - 

Reassessment of the historical seismic activity with major 

impact on S. Miguel island (Azores). Natural Hazards and 

Earth System Sciences 3, 615-623. 

Trota, A. (2008) - Crustal deformation studies in S. Miguel 

and Terceira islands (Azores). Volcanic unrest evaluation 

in Fogo/Congro area (S. Miguel). PhD thesis, Azores 

Univ., 281p. 

Wallenstein, N. (1999) – Estudo da história recente e do 

comportamento eruptivo do Vulcão do Fogo (S. Miguel, 

Açores). Avaliação preliminar do “hazard”. PhD thesis, 

Azores Univ., 266p. 

25


EARTHQUAKE 

ARCHAEOLOGY 



FAULT TECTONICS REGARDING THE NEOTECTONIC PERIOD AND INFLUENCE OF 

TECTONIC STRUCTURES ON GLACIAL PROCESS IN AREAS OF THICK QUATERNARY 

COVER 

Jolanta yžien 

Lithuanian Geological Survey. Konarskio, 35. 03123-Vilnius. LITHUANIA. Email: jolanta.cyziene@lgt.lt 

Abstract (Fault tectonics regarding the neotectonic period and influence of tectonic structures on glacial process in 

areas of thick Quaternary cover): Through the Quaternary period Lithuania has been covered by continental ice sheets 

originated in Fennoscandia which corresponds to all glaciations known so far in the Eastern Europe, this causing very complicated 

structure of the Quaternary cover. This paper discusses the coincidence of outlines of main morphological features with regional 

tectonic structures and also whether tectonic processes influence on Quaternary development. To understand the tectonic 

processes influence on Quaternary sediments, neotectonic structures and movements were estimated on the basis of studies of 

the sub-Quaternary relief, Quaternary succession, modern relief and drainage system. A comparison of sub-Quaternary relief and 

present surface with tectonic structure and neotectonic activity of studied territories has revealed a frequent coincidence of linear 

and areal geological and geomorphological objects. 

Key words: Quaternary, (neo)tectonics, Baltic, palaeoinsicions 

INTRODUCTION 

The territory of Lithuania and adjacent areas in the 

Baltic Sea Region, located in the south-western 

margin of the East European Craton, could be 

considered as a region of low seismic activity due to 

the Early Precambrian crust and distant location to 

the active tectonic regions. Nevertheless, more than 

40 seismic events (historical and instrumental) with a 

M4.5 and intensities up to VI-VII (MSK-64 scale) 

were reported in the Baltic countries and adjacent 

territories since year 1616 (Pasa, 2007), and two 

strong earthquakes that took place in the Kaliningrad 

region on 21 September 2004, the magnitudes being 

4.4 and 5.0. Recent seismic activity of Baltic Sea 

Region often related to glacio-isostatic rebound of the 

Fennoscandian Shield, but also could be triggered by 

plate-scale North Atlantic ridge-push forces (Pascal 

et al., 2010). The territory of Lithuania is one of 

classical regions with Quaternary cover formed 

during continental glaciations. The average thickness 

of Quaternary cover in Lithuania is approximately 130 

m and varies from 10-30 m in the northern part of 

country – the area of prevailing glacial erosion – up 

to 200-300 m in marginal highlands and the buried 

valleys or palaeoincisions. The processes of 

accumulation, erosion during the glaciations and icefree 

periods, developed under the influence of 

neotectonic movements, have created the wide 

variety of Quaternary sediments and landforms. 

These processes made great impact creating the 

present shape of the sub-Quaternary surface and 

modified tectonic structures. Structural units of the 

landscape generated by a combination of tectonic 

activity and climate (morphostructures) often coincide 

with deep pre-Quaternary structures (Šliaupa, 1998) 

The coincidence of the faults defined by geophysical 

and well data to morphotectonic lineaments was 

considered as strong evidence, determining its 

“activity” during the glacial and post-glacial times 

(Šliaupa, 2003). To understand the tectonic 

processes influence on Quaternary sediments, 

neotectonic structures and movements were 

estimated on the basis of studies of the sub- 

Quaternary relief, Quaternary succession, modern 

relief and drainage system. This paper presents 

discussions on the possibilities to detect 

(neo)tectonic characteristics in the regions with 

Quaternary cover formed during continental 

glaciations and assumptions for the morphotectonic 

evidence. 


Lithuania is situated within the Baltic sedimentary 

basin. The sedimentary succession of Lithuania is 

subdivided into four major structural-sedimentary 

complexes: Baikalian, Caledonian, Hercynian and 

Alpine and consists of Upper Vendian to Quaternary 

sedimentary rocks resting on an Early Proterozoic 

crystalline basement. The thickness of the 

sedimentary cover ranges from 0.2 km in the East to 

2.3 km in west Lithuania. All of the complexes are 

separated by unconformities within the sedimentary 

succession that represents periods of non-deposition 

and erosion. A set of faults are recognized in the 

sedimentary cover that are most distinct in the 

Caledonian complex. The displacements of some 

reverse faults exceed 200 m in western Lithuania 

(e.g. Telšiai fault). The oldest traces of the tectonic 

activity are recorded in the sub-Jotnian rocks that 

were preserved in a small area in west Lithuania. A 

limited extensional faulting took place in Vendian and 

Cambrian times, as it was identified in seismic 

26


EARTHQUAKE 

ARCHAEOLOGY 



profiles in western and central Lithuania. This 

extensional event(s) is related to the initial stages of 

establishment of the passive margin due to breaking 

apart of the Rodinian supercontinent. In western 

Lithuania these basement blocks were draped by 

Cambrian rocks, and they are overlain by Vendian 

deposits in the eastern Lithuania. It was rather quiet 

during Ordovician and Silurian times, no structuring 

was documented except some gentle flexuring. The 

tectonic forces drastically increased in latest Silurian 

– earliest Devonian times relating to far-field stress 

transmission generated by Scandian orogeny in 

Scandinavian Caledonides. The dense family of 

compressional and transpersional faults was 

established in western Lithuania, while faulting was 

only scarce in the eastern half of Lithuania that is 

accounted to stronger lithosphere and longer 

distance the stress source. The W-E and WSW-ENE 

oriented faults show transpressional geometries, 

whereas the NE-SW striking faults are compressional 

features (Šliaupa et al., 2002). The morphology of 

faults is rather variable, ranging from single-plain 

fault to complex flower structure and terraced fault 

sets. The faulting associated to formation of local 

uplifts along the hanging walls of the faults. The 

Telšiai reverse fault, formed in transpressional 

regime, is the largest tectonic feature in the 

sedimentary cover of Lithuania. This fault trends W-E 

for a few hundred kilometres, it shows variable 

geometry along the strike that associate with 

changing abundance and scale of local uplifts. Faults 

are rather rare in the younger structural complexes 

where flexures are most common features of tectonic 

displacement. Structural-sedimentary complexes 

differ by their geological composition and 

independently by the structural pattern. Especially 

high variety of genesis and lithological composition of 

the Quaternary deposits is reflected in drastic 

changes both in vertical and lateral distribution of 

Quaternary sediments. Surface formations and 

geomorphological features in a large part of 

Lithuanian territory were formed during the Late 

Nemunas (Late Weichselian) Glaciation. 

ASSUMPTION FOR THE MORPHOTECTONIC 

EVIDENCE 

The studies focusing on mutual relationships 

between glacial landforms and tectonic structures in 

areas glaciated in the Pleistocene usually take into 

account two aspects. The first one concerns the 

influence exerted by pre-glacial tectonic structures on 

the behaviour of the ice-sheet, controlling thereby 

same glacial landforms. One of the issues under 

consideration is the influence of fault zones 

reactivated by ice-sheet load upon location and 

course of subglacial tunnels. The second aspect is 

related to post-glacial neotectonic movements, the 

most prominent manifestations of which are 

tectonically-controlled erosional scarps, either 

undermining or displacing glacial landforms. 

Essential question concerns coincidence of outlines 

of main morphological features with regional tectonic 

structures and also whether tectonic processes 

influence on Quaternary development, and in which 

way. 

Neotectonic movements of the Earth’s crust are 

reflected by deformation processes often 

accompanied by tectonic activity. Number of 

publications reported that the last deglaciation of 

northern Fennoscandia was accompanied by a high 

seismic activity. The earthquakes triggered landslides 

in glacial till, seismically-induced soft sediment 

deformation structures, “seismites”, are common in 

trench exposures in the vicinity of the faults in 

northern Sweden. Deformation of sandy-silty 

sediments, potentially caused by earthquakes, have 

also been encountered in central and southern 

Sweden (e.g., Mörner, 2004), but are less common 

than in the north. The deglaciation history in 

Lithuania is longer, but there is so far not recorded 

and no published evidence of paleoseismic events. It 

must be pointed out that the late- to postglacial fault 

scarps identified in northern Sweden are all 

developed in the Precambrian crystalline basement, 

and mainly in rocks of Proterozoic age. 

Morphologically prominent faults occur also in the 

Caledonian bedrock in the mountain range, but so far 

no recent fault movements along any of these 

features have been indicated (Lagerback & Sundh, 

2008). The geological setting of Lithuania is very 

different compared to Sweden. The study area is 

covered entirely by Quaternary sediments of glacial 

origin (average thickness of sediments is 130 m). 

Two major types of faults prevail in Lithuania, i.e. the 

oldest, defined only in the Precambrian crystalline 

basement and do not dissecting the sedimentary 

cover and younger that penetrate into the sediments 

overlying the crystalline basement. In comparison to 

Sweden, no faults so far have been detected in 

Quaternary succession. Fault tectonics regarding the 

neotectonic period in Lithuania, as well as in all Baltic 

region, is rather subtle problem. It is difficult to 

determine the influence of tectonic structures on 

glacial process in the Pleistocene and testing its 

influence on the process of the mass movements in 

glaciotectonic and neotectonic young-alpine 

structures. Tectonic deformations of the sub- 

Quaternary relief during Neogene-Quaternary or 

Quaternary period, i.e. a part of a tectonic and 

denudation factors imprinting the sub-Quaternary 

surface, are the key problem. Numerous evidences 

reported from the Baltic Region indicate that steps, 

elevations and depressions of the sub-Quaternary 

surface are partly of tectonic nature (Šliaupa & 

Popov, 1998). 

Large-scale landforms and neotectonic movements 

were investigated in close connection to structural 

features of the pre-Quaternary and Quaternary 

deposits. Those movements were predetermined at 

large by ancient pre-Quaternary tectonic structures 

showing inherited trends of movements. In Lithuania 

the commonly used term “neotectonic fault zone” 

concerns, in fact, mostly a system of linear elements 

defined by remote sensing, morphometric, and 

geomorphological-structural investigations. The 

influence of linear tectonic zones is shown in a sharp 

change of the composition and structure of the 

Quaternary cover. Based on available data of 

geological mapping at scale 1:50 000 (coverage c.a. 

27


EARTHQUAKE 

ARCHAEOLOGY 



48 % of the country), especially from the areas with 

dense network of boreholes, several factors having 

morphotectonic implications must be pointed out: 

- block structures of Quaternary cover; 

- palaeoincisions of pre-Quaternary surface and 

inside of Quaternary cover; 

- linear structures of present topography; 

- river valleys (tunnel valleys); 

- ravines; 

- palaeolacustrine basins and zones of distortions; 

- glaciodislocations, rafts of pre-Quaternary rocks; 

- intrusions of mineralised water and springs 

(yžien & Satknas, 2008). 

DISCUSSION 

A particular role belongs to deep palaeoincisions of 

sub-Quaternary surface and tunnel valleys of present 

topography. The palaeoincisions are distinct feature 

of pre-Quaternary surface and particularly are proper 

for the Baltic Highland reaching even 280 m in depth. 

Network of especially deep palaeoincisions is 

determined in the Moltai Lakeland (north of Vilnius). 

Genetically palaeoincisions of pre-Quaternary 

surface are analogous to the tunnel valleys, which 

were formed during subglacial erosion by meltwater 

under the glaciodynamic pressure. In the cases of 

clear correspondence of palaeoincisions with tunnel 

valleys, their morphotectonic implication could be 

concluded (yžien & Satknas, 2008). However, 

reliable determination of spatial form and presence of 

network of palaeoincisions requires very dense 

network of boreholes and is dependent on ways of 

interpretation and interpolation of topography of pre- 

Quaternary surface (e.g. Šliaupa et al., 1999). 

Therefore, different patterns of forms and networks of 

palaeoincisions are presented by different 

researchers. 

The problem of the genesis of palaeoincisions of sub- 

Quaternary surface in Lithuania is of great practical 

importance. The genesis of these particular forms is 

discussed taking into account all factors that could 

influence the formation of palaeoincisions in the 

environment of continental glaciations. Analysis of 

palaeohydrography during the interglacial periods 

shows that basis of erosion could not be the reason 

for incision of river valley 100-120 m below the 

present sea level. The basis of fluvial erosion at the 

end of the Neogene-beginning of the Quaternary 

most probably was not lower than 50-60 m above the 

present sea level and this could not affect incision or 

river valleys deeper than +50 m (Satknas, 2000). 

Bottom of deepest palaeoincisions of Vilnius environs 

occur at an altitude of 10-56 m below the present sea 

level. According to the age of tills, occurring in 

vicinities of the studied palaeoincisions, these forms 

have been formed during the Dainava (Elsterian) and 

Žemaitija (Saalian) glaciations. The lithology and 

rhythmic structure of the material filling the 

palaeoincisions show that they have been formed as 

a consequence of subglacial erosion under glacial 

pressure. Morphological similarity of palaeoincisions 

and ravines of present topography indicated the 

same origin of both these forms. The ravine Tapeliai 

(10 km north-east of Vilnius) was studied, using 

special borehole data. Sand and gravel (total 

thickness 44 m) filling the ravine has a rhythmic 

structure characteristic of the palaeoincisions of sub- 

Quaternary surface. This property also shows the 

general similarity of ravines of present topography 

and palaeoincisions (Satknas, 2008). 

Tunnel valleys of present topography are revealed on 

the geomorphological map (Guobyt, 2000) and their 

good correspondence with photo-lineaments are 

concluded (Guobyt, 1995). Besides that, the tunnel 

valleys in most cases are interpreted as 

neotectonically active linear zones (Šliaupa, 2005). 

The coincidence of palaeoincisions and the tunnel 

valleys in many cases still has to be confirmed by 

boreholes, which are generally absent in the tunnel 

valleys. Therefore, due to lack to the direct data the 

palaeoincisions in places seems hardly correlating 

with tunnel valleys (yžien & Satknas, 2008). On 

the other hand, it still remains underestimated 

presence of paleoincisions that do not reach the 

surface of pre-Quaternary rocks. 

A comparison of sub-Quaternary relief and present 

surface with tectonic structure and neotectonic 

activity of territories has revealed a frequent 

coincidence of linear and areal geological and 

geomorphological objects. 

Acknowledgements: Thanks to the Marie Curie 

Programme „Transfer of Knowledge“ in the Sixth 

Framework Programme for the support of the project 

"Morphotectonic Map of European Lowland Area”. 

References 

yžien, J., Satknas, J. (2008). Morphotectonic of 

Quaternary landforms, examples from Baltic region. 

Volume of abstracts of the third conference of MELA 

project Cartographical approach of the morphotectonic of 

European lowland area, 20-23. 

Guobyt, R. (1995). Geological mapping of the Moltai 

area, 1:50 000. Geological Report No. 5478, Lithuanian 

Geological Survey (LGT). 

Guobyt, R. (2000). Geomorphological map of Lithuania, 

1:200 000. Geological Report No. 4378, Lithuanian 

Geological Survey (LGT). 

Lagerbäck, R., Sundh, M. (2008). Early Holocene faulting 

and paleoseismicity in northern Sweden. Research Paper 

C 836, Geological Survey of Sweden, 80 p. 

Mörner, N.-A. (2004). Active faults and paleoseismicity in 

Fennoscandia, especially Sweden. Primary structures 

and secondary effects. Tectonophysics, 380, 139-157. 

Pasa, A. (2007). Seismological observation in Lithuania. 

Volume of abstracts of the international workshop 

Seismicity and seismological observations of the Baltic 

sea region and adjacent territories, September 10-12, 

2007, Vilnius, Lithuania, LGT, 66-69. 

Pascal, C., Roberts, D., Gabrielsen, R.H. (2010). Tectonic 

significance of present-day stress relief phenomena in 

formerly glaciated regions. Journal of the Geological 

Society, 167, 363-371. 

Satknas, J. (2000). Formation of palaeoincisions in the 

environment of continental glaciations: an example of 

Eastern Lithuania. Geologija, 31, 52-65. 

Satknas, J. (2008). Morphotectonic implication of tunnel 

valleys in areas with thick Quaternary cover. Volume of 

abstracts of the third conference of MELA project 

28


EARTHQUAKE 

ARCHAEOLOGY 



Cartographical approach of the morphotectonic of 

European lowland area, 42-45. 

Šliaupa, A. (1998). Neotectonic structures of Lithuania and 

Adjacent Territories. Litosfera, 2, 37-46. 

Šliaupa, A. (2005). Neotectonic map of Lithuania and 

adjacent areas, 1:1 000 000. Evolution of geological 

environment in Lithuania, Vilnius, Lithuanian Geological 

Survey, Report, CD. 

Šliaupa, S., Popov, M. (1998). Linkage between Basement 

and Neotectonic Linear Structures in Lithuania. Litosfera, 

2, 23-36. 

Šliaupa, S. (2002). Kinematic features of the Telšiai fault in 

Western Lithuania: structural and permeability prognosis.. 

Geologija, 38, 24-30. 

Šliaupa, S. (2003). Geodynamic evolution of Baltic 

sedimentary basin. Doctoral thesis, Geological Institute, 

Vilnius, 207 p. 

29


EARTHQUAKE 

ARCHAEOLOGY 



PLIO – PLEISTOCENE TECTONIC ACTIVITY IN THE SOUTHWEST OF PORTUGAL 

Figueiredo, Paula M. (1, 2), Cabral, João (1,2), Rockwell, Thomas K. (3) 

(1) IDL, Instituto Dom Luiz, pmfigueiredo@fc.ul.pt 

(2) Geology Department, Science Faculty, Lisbon University 

(3) Geological Department, San Diego State University 

Abstract: Southwest Portugal, located close to the Eurasia-Núbia plate boundary, is characterized by moderate 

seismicity, although strong events may occur, as in 1755 (Mw8), 1969, (Mw 7.9), and more recently in 2007 (Mw 

6.3) and 2009 (Mw 6.1), all located in the offshore. No historical earthquakes with onshore rupture are known for 

this region. Inland, recent neotectonic and paleoseismological studies corroborate seismogenic activity during the 

Pleistocene, at the S.Teotónio–Aljezur–Sinceira Fault System, suggesting that these structures are active and are 

potential sources for moderate seismicity. At the coastline, several features such as old beach sediments, paleo 

abrasion platforms and paleo cliffs were recognized. A sequence of poorly preserved surfaces with thin deposits 

may correspond to Pleistocene marine terraces, suggesting a higher uplift rate than expected for this region. 

Key words: Pleistocene, Uplift, active tectonics, Portugal 

INTRODUCTION 

Southwestern Portugal is located close to the 

Eurasia-Nubia plate boundary, near the Azores- 

Gibraltar fracture zone. East of the Gloria transform 

fault, this boundary becomes complex and diffuse, 

where deformation is distributed across a few 

hundred kilometres wide zone, related to the NW-SE 

convergence of Iberia and Nubia at a rate of ca. 4-5 

mm/yr. In this area is located the inferred 

seismogenic source zone of the 1755 earthquake 

and tsunami (estimated Mw 8), and also of the Mw 

7.9 1969 event. (Zittelini et al., 2009). 

Fig. 1: Main Neotectonic structures located in southwern 

Portugal. Onshore structures: APF, Alentejo-Placencia 

Fault; STASFS, São Teotónio-Aljezur-Sinceira Fault 

System. Offshore Structures: PSF, Pereira de Sousa Fault; 

MPF, Marquês de Pombal Fault; HSF, Horseshoe Fault; 

GBF, Guadalquivir Bank Faults. 

and other faults trending close to E-W, such as the 

Guadalquivir Bank and SWIM faults (Figure 1). To 

understand and study the recent tectonic activity in 

this sector of Iberia, it is necessary to study the 

individual inland structures but also inland 

deformation evidences that may be related to some 

of the referred offshore active structures. This 

abstract depicts both perspectives. 

DISCUSSION 

The São Teotónio –Aljezur- Sinceira Fault System 

The São Teotónio–Aljezur–Sinceira fault system 

(STASFS) corresponds to the nearest inland brittle 

structure that may correlate to the ongoing plate 

boundary deformation in the offshore. It extends 

NNE-SSW for 50 km, parallel and close to the 

southwest Portuguese coast (Figure 1), and 

comprises left-lateral strike-slip faults, deforming a 

large regional abrasion platform ca. 10 km wide, 

considered of probable late Miocene age, and which 

was reoccupied during the Pliocene and the 

Pleistocene. Four small Cenozoic tectonic basins, 

filled with Miocene to Pleistocene sediments, occur 

along the STASFS. 

Post-Pliocene vertical displacements of up to 100 m 

may have occurred related to tectonic activity in the 

STASFS, but generally these only reach a few tens 

of meters. Dias (2001) estimated a slip-rate of ca. 

0.03-0.06 mm/yr, based on the vertical offset of 

morphological features. However, as the main slip on 

these structures is strike-slip, those values are a 

minimum estimative. Dias (2001), taking into account 

This seismogenic source zone is characterized by 

several structures trending NNE-SSW to NE-SW, as 

the Marquês de Pombal and the Horseshoe faults, 

30


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 2: South wall of trench ALF1, located at the Alfambras basin. Description of geology included in the figure. 

the length of the known fault also estimated a 

maximum magnitude of Mw 7, an average co-seismic 

displacement of 1 m and a recurrence period of 

19210 - 32017 yr, based upon Wells & Coppersmith 

empirical relationships. 

Recently, through detailed geomorphologic and field 

work studies, several paleoseismic sites were 

selected. At the Alfambras basin, two trenches were 

excavated: in the ALF-1 trench (Figure 2), one of the 

active (?) fault branches was identified, although no 

paleo-earthquakes were individualized. This fault 

deforms post-Miocene sediments that may be 

Pleistocene. OSL samples were collected to better 

constrain the ages. A paleosoil, probably 700 ka old, 

is also faulted showing a vertical displacement of ca. 

1 m. We saw no deformation within the upper soil 

units; therefore we could not state an obvious 

correlation between the fault trace and the 

topography which would imply a more recent activity. 

Since no piercing points were identified, no lateral 

displacement was quantified at this site. 

existence of several fault traces near the topographic 

surface. 

The Framangola site (at Alfambras basin) is located 

close to a feature recognized as a shutter ridge and 

two trenches were opened at the alluvial plain to 

investigate the Holocene sequence. Unfortunately, 

the very coarse sediments and the high water table 

did not allow us to safely investigate this trench site 

at a deeper level: a thick gravel sequence was 

recognized as likely to be related to several 

landslides triggered during the 1755 earthquake 

based upon archaeological artefacts. 

3 

2 

1 

Fig.4. Geoelectric tomography profile (using Schlumberger 

method) at Framangola site at the alluvial plain, strongest 

contrast coincides with previously identified fault trace; the 

trench units identified also correlate with the tomography (1) 

Clays and fine silts; (2) Very coarse Gravel (3) Silty sandy & 

fine conglomerate 

Fig.3: Geoelectric tomography profile (using 

Schlumberger method) at Monte Ferreiros site, strong 

contrast coincides with geomorphologic trace interpreted 

in the lower image. 

More paleoseismic sites were selected farther north 

in the Aljezur and Alfambras basins, where 

geoelectric tomography profiles (Figures 3 & 4) up to 

50 m depth were obtained corroborating the 

At its northern end, near São Teotónio, the STASFS 

joins the Alentejo-Placencia fault (APF), the iberian 

fault with higher length. In this area the tectonic 

deformation becomes distributed, suggesting than 

APF may have several splays along this intersection 

and it is difficult to scrutinize the relationship between 

both structures. No Holocene or late Pleistocene 

deposits were recognized in association with the 

identified fault segments, which increase the 

uncertainty concerning the recent activity. Presently it 

is still not clear whether the STASFS extend to the 

north of the APF. 

31


EARTHQUAKE 

ARCHAEOLOGY 



Towards the south, at the southern sector of the 

STASFS, south of the Sinceira basin, the local 

morphology does not evidence any significant Plio- 

Pleistocene vertical deformation related with this fault 

system, and it seems that this fault system may splay 

into several faults. In fact, several parallel faults 

occur in the Mesozoic bedrock sediments where 

limestones dominate. Several karst pits are filled with 

Plio-Pleistocene sands (Faro-Quarteira sands) 

mainly at the coastal section Martinhal – Zavial 

where the STASFS should intersect the coastline. 

The elongated shape those karst sometimes present, 

suggested a structural control by a previous 

structural fabric on the karst development (Dias and 

Cabral, 2002). 

The Coastal region 

The southwestern Portuguese littoral is characterized 

by coastlines with two distinct morphologies: one 

trending N-S, presenting high cliffs in Paleozoic 

schists and greywackes that reach up to 100 m 

height, with several overhanging fluvial valleys and a 

narrow abrasion platform, and another trending E-W, 

formed mainly in Mesozoic limestones, forming an 

irregular and lower coastline, with karst wells filled 

with Plio-Pleistocene sands (generally of the Faro- 

Quarteira Sands regional stratigraphic unit) as 

referred above. 

abrasion platform considered late Miocene in age 

and reoccupied during Pliocene and Pleistocene 

times (figure 5). 

The highest beach sediments we were able to 

identify are at Fonte Santa, circa 350 m elevation, 

and underlying aeolionites considered to be Pliocene. 

In the western coast, north of Sagres, where the cliffs 

are very abrupt and the drainage is strongly incised, 

we recognized a raised beach consisting of a coarse, 

pebbly layer underlying beach sands at 76 m height 

(Figures 5 & 6). Two samples of sand were collected 

for OSL dating. The basal unconformity surface 

apparently dips gently southwards (~2º). 

Further to the north, at the Castelejo beach (Figure 5 

& 7), we identify an aeolionite with several paleosoils 

and colluviums that overlies a marine abrasion 

platform on Palaeozoic schists circa 2 m elevation, 

almost coincident with the modern one. We collected 

OSL samples at the base of this sequence with a 

resulting age of ca.64 ka. Underlying the aeolionite 

and overlying the abrasion platform two 

conglomerate beach deposits with a thin beach sand 

layer between them were identified and OSL samples 

were collected. We interpreted this to represent 

aeolian deposition during MIS 4, when sea level was 

lowered to expose the offshore sandy marine 

sediments, and the underlying platform to be the late 

stage 5 (MIS 5a) marine terrace. Correlative 

sediments were also identified further to the north, at 

Amoreira beach. 

ae 

Fig.5 Digital Terrain Model of the Sagres region, southwest 

STASFS showing the location of sites referred in the text. 

bc 

bs 

bc 

Fig. 6 Pleistocene beach sediments resting over a poorly 

preserved abrasion surface cut on Triassic sediments at 65- 

76 m height at Telheiro (adapted from Dias, 2001). Beach 

sands overlie a basal pebbly deposit. At the top of the 

sequence occur several eolionites apparently with distinct 

paleosoils, suggesting several eolic sedimentation events. 

Detailed field surveys were conducted in both 

coastlines in order to recognize paleo shore-line 

features and beach deposits in a wide raised 

Fig. 7 Pleistocene beach deposit lying over an abrasion 

platform cut on Carboniferous schists, at 2 m height (bc - 

beach conglomerate; bs – beach sand). Above this deposit 

outcrops a sequence of strongly carbonate cemented 

eolianites, likely to be MIS 4 (ae). 

At the E-W trending southern coast, detailed 

geomorphologic studies combined with field survey 

and paleosoils characterization, suggest that the 

culminant abrasion platform, which is well bc preserved 

near Sagres, dipping gently to the Southeast, may 

32


EARTHQUAKE 

ARCHAEOLOGY 



actually correspond to a sequence of closely spaced, 

poorly preserved, middle Pleistocene marine 

terraces. 

A raised abrasion platform cut on Mesozoic 

limestones, at ~12 m elevation, and the 

corresponding paleo-cliff were recognized at the 

Ingrina beach, though no beach sediments were 

encountered (Figure 8). This raised abrasion platform 

at Ingrina, probably corresponds to the MIS 5e 

abrasion surface. 

poorly expressed, early to middle Pleistocene 

terraces. OSL samples and marine shells were 

collected for dating at Telheiro, where beach deposits 

are present at 76 m elevation. In the same region at 

Castelejo, we dated an aeolionite that directly 

overlies a marine abrasion platform at 2 m elevation, 

with a resulting age of ca. 64 ka. We interpreted this 

to represent aeolian deposition during MIS 4, when 

sea level was lowered to expose the offshore sandy 

marine sediments, and the underlying platform to be 

the late stage 5 (MIS 5a) marine terrace. 

The 12 m terrace at Ingrina, cut across carbonate 

rocks, probably corresponds to the MIS 5e abrasion 

surface. If this uplift rate is applicable for the entire 

Quaternary, this implies a late Miocene to Pliocene 

age of the aeolionites and marine deposits at Fonte 

Santa, which lie at an elevation of nearly 350 m. 

These observations imply a long-term uplift rate of 

about 0.06 mm/yr. 

These are still preliminary results, to confirm with 

further studies. 

Fig. 8. Pleistocene raised abrasion platform at Ingrina 

beach, at 12m elevation. 

West of Ingrina beach, there is a surface slightly 

under 50 m height, that probably corresponds to an 

abrasion surface, with several residual reliefs up to 5 

m high. This surface is overlain by a thin clayey sand 

deposit, which is generally covered by abundant 

rounded quartz pebbles, thus suggesting that the 

surface may correspond to a marine terrace at 45-50 

m. Samples were collected for OSL dating. A poorly 

preserved surface around 60-65 m may correspond 

to the surface identified at Telheiro, but no beach 

sediments were recognized. 

All these surfaces are covered by several eolianite 

bodies, likely to have been reactivated trough time. 

CONCLUSIONS 

Trenches excavated across the São Teotónio– 

Aljezur–Sinceira fault system (STASFS), which 

extends NNE-SSW for 50 km parallel to the 

southwest Portuguese coast, exposed faulted 

alluvium that is inset below the regional marine 

abrasion surface. Sparse age control based on 

paleosol development, along with the geomorphic 

position of these alluvial terrace deposits relative to 

the marine deposits, suggests that this faulting 

occurred in the middle to late Pleistocene timeframe, 

although we have yet to find evidence of Holocene 

activity. 

We mapped a raised abrasion platform at an 

elevation of ~75 m near Sagres, which dips gently to 

the SE and may represent part of a sequence of 

We attribute the observed deformations to the 

continued NW-SE 4 to 5 mm/yr convergence of Iberia 

and Nubia, with the STASFS accommodating some 

of the ongoing plate boundary activity. 

Acknowledgements : This work was funded by Fundação 

da Ciência e Tecnologia, through a PhD scholarship 

(SFRH/BD/36892/2007) and Research Project 

“Paleoseismological study of active faults in Mainland 

Portugal” (PTDC/CTE-GIN/66283/2006), co-financed by 

FEDER. 

References 

Cabral, J. (1995). Neotectónica em Portugal Continental, 

Mem. Nº31, Inst.Geol.Min., 265 p. 

Carrilho, F. (2005). Estudo da Sismicidade do Sudoeste de 

Portugal Continental, Diss. Mestrado, Dep.Fisica, Univ. 

Lisboa, 172 p. 

Dias, R. (2001). Neotectónica da Região do Algarve, 

Dissertação de Doutoramento, Univ. Lisboa, 369 p. 

Dias, R. P., Cabral, J. (2002): Interpretation of recent 

structures in an area of cryptokarst evolution - 

neotectonic versus subsidence genesis. Geodinamica 

Acta, 15 (4), 233-248. 

Manuppella, G. (Coord.) (1992) - Carta Geológica da 

Região do Algarve, escala 1/100 000., Serviços 

Geológicos de Portugal. 

Wells, D., Coppersmith, K. (1994). New empirical 


width, rupture area and surface displacement. Bulletin 

of the Seismological Society of America, Vol.84 (4), 

974-1002. 

Zitellini N., Gracia E., Matias L., Terrinha P., Abreu M.A., 

DeAlteriis G., Henriet J.P., Danobeitia J.J., Masson 

D.G., Mulder T., Ramella R., Somoza L., Diez S. (2009) 

- The quest for the Africa–Eurasia plate boundary west 

of the Strait of Gibraltar, Earth and Planetary Science 

Letters Vol.280, 1-4, 15 April 2009, 13-50. 

33


EARTHQUAKE 

ARCHAEOLOGY 



GEODETIC EVIDENCE OF THE CONTROL OF A MAJOR INACTIVE TECTONIC 

BOUNDARY ON THE CONTEMPORARY DEFORMATION FIELD OF ATHENS (GREECE) 

Michael Foumelis (1, Ioannis Fountoulis (2), Ioannis D. Papanikolau (3, 4), Dimitrios Papanikolaou (2) 

(1) Department of Geography, Harokopio University of Athens, 70 El. Venizelou Str., Kallithea, 176 71, Athens, Greece, 

Email: mfoumelis@hua.gr 

(2) Department of Dynamics Tectonics & Applied Geology, Faculty of Geology and Geoenvironment, National and Kapodistrian 

University of Athens, Panepistimioupolis, Ilissia, 157 84, Athens, Greece 

(3) Laboratory of Mineralogy & Geology, Department of Geological Sciences and Atmospheric Environment, Agricultural 

University of Athens, 75 Iera Odos Str., 118 55, Athens, Greece 

(4) AON Benfield UCL Hazard Research Centre, Department of Earth Sciences, University College London, WC 1E 6BT, 

London, UK 

Abstract (Geodetic evidence for control of a major inactive tectonic boundary on contemporary deformation field of 

Athens (Greece)): A GPS-derived velocity field from a dense geodetic network established in the broader area of Athens is 

presented, whereas local variations of strain rates across a major inactive tectonic boundary separating metamorphic and nonmetamorphic 

geotectonic units are also highlighted. An apparent differentiation of the eastern part of Athens plain with negligible 

deformation rates, from the western part where relatively higher strain rates are observed, indicate its control of the above 

mentioned boundary on the contemporary deformation field of the region. These findings are in agreement with previous 

geological observations, however, due to the dense local GPS network it was fatherly possible to localize and quantify the effect of 

such a major inherited tectonic feature on the deformation pattern of the area. 

Key words: GPS velocities, strain rates, tectonics, Athens Basin 

INTRODUCTION 

Detailed instrumental observations of the tectonic 

movements in Athens Basin by geodetic or other 

methods are absent. The contribution of previous 

geodetic GPS studies to examine the kinematic field 

of Attica are limited to observations from regional 

networks, designed to monitor large-scale rather than 

local tectonic movements (Clarke et al., 1998; Veis et 

al., 2003). With a limited number of stations within 

the region, a general picture of the motion is gained, 

while changes within are hardly addressed. 

In the present study a comprehensive GPS-derived 

velocity field for the broader area of Athens is 

presented. Variations of strain rates across a major 

tectonic boundary occurring in the region are 

highlighted and implication on the contemporary 

kinematics and dynamics of the region are 

discussed. 


The Athens basement belongs to alpine formations 

outcropping in the mountains and the hills of the 

area. Recent post-alpine sediments (syn-rift deposits) 

often cover the slopes of the mountains as well as 

areas of low altitude. 

The area presents a complex alpine structure 

comprising mainly by Mesozoic metamorphic rocks of 

Attica geotectonic unit, occurring at Pendeli and 

Hymmetus mountains and Mesozoic nonmetamorphic 

rocks of the Eastern Greece 

geotectonic unit, occurring at Parnitha, Poikilo and 

Aegaleo mountains. The boundary between the 

metamorphic and non-metamorphic geotectonic 

units, although generally accepted to be of tectonic 

origin, its exact geometric and kinematic 

characteristics are yet to be determined, since no 

direct geological mapping could be undertaken. The 

entire tectonic structure within the area is covered by 

an allochtonous system, called “Athens schists”, 

tectonically overlaid on the two previously mentioned 

units, as well as Neogene and Quaternary deposits. 

It is traced northwards from the Aegean coast of 

Southern Evia, through Aliveri to Kalamos in 

northeast Attica and continues to the southwest into 

the plain of Athens. Within the area of interest its 

locations coincide approximately with the riverbed of 

Kifissos R. (Papanikolaou et al., 1999; Mariolakos & 

Fountoulis, 2000; Xypolias et al., 2003) (Fig. 1), also 

confirmed by geophysical investigations at the 

northern part of the basin (Papadopoulos et al., 

2007). Results of seismic tomography indicate the 

presence of abnormally high seismic velocities in the 

central part of the basin, most likely related to this 

major boundary, extending towards the southeast at 

Saronikos Gulf (Drakatos et al. 2005). 

According to Papanikolaou & Royden (2007) this 

boundary represents a broad extensional detachment 

with significant portion of dextral shear, whereas 

opinions of a right-lateral strike slip fault zone have 

also been reported (Mariolakos & Fountoulis, 2000; 

Krohe et al. 2009). Considering a depth of about 30 

km for the metamorphics (Lozios, 1993), it is clear 

that this tectonic boundary has accommodated more 

34


EARTHQUAKE 

ARCHAEOLOGY 



than 25 km of displacement. It was active throughout 

Late Miocene times and gradually became inactive 

during Early Pliocene (Papanikolaou & Royden, 

2007). However, it forms a major boundary that 

separates the E-W trending higher slip-rate active 

faults in the western part of Attica from the NW-SE 

trending lower slip-rate faults in the eastern part 

(Mariolakos & Papanikolaou, 1987; Papanikolaou et 

al. 2004). 

Fig. 1: Simplified neotectonic map of Attica showing the 

approximate location of the major tectonic boundary 

separating metamorphic and non-metamorphic alpine rocks 

(modified from Papanikolaou et al. 1999). 

each station for a period of at least four hours. In an 

effort to achieve optimal results, selected stations 

were occupied for several days per epoch 

(independent sessions). To avoid large tropospheric 

errors an initial elevation cut-off angle of 10 was 

used. 

Collected data were processed using Leica Geo 

Office v.1.1 and Bernese ver. 4.2 (Beutler et al., 

2001). The realization of the reference frame was 

performed using the coordinates and velocity of 

Dionysos (DION) continuous GPS station, located on 

the metamorphic alpine basement. DION was tied to 

the ITRF2000 at epoch 2005.0 by almost a decade of 

observations from numerous sites of the EUREF 

permanent network (Prof. D. Paradissis, personal 

communication). It can be argued that connecting the 

local network to the ITRF through DION reference 

station would be sufficient, taking into account the 

network extend. Details on data collection and 

processing could be found in Foumelis (2009). 

GPS NETWORK ESTABLISHMENT 

Given the lack of previous instrumental observations, 

the design of the geodetic network was primarily 

focused on the investigation of the local tectonic 

regime. The minimum number of survey points 

required is imposed by the tectonic complexity of the 

region and the degree of fragmentation of the crust. 

The established Athens Geodetic Network (AGNET) 

consisted of a total number of 41 campaign GPS 

sites (Fig. 2) including already available benchmarks 

of the Hellenic Military Geographical Service 

(HMGS), as well as sites previously installed by the 

Hellenic Mapping and Cadastral Organization 

(HEMCO) and the National Technical University of 

Athens (NTUA). Continuous (real-time) GPS station 

operate in the region by Metrica company (MET0), 

National Observatory of Athens (NOA1) and National 

& Kapodistrian University of Athens (UOA1), and 

despite their relatively limited observations at the 

time, they where considered in the analysis as well. 

The network covers essentially both the basins of 

Athens and Thriassio as well as their bordering 

mountain ranges, showing a relatively uniform spatial 

distribution. With an average distance between 

stations of approximately 5 km, a sufficient sampling 

of local deformation field is accomplished. 

GPS MEASUREMENTS AND ANALYSIS 

GPS campaigns were carried out from 2005 to 2008 

(3.2 yr) following an annual re-occupation strategy. 

The benchmarks of the HMGS were first measured 

during network establishment and together with 

selected GPS sites once more on 2008. 

Measurements were conducted using LEICA 

geodetic GPS receivers equipped with SR299/399, 

AT202/302 and Ach1202Pro antennas. Carrier phase 

observations were recorded every 10 seconds from 

Fig. 2: Annual GPS velocities of broader Athens area, 

relative to DION, for the period 2005-2008. The error 

ellipses represent the 1-sigma confidence region. Velocities 

of E067 and G20 benchmarks from Veis et al. (2003), and 

NOA1 from the EUREF website, after transformation to the 

specific ITRF. 

Repeated campaign observations allow the 

determination of the displacement vector as a 

function of time. The estimation of velocities and the 

corresponding errors was carried out on a statistical 

basis, by analysis of time series of each individual 

component of motion, by least square adjustment. 

Uncertainties were determined using the average 

scatter of residuals of the linear regression, providing 

more realistic error estimates. The estimated GPS 

velocity field is presented in a local DION-fixed 

reference frame in order to allow the recognition of 

local scale displacement patterns (Fig. 2). Site 

velocities from previous geodetic studies (E067 & 

G20) as well as EUREF solutions (NOA1) were also 

considered for the sake of completeness. 

35


EARTHQUAKE 

ARCHAEOLOGY 



STRAIN RATES 

In order to provide results independent from the 

choice of the reference frame, strain analysis was 

performed by the grid_strain Matlab TM software 

package (Teza et al., 2008). It allows the definition of 

the deformation pattern by providing the intensity and 

direction of principal components of strain tensor 

together with corresponding errors, by means of a 

modified linear least-squares (LS) inversion, under 

the hypothesis of uniform strain field condition. Inputs 

for calculating strain were horizontal GPS annual 

velocities and their corresponding errors. In this 

sense results express the linear strain rates in the 

region. 

For the purpose of the analysis, GPS sites located on 

the mountains bordering the Athens Basin, 

specifically on the metamorphic basement of Pendeli 

and Hymettus mountains to the East (APR, ARG, 

HYM, NER and TAT) and on the non-metamorphic 

formations of eastern Parnitha Mt. and Aegaleo Mt. 

(E067, CHS, KOR, PKL and PRM) were selected. 

The analysis involved initially the calculation of a 

single strain tensor based on all selected stations 

and then, by gradual segmentation of the area for a 

more detailed investigation of spatial variations of the 

deformation regime. All calculations are referred to 

the center of mass of each set of sites considered. 

Fig. 3: Principle axes of the strain rate tensor for the area of 

interest, calculated from velocities of selected GPS sites, in 

background contour lines of 20m interval. 

From single strain tensor calculations, an extension 

of 0.27 0.06 strain/yr along a NNW–SSE direction 

(N 347) is shown, with a negative eigenvalue 

(compression) for the minimum principal axes (Fig. 

3). It is nevertheless evident that a single strain 

tensor is insufficient to express adequately the 

apparent heterogeneity of the local displacement 

field. Further examination of the strain field (Fig. 4) 

indicate negligible compressional rates at the 

southern part of the basin compared to the dominant 

extensional regime of relatively higher strain rates 

(0.91 0.09 strain/yr) at the northern part between 

Pendeli and Parnitha ranges. 

A more detailed consideration of the strain field 

between the two geotectonic units was performed by 

triangulation of the selected GPS sites (Fig. 5). 

Herein, it is interesting to note the major 

differentiation between the western and the eastern 

parts of Athens Basin with significantly lower strain 

rates in the latter. Moreover, the gradual increase of 

the extension rates at the western part of Athens 

plain moving to the North is clearly depicted, while a 

counterclockwise rotation of the maximum principle 

axis of the strain tensor is also observed. The 

compressional regime at the southeastern part of the 

basin should be underlined. 

DISCUSSIONS 

The stress field in the region is characterized by 

extension in a NNE–SSW direction, also confirmed 

by regional geodetic measurements (Veis et al., 

2003). However, the 5-km spacing of the geodetic 

network allowed investigating local variations of the 

strain rates. 

Fig. 4: Principle axes of the strain rate tensors within Athens 

Basin. Dashed lines indicate local estimates around which 

GPS data are poorly distributed from a geometrical point of 

view. 

A differentiation of strain rates across the inactive 

tectonic boundary is evident with significantly higher 

rates at the western part of Athens Basin. Given its 

inactive characteristics, a passive control should be 

considered. Such behavior has also been mentioned 

during the Athens 1999 earthquake from SAR 

interferometric observations of the spatial expansion 

of the co- and post-seismic displacement field 

(Foumelis et al., 2009). 

36


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 5: Detailed strain analysis by different triangulations of 

selected GPS sites. Principle axes of the strain rate tensors 

are calculated at the center of mass of each triangle. 

The broader area is essentially a transitional zone 

between the Corinth Gulf and Beotia to the west, 

characterized by E-W trending active faults with 

significant seismic activity and those of southern 

Attica and Cyclades islands to the east, showing low 

deformation rates (Mariolakos & Papanikolaou, 1987; 

Papanicolaou & Lozios, 1990). Thus, the observed 

high strain rates at the northern part of the basin 

should be attributed to the high crustal velocities 

observed within Parnitha Mt. an area controlled 

mainly by E-W trending active fault zones (Ganas et 

al. 2005; Papanikolaou & Papanikolaou, 2007) 

although the role of NE-SW trending faults should be 

important as well (Mariolakos & Fountoulis, 2000). 

Acknowledgements: The authors would like to 

acknowledge Prof. E. Lagios for his support and supervision 

during GPS measurements, Prof. D. Paradissis for his 

suggestions, Dr. V. Sakkas for processing of UOA1 data 

and METRICA company for MET0 station data provision. 

References 

Beutler, G., H. Bock, E. Brockmann, R. Dach, P. Fridez, W. 

Gurtner, U. Hugenntobler, D. Ineichen, J. Johnson, M. 

Meindl, L. Mervart, M. Rothacher, S. Schaer, T. Springer 

& R. Weber, (2001). Bernese GPS Software Version 4.2. 

U. Hugentobler, S. Schaer & P. Fridez (eds.), 

Astronomical Institute, University of Berne, Switzerland, 

515p. 

Clark, P.J., R.R. Davies, P.C. England, B. Parsons, H. 

Billiris, D. Paradissis, G. Veis, P.A. Cross, P.H. Denys, V. 

Ashkenazi, R. Bingley, H.-G. Kahle, M.V. Müller, & P. 

Briole, (1998). Crustal strain in central Greece from 

repeated GPS measurements in the interval 1989-1997. 

Geophysical Journal International 135 (1), 195-214. 

Drakatos, G., V. Karastathis, J. Makris, J. Papoulia & G. 

Stavrakakis, (2005). 3D crustal structure in the 

neotectonic basin of the Gulf of Saronikos (Greece). 

Tectonophysics 400, 55-65. 

Foumelis, M., (2009). Study of crustal deformation in the 

broader area of Athens by means of Differential GPS and 

SAR Interferometry. Dissertation thesis, Department of 

Geology and Geoenvironment, University of Athens, 

358p (in Greek). 

Foumelis, M., I. Parcharidis, E. Lagios & N. Voulgaris, 

(2009). Evolution of post-seismic ground deformation of 

Athens 1999 earthquake observed by SAR 

interferometry. Journal of Applied Geophysics 69, 16-23. 

Ganas, A., S. Pavlides & V. Karastathis, (2005). DEMbased 

morphometry of range-front escarpments in Attica, 

central Greece and its relation to fault slip rates. 

Geomorphology 65, 301-319. 

Krohe, A., E. Mposkos, A. Diamantopoulos & G. Kaouras, 

(2009). Formation of basins and mountain ranges in 

Attica (Greece): The role of Miocene to recent low-angle 

normal detachment faults. Earth-Science Reviews 98 (1- 

2), 81-104. 

Lozios, S., (1993). Tectonic analyisis of the metamorphic 

rocks in NE Attica (in Greek). Dissertation thesis, 

Department of Geology, University of Athens, 299p (in 

Greek). 

Mariolakos, I. & I. Fountoulis, (2000). The Athens 

earthquake September 7, 1999 neotectonic regime and 

geodynamic phenomena. Ann. Geol. Pays Hellen. 38 (B), 

165-174. 

Mariolakos, I. & D. Papanikolaou, (1987). Deformation 

pattern and relation between deformation and seismicity 

in the Hellenic Arc. Bull. Geol. Soc. Greece 19, 59-76 (in 

Greek). 

Papadopoulos T.D., N. Goulty, N.S. Voulgaris, J.D. 

Alexopoulos, I. Fountoulis, P. Kambouris, V. Karastathis, 

C. Peirce, S. Chailas, J. Kassaras, M. Pirli, G. Goumas & 

E. Lagios, (2007). Tectonic structure of Central-Western 

Attica (Greece) based on geophysical investigations - 

Preliminary Results. Bull. Geol. Soc. Greece 40 (3), 

1207-1218. 

Papanikolaou, D., E. Bassi, H. Kranis & G. Danamos, 

(2004). Paleogeographic evolution of the Athens basin 

from upper Miocene to Present. Bull. Geol. Soc. Greece 

36 (2), 816-825 (in Greek). 

Papanikolaou, D., E. Lekkas, Ch. Sideris, I. Fountoulis, G. 

Danamos, Ch. Kranis, S. Lozios, I. Antoniou, E. 

Vassilakis, S. Vasilopoulou, P. Nomikou, I. Papanikolaou, 

E. Skourtsos & K. Soukis, (1999). Geology and tectonics 

of Western Attica in relation to the 7-9-99 earthquake. 

Newsletter of E.C.P.F.E., issue no. 3, 30-34. 

Papanikolaou, D. & S. Lozios, (1990). Comparative 

neotectonic structure of high (Korinthia-Beotia) and low 

rate (Attica-Cyclades) activity. Bull. Geol. Soc. Greece 

16, 47-65 (in Greek). 

Papanikolaou, D. & I. Papanikolaou, (2007). Geological, 

geomorphological and tectonic structure of NE Attica and 

seismic hazard implications for the northern edge of 

Athens plain. Bull. Geol. Soc. Greece 40 (1), 425-438. 

Papanikolaou, D. & L. Royden, (2007). Disruption of the 

Hellenic Arc: Late Miocene extensional detachment faults 

and steep Pliocene-Quaternary normal faults – or – What 

happened to Corinth ? Tectonics 26, TC5003, 

doi:10.1029/2006TC002007. 

Teza, G., A. Pesci & A. Galgaro, (2008). Grid_strain and 

grid_strain3: Software packages for strain field 

computation in 2D and 3D environments. Computers & 

Geosciences 34, 1142-1153. 

Xypolias, P., S. Kokkalas & K. Skourlis, (2003). Upward 

extrusion and subsequent transpression as a possible 

mechanism for the exhumation of HP/LT rocks in Evia 

island ( Aegean Sea , Greece ). Journal of Geodynamics 

35, 303-332. 

Veis et al. (2003). Tectonic displacements along the 

Aegean and the Alkyonides- Perachora-Parnitha triangle. 

Applied Research Project, Earthquake Planning and 

Protection Organization. Athens, 74p (in Greek). 

37


EARTHQUAKE 

ARCHAEOLOGY 



NEOTECTONICS AND COMPARISON OF THE ENVIRONMENTAL SEISMIC INTENSITY 

SCALE (ESI 2007) AND THE TRADITIONAL SCALES FOR EARTHQUAKE INTENSITIES 

FOR THE KALAMATA (SW GREECE) EARTHQUAKE (MS=6.2R, 13-09-1986) 

Fountoulis, Ioannis (1, Mavroulis, Spyridon (1) 

(1) National and Kapodistrian University of Athens, Faculty of Geology and Geoenvironment, Department of Dynamic Tectonic 

and Applied Geology, Panepistimiopoli, Zografou 15784 Athens, GREECE. Email: fountoulis@geol.uoa.gr , smavroulis@yahoo.gr 

Abstract: The Kalamata (13-09-1986, Ms=6.0R, SW Peloponnese) earthquake can be classified as a medium to small scale 

event based on the tectonic structures that triggered the earthquake and the effects caused on human, structural and natural 

environment. The aim of this paper is to present the geotectonic and seismotectonic regime of the earthquake affected region 

based on field data along the seismic fault zone and an attempt is made towards the: (i) estimation of the intensity values 

according to the European Macroseismic Scale (EMS 1998) and the Environmental Seismic Intensity Scale (ESI 2007) and the 

determination of their geographical distribution in a macroscale, (ii) interpretation of the intensity values data and their distribution 

according to the seismotectonic, geodynamic and geotechnical regime, and (iii) conduction of a comparative evaluation review on 

the application of both EMS 1998 and ESI 2007. The application of both EMS 1998 and ESI 2007 and the comparative evaluation 

of the results indicate that the estimated values of EMS 1998 and ESI 2007 were almost in agreement, despite the fact that the 

geographical locations of assessment data were different suggesting that the application and use of both scales appears to 

represent a useful and reliable tool for seismic hazard estimation. 

Key words: Kalamata, earthquake, neotectonics, environmental effects 

INTRODUCTION 

Kalamata is located very close (< 70km) to the 

Hellenic (Ionian) Trench region in which the 

subduction zone of the African plate beneath the 

European (Aegean) one exists and thus is one of the 

most seismically active areas of Europe (Figure 1). 

On 13 September 1986, a shallow depth (< 10km) 

earthquake struck the wider Kalamata area resulting 

in 20 casualties, extensive damages and many 

environmental effects. The epicenter of the main 

earthquake was located about 10km NNE of the city 

of Kalamata, and its magnitude was Ms=6.2 

(Papazachos, et al., 1988). Two days later, a second 

shock of Ms=5.4R (Papazachos, et al., 1988) 

occurred closer to the Kalamata city at the same 

depth. The focal mechanism of the main shock 

shows an E-W extension (Lyon-Caen et al., 1987; 

Papazachos, et al., 1988). 

Seismological studies of Papazachos et al. (1988), 

and Lyon-Caen et al. (1988) indicated that 

aftershocks defined two clusters and an about 450 

west-dipping fault plane. The foci depths of the 

seismic sequence were ranging between 11 and 

0.9km. Based on the variety of orientations and dips 

calculated for the sub-faults activated during the 

aftershock sequence, since the analysis of the 

northern cluster indicates the existence of two types 

of orientation, which are dipping in four different 

angles and the southern cluster is characterized by 

an almost uniform behaviour activated later in the 

sequence, Tselentis et al. (1989) concluded that the 

area is tectonically very complex which is in 

agreement with the neotectonic structure described 

by Mariolakos et al. (1989; 1992, 1993) and 

Mariolakos & Fountoulis (1998). 

Stiros and Kontogianni (2008) applied two first-order 

leveling traverses crossing the wider Kalamata area 

and measured subsidence of about 7cm NE of the 

Kalamata city in epicentral area of the southern 

cluster. The Kalamata earthquake produced a 

maximum intensity VIII+ on the IMM or EMS 1992 

scale (Elnashai et al., 1987; Gazetas et al., 1990), 

while Panou et al. (2004) based on building damages 

estimated the intensity up to IX - X for the city of 

Kalamata. 

The aim of this paper is to present the geotectonic 

and seismotectonic regime of the earthquake 

affected region based on field data along the seismic 

fault zone and an attempt is made towards the: (i) 

estimation of the intensity values in terms of the 

European Macroseimic Scale (EMS 1998; Grünthal, 

1998) and Environmental Seismic Intensity Scale 

(ESI 2007; Michetti et al., 2007) and the 

determination of their geographical distribution in a 

macroscale, (ii) interpretation of the intensity values 

data and their distribution according to the 

seismotectonic, geodynamic and geotechnical 

regime, and (iii) conduction of a comparative 

evaluation review on the application of both EMS 

1998 and ESI 2007. 

GEOLOGY - TECTONICS - NEOTECTONICS - 

FAULT ZONES - FAULTS 

In the broader Kalamata area the following four 

alpine geotectonic units from the lower to the upper 

occur (Psonis, 1986; Mariolakos et al., 1993): (a) the 

38


EARTHQUAKE 

ARCHAEOLOGY 



Mani unit consisting mainly of marbles, (b) the Arna 

unit consisting of quartzites and phyllites, (c) the 

Tripolis unit which consists of neritic carbonates and 

flysch formation and (d) the Pindos unit consisting of 

thin-bedded pelagic carbonates and clastic 

formations. From the structural point of view, the four 

above-mentioned geotectonic units form a 

succession of three nappes. The Mani unit (slightly 

metamorphosed) is considered to be the relatively 

autochthonous one. The Arna unit overthrusts the 

Mani unit, the Tripolis unit (second nappe) 

overthrusts the Arna unit and the Pindos unit (third 

nappe) overthrusts the Tripolis unit (Figure 2). The 

Late Pliocene-Early Pleistocene marine deposits 

consist of marls, sandstone and conglomerates 

(Marcopoulou-Diacantoni et al. 1989; Mariolakos et 

al., 1993). The Middle-Late Pleistocene deposits 

consist mainly of red colored siliceous sandssandstones 

and conglomerates. Alluvial deposits, 

clastic material and talus represent the Holocene. 

were observed within this graben. The marginal fault 

zones consist of many faults, which are not 

continuous and differ on strike even when they 

belong to the same fault zone, as they form 

conjugate fault systems. 

Fig. 2: Simplified geological map showing the four alpine 

geotectonic units overthrust one on top of the other, as well 

as the post-alpine sediments of the region of the Kalamata 

area. 1: Holocene deposits; 2: Continental deposits; 3: Early 

Pleistocene marine deposits; 4: Pindos unit; 5: Gavrovo- 

Tripolis unit; 6: Arna unit; 7: Mani unit; 8: Overthrust; 9: 

Fault; 10: Detachment fault. The numbers in the black 

circles correspond to the smaller order neotectonic 

macrostructures of the Kato Messinia sub-graben 1: 

Asprochoma-Koutalas horst, 2: Dimiova-Perivolakia graben, 

3: Kalathion Mt. horst, 4: Altomyra semi-graben, 5: Kambos 

graben, 6: Vardia-Koka horst, 7: Kitries-Mantinia subgraben, 

XFZ:Xerilas Fault zone, NFZ:Nedon Fault Zone. 

Fig. 1: The second order neotectonic macrostructures within 

the first order neotectonic macrostructure of the Kalamata- 

Kyparissia graben. The numbers correspond to the 

following second order neotectonic macrostructures: 1: Kato 

Messinia graben, 2: Meligalas horst, 3: Ano Messinia 

graben, 4: Dorion basin, 5: Kyparissia-Kalo Nero graben 

The meizoseismal area is located at the eastern 

margin of the Kalamata - Kyparissia graben and 

constitutes the northward prolongation of the Gulf of 

Messinia (Figure 1). Large and composite fault zones 

define its margins and second order macrostructures 

are observed within as well as at the margins 

representing smaller grabens and horsts (Figure 1) 

(Mariolakos & Fountoulis, 1998). The E-W striking 

Dimiova - Perivolakia graben is bounded by the Kato 

Karveli - Venitsa fault zone to the north, by the 

Arahova to the east, by the Xerilas fault zone (XFZ) 

to the south and by the Nedon fault zone (NFZ) to the 

west (Figure 2). This macrostructure constitutes one 

of the most interesting minor order neotectonic 

macrostructures because of the occurrence of the 

Pindos unit. Mariolakos et al. (1989) interpreted the 

kinematic regime of this macrostructure suggesting 

that this graben rotates around an N-S axis located at 

the area of Arahova westwards. At the western part 

of the fault zone the total throw is more than 2.000m 

(Mariolakos et al., 1986; Mariolakos et al., 1989). The 

most of the environmental effects and damages 

caused during the seismic activity of September 1986 

SPATIAL DISTRIBUTION OF ENVIRONMENTAL 

EFFECTS 

During the above-mentioned seismic activity, fault 

reactivation (seismic faults), new faulting and seismic 

fracturing were observed (the latter distinguished by 

no displacement) (Figure 3). The reactivated faults 

strike in different directions (N-S, E-W, NNE-SSW) 

and the throw of the faults due to the reactivation is 

generally small (max=20cm) and of normal character. 

The maximum throw has been observed at a seismic 

fault caused by the main aftershock Ms=5.6 R. 

Numerous seismic ruptures trending N-S, NNE-SSW, 

NE-SW, E-W and NW-SE were mapped in the 

affected area, in most cases in en echelon 

arrangement (Mariolakos et al., 1989). These seismic 

fractures presented a vertical displacement of several 

mm up to 25-30cm and they often presented a 

horizontal component showing sinistral or dextral 

displacement. 

The majority of rock falls were observed in several 

sections along the slopes of the Tzirorema, Karveli 

and Xerilas streams and the Nedon river valleys as 

well as in the wider area of Eleochori, Karveli and 

Ladas villages (Figure 3). They were observed in 

areas characterized by steep slopes (> 50 per cent) 

and they were related almost everywhere to small or 

39


EARTHQUAKE 

ARCHAEOLOGY 



large faults with some of them reactivated during the 

earthquake and others not. 

because those areas belong to different neotectonic 

macrosturctures that were not reactivated during the 

earthquakes of 1986 (central region of the tectonic 

basin of Kato Messinia and tectonic horst of Kalathio 

Mt respectively). 

Fig. 3: The spatial distribution of the environmental effects 

observed during the Kalamata earthquake sequence (based 

on data from Mariolakos et al., 1992; Gazetas et al., 1990; 

Fountoulis, 2004; Stiros & Kontogianni, 2008). 

SPATIAL DISTRIBUTION OF DAMAGES 

The damages were limited to an area of triangular 

shape, which is defined to the south by the fault zone 

of the Xerilas River, to the east by the fault zone of 

Nedousa - Arahova and to the west by the fault zone 

of the Nedon River (Figure 3). No damages were 

recorded to the west of the Nedon fault zone and 

south of the Xerilas fault zone and especially in areas 

where the geological basement has the same 

seismo-geological behavior as those in the city of 

Kalamata and Eleohori village, which caused serious 

damage. Based on field observations, the damage is 

not determined only by the age, type, height and 

other characteristics of buildings. There were cases 

with two nearly identical constructions in the same 

area; one remained intact while the other was 

destroyed. In other cases the building destruction is 

linked to zones of seismic fracturing that were 

observed in the construction basement. Of course, 

this is not the rule. In many other cases the building 

destruction is linked to zones of seismic fracturing 

that were observed in the construction basement. Of 

course, this is not the rule. 

Fig. 4: (A) EMS 1998 intensity distribution of the Kalamata 

earthquake sequence (based on data from Gazetas et al., 

1990; Panou et al., 2004). (B) EMS 1998 intensity 

distribution of the Kalamata earthquake sequence for 

Kalamata city (based on data from Panou et al., 2004). 

CONCLUSIONS 

Taking into account the aforementioned we can draw 

the following conclusions: 

The damages were limited to the area that can be 

regarded as a transitional area between the tectonic 

basin Kalamata - Kyparissia and the tectonic horsts 

of Asprohoma - Koutala to the north and the Kalathio 

Mt. to the south. On the contrary, in Messini and in 

Verga, damages of that size were not observed 

Figure 5: ESI 2007 Intensity distribution based on data of 

Figure 3. 

Rock falls were observed mainly in the tectonic 

graben that was activated and also north of it, at 

Tzirorema. On the other hand, on the steep slopes of 

the Kalathio Mt. that belong to the homonymous 

40


EARTHQUAKE 

ARCHAEOLOGY 



neotectonic macrostructure, which was not 

reactivated, no rock falls were observed. 

An important factor in the distribution of the damages 

and rock falls in the greater area was the reactivation 

of old faults or the creation of new soil ruptures. In 

this way, the fact that the destruction of buildings was 

observed in Giannitsanika (higher intensity in Figure 

4) and not near the coast can be explained, although 

the foundation ground - red siliceous clastic formation 

- in the first case theoretically presents better 

geotechnical characteristics in comparison to the 

loose coastal deposits. 

The ESI 2007 scale appears to fit better than the 

EMS scale in the neotectonic regime of the area as 

its boundaries coincide better with the boundaries of 

the activated graben and the observations we have 

done concerning the distribution of the environmental 

effects. The application of both EMS 1998 and ESI 

2007 and the comparative evaluation of the results 

indicate that the estimated values of EMS 1998 and 

ESI 2007 were almost in agreement, despite the fact 

that the geographical locations of assessment data 

were different suggesting that the application and use 

of both scales appears to represent a useful and 

reliable tool for seismic hazard estimation. 

References 

Elnashai, A., Pilakoutas, K., Ambraseys, N. & Lefas, I. 

(1987). Lessons learnt from the Kalamata (Greece) 

earthquake of 13 September 1986. European Earthquake 

Engineering, 1, 11-19. 

Fountoulis, I. 2004. The neotectonic macrostructures and 

the geological basement, the main factors controlling the 

spatial distribution of the damage and geodynamic 

phenomena resulting from the Kalamata (13 September 

1986) and Athens (7 September 1999) earthquakes. WIT 

Press: In Earthquake Geodynamics: Seismic Case 

Studies. Edited by: E. L. Lekkas, Series: Advances in 

Earthquake Engineering, vol. 12, p. 45-63. 

Gazetas, G., Dakoulas, P. & Papageorgiou, A. (1990). Local 

soil and source-mechanism effects in the 1986 kalamata 

(Greece) earthquake. Earthquake Engineering and 

Structural Dynamics, vol. 19, 431-456. 

Grünthal, G. 1998. (ed.): European Macroseismic Scale 

1998 (EMS-98). Cahiers du Centre Européen de 

Géodynamique et de Séismologie 15, Centre Européen 

de Géodynamique et de Séismologie, Luxembourg, 99 

pp. 

Lyon-Caen, H., Armijo, R., Drakopoulos, J., Baskoutas, J., 

Delibassis, N., Gaulon, R., Kouskouna, V., Latoussakis, 

J., Makropoulos, K., Papadimitriou, P., Papanastassiou, 

D. & Pedotti, G. (1988). The 1986 Kalamata (South 

Peloponnesus) earthquake: Detailed study of a normal 

fault, evidences for east-west extension in the Hellenic 

arc, J. Geophys. Res. 93, 14.967–15.000. 

Marcopoulou-Diacantoni, A., Mirkou, M.R., Mariolakos, I., 

Logos, E., Lozios, S. & Fountoulis, I. (1989). Stratigraphic 

observations in the post alpine deposits in Thouria-Ano 

Amfia area (Messinia Province, Greece) and their 

neotectonic interpretation. Bull. Geol Soc. Greece, 

XXIII/3, pp. 275-295, (in Greek). 

Mariolakos, I., Sabot, V., Alexopoulos, A., Danamos, G., 

Lekkas, E., Logos, E., Lozios, S., Mertzanis, A. & 

Fountoulis, I. (1986). Microzonic study of Kalamata 

(Geology, Tectonics, Neotectonics, Geomorphology). 

Report of EPPO (Earth Planning Protection 

Organization), (in Greek). 

Mariolakos, I., Fountoulis, I., Logos, E. & Lozios, S. (1989). 

Surface faulting caused by the Kalamata (Greece) 

earthquakes (13.9.1986). Tectonophysics, 163, pp. 197- 

203. 

Mariolakos, I., Fountoulis, I. & Nassopoulou, S. (1992). The 

influence of the neotectonic macrostructures, fractures 

and the geological basement in the distribution of the 

damages in the Kalamata earthquake (13-9-1986). Proc. 

1st Greek Congress on Antiseismic Engineering and 

Technical Seismology Technical Chamber of Greece, 1, 

pp. 55-68, (in Greek). 

Mariolakos, I., Schneider, H., Fountoulis, I. & Vouloumanos, 

N. (1993). Paleogeography, sedimentation and 

neotectonic implications at the Kambos depression and 

Kitries Bay area (Messinia, Peloponnese, Greece). Bull 

Geol. Soc. Greece, VIII/1, pp. 397-413, 1993. 

Mariolakos, I. & Fountoulis, I. (1998). Is it safe to built on 

fault surfaces in a seismically active area? Proc. 8th 

International IAEG Congress, pp. 665-670, Balkema. 

Michetti A.M., Esposito, E., Guerrieri, L., Porfido, S., Serva, 

L., Tatevossian, R., Vittori, E., Audemard, F., Azuma, T., 

Clague, J., Comerci, V., Gürpinar, A., Mccalpin, J., 

Mohammadioun, B., Mörner, N.A., Ota, Y. & Roghozin, E. 

(2007). Environmental Seismic Intensity Scale 2007 - ESI 

2007. In: Guerrieri L. and Vittori E. (Eds.), Memorie 

Descrittive della Carta Geologica d’Italia, 74, 7-54, 

Servizio Geologico d’Italia – Dipartimento Difesa del 

Suolo, APAT, Roma, Italy. 

Panou, A., Theodoulidis, N. & Savvaidis, A. (2004). 

Correlation between (H/V) ratio and seismic damage: The 

case of the city of Kalamata. In SESAME project, Report 

of the WP04, H/V Technique: Empirical Evaluation, 

Project No. EVG1-CT-2000-00026. 

Papazachos, V., Kiratzi, A., Karacostas, B., 

Panagiotopoulos, D., Scordilis, E. & Mountrakis, D. 

(1988). Surface fault traces, fault plane solution and 

spatial distribution of the aftershocks of the September 

13, 1986 earthquake of Kalamata (Southern Greece). 

PAGEOPH, vol. 126, No. 1, 55-68. 

Psonis, K. 1986. Geological map of Greece, Kalamata 

sheet, scale 1/50,000, Geological Survey of Greece: 

Athens. 

Stiros, S. & Kontogianni, V. (2008). Modelling of the 

Kalamata (SW Greece) earthquake faulting using 

geodetic data. Journal of Applied Geodesy, 2, 179-185; 

DOI 10.1515/JAG.2008.020 

41


EARTHQUAKE 

ARCHAEOLOGY 



QUANTIFICATION OF RIVER VALLEY MAJOR DIVERSION IMPACT AT KYLLINI 

COASTAL AREA (W. PELOPONNESUS, GREECE) WITH REMOTE SENSING 

TECHNIQUES 

Fountoulis Ioannis D. (1), Vassilakis Emmanuel (1), Mavroulis Spyridon (1), Alexopoulos John (2), Erkeki Athanasia (3) 

(1) School of Geology and Geoenvironment, Department of Dynamics, Tectonics and Applied Geology, National and Kapodistrian 

University of Athens, 15784, Greece. Email: evasilak@geol.uoa.gr 

(2) School of Geology and Geoenvironment, Department of Geophysics and Geothermy, National and Kapodistrian University of 

Athens, 15784, Greece. 

(3) School of Geology and Geoenvironment, Laboratory of Natural Hazards, National and Kapodistrian University of Athens, 

15784, Greece. 

Abstract: The effects of the geological, tectonic and neotectonic structure and the impact of the human presence and activity on 

the drainage network of Pineios river are presented here in order to determine the causes of its diversion and the implications to 

the shoreline. We used, analyzed and evaluated (a) geomorphological, geological, tectonic and neotectonic data of the study area, 

(b) historical information and archaeological findings from buried and eroded archaeological sites of the wider study area, (c) 

published data related to drill cores and radiocarbon dates, and (d) remote sensing datasets, as satellite and aerial photos of 

different capturing periods, as well as real-time kinematic differential GPS measurements for the definition of the current shoreline. 

It is concluded that the detected shoreline displacements and drainage diversions are the result of the combination of active 

tectonics and human activity during the last 100 kyrs. 

Key words: Kyllini peninsula, Pineios river, RTK DGPS, river evolution, coastal erosion 

INTRODUCTION 

The Pineios River development and history takes 

place in one of the most tectonically and seismically 

active areas in Greece. The intense and continuous 

tectonic activity in the area is highly related to its 

location on the external part of the Hellenic arc and 

adjacent to the convergent boundary where African 

plate is subducted beneath the Aegean as well as the 

diapirism of near surface evaporitic domes. The 

highest seismicity levels recorded in the area 

(Hatzfeld et al., 1990) as well as the generation of 

many historic strong earthquakes confirm the 

neotectonic observations, which show that the area is 

undergoing a complicated tectonic deformation. 

d. 0,16 to 0,48 mm/yr for the eastern (inland) part of 

Kyllini peninsula (125 kyrs) 

The most important fault zones in the study area are 

the Panopoulo fault zone (Panopoulo FZ), Pineios 

fault zone (Pineios FZ) and the strike - slip fault zone 

that gave rise to the Andravida earthquake (08-06- 

2008, ML=6,5). These major faults form several 

neotectonic blocks in the study area including the 

Gastouni graben (hangingwall of Pineios fault zone), 

the uplifted area of Varda (footwall of Pineios fault 

zone) and the Kyllini horst (Fig. 1). 

GEOCHRONOLOGICAL INTERPERETATION 

In order to determine the effect of the ongoing active 

tectonics on the Pineios River diversion during the 

late 18 th or the early 19 th century, we calculated 

relative uplift rates for several sites of the study area 

based on 

230 Th/ 238 U dating of corals made by 

Stamatopoulos et al. (1988) and dating of marine 

deposits in Kyllini peninsula estimated by Mariolakos 

et al. (1988): 

a. 0,39 mm/yr for Psari area (103 kyrs) 

b. 0,50 mm/yr for Neapoli area (118 kyrs) 

c. 0,67 mm/yr for Aletreika area (209 kyrs) 

Fig. 1: Sketch map of the contemporary Pineios river deltaic 

area (at the hanging wall of the Pineios fault) and the former 

deltaic area at the footwall of the same fault. The estimated 

shorelines for the Roman and Neolithic periods are shown. 

The archaeological sites and the sampling sites of the 

geochronological analysis are also noted, along with the 

calculated uplift rates for the last 100kyrs. 

42


EARTHQUAKE 

ARCHAEOLOGY 



The general conclusions after the interpretation of the 

geochronological data are: 

a. The maximum relative uplift rate (0,67 mm/yr) 

characterizes an area (Aletreika) located on the 

footwall side and very close to the Pineios FZ. 

b. The relative uplift rate of the Gastouni graben 

(hangingwall of Pineios FZ, 0,19 mm/yr) is less 

even than the lowest value of the Pineios FZ 

footwall relative uplift rate (0,39 mm/yr). 

c. The northeastern part of Kyllini peninsula has 

higher relative uplift rate (0,48 mm/yr) than the 

southeastern part (0,30 mm/yr) 

d. The maximum relative uplift rate of the footwall of 

Pineios FZ is significantly higher than the 

maximum relative rates of the eastern part of 

Kyllini peninsula and the Gastouni graben. 

Lagoon recorded by Kontopoulos and Koutsios 

(2010) we note that: 

a. the shoreline in the Pineios delta advanced by 3,5 

km into the sea in the 6.350 yrs period from 

Neolithic (8.500 yrs BP) to Roman (2.150 yrs BP) 

period, which shows a coastal progradation rate 

of the order of 0,55 m/yr, and 

b. the shoreline in the Pineios delta retreated by 

1,75 km in the 2.150 yrs period from Roman 

period (2.150 yrs BP) to present which shows a 

retrogradation rate as 0,81 m/yr from Roman 

period to present. 

HISTORICAL COASTLINE DATA 

It is more than obvious that the major percentage of 

the coastline displacements in the study area, during 

the last 8kyrs, are related to active structures and the 

tectonic instability as this affects the alongshore 

redistribution of sediments from the Pineios delta. 

After the organization of all the available geological 

and historical data we were able to estimate and 

reconstruct the paleo-coastline in several periods for 

the last 100kyrs (Fig. 2). It is quite easy to accept that 

during Tyrrhenian most of the area of Kyllini was 

under the water since the marine sediments were 

deposited. 

The palaeo-delta of Pineios River was developed N 

of Kyllini peninsula before and during Neolithic 

period. The Neolithic and Helladic shoreline was 

located 3,5 km onshore from the present shoreline. 

During the Roman period, Pineios River flowed 

directly S of the Kotychi lagoon forming a levee, 

which is now abandoned, eroded and stands as a 

low sea cliff. An acceleration of coastal deposition 

and consequently delta propagation took place. The 

Roman shoreline was 1,5 km seaward from the 

present shoreline. During the Othoman period, 

Pineios occupied the channel 5 km S of Kotychi 

lagoon forming another levee standing well above the 

floodplain at the shoreline and indicating coastal 

retreat. This channel is in the process of filling. The 

minimum age of this levee is about 200 yrs BP. 

The Pineios diversion to the south of Kyllini peninsula 

took place during the late 18 th century. Following this 

diversion, the pre-18 th -century-A.D. Pineios River 

delta shoreline in now undergoing marine 

transgression and intense coastal erosion, as is to be 

expected in a former delta now essentially starved of 

new sediment. The pre-18 th -century-A.D. northern 

channels of Pineios River and few smaller streams 

can still be seen in their courses to the northwest, 

now dry. The dominant geomorphic processes in the 

modern delta of Pineios River are progradation and 

aggradation with large volumes of river sediment. 

Based on the palaeogeographic reconstructions 

developed by Kraft et al. (2005) and the late 

Holocene environmental changes from Kotychi 

Fig. 2: Shoreline displacements in the study area during the 

last 8 kyrs. 

REMOTE SENSING CONTRIBUTION 

In order to determine whether or not progradation or 

retrogradation took place in Pineios former and 

current deltas in recent years, we initially mapped the 

shorelines at different times in the 27-year-period 

from 1972 to 1999 using (a) topographic maps at 

1:5.000 scale (1972), (b) two datasets of aerial 

photos (1987, 1996), (c) satellite images (1999). 

Then, these data were compared with the present 

shoreline (2011), which was traced with the use of 

real-time kinematic differential GPS. 

The initial phase was to collect the available remote 

sensing data and create a time series of images 

along the contemporary coastline. The oldest data 

available were the topographic maps acquired from 

the Geographic Agency of the Hellenic Army that was 

also based on photogrammetry techniques on 

previously acquired aerial photographs. 

Using 42 air photographs acquired during 1987 we 

generated an ortho-mosaic for the same year. During 

this photogrammetric procedure a high resolution (2- 

meters) DEM was produced, and used for the ortho- 

43


EARTHQUAKE 

ARCHAEOLOGY 



rectification of a 15-meter resolution Landsat-7 

ETM+, panchromatic image. All the data were 

registered with an ortho-mosaic produced by 1996 

aerial photographs (Fig. 3). 

Next, by using image interpretation techniques we 

traced the coastline in the different periods. The 

difficulty was to identify the exact points of contact 

between the seawater and the land. This was made 

by equalizing the image histogram and in some 

cases applying a threshold value. The use of the 

panchromatic part of the spectrum for all the 

collected remote sensing data provides the 

homogeneity of the methodology. 

there is no systematic progradation or retrogradation 

in these delta fronts according to the data covering 

the last 40-year-period from 1972 to 2011 (Fig. 5). 

Nevertheless, there are parts of the coastline, 

especially where the Roman and Othoman levees 

used to function, that most of 50 meters of the beach 

have been eroded. 

Fig. 5: Synthetic image with all the traced coastlines at a 

part of the study area where Pineios river used to flow into 

the sea before 1800’s. 

CONCLUSIONS 

Fig. 3: Parts of the digital data used for the interpretation of 

recent coastline displacement. 

Establishing 4 GPS bases along the shore and use 

the technology of real time kinematic GPS point 

acquisition completed the methodology. The 

accuracy of the present coastline was very good as 

the specifications of the equipment give less than 

10cm (Fig. 4). 

Fig. 4: Using high accurate RTK GPS measurements for the 

tracing of the present coastline. 

The combination of all the traced coastlines on the 

remote sensing data with the RTK GPS recorded 

coastline have shown that both the former and the 

current delta fronts of Pineios River are divided into 

various sub-areas characterized by different type, 

phase and rate of shoreline displacement. Moreover, 

It is obvious that the western part of Pineios drainage 

basin is developed in an area (Gastouni graben), 

which is uplifted with lower relative uplift rate in 

comparison with the other surrounding areas. Hence, 

the Lower Pineios River was and is forced to flow in 

this graben, close and parallel to Pineios FZ. 

Furthermore, the age of Pineios FZ initiation 

progressively decreases from E to W. A similar 

decrease from E to W is also observed in the throw 

of Pineios FZ. The throw of the western part was 

gradually increased until a critical point in time 

(probably during 18 th century A.D.) when the relative 

uplift rate of the Pineios FZ footwall was larger than 

the relative uplift rate of the hanging wall. Since then, 

Pineios River was blocked, not able to flow N-wards 

and over the morphology escarpment formed by the 

fault and consequently enforced to shift S-wards. 

Moreover, the combined uplift movement of the 

footwall of Pineios FZ in the E and the northeastern 

(inland) part of the Kyllini peninsula in the W resulted 

in the slightly uplifted margin of the northwestern part 

of Gastouni graben, the block of the northwards flow 

of Pineios River and the initiation of the southwards 

flow of the river. 

This natural trend of Pineios southward diversion 

during 18 th century was supported and enforced by 

the human activity in the study area and especially by 

the construction of the ancient retaining wall of 

Pineios River (Papaconstantinou, 1991) during the 

Hellenistic period (2.330-2.150 B.C.) in order to 

protect the northern banks from the destructive river 

action. 

44


EARTHQUAKE 

ARCHAEOLOGY 



The study area is undergoing intense and differential 

tectonic deformation, which has continued since 

Pliocene. 

The areas of maximum thickness are constantly 

subsiding during the sedimentation phase and strictly 

related to the Pineios delta and river sediment loads 

and transport. Moreover, the southern area presents 

higher subsidence rates than the northern one. 

The study area is also divided to three subareas that 

uplift with different rates (Fig. 6): (i) the footwall of 

Pineios fault zone (0,39-0,67 mm/yr), (ii) the 

Gastouni graben (0,19 mm/yr), (iii) the eastern 

(inland) part of Kyllini peninsula (0,36-0,48 mm/yr). 

The western part of Pineios basin is corresponding to 

the subarea with the lowest relative uplift rate (0,19 

mm/yr, Gastouni graben). 

Fig. 6: Areas of uplift and subsidence around the Pineios 

river former and contemporary deltas. 

The diversion of Pineios River to the south of Kyllini 

peninsula during the 18 th century is a case of fluvial 

antecedence upon the slightly uplifted margin of the 

Gastouni graben and is the result of the gradually 

increase of the throw along the western part of the 

Pineios fault zone during historic times marked by 

strong and destructive earthquakes during the late 

18 th or early 19 th century A.D. This natural process 

and trend is supported and enforced by the human 

activity in the study area during historic times as it is 

revealed by significant human constructions in the 

area. 

References 

Fountoulis, I., 1994. Neotectonic evolution of the Central- 

Western Peloponnese. Ph.D Thesis, Faculty of Geology, 

National and Kapodistrian University of Athens, GAIA 7, 

386pp. (in Greek, abridged English version) 

Hatzfeld, D., Pedotti, G., Hatzidimitriou, P., Makropoulos, 

K., 1990. The strain pattern in the western Hellenic arc 

deduced from a microearthquake survey: Geophys. J. 

Int., v. 101, p. 181-202. 

Kamberis, E., 1987. Geology and oil-geologic study of the 

post alpine sediments of NW Peloponnese. PhD Thesis. 

National Technical University of Athens, Greece (in 

Greek). 

Kontopoulos, N., Koutsios, A. 2010. A late Holocene record 

of environmental changes from Kotihi lagoon, Elis, 

Northwest Peloponnesus, Greece, Quaternary 

International, v. 225, 2, p. 191-198. 

Kraft, J.C., Rapp, (Rip), G., Gifford, J.A., Aschenbrenner, 

S.E., 2005. Coastal change and Archaeological Setting in 

Elis. Hesperia 74, 1-39. 

Lekkas, E., Papanikolaou, D., Fountoulis, I., 1992. 

Neotectonic Map of Greece, Pyrgos - Tropaia sheets, 

scale 1:100.000, Research project of the University of 

Athens, Department of Geology, Division of Dynamic, 

Tectonic, Applied Geology, Athens 

Lekkas, E., Papanikolaou, D., Fountoulis, I. 1995. The 

Pyrgos earthquake - The geological and geotechnical 

conditions of the Pyrgos area (W. Peloponnese, Greece). 

XV Congress of the Carpatho-Balcan Geological 

Association, Seminar on active faults, Geol. Soc. Greece, 

42-46, Athens. 

Mariolakos, I., Lekkas, E., Danamos, G., Logos, E., 

Fountoulis, I., Adamopoulou, E., 1988. Geological - 

Tectonic Study of the Earthquake affected areas in Elis 

Prefecture (Kyllini Peninsula). Research project of the 

University of Athens, Department of Geology, Division of 

Dynamic, Tectonic, Applied Geology, 108 p., Athens 

1989 (in Greek). 

Mariolakos, I., Papanikolaou, D., Lagios, E. 1985. A 

Neotectonic Geodynamic Model of Peloponnesus Based 

on Morphotectonics, Repeated Gravity Measurements 

and Seismicity. Geol. Jb., v. B 50, p. 3-17. 

Mavroulis, S. 2009. Fault activity assessment in NW 

Peloponnesus – The Andravida Earthquake 

(08/06/2008). Msc Thesis, Inter-University Post-Graduate 

Studies Programme on “Prevention and Management of 

Natural Disasters”, Department of Geology and Geo- 

Environment of the National and Kapodistrian University 

of Athens, and Department of Geoinformatics and 

Topography of the Technological Educational Institute of 

Serres. 622 p. 

Mavroulis, S., Fountoulis, I., Lekkas, E., 2010. 

Environmental effects caused by the Andravida (08-06- 

2008, ML = 6.5, NW Peloponnese, Greece) earthquake. 

In: A. Williams, G. Pinches, C. Chin, T. McMorran and C. 

Massey, Editors, Geologically Active: 11 th IAEG 

Congress, Taylor & Francis Group, Auckland, New 

Zealand (2010), pp. 451-459. 

Papaconstantinou, E., 1991. Architectonic elements of the 

ancient Agora of Elis in the retaining wall of the Pineios 

River. First International Symposium on Achaia and Elis 

in Antiquity, Athens, p. 329-334 (in Greek). 

Papanikolaou, D., Fountoulis, I., Metaxas, C., 2007. Active 

faults, deformation rates and Quaternary paleogeography 

at Kyparissiakos Gulf (SW Greece) deduced from 

onshore and offshore data, Quatern. Int. 171-172, pp. 14- 

30. 

Stamatopoulos L., Voltaggio, M., Kontopulos, N., Cinque, 

A., La Rocca, S., 1988. 230 Th/ 238 U dating of corals from 

Tyrrhenian marine deposits of Varda area (North-western 

Peloponnesus), Greece. Geogr. Fis. Dinam. Quat., v. 11, 

p. 99-103. 

45


EARTHQUAKE 

ARCHAEOLOGY 



COULD LARGE PALAEOEARTHQUAKES BREAK GIANT STALACTITES IN 

CACAHUAMILPA CAVE? (TAXCO, CENTRAL MÉXICO) 

Garduño-Monroy VH. (1), R. Pérez-López (2)*, MA Rodríguez-Pascua (2), J. García Mayordomo (2) 

I. Israde-Alcántara(1) and J. Bischoff (3) 

(1) Universidad Michoacana San Nicolás de Hidalgo. Morelia. México. Email: vgmonroy@umich.mx 

(2) IGME – Instituto Geológico y Minero de España, Área de Investigación en Riesgos Geológicos. Madrid, España. Email: 

r.perez@igme.es , m.rodríguez@igme.es 

(3) United States Geological Survey. 345 Middlefield Road, MS 211, Menlo Park, CA 94025. EE.UU. jbischoff@usgs.gov 

* Corresponding author 

Abstract (Could large palaeoearthquakes break giant stalactites in Cacahuamilpa cave? (Taxco, central Mexico)): the 

Cacahuamilpa cave, an outstanding karstic system located in the nearby of Taxco, Central Mexico, shows several damaged 

dripstones: stalagmites and stalactites. This cave is determined by large E-W horizontal phreatic galleries, with large dripstones, 

mainly stalagmites, stalactites and travertine columns. The size of these speleothems reaches a maximum value of 30 m for 

stalagmites (The Champagne Bottle) and 10 m for stalactites. Oriented and fallen stalagmites and broken stalactites were dated 

by U-Th technique, and we have obtained a probable age of 1540 BP. We also have modelled the earthquake magnitude for 

breaking these stalactites and a minimum magnitude M7 was estimated. The cave is located 250 km away from the Middle 

American Trench, responsible of large earthquakes as the Michoacán, M8.1 in 1985. Besides these large subduction zone 

earthquakes, there are also large active intraplate normal and strike slip faults, but with low reoccurrence intervals that can 

generate large earthquakes in the region. 

Key words: speleoseismology, stalagmites, earthquake, Mexico 

THE CACAHUAMILPA CAVE SYSTEM 

Cacahuamilpa cave (CC) is located within the 

Guerrero State of Mexico, near of Taxco city, wellknown 

for its silver mining industry. This karst system 

is located within the Ixtapan Valley, a NW-SE 

elongated valley, 60 km long and 40 km wide. The 

Nevado de Toluca volcano (4558 m asl) is the 

highest peak and the karst is determined by two main 

fluvial channels: the San Jeronimo (running N-S) and 

the Chontalcoatlán (E-W). Both rivers meet under the 

La Corona hill and both appear outside of the 

carbonitic massif and from their hypogeum tunnelling 

at Dos Bocas. The lowest topographic point 

corresponds with the cave entrance of 

Cacahuamilpa, 1000 m asl. The Amacuzac River 

rises from the cave pit (Fig. 1). The CC belongs to 

the La Estrella Karstic System (LAKS). LAKS is 

configured by several caves in the surrounding of 

Cacahuamilpa: Cuevas Pacheco, Cueva Agua 

Brava, Gruta de Acuitlapán and Cuevas de La 

Estrella, among other minor caves. 

The geology of the cave is described as the Morelos 

Unit, a stratified limestone and dolostone deposit of 

Lower Cretaceous (Fries, 1960), with a maximum 

thickness of 900 m. Furthermore appears Albian 

limestone (Xochicalco and Cuautla Units) and the 

Mezcala Unit, formed by sandstone with interlayered 

black limestone (Fries, 1960). Quaternary deposits 

are volcanic andesitic rocks and basalts and the 

youngest ones are Holocene travertines. 

Hydrothermal activity related with active volcanoes 

(i.e. Nevado de Toluca) is described in Ixtapan and 

Tonatico (35º-40ºC)(Fries, 1960). 

The cave topography is mainly determined by a large 

horizontal phreatic tube (Fig. 2)(Bonet, 1971), with 

semi circular cross section. Large dripstones appear 

along the cave: (1). El Chivo, travertine; (2) and (3): 

La Tortuga and El Guerrero, cracked and toppled 

flowstone and columns. (4) El Guerrero dripstones. 

(5) The Botella de Champán, the greatest stalagmite 

36 m high, (6) The Calendario Azteca (Sun Stone) 

and the Volcán (7), a cone-shaped large flowstone. 

This work aims to demonstrate the possibility that 

large earthquakes (with magnitude M>7) could break 

giant speleothemes, and giving the fingerprint to 

obtain large palaeoearthquakes in the geological 

record that affected the area. 

Fig. 1: Overview of the carbonatic massif of Cacahuamilpa. 

Main fluvial channels run from NW- to SE. The topography 

of the cave is related with the hydraulic gradient. Yellow 

arrows indicate the hydraulic gradient. 

DAMAGED SPELEOTHEMS 

Spelaeoseismology is a new branch of 

palaeoseismology, which studies ancient 

earthquakes from the karstic record in caves (Lacave 

and Koller, 2004; Kagan et al., 2005; Pérez-López et 

46


EARTHQUAKE 

ARCHAEOLOGY 



al., 2009). Earthquakes are well recorded in caves by 

multiple damage affecting dripstones, travertine, also 

by ceiling collapse and faulting. Furthermore, the age 

of damage could be well constrained by using 

techniques such as U-Th disequilibrium of 

speleothems (e.g. Kagan et al., 2005). 

suggest earthquakes larger than M 7 close to the 

cave (50 km away). We have used the technique 

described in Maestro et al. (2011). In both cases the 

fallen structures have the same lithological 

composition (carbonates) and similar geometric 

relationships (H/D factor). However, in this case no 

strong water-flows are related to the toppled 

stalagmites. Nevertheless, the natural vibration of 

such large dripstone could be the responsible of the 

broken pieces (Cadorin et al. 2001). Assuming 

gravitational instability, the stalagmites would be 

oriented to the slip side instead of perpendicular 

(Fig.3). 

SEISMIC SOURCES 

Fig. 2. Cave topography in plant of Cacahuamilpa. 

Numbers indicate sites where broken speleothems are 

described. Red stars show those sites where sampled 

were collected. 1. El Chivo, 2 y 3. La Tortuga. 4. El 

Guerrero. 5. Botella de Champán. 6. Calendario Azetca o 

piedra del Sol. 7. El volcán. After cave topography from 

Bonet, 1971. 

Different damaged speleothems (seismothems) 

appear along the cave: broken stalagmites, 

stalactites, cracked columns and travertines. The 

Calendario Azteca stalactite, or Sun Stone, 

represents one of the most spectacular dripstone 

ever described in caves around the world. The total 

length estimated is 13 m approximately, and it shows 

an outstanding size for a hanging structure, 3.5 m of 

diameter and estimated longitude of 12 m. This E-W 

trending stalactite appears broken in six slices of 

concentric carbonate mega stones from the soda 

straw line (cm). Despite that this stalactite represents 

a national natural monument several samples were 

collected with the aim to obtain the age and the 

carbonate precipitation rate for dripstones in this time 

interval. At this moment data are still processing at 

the laboratory. 

The convergence between the Cocos plate and the 

North American plate at 53 mm/yr (GPS velocity, De 

Mets et al. 1990; Pardo & Suarez, 1995), configures 

and active area dominated by subduction 

earthquakes along the Middle American Trench 

(MAT) (Fig. 4). During the last century, thirteen 

instrumental earthquakes with magnitude greater 

than 7 have been recorded (1967-2011) in the trench 

zone of the MAT, and in a direction perpendicular to 

the cave (ca 250 km away). Moreover, the great 

Oaxaca earthquake of 1931 (M 8.1), the historic 

earthquake of Michoacán (1858) of 7.5 (Singh et al,. 

1985) and the Acambay earthquake of M7 (1912), 

suggest a complex relationship between subduction 

earthquakes and intraplate earthquakes due to 

normal lithospheric faulting. 

ABUNDANCE g/gsample Activity (dpm/s) 

234U 6.90E-11 ±1.4845E-13 0.95341 ±0.0020526 

238U 9.52E-07 ±1.7109E-09 0.7105 ±0.0012766 

230Th 2.95E-13 ±8.7190E-15 0.013429 ±0.00039744 

232Th 1.66E-09 ±4.9300E-11 0.00040293 ±0.000012 

Activity Ratio 

234U/238U 1.34E+00 ±0.0037630 AGE yr BP 1548±47 

230Th/234U 1.41E-02 ±0.00041796 

230Th/232Th 3.30E+01 ±1.4 

232Th/238U 5.67E-04 ±1.6921E-05 

230Th/238U 1.89E-02 ±5.6E-04 

Table 1. Radiometric dating by U-Th disequilibrium of 

the growth stalagmite post breaking (see Fig. 3 down). 

The sample wt is 0.52 g. Sample CAVO11-09. 

Other spectacular seismothems and fallen 

stalagmites are similarly oriented and could be 

related with the same phenomenon (Fig.2, 7 th site, El 

Volcán)(Fig.3). In this site, 4 stalagmites ranging 

between 4.2 and 1.7 m, with a diameter between 

0.57 and 0.4 m, appear broken and oriented E-W. 

We have dated the breaking by U-Th on the 

subsequent growing post the oriented toppling (Table 

1). 

The natural period of vibration of stalagmites for H/D 

ranging between 3 and 4 and acceleration measures 

Fig. 3. Up: Broken and oriented stalagmites located at El 

Volcán (see Fig. 2). Down: Photo interpretation of the 

fallen speleothems. The axis orientation is E-W, 

approximately. Red dots indicate sample collected points 

and blue ones points to be sampled in the future field 

work. Black and grey triangles indicate the maximum 

slope direction. 

47


EARTHQUAKE 

ARCHAEOLOGY 



The Michoacán earthquake of the 19 th of September, 

1985 was the more devastated with a magnitude of 

8.1. This earthquake caused 9.500 fatalities, about 

30.000 people were injured and more than 100.000 

people were homeless, whereas severe damage was 

caused in buildings of Mexico DF and in different 

states of central Mexico also 

(http://earthquake.usgs.gov/earthquakes/world/event 

s/1985_09_19.php). 

Therefore, both the MAT subduction earthquakes as 

the large intraplate normal and strike-slips faults (e.g. 

the Morelia-Acambay Fault System, The Oaxaca 

Fault zone)(Garduño-Monroy et al., 2009), could 

generate large earthquakes to affect large 

speleothems within the Cacahuamilpa cave. It is 

interesting to describe large palaeoearthquakes 

triggered by intraplate faults in this zone due to, at 

present there is no instrumental record for such sized 

earthquakes. Therefore, infrequent large intraplate 

earthquake could be hibernating and palaeoseismic 

studies in the surroundings will be worthy to fill up the 

gap in the historical records. 

RESULTS 

Preliminary results suggest a potential 

palaeoearthquake dated c.a.1548 ±48BP, obtained 

from U-Th disequilibrium dating technique (Table 1.). 

The magnitude of the potential earthquake has been 

speculated from the geometry and mechanical 

properties of the stalagmites. Results suggest M>7. 

The geometry factor is defined by the relationship 

between the high and width (H/D), and the 

mechanical properties assuming a crystalline 

carbonate. Hence, we can conclude that: 

(1) The broken and oriented stalagmites within the 

Cacahuamilpa cave could be related with 

palaeoearthquakes. 

(2) The large size of damaged speleothems could 

be related with large earthquakes (M>7), in 

agreement with the tectonic framework of the 

Pacific side of Mexico: the convergence between 

Cocos and North American plates. 

(3) Palaeoearthquakes in this zone could be related 

to either subduction earthquakes or large 

intraplate earthquakes but with long 

reoccurrence intervals. 

The next step is to assign the potential seismic 

source. More accurate data are required to 

reconstruct the complete history of the broken 

speleothems in Cacahuamilpa. However, the 

earthquake hypothesis appeared as the strongest 

cause for this large damage. The interest of this work 

is for finding large shallow earthquakes (>M7) in a 

zone which is located 250 km far from the Middle 

American Trench. 

Fig. 4. Tectonic frame of Mexico. Most of earthquakes are related to the subduction between the Cocos plate and North 

American plate. Earthquakes represented here were obtained from Harvard online catalog (http://www.globalcmt.org/). After 

Pérez-López et al. (2011). TMVB: Trans Mexican Volcanic Belt 

48


EARTHQUAKE 

ARCHAEOLOGY 



Acknowledgements: We are very grateful to Parque 

Nacional Grutas de Cacahuamilpa to give us all 

permissions to access to all parts of the cave, for extract 

samples and provide people to support us during the 

underground field work. Thanks are given to the CONACYT 

ministerial Mexican project entitled: Estudio tectónico, de 

paleosismología y de efectos sísmicos arqueológicos en 

lagos del Holoceno y actuales del Cinturón Volcánico 

Transmexicano y del bloque Jalisco. 

References 

Bonet, F. (1971). Espeleología de l aRegión de 

Cacahuamilpa. Universidad Autónoma Nacional 

Mexicana Instituto Geológico. Boletín nº 90. 98 p. 

Cadorin, J.-F., Jongmans, D., Plumier, A., Camelbeeck, T., 

Delaby, S. and Quinif, Y. (2001). “Modelling of 

speleothems failure in the Hotton cave (Belgium). Is the 

failure earthquake induced?” Netherlands Journal of 

Geosciences 80(3–4): 315–321. 

DeMets, C., R.G. Gordon, D.F. Argus, and S. Stein, 

(1990).Current plate motions. Geophysics Research 

Journal International 101: 425-478. 

Garduño-Monroy V. H., R. Pérez-López, I. Israde-Alcantara, 

M. A. Rodríguez-Pascua, E. Szynkaruk, V. M. 

Hernández-Madrigal, M. L.García-Zepeda, P. Corona- 

Chávez, M. Ostroumov, V. H. Medina-Vega, G. García- 

Estrada, O. Carranza, E. Lopez-Granados and J. C. Mora 

Chaparro. (2009).Geometry and Paleoseismology of the 

Southwestern Part of the Morelia-Acambay Fault System, 

Central Mexico. Geofisica Internacional 48(3): 319- 335. 

Kagan, E. J., Agnon, A., Bar-Matthews, M. & Ayalon, A. 

(2005). Dating large, infrequent earthquakes by damaged 

cave deposits. Geology 33(4), 261–264. 

Fries, C. (1960). Geología del Estado de Morelos y de 

partes adyacentes de México y Guerrero, región central 

meridional de México. Instituto Geológico de México, 

Boletín nº 60(9), 236p. 

Lacave C. & M. G. Koller . (2004). What Can Be Concluded 

About Seismic History From Broken And Unbroken 

Speleothems? Journal of Earthquake Engineering, Vol. 8, 

No. 3 (2004) 431–455. 

Maestro A, G. Jané, J. García-Mayordomo, B. Fernández- 

Revuelta, M.A. Rodríguez-Pascua, J.J. Martínez-Díaz & 

R. Pérez-López. (2011). Cause Of The Rupture And 

Distribution Of Broken Submarine Carbonate Chimneys 

In The Gulf Of Cádiz (Southwestern Spain). Quaternary 

Int. In Press. 

Pardo, M. & Suárez, G. (1995). Shape of the subducted 

Rivera and Cocos plates in southern Mexico: seismic and 

tectonic implications. Journal of Geophysical Research 

100: 12,357-12,373. 

Pérez-López R., Rodríguez-Pascua, M.A., Giner-Robles, 

J.L., Martínez-Díaz, J.J., Marcos-Nuez, A., Silva, P., 

Bejar, M. and Calvo, J.P. (2009). Spelaeoseismology 

and palaeoseismicity of the “Benis Cave” (Murcia, SE of 

Spain): coseismic effects of the 1999 Mula earthquake 

(mb 4.8). Geological Society of London, Special 

Publications, 316: 207-216. 

Pérez-López R., D. Legrand, V.H. Garduño-Monroy, M.A. 

Rodríguez-Pascua , J.L. Giner-Robles. (2011). Scaling 

laws of the size-distribution of monogenetic volcanoes 

within the Michoacán-Guanajuato Volcanic Field 

(Mexico). Journal of Volcanology and Geothermal 

Research 201: 65–72. 

Singh, S. K., Gerardo Suárez & T. Domínguez. (1985). The 

Oaxaca, Mexico, earthquake of 1931: lithospheric normal 

faulting in the subducted Cocos plate. Nature 317:56-58. 

49


EARTHQUAKE 

ARCHAEOLOGY 



THREE-DIMENSIONAL INVESTIGATION OF THE AD 1621 PEDRO MIGUEL FAULT 

RUPTURE FOR DESIGN OF THE PANAMA CANAL’S BORINQUEN DAM 

Gath, Eldon M. (1 and Tania Gonzalez (1) 

(1) Earth Consultants International, 1642 E. 4 th Street, Santa Ana, California, 92701, USA. Email: gath@earthconsultants.com 

Abstract (Three-dimensional investigation of the AD 1621 Pedro Miguel fault rupture fro desing of the Panama canal`s 

Bronquen dam): Using a series of trenches, we excavated the Pedro Miguel fault in 3-D to measure the displacement magnitude 

and kinematics of the AD 1621 rupture. The purpose of the study was to use the Most Recent Event event as a proxy for a future 

rupture of the fault through the foundation of Borinquen Dam, a major new component of the Panama Canal Expansion. The study 

used a small, fault-affected, cobble-filled channel as the target for displacement measurements. Hundreds of ground survey 

points were obtained for contacts, faults, the channel thalweg and margins. The channel is offset 3.0±0.2 m right-laterally, and 

0.5±0.5 m reverse-vertically, with the vertical component occurring only within a few meters of the main fault tip. The fault rupture 

is expressed as a low-angle, one-sided, transpressive flower structure, exploiting weak bedding planes to propagate an enechelon 

stepping rupture across the landscape. Mitigation of this rupture will be an important requirement for the dam designers. 

Key words: Panama, paleoseismology, dam design, fault rupture 

Introduction 

As part of the Panama Canal’s Expansion Project, a 

four-segment earthen dam is being designed to form 

an ~7 km-long waterway to bypass the existing 

Miraflores and Pedro Miguel Locks (Fig. 1). 

Borinquen Dam will retain the Gatun Lake water 

elevation ~11 m above the Miraflores Lake elevation, 

and as such, must be designed to resist seismic 

loads. Extensive prior paleoseismic investigations 

(Rockwell et al., 2010) of the Pedro Miguel fault have 

shown the dam must also be designed to resist fault 

rupture. 

500±100 years, and its Most Recent Event (MRE) on 

May 2, 1621. Attempts to constrain the 

displacements from the MRE resulted in only 

minimum values of ~2 m near the dam site, but a well 

constrained 2.8 m at the fault’s northern end where it 

severed the Camino de Cruces (Gath and Rockwell, 

2009). The purpose of this latest investigation (ECI, 

2010) was to attempt to reconcile these two 

displacement results at the dam’s location. 

Fig. 1: Aerial view, looking NW, of the Pacific approach to 

the Panama Canal, showing the proposed location of the 

new locks, channel, and Borinquen Dam. The northerlytrending 

Pedro Miguel fault cuts through the proposed dam 

foundation about midway between the existing Miraflores 

and Pedro Miguel locks. 

The Pedro Miguel fault is a right-lateral strike-slip 

fault that passes through the planned Borinquen 

Dam’s foundation (Fig. 1). In earlier work, we (ECI, 

2007, 2008, 2010) determined that the fault has had 

multiple Holocene ruptures, a late Quaternary slip 

rate of ~5 mm/yr, a Holocene recurrence interval of 

Fig. 2: The investigation’s target was a small channel that 

appeared to be offset 3-4 meters. The fault (red line) was 

inferred to extend through the area, right-laterally offsetting 

a channel that approaches from the photo’s upper right 

corner and exits along the bottom left (shown by the 

geologists and the blue lines). 

From earlier studies and recent construction 

exposures, we knew the fault location to within a few 

meters (Fig. 2). The purpose of this paper is to 

present the technical details of the study’s findings, 

and to also present and discuss the methodology of 

the investigation, including the planning, execution, 

findings, and modifications that were made along the 

way, because there were many. 

50


EARTHQUAKE 

ARCHAEOLOGY 



Investigation 

The investigation began with a simple plan (Fig. 3) to 

expose the fault on both sides of the small 

geomorphically defined area, and then characterize 

the channel geometry on both sides of the fault. 

Once the site was delineated, the channel margins 

could be slowly excavated towards the fault trace 

using a series of thin slices. Unfortunately, the initial 

fault-perpendicular trench (T-48 in Fig. 4) did not 

expose the fault where expected, and we failed to 

recognize the significance of a fault that was exposed 

elsewhere in the trench. The second fault-locator 

trench (C in Fig. 4) immediately filled with water from 

a sudden storm and was abandoned for four days 

until it could be pumped out. The two fault-parallel 

trenches (A and left part of B in Fig. 4) intended to 

define the channel geometry into and out of the fault 

did not expose any channel deposits or channel 

morphology. After four initial trenches, our 

investigation was definitely in trouble, with no fault 

and no channels to show for the work done to that 

point. 

excavations did not remove the geologic data and 

relationships that were vital to the measurement of 

the channel displacements. This was accomplished 

by excavating from the outer edges of the site 

inward, and by always keeping a mental map of the 

site and the goal, a preserved and measurable 

channel offset. 

N 

N 

Fig. 4: Schematic layout of our final trenching study. 

Trenches are shown as rectangles, the faults are in red, and 

the channel structure is shown by the blue lines, trending 

across the middle of the study area. The complex nature of 

the fault rupture pattern meant that the initially simple 

geomorphic offset inferred from the pre-trenching landscape 

was incorrect. The channel was effectively trapped within 

the fault zone, and each transpressive “petal” of the fault 

offset the channel progressively. The channel that we 

trenched first appears to have been man-made, to facilitate 

surface drainage to a culvert under Borinquen Road. 

Fig. 3: The original plan was to locate the fault (red line) on 

opposite sides of the displaced channel (blue arrows) by 

trenching perpendicular to the fault (large gray rectangles), 

then locate the channel margins by trenching parallel to the 

fault and perpendicular to the channel form (small gray 

rectangles). Using hand excavations and continuous survey 

control, we would then excavate the channel margins 

progressively closer to the fault, until they were in faultcontact 

on both sides of the fault. 

Fig. 4 diagrammatically shows the trenches that were 

finally excavated as we tried to sort out the details of 

the site and salvage some data for use by the dam 

designers. Once T-48 and 48-C failed to expose the 

faults where expected (Fig. 4), and trenches 48-A 

and B failed to expose the channels where expected, 

we lengthened 48-B until we found both the fault and 

the channel (Fig. 6). Fortunately, the channel was 

still fully contained on the hanging wall of the fault, 

and was not yet in fault contact, so we had not 

removed that important interaction point with our 

excavation. 

Fig. 5 shows a modification of Fig. 3 to reflect the 

pattern of the faults and channels at the site, as 

defined by the final trenches. The challenge was to 

continue the excavations but be careful that the 

N 

Fig. 5: Pedro Miguel fault trenching site immediately south 

of the old Borinquén Road (base of photo), following brush 

removal, but before trenching started. The fault and channel 

locations, as interpreted from the geomorphology, are 

shown with the lighter, dashed lines, whereas the actual 

fault and channel locations found after trenching are shown 

diagrammatically with the bold and solid lines. 

Fig. 7 shows the channel on the hanging wall above 

the fault, whereas Fig. 8 shows the structural 

complexity of the fault zone that forced us to evolve 

the initial investigation plan to accommodate the 

unexpected. The extreme low angle of the fault 

acted as a bulldozer of the surface soils pushing 

them out and over the channel alluvium, but this also 

51


EARTHQUAKE 

ARCHAEOLOGY 



served to bury, and thereby preserve, the channel 

deposits under the fault petal. 

The purpose of digging 48-K (Fig. 9) was to continue 

the exposure of 48-B up-dip of the fault to get as 

close as possible to the spot where the fault first cut 

the base of the channel. With the fault dipping to the 

west and the channel flowing to the east, this could 

occur suddenly. In Fig. 9 it appears that the deepest 

part of the channel is touching the fault, but there are 

still 3-5 cm of separation. In 48-M (Fig. 10) however, 

the base of the channel is the fault, and the SE 

channel margin is completely removed by the fault. 

Thus, our channel margin’s northern piercing point 

lies between 48-K and 48-M (±2 m), and the 

thalweg’s piercing point lies within 48-M (±0.5 m). 

Fig. 6: Trench 48-B - the discovery trench, showing the fault 

(right) and the channel deposits (left), looking out over the 

Panama Canal in the background. 

Fig. 9: Interpreted image of the end of Trench 48-K, with 48- 

B (Fig. 7) ~ 1 meter on the other side (to the south) of the 

trench. This image shows the channel deposits still above 

the fault trace. 

Fig. 7: Interpreted image of part of Trench 48-B showing the 

low-angle fault and the channel deposits on the hanging 

wall. The area shown as a “clay extrusion” is interpreted to 

be a weathered mole track from the MRE as it is intruded 

into, and deformed, the modern surface soils.. 

Fig. 10: Looking easterly at the interpreted image of the end 

of Trench 48-M, excavated ~1 m to the left (east) of 48-K 

(Fig. 7). This photo shows the base of the channel deposits 

now in fault contact in the head of the trench, and truncated 

by the fault on the right side. 

Fig. 8: Interpreted image of Trench 48-B, looking back 

towards Fig. 7, showing the low-angle fault petals where 

they have broken upwards to the surface, and the lack of 

alluvial deposits on the eastern (right) wall. 

In addition to the fault complexity shown in Fig. 8, it is 

important to note that there were no alluvial deposits 

visible on the fault’s footwall in the northern wall of 

the trench. However, trenches 48-F, G, H, J, & M all 

exposed channel deposits and cobbles (Figs 4, 10 

and 11). This is because the channel margin on the 

footwall lies 0.5-1.0 m north of the north face of 

Trench 48-B. Thus Trench 48-B missed taking out 

52


EARTHQUAKE 

ARCHAEOLOGY 



the southern margin of our target channel by less 

than 1 m. With the southern margin so tightly 

constrained, the most accurate offset measurements 

came from that side. 

channel by the most recent earthquake. Using 15 

trenches and hundreds of surveyed data points, we 

were able to constrain the MRE rupture to 3.0±0.2 m 

of right-lateral displacement, and 0.5±0.5 m of 

localized reverse-slip uplift at the surface tip of the 

fault. Because the fault is expressed through the 

dam as an en-echelon stepping, transpressional 

flower structure that exploits the weak bedding 

planes of the near-surface strata, it will be a difficult 

fault to mitigate in the design of Borinquen Dam (Fig. 

13). 

N 

Fig. 11: Interpreted image of Trench 48-H, with 48-B (Fig. 8) 

exposed through the window at the end of the trench. This 

image shows the limits of the channel deposits on the 

footwall side of the fault, and shows a secondary fault petal 

on the right wall vertically truncating the cobble deposits 

within the excavated width of the trench. 

Fig. 13: Map of the Pedro Miguel and Miraflores faults 

through the Borinquen Dam area (hachured). Areas where 

we have conducted paleoseismic trenching are shown with 

the green squares; the green square directly over the dam 

location is the location of this study. 

Acknowledgements 

Thanks to Kay St. Peters and Barrett Salisbury for excellent 

field assistance and data mapping. Appreciation is due to 

the Autoridad del Canal de Panama for permission to 

conduct this study and to Ms. Pastora Franceschi for 

arranging all the details. 

Fig. 12: Geologically interpreted map of the channel offsets 

based on hundreds of survey points collected. The yellow 

area reflects the full margins of the sandy channel deposits 

whereas the blue area defines the cobble-filled channel 

thalweg. The northern margin of the channel is more poorly 

constrained (±2 m), while the southern margin and thalweg 

margins are constrained to less than ±1 m. The right-lateral 

offset across the fault is measured at 3.0±0.2 m. A 0.5 m 

vertical offset occurred at the tip of the main fault petal, 

resulting in the upward bowing and erosional removal of the 

channel thalweg, but this uplift is localized to only a few 

meters from the fault tip and is not present away from the 

fault. Farther east, the fault tip over-rides and protects the 

alluvial channel deposits. 

Conclusions 

Although the study was much more complicated than 

initially planned, we were successful in locating the 

fault, in exposing a young channel that was offset 

across the fault, and in measuring the offset of that 

References 

Earth Consultants International (ECI), 2007, Paleoseismic 

Trenching of the Pedro Miguel Fault in Cocolí, Located 

Immediately Southwest of the Panamá Canal, Panamá; 

consulting report for the Autoridad del Canal de Panamá 

(ACP), Project No. 2614.02. 

ECI, 2008, Quantitative Characterization of the Pedro 

Miguel Fault, Determination of Recency of Activity on the 

Miraflores Fault, and Detailed Mapping of the Active 

Faults Through the Proposed Borinquén Dam Location; 

consulting report for the ACP, Project No. 2708.05. 

ECI, 2010, Additional Trenching of the Pedro Miguel and 

Miraflores faults in Cocoli, immediately southwest of the 

Panamá Canal; consulting report for the ACP, Project No. 

3012. 

Gath, E.M. and Rockwell, T.K., 2009, Coseismic offset of 

the Camino de Cruces confirms the Pedro Miguel fault as 

the cause of the AD 1621 Panamá Viejo earthquake; in 

Pérez-Lopez, R. and others (eds), Archeoseismology and 

Palaeoseismology:: 1 st 

INQUA-IGCP-567 Int. Workshop 

on Earthquake Archaeology and Palaeoseismology, 

Baelo Claudia, Spain, p. 32-34. 

Rockwell, T. and others, 2010, Neotectonics and 

paleoseismology of the Limón and Pedro Miguel faults in 

Panamá: Earthquake Hazard to the Panamá Canal: B. of 

the Seis. Soc. America, v. 100, p. 3097-3129. 

53


EARTHQUAKE 

ARCHAEOLOGY 



VELOCITY FIELD IN BULGARIA AND NORTHERN GREECE FROM GPS CAMPAIGNS 

SPANNING 1993-2008 

Georgiev, Ivan (1, Dimitar Dimitrov (1), Pierre Briole (2), Emil Botev (3) 

(1) Department of Geodesy, National Institute of Geophysics, Geodesy and Geography, Bulgarian Academy of Sciences, Sofia 

1113, Acad. Bonchev Str., Bl. 3, BULGARIA, E-mail: ivan@bas.bg 

(2) Ecole Normale Superieure, Paris 75230, 45 rue d’Ulm, FRANCE 

(3). Department of Seismology, National Institute of Geophysics, Geodesy and Geography, Bulgarian Academy of Sciences, Sofia 

1113, Acad. Bonchev Str., Bl. 3, BULGARIA 

Abstract (Velocity Field in Bulgaria and Northern Greece from GPS Campaigns Spanning 1993-2008): We 

use GPS observations during the period from 1993 to 2008 to obtain the velocities of about 100 points in Bulgaria 

measured in more than 30 campaigns. Along with data from 21 points in Northern Greece, measured between 1999 

and 2008, the results constrain the present-day motion in the region. Points in FYROM and Albania are also 

included in the processing to outline the picture of recent motions from Black to Adriatic Sea. The velocity field 

shows overall motion to the south relative to stable Eurasia and increasing motion from north to south confirming 

the extensional regime in Southwest Bulgaria and Northern Greece. The results can be used to constrain the north 

boundary of South Balkan extensional region. 

Key words: GPS, velocity field, Bulgaria and Nortern Greece, South Balkan Extensional Region 

MOTIVATION 

The region of South Bulgaria, especially Southwest 

Bulgaria and the Rhodopes Mountains, is the most 

active tectonic and seismotectonic area of the 

country with proved recent active tectonic structures 

and crustal movements. The strongest earthquake in 

continental Europe in the last two centuries occurred 

in the Krupnik-Kresna region. The territory belongs to 

the southern part of the Central-Balkan neotectonic 

region – a zone with recent extension of the crust 

and with complex interaction between the horizontal 

and vertical movements of the tectonic structures 

(Zagorchev, 2001). The geological and geophysical 

data confirm the recent activity of the fault structures 

formed during the Late Neogene and the Quaternary. 

The area is located in the northern part of the North 

Aegean region and is strongly affected by its 

tectonics and high seismicity. This is the main reason 

for establishing in the early 2000 a relatively dense 

geodynamic GPS network in Southwest Bulgaria for 

long-term monitoring of recent crustal movements. 

The network is repeatedly measured in the period 

2001 - 2008. Along with GPS data from the new 

National GPS network of Bulgaria and a few EUREF 

campaigns spanning the period 1993 - 2008 we 

obtained velocities of about 100 points in Bulgaria. 

We use also GPS data from several campaigns in 

Northern Greece, few points in FYROM and Albania 

to obtain the velocity field and to constrain tectonics 

in the region. The obtained results can be used to 

constrain the northern boundary of the South-Balkan 

extensional province. 

GPS DATA 

GPS data used in the solution include campaigns on 

the territory of Bulgaria, Northern Greece, FYROM 

and Albania and comprise 158 points. In Bulgaria 

GPS data span the period 1993 – 2008. The data 

from the nineties are mainly from the EUREF 

campaigns in Bulgaria. Most of the data are from 

extensively measured geodynamic network in 

Southwest Bulgaria and points from the new National 

GPS network. The typical duration for each point 

measurement in the successive campaigns is 48 

hours. The points with long observation history are 

98. In Northern Greece we analyzed data from 21 

points collected in 1999, 2000 and 2008. The data 

from 2008 are from joint Bulgarian-Greek GPS 

campaign with 48 to 72 hours measurements. Data 

from 5 points in FYROM are included in the solution 

from campaigns in 1996, 2000 and 2008. The data 

from 1996 and 2000 are measured in two EUREF 

campaigns, courtesy given by the Agency of 

Cadastre and the 2008 campaign is performed jointly 

by Department of Geodesy, Sofia, and Faculty of 

Civil Engineering, University “Ss. Cyril and Metodius”, 

Skopje. Albanian data comprise 34 stations in four 

campaigns from 2003 till 2009 and are committed by 

the Laboratory of Alpine Belts Geodynamics, 

University of Grenoble, France. 

All the data are processed and analyzed by the 

Bernese 5.0 software. The reference frame is 

ITRF2005 and the processing follows the EUREF 

standards (see for example Boucher and Altamimi, 

2007). The reference and kinematic frame are 

defined by points of the European Permanent 

Network (EPN) included in the solution. The details of 

54


EARTHQUAKE 

ARCHAEOLOGY 



the processing strategy can be found in Georgiev 

(Georgiev et al., 2007). 

RESULTS AND DISCUSSION 

Fig. 1 shows the obtained velocities of all 158 points 

in Bulgaria, Northern Greece, FYROM and Albania 

relative to (stable) Eurasia. The results filled in the 

gap in velocity field in the Eastern Mediterranean to 

the north of the extensively studied Marmara and 

Aegean regions. Results from other authors are also 

presented on the figure (Reilinger et al., 2006, 

Hollenstein et al., 2008, Floyd et al., 2010) to show 

the position of Bulgarian and Northern Greece 

territory in the East Mediterranean geodynamic 

settings. The velocities of the Albanian stations and 

points in Bulgaria (red vectors, points mainly from the 

National GPS network) show the velocity field in 

east-west direction from Black to Adriatic Sea. 

Fig. 1: Velocity field in Bulgaria, Northern Greece and surrounding areas relative to Eurasia. The South Balkan extensional region 

is outlined according to Burchfiel et al. (Burchfiel at al., 2006) 

mm/y in Central-west Bulgaria (Balkan Mountain) to 

10 mm/y for the region of Chalkidiki. On the northsouth 

profile from the Balkan Mountain till the 

Chalkidiki peninsula the north component of the 

estimated velocities are drawn (Fig. 2). The 

increasing velocities from north to south are clearly 

seen on this 400 km long profile confirming the northsouth 

extensional regime in the region (see also 

McClusky et al., 2000, Hollenstein et al. 2008). 

Fig. 2: North-south velocity gradient from Central-west 

Bulgaria (Balkan Mountain) to the region of Chalkidiki 

The obtained velocities in Southwest Bulgaria and 

Northern Greece indicate an overall motion to the 

south. The results show an increasing rate from 2 

One point deviates from the overall south motion in 

Southwest Bulgaria and it is clearly seen on the 

velocity gradient profile (Fig. 2). This is a point from 

the local geodynamic network of the Department of 

Geodesy around the Krupnik fault, located on its 

northern edge. The result is in good agreement with 

the dipping of the fault obtained by geologic and 

seismotectonic data. 

The points in northern and eastern part of the 

Bulgarian territory (red vectors) are not dense 

enough to draw unquestionable conclusions. But 

from the obtained results it is clear that North 

Bulgarian territory, north from the Balkan Mountain, is 

55


EARTHQUAKE 

ARCHAEOLOGY 



part of Eurasia with velocities practically near to zero. 

On the Fig. 1 are shown the slightly changed borders 

of the South Balkan extensional region proposed by 

Burchfiel et al. (Burchfiel at al., 2006). According to 

this model the northern border of the extensional 

region follow approximately the Balkan Mountain. 

Our data as a whole, although not sufficiently dense 

in Southeast Bulgaria, seems to confirm this 

hypothesis. The question which remains is the 

behavior of few points located on the southeastern 

Black sea coast showing north-northeast motion. We 

need a denser network in this region to clear these 

two questions – to confirm the northern border of the 

South Balkan extensional region and to resolve the 

motion along the Black Sea coast. 

Acknowledgements: We gratefully acknowledge the 

support of all agencies in the countries for the 

measurements: Military Geographyc Service and Agency of 

Geodesy, Cartography and Cadastre, Sofia, Bulgaria; 

Faculty of Civil Engineering, University “Ss. Cyril and 

Metodius”, Skopje, FYROM; Aristotle University of 

Thessaloniki, Greece and Laboratory of Alpine Belts 

Geodynamics, University of Grenoble, France. 

References 

Boucher, C., Altamimi, Z. (2007). Memo: Specifications for 

reference frame fixing in the analysis of a EUREF GPS 

campaign, Version 27-03-2007. 

Burchfiel, B.C., King, R.W., Todosov, A., Kotzev, V., 

Durmurdzanov, N., Serafimovski, T., Nurce, B. (2006), 

GPS results for Macedonia and its importance for the 

tectonics of the Southern Balkan extensional regime. 

Tectonophys 413 (3-4), 239-248. 

Floyd, M. A., et al. (2010), A new velocity field for Greece: 

Implications for the kinematics and dynamics of the 

Aegean. J. Geophys. Res., 115 (10), B10403, 

doi:10.1029/2009JB007040. 

Georgiev I., Gabenski, P., Gladkov, G., Tashkov, T., 

Danchev, P., Dimitrov D. (2006). NATIONAL GPS 

NETWORK: Processing the observations from the 

Main order. Geodesy N 18, Special edition, Military 

Geographic Service of the Bulgarian Army, 209p. 

Georgiev I., (2009). National and Permanent GPS networks 

of Bulgaria – processing of the observation, analysis and 

application in geodynamic Thesis to obtain the degree 

Doctor of Science”, Department of Geodesy, National 

Institute of Geophysics, Geodesy and Geography, Sofia, 

Bulgaria, 258p. 

Hollenstein et al. (2008). Crustal motion and deformation in 

Greece from a decade of GPS measurements, 1993– 

2003. Tectonophysics 449, 17–40. 

Matev, K. (2011). GPS constrains on current tectonics of 

Southwest Bulgaria, Northern Greece and Albania. 

Thesis to obtain the degree “Doctor” of University of 

Grenobe, Laboratory of Alpine Belts Geodynamics, 

University of Grenoble, France, Department of Geodesy, 

National Institute of Geophysics, Geodesy and 

Geography, Sofia, Bulgaria, 203p. 

McClusky, S., Balassanian, S., Barka, A. Demir, C. 

Ergintav, S., Georgiev, I., Gurcan, O., Hamburger, 

O., Hurst, K, Kahle, H., Kastens, K., Kekelidze, G., 

King, R,, Kotzev, V., Lenk, O., Mahmoud, S., 

Mishin, A., Nadariya, M., Ouzounis, A.. Paradissis, 

D., Peter. Y., Prilepin, M., Reilnger, R., Sanli, I., 

Seeger, H., Tealeb, A., Toksoz, M., Veis, G. (2000). 

Global Positioning System Constraints on Plate 

Kinematics and Dynamics in the Eastern 

Mediterranean and Caucasus. Journal of 

Geophysical Research - Solid Earth, Vol. 105, No.B35, 

5695-5719. 

Reilinger, R., et al. (2006), GPS constraints on continental 

deformation in the Africa-Arabia-Eurasia continental 

collision zone and implications for the dynamics of plate 

interactions, J. Geophys. Res., 111, B05411, 

doi:10.1029/2005JB004051. 

Zagorchev, I. (2001). South Western Bulgaria, Geological 

Guidebook. Sofia, Bulg. Acad. Sci., 98p. 

56


EARTHQUAKE 

ARCHAEOLOGY 



ACROCORINTH - GEOLOGICAL HISTORY AND THE INFLUENCE OF 

PALEOSEISMIC EVENTS TO RECENT ARCHAEOLOGICAL RESEARCH 

Gielisch, Hartwig (1) 

(1) EurGeol. Dr. Hartwig Gielisch, DMT GmbH & Co. KG, Am Technologiepark 1, 45307 Essen Germany 

Phone, +49 201 172 1888, email: hartwig.gielisch@dmt.de 

Abstract (Acrocorinth - Geological history and the influence of paleoseismic events to recent archaeological research): 

The Acrocorinth is the result of two major tectonical events. From Upper Cretaceous to Eocene the Acrocorinth limestone, origins 

from the Mesozoic carbonate platforms, were mixed with cherts of the Böotian Ocean in an accretionary wedge of the Central 

Hellenic Melange Belt. The melange units were covered by Neogene Acrocorinth conglomerates. Pliocene Corinthian marls cover 

all older units. In the Pleistocene the hill was pushed up very rapidly through the Tertiary units. Pleistocene marine terraces occur 

in 220 m a. s. l. at Acrocorinth of Pre-Paleotyrrhean age, thus verifying an uplift of more than 200 m in the last 400,000 years. All 

sediments are covered by the several sheets? of debris, the results of several major paleoseismic events. These breccias contain 

as components Mesozoic and Cainozoic rocks occurring at the Acrocorinth and cover probably? the remaining nine temples 

described by Pausanias. In order to find the missing temples it is essential to know, which part of the sediments is in-situ or are 

part of landslides triggered by paleoseismic events. 

Key words: conglomerates, rapid uplift, debris breccias, covered temples 

INTRODUCTION 

Acrocorinth is located in the northern Argolis south of 

the Corinthian golf and directly south of the ruins of 

ancient Corinth. The geological history of the hill is 

subdivided in two major tectonical events: the 

genesis of a melange during a continent-continent 

collision in the lower Tertiary and a rapid uplift in the 

younger Pleistocene. Several paleoseismic events 

form the actual shape of the hill. 

GEOLOGICAL HISTORY OF THE ACROCORINTH 

The geological history of Acrocorinth starts in the 

middle Mesozoic. Neritic lime- and dolostones 

formed on the Parnassus-carbonate platform as part 

of the Centralhellenic Systems, while in the center of 

the Böotian Ocean, a sub basin of the Pindos Ocean, 

pelagic cherts and ophiolitic sequences were built. 

During the closing of the Böotian Ocean carbonates, 

cherts and ophiolites were tectonically mixed as part 

of an accretion in the subduction trench thus forming 

a mélange belt between the Parnassus- and 

Pelagonian-plates (GIELISCH, 1993, 1994). The 

youngest components of these mega-breccias are of 

Eocene age. The alpine orogenesis of Greece is 

completed in the Lower to Middle Miocene. 

During the Upper Miocene the alpine Hellenides were 

fragmented by a phase of intense brittle faulting. The 

major systems of the Greek intra-mountainous 

graben-systems developed during the Upper 

Miocene and the Lower Pliocene. 

During and directly after these phases of brittle 

faulting the Acrocorinth conglomerates have been 

deposited over the Mesozoic units. Components of 

the conglomerates comprise mainly local and 

regional geological units (carbonates, cherts, 

serpentinites of the ophiolites, quartzites) (RICHTER 

et al., 1992). Exotic clasts like olivine-gabbros, 

gap 

limestones 

(Triassic) 

v 

v 

v 

v 

v 

v 

v 

v 

v 

v 

v 

flyschsandstones 

(Cretaceous- 

Paleogene) 

gap 

gap 

v 

v 

v 

ophiolites 

(Jurassic) 

Acrocorinth 

limestones 

(U. Jurassic- 

L. Cretaceous) 

v 

v 

v 

Sheet of debris 

(Quarternary) 

Corinthian Marls 

(Neogene) 

Acrocorinth 

conglomerates 

(Lower Neogene) 

radiolarites, 

cherts 

pyroclastic 

intercalations 

(Jurassic) 

Fig. 1: Schematic profile of sequences in the Corinthian 

area showing the stratigraphical position of the main 

geological units 

57


EARTHQUAKE 

ARCHAEOLOGY 



marbles and other metamorphic rocks show a 

connection to source rocks, which are nowadays 

located further away. Following after the Acrocorinth 

conglomerates, the Corinthian Marls formed in the 

Neogene and cover all other geological units (s. fig. 

1). 

In the beginning of the Pliocene the Graben of 

Corinth started to sink. As compensation for the 

deepening of the graben the Parnassus-Ghiona 

Mountains rose in the north, while in the south the 

Trapazona Mountains and isolated areas north, 

including Acrocorinth have been up-lifted pushing 

through the Neogene and Quaternary sediments. At 

Acrocorinth marine terraces are developed at 220 m 

a.s.l.. The age of the terraces is Pre-Paleotyrrhean 

(460,000 to 400,000 a) thus verifying an uplift of the 

hill of some 200 m in the last 400,000 years (SEDAT, 

1986). 

Today the components of the Mesozoic mélange 

form the top of the Acrocorinth hill as well as a 

smaller hill west of Acrocorinth, (Penteskufion). 

Acrocorinth conglomerates occur east and west of 

the hill, pushed up during the up-lift of Acrocorinth, 

with circular dip directions outwards on the flanks. 

West of the Penteskufion the conglomerate also 

occurs, dipping also into a westerly direction. In the 

eastern and western areas the conglomerates are 

covered by Corinthian Marls. In the south of 

Acrocorinth, conglomerates do not outcrop. Here the 

Corinthian Marls are directly in contact with Mesozoic 

units. On the northern flank conglomerates do also 

not outcrop. The outcropping units on the northern 

flank of the hill are interpreted to be breccias with 

angular components and not conglomerates. 

last 2000 years. Following PHILLIPSON (1892) and 

v. FREYBERG (1952) three bigger earth quick 

destroyed the area in historical times: 420 B.C., 77 

A.C. and 551 A.C. 

One of the biggest was the earthquake of 551 A.D. 

This earthquake affected mainly Acrocorinth and 

thereafter the acropolis on the Acrocorinth remained 

inhabited until the 8th century. The northern flanks of 

the hill collapsed and a mixture of natural rocks and 

cultural debris formed talus deposits, which cover the 

main parts of the northern flanks (s. Fig. 2). These 

slide breccias have been cemented in the following 

centuries and can therefore, be misinterpreted as 

conglomerates. However, the angular to sub-angular 

character of the clasts point to an origin as breccias 

in contrast to the well-rounded clasts of the 

Acrocorinth conglomerates. In summary, it was 

possible it identify four different generations of slide 

breccias: 

I. Very coarse sheet of debris: components sharpedged, 

up to 50 cm diameter, lightly cemented with 

powdery, carbonate matrix. This breccia shows a 

fining upwards, the components become smaller to 

the top of the unit. The unit is older than unit II. No 

other dating was possible. 

PALAEO-EARTHQUAKES 

In ancient times, beginning from the Mycenaean 

period until the Turkish invasion, the Acrocorinth was 

the acropolis and fortress of the ancient city of 

Corinth. Nearly every civilization left archaeological 

remains on the top and the flanks of the hill. 

Following PAUSANIAS (II. 4.6) in Roman times (A.D. 

160) ten smaller temples and sanctuaries 

(BOOKIDIS & STROUD (1987) were located on the 

northern flank of the hill: 

“:::[2.4.6] As you go up this Acrocorinthus you see 

two precincts of Isis, one if Isis surnamed Pelagian 

(Marine) and the other of Egyptian Isis, and two of 

Serapis, one of them being of Serapis called “in 

Canopus.” After these are altars to Helius, and a 

sanctuary of Necessity and Force, into which it is not 

customary to enter…..” PAUSANIAS (II. 4.6) 

The Demeter and Persephone temple, one of these 

sanctuaries, has been excavated in the 1970ies and 

80ies by the American School of Classical Studies. 

Locations of the remaining 9 temples or sanctuaries 

are unknown. But where are the ruins of the 

sanctuaries? 

The Graben of Corinth is a seismic high risk area. 

Several major earthquakes affected this area in the 

Fig.2: Sheets of debris at the Northern flank of the 

Acrocorinth: a: base of unit IV; b: base of unit I 

II. Fine-coarsed sheet of debris: components angular 

rounded with diameters of 5 to 10 cm. The 

components are firmly cemented by carbonate matrix 

and building pedocretes. The unit is older than 700 

B.C.. The small theatre in the temple of Demeter was 

cut into this unit and this temple was built around 700 

B.C.. 

III. Medium-coarsed sheet of debris: components 

angular of up to 20 cm diameter. The components 

are firmly cemented by a carbonate matrix. Also here 

pedocrete genesis is possible to observe. This 

breccia is young the unit II, because it overlies unit II. 

The breccia contains in addition to the natural 

components bricks and human processed stones. 

This unit could be the sheet of debris of the earth 

quick from 420 B.C., because it overlies the theatre 

of the temple of Demeter. 

58


EARTHQUAKE 

ARCHAEOLOGY 



IV. Fine-to medium coarsed sheet of debris: 

components sharp-edged, up to 10 cm diameter, 

lightly cemented with terra fusca and terra rossa. The 

base of the breccia shows bricks, destroyed pottery 

and traces of cinders (s. Fig. 3). The breccia contains 

a lot of terrestrial Stylommatophora shells. In the year 

146 B.C. the Romans destroyed Corinth fighting 

against the Archaean League. The city was 

completely razed and Corinth was nearly 100 year’s 

uninhabitated. Julius Ceasar refounded the city as 

Colonia Laus Iulia Corinthiensis in 44 B.C. shortly 

before his assassination. This horizon could be the 

level of the Roman ravage in the year 146. B.C.. In 

this case the unit IV is younger than 146 B.C.. 

Pausanias visited the area around 160 A.D., 83 years 

after the major earth quake from 77 A.D. The unit IV 

conforms to a typical rock fall landslide and based on 

this study it is the final bigger unit becoming exist as 

result of such an event. Hence unit IV could be the 

sheet of debris of the 551 A.D. seismic event. 

Following another earthquake on the 21st February 

1858 the city was given up and rebuilt 6 km 

northwest of the ancient settlement. 

CONCLUSIONS 

BOOKIDIS & STROUD (1987) describe the northern 

flank sediments as conglomerates. These 

conglomerates at Acrocorinth are of a Neogene age 

and excavations in these sediments will probably not 

lead to any successes looking for remains from the 

Roman period. 

Recent studies indicate that the the northern flanks of 

Acrocorinth are covered by breccias resulting from 

palaeoseismic events. Sedimentary characteristics 

do indicate an origin as cemented talus deposits, 

some of which triggered by paleo-earthquakes. 

These breccias are most probably younger than the 

Roman period because of cultural debris. A Neogene 

age can be excluded. 

Archaeological excavations in and below these 

recent sediments may locate the remaining of 

sanctuaries described by PAUSANIAS 160 A.D. 

References 

Fig. 3: Unit IV a: shells of Stylommatophora; b: level of the 

Roman ravage; c: conglomerates of the Corinthian Marls 

Corinth and parts of the fortifications on Acrocorinth 

were rebuilt during the rule of Justinian I, Caesar of 

Byzantium. In this time the Byzantine Empire was 

nearly 200 years converted to Christianity and there 

was no reason to rebuild ancient heathen temples. In 

order to locate the remaining ruins of the 9 missing 

temples it is necessary to identify the level between 

unit III and unit IV and start excavations at this level. 

Bookidis, N. & Stroud, R. (1987). Demeter and Persephone 

in ancient Corinth. 33 p., American School of Classical 

Studies, ISBN 0876616716, Princton, New Jersey. 

Freyberg, B. v. (1952). Der Bau des Isthmus von Korinth.- 

Ann. geol. Pays Hell., 4, p. 157-188, Athens. 

Gielisch, H. Dragastan, O. & Richter, D.K. (1993). Lagoonal 

to tidal carbonate sequences of Upper Jurassic/Lower 

Cretaceous age in the Corinthian area: Melange blocks of 

the Parnassos Zone.- Bull. Geol. Soc. Greece, Vol. XXVI- 

II/3, 663676, Athen. 

Gielisch, H. (1994). Mikrofazies und stratigraphische 

Stellung des Jura-Kreide-Übergangsbereichs der 

Parnass-Kiona-Zone zwischen Mittelgriechenland und 

Argolis.- Dissertation, 223 p., Bochumer geol. u. 

geotechn. Arb., Bd. 43, Bochum. 

Pausanias (II, 4.6) from http://www.theoi.com/ 

Text/Pausanias1A.html 

Philippson, A. (1892). Der Peloponnes – versuch einer 

KLandeskunde auf geologischer Grundlage. 642 p., 

Berlin. 

Sedat,R (1986). Das Stufenland von Korinth 

(Griechenland).- 183 p., Diss., Bochum. 

Richter, D.K., Dragastan, O. & Gielisch, H. (1992): 

Microfacies, diagensis and biostratigraphy of the 

Jurrassic/Cretaceous lagoonal Acrocorinth-Limestone 

(Parnassos Zone NE-Peloponnesos, Greece).- 

Bochumer geol. u. geotechn. Arb., 39, 1-70, Bochum. 

59


EARTHQUAKE 

ARCHAEOLOGY 



INTERPRETING OFFSHORE SUBMERGED TSUNAMI DEPOSITS: 

AN INCOMPLETELY COMPLETE RECORD 

Beverly N. Goodman-Tchernov (1) 

(1) Moses Strauss Department of Marine Geosciences, Leon Charney School of Marine Sciences, University of Haifa, Mt. 

Carmel, Israel 31905. bgoodman@univ.haifa.ac.il 

Abstract (Interpreting Offshore Submerged Tsunami Deposits: An Incompletely Complete Record): The complications and 

challenges of understanding and interpreting offshore submerged tsunami deposits is presented here and discussed in relation to 

recently and previously published tsunami studies. Regular annual storm activity impacts the shallow uppershelf and littoral zone, 

and rarer larger storms can impact deeper water depths. These events combined, with subwater surface currents, can regularly or 

periodically alter the seabottom, manipulating and reworking previously laid tsunamigenic horizons. Because of this, offshore 

deposits must be carefully considered within their meteorological and coastal morphological contexts. The differentiation between 

storm and tsunami deposits has long been a primary concern for understanding the sedimentological signature of paleotsunami 

deposits. Here, examples of paleotsunami deposits offshore on the Mediterranean coast of Israel will be presented with a 

discussion of how to address the issue of ‘completeness’ of the offshore record in order to properly place these evidence in the 

overall context of tsunami records. 

Key words: tsunami, sedimentology, uppershelf, Mediterranean 

INTRODUCTION 

Recent research offshore from Caesarea, Israel 

(Reinhardt et al. 2006, Goodman-Tchernov et al. 

2009, Fig. 1) argues for the presence of preserved 

tsunamites from multiple paleotsunami events. The 

specific horizons were identified based on the 

presence of tsunamigenic characteristics anchored in 

granulometry, shell taphonomy, sedimentological 

structures, erosional features, and terrestrial and 

marine mixing. While these horizons are well 

documented and laterally extensive, there are 

variations in the thickness and some events 

mentioned in historical documentation that are not 

present. The question arises whether the physical 

record is incomplete, or the historical records 

fabricated. The offshore record is highly problematic 

due to regularly occurring high-energy storm events 

which can disturb and rework remains from tsunami 

events (Dawson 2007, Shanmugam 2011). In 

addition, this reworking varies by a multitude of 

factors including depth, seafloor sedimentology and 

morphology, ultimately leaving a heterogeneous 

footprint. A deposit of one composition might survive 

intact due to width, material hardness, or mechanical 

strength, while another is easily erased. This 

absence and presence is seen when comparing 

across cores. In Caesarea, current research is aimed 

at tackling the complicated story in the near offshore 

environment through detailed analysis of sediment 

cores and the comparative study of modern storm 

impact and deposits. 

STORM IMPACTS 

In mid-December 2010, a large storm hit the Israeli 

coastline with measured wave heights of 13.7 

meters, just after a comparative storm/non-storm 

study was launched. A storm of this magnitude is 

only known to have a reoccurrence cycle of 2-3 times 

a century. The impact of the storm on the coastline 

was significant, dismantling and transporting large 

sections of concrete from a modern wave breaker, 

scouring the base of the ancient aqueduct, stripping 

the upper layer of sand from the seabottom in many 

parts of the harbor, and scraping 1-2 meters off of the 

face of coastal cliffs (see Fig. 2). A small underwater 

excavation survey and sediment study was 

completed in the summer and fall of 2010, preceding 

the storm. Following the storm, the same areas were 

Fig. 1: Caesarea, Israel. Aerial photograph of the 

coastline with the submerged ancient harbor. Tsunamis 

in antiquity struck the harbour, likely contributing to the 

harbour’s ultimate demise. 

60


EARTHQUAKE 

ARCHAEOLOGY 



revisited. This rare large storm had a great effect on 

the areas studied, and the deposits give a glimpse 

into the degree of impact that large storms may have 

had in antiquity, providing a reference point for 

differentiating between normal sea conditions, large 

storms, and tsunamis. 

Fig. 2: December 15, 2010, Caesarea, Israel. Two days 

after height of major decadal storm. Rubble eroded from 

buildings and concrete slabs moved from the outer 

breakwater cover a structure on the harbour pier. 

Normal, storm, and tsunami conditions were defined 

in sediment cores with the use of particle size 

distribution contour mapping (Goodman-Tchernov et 

al. 2009, see Fig. 3). Now, the recent large storm has 

provided a modern analog for comparison. 

DISCUSSION 

Offshore sedimentological research presents a 

unique preservation scenario, one in which there is 

no one-for-one record as one might find in certain 

varved lake sediments. The finds offshore of 

Caesarea are surprising, in the regard that any 

record remains in 20 meters water depth and 

shallower. Observations from the recent large storm 

clearly demonstrate that storms are capable of 

altering the record substantially, and it is therefore 

significant that finds demonstrate the presence of 

some preservation. What does remain, could be 

argued, is telling about the types of materials that are 

capable of withstanding later storm impacts. It could 

also be argued that the presence of events in the 

historical record that lack physical correlation in the 

offshore sediments should not be dismissed. Such 

absences might relate to the size or characteristic of 

the tsunami deposit—that it was not substantial 

enough to survive later storms. Future work might 

investigate cores from deeper waters, beyond any 

storm influence (~30m water depth) to see whether 

indications of those ‘missing’ events are present. 

Offshore tsunami deposits, particularly in the zone of 

storm-influence, do not represent a complete 

depositional sequence, rather, they are a 

representation of specific sets of circumstances 

dictated by deposit composition, thickness, and 

surrounding geomorphological conditions that 

nonetheless provide an important and independent 

window into paleotsunami impacts. 

Fig. 3: Particle size distribution contour mapping used to 

differentiating between storm and tsunami events (from 

Goodman-Tchernov et al. 2009). ‘S’ indicates storms, ‘E’ 

indicates tsunami events. The lack of storm events in core 2 

is interpreted as the product of water depth and distance 

from the shore. 

Acknowledgements: Appreciation expressed to Israel 

Antiquities Authority (K. Sharvit), Hatter Laboratory for 

Coastal Archaeology (M. Artzy), Israel Science Foundation, 

Israel Ministry of Infrastructure, and H. Dey. 

References 

Dawson, A. G., and I. Stewart (2007), Tsunami deposits in 

the geological record, Sediment.Geol., 200(3-4), 166- 

183. 

Goodman-Tchernov, B. N., H. W. Dey, E. G. Reinhardt, F. 

McCoy, and Y. Mart (2009), Tsunami waves generated 

by the Santorini eruption reached Eastern Mediterranean 

shores, Geology, 37(10), 943-946. 

Reinhardt, E. G., B. N. Goodman, J. I. Boyce, G. Lopez, P. 

van Hengstum, W. J. Rink, Y. Mart, and A. Raban 

(December, 2006), The tsunami of 13 December A.D. 

115 and the destruction of Herod the Great's harbor at 

Caesarea Maritima, Israel, Geology, 34(12), 1061-1064. 

Shanmugam, G. (2011), Process-sedimentological 

challenges in distinguishing paleo-tsunami deposits, 

Natural Disasters. 

61


EARTHQUAKE 

ARCHAEOLOGY 



EARTHQUAKE ENVIRONMENTAL EFFECTS, INTENSITY AND SEISMIC HAZARD 

ASSESSMENT: THE EEE CATALOGUE (INQUA PROJECT #0418) 

Guerrieri, Luca (1, Anna Maria Blumetti (1), Elisa Brustia (1), Eliana Esposito (2), Mauro Lucarini (1), Alessandro M. Michetti (3), 

Sabina Porfido (2), Leonello Serva (1) & Eutizio Vittori (1), & the INQUA TERPRO Project #0811 Working Group 

(1) Dipartimento Difesa del Suolo, Servizio Geologico d’Italia, ISPRA, Via V. Brancati 48, 00144 Roma, ITALY. Email: 

luca.guerrieri@isprambiente.it 

(2) Istituto per l’Ambiente Marino e Costiero, CNR, Calata Porta di Massa - 80133 Napoli, ITALY 

(3) Dipartimento di Scienze Chimiche e Ambientali, Università dell'Insubria, Via Valleggio, 11, 22100 Como, ITALY 

Abstract (Earthquake Environmental Effects, intensity and seismic hazard assessment: the EEE Catalogue (INQUA 

Project #0418): Earthquake Environmental Effects (EEE) are the effects produced by an earthquake on the natural environment, 

either directly linked to the earthquake source or triggered by the ground shaking. These include surface faulting, regional uplift 

and subsidence, tsunamis, liquefaction, ground resonance, landslides, and ground failure phenomena. 

The EEE catalogue is a data collection of Earthquake Environmental Effects from modern, historical and paleoseismic 

earthquakes compiled at global level by the INQUA TERPRO Project #0811 WG. 

The damages caused by recent catastrophic seismic events have been mostly linked to the vulnerability of physical environment 

enhancing the crucial role of EEEs, including tsunamis, for seismic hazard purposes. Therefore, these events have confirmed that 

the EEE Catalogue is an essential tool to complete traditional SHA based on PGA maps, since it allows to identify the natural 

areas most vulnerable to earthquake occurrence and to objectively compare in time and in space the earthquake intensity through 

the ESI scale. 

Key words: Earthquake Environmental Effects, ESI intensity scale, seismic hazard. 

INTRODUCTION 

Earthquake Environmental Effects (EEE) are the 

effects caused by an earthquake on natural 

environment, including surface faulting, regional uplift 

and subsidence, tsunamis, liquefactions, ground 

resonance, landslides and ground failure, either 

directly linked to the earthquake source or provoked 

by the ground shaking (Michetti et al., 2007). 

Most of the damage resulting from moderate to large 

earthquakes is typically related with the vulnerability 

of the physical environment. This point has been 

sadly and dramatically confirmed by the two relevant 

seismic events occurred in the last months in 

countries with strong economy and modern building 

codes, i.e., the February 22, 2011, Mw 6.3 

Christchurch and the March 11, 2011, Mw 9.0 East 

Japan earthquakes. 

The March 11, 2011, earthquake occurred in the 

Pacific Ocean near the coast of NE Japan. Most of 

the damage in terms of dead toll (more than 20,000 

people) and destruction was caused by the huge 

tsunami (run up values larger than 38 m; more than 5 

km inland penetration in the Sendai coastal plain). 

Comparatively, the amount of damage induced by 

the vibratory ground motion itself was modest. The 

size of the 2011 tsunami was fairly larger than those 

recorded in the affected area in the last century, but 

comparable with the tsunamis affecting the same 

areas in 869 A.D. recently well documented by 

means of geological and paleoseismic studies (Sawai 

et al., 2007; HERP, 2009). 

Nevertheless, the relevance of EEE has been also 

shown by the moderate-size event occurred on 

February 22, 2011, very close to the town of 

Christchurch, New Zealand. 

Fig. 1: – Geological record of past tsunamis at Watari and 

Yamamoto, Myagi prefecture (Sawai et al., 2007). The 

characteristics and spatial distribution of these deposits 

allowed to identify ancient tsunamis comparable with the 

March 11, 2011 event. 

62


EARTHQUAKE 

ARCHAEOLOGY 



This was basically an aftershock of the 2010 

September 4, Mw 7.0 event. However, while the main 

shock did not cause victims or large damage, the 

2011 event caused the death of at least 75 people 

and the destruction of several districts in the 

Christchurch area. This scenario of bad damage was 

mainly linked to local effects of site amplification, 

which depends largely on the stratigraphic 

characteristics (geological history) of the recent 

deposits over which the town is built; as well, the 

same sediments are particularly susceptible to 

liquefaction. As a consequence, also houses 

designed in agreement with the local seismic codes 

have collapsed, due to liquefaction within the 

foundation soils. 

Both events, although not comparable in size and 

geodynamic setting, have confirmed once again i) the 

relevance of earthquake environmental effects as a 

major source of seismic hazards, in addition to 

vibratory ground motion; ii) the need of re-evaluating 

the significance of macroseismic intensity as 

unequivocal measurement of the final earthquake 

impact on the local natural and built environment, 

and iii) the relationships between epicentral intensity, 

earthquake size, and source geometry. As a matter 

of fact, intensity is a parameter able to describe a 

complete earthquake scenario, based on direct field 

observation. Thus, the knowledge of the 

characteristics and spatial distribution of EEE 

induced by past earthquakes will strongly improve 

standard seismic hazard assessments, that typically 

consider only the vibratory ground motion hazard. 

THE EEE CATALOGUE 

Nowadays, a significant amount of data about 

Earthquake Environmental Effects is available for a 

very large number of recent, historical and paleoearthquakes. 

However available information is 

located in several different sources (scientific papers, 

historical documents, professional reports), and often 

difficult to access. 

The EEE Catalogue has been promoted with the aim 

to properly retrieve the available information about 

EEE at global level and archive it into a unique 

database, in order to facilitate their use for seismic 

hazard purposes. Its implementation has been 

endorsed at global level by the INQUA TERPRO 

Project #0811, through a Working Group coordinated 

by ISPRA - Geological Survey of Italy. 

The EEE catalogue collects the characteristics, size 

and spatial distribution of coseismic effects on nature 

in a standard way from modern, historical and 

paleoearthquakes. For each event, we have 

assessed epicentral and local intensities based on 

EEE data through the ESI 2007 scale (Michetti et al., 

2007), that integrates and completes the traditional 

macroseismic intensity scales, allowing to assess the 

intensity parameter also where buildings are absent 

or damage-based diagnostics saturates. This 

procedure has allowed an objective comparison in 

terms of earthquake intensity, for events occurred in 

different areas and/or in different periods. 

The information is collected at three levels of 

increasing detail (Earthquake, Locality, Site). Also 

available imagery documentation (photographs, 

video, sketch maps, stratigraphic logs) can be 

uploaded into the database.The quality of the 

database in terms of completeness, reliability, and 

resolution of locations is strongly influenced by the 

age of the earthquake so that it is expected to be 

very variable. 

Nevertheless, even where the information is less 

accurate (historical earthquakes), the documented 

effects are typically the most relevant i.e. most 

diagnostic for intensity assessment. Similarly, the 

information from paleoseismic investigations, 

although poorly representative of the entire scenario, 

still includes significant data (i.e. local coseismic fault 

displacements) very helpful of a minimum size of the 

earthquake. 

A first official release of the EEE Catalogue has been 

done in the frame of the XVIII INQUA Congress, held 

in Bern in July 2011. However, the implementation of 

the EEE catalogue is always in progress at 

http://www.eeecatalog.sinanet.apat.it/login. Data can 

be explored on a public interface (Fig. 2) based on 

Google Earth at http://www.eeecatalog.sinanet. 

apat.it/terremoti/index.php. Earthquake records 

validated by the Scientific Committee of the Project 

can be also downloaded from the site. 

THE ADDED VALUE 

The major added value of the EEE Catalogue in 

terms of seismic risk is the possibility to explore the 

scenarios of environmental effects induced by past 

earthquakes and therefore identify the areas where 

the anthropic settlements and infrastructures are 

more exposed to this source of potential hazard. To 

this end, a good accuracy of EEEs location becomes 

crucial. Typically, EEEs from recent earthquakes are 

mapped with good accuracy immediately after the 

event. Nevertheless, even for some historical 

earthquakes it is possible to retrieve with very good 

detail this information. It is the case of the December 

28, 1908 Messina Straits earthquake and consequent 

tsunami (Fig. 3), where the EEE Catalogue allows to 

locate the earthquake/tsunami effects over the 

present urban texture with a spatial resolution of a 

few meters, pointing out the areas characterized by 

the highest risk. Furthermore, the EEE Catalogue 

allows to reveal possible trends in the spatial 

distribution of primary and secondary effects. For 

example, Fig. 4 shows the spatial distribution of 

EEEs induced by the October 8 2005, Muzaffarabad, 

Pakistan, earthquake (Ali et al., 2009): it is quite 

evident that the location and amount of surface 

faulting is consistent with the spatial distribution of 

coseismic slope movements, in terms of both areal 

density and size. 

A similar result is shown by the spatial distribution of 

EEEs induced by the 1811-1812 New Madrid, 

Missouri, earthquakes, mapped in Fig. 5. Indeed, the 

most relevant primary and secondary effects are 

located along the Mississippi valley near New 

63


EARTHQUAKE 

ARCHAEOLOGY 



Madrid, consistently with the surface projection of the 

causative faults, and unquestionably provide 

diagnostic elements for assessing an epicentral 

intensity equal to XI. 

Fig. 2: The public interface of the EEE Catalogue, developed on Google Earth 

http://www.eeecatalog.sinanet.apat.it/terremoti/index.php. 

Fig. 3: EEEs induced by the December 28, 1908 Messina Straits, Italy, earthquake in the area of the Messina harbour. If 

information from contemporary sources is very precise, it is possible to use the EEE Catalogue also for local seismic 

microzonation. 

FINAL REMARKS 

The recent catastrophic earthquakes occurred in 

Japan and New Zealand have clearly pointed out that 

traditional seismic hazard assessment based only on 

vibratory ground motion data need to be integrated 

with information about the local vulnerability of the 

territory to earthquake occurrence. 

The collection of Earthquake Environmental Effects 

provided by the EEE Catalogue aims at identifying 

the areas most vulnerable to earthquake occurrence. 

This information must complement traditional SHA 

based on PGA maps. 

Moreover, based on EEE characteristics, size and 

spatial distribution it is possible i) to assess the 

earthquake intensity through the ESI scale, and ii) to 

objectively compare the earthquake intensity of 

events occurred in different areas and/or in different 

periods. 

64


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 4 :Surface faulting and slope movements induced by the October 8, 2005 Muzaffarabad earthquake. 

Fig. 5 : EEE induced by the December 16 1811 New Madrid, Missouri, earthquake. Primary and secondary effects indicative of 

intensity XI in the ESI 2007 scale are located in the epicentral area along the Mississippi Valley. 

Acknowledgements: 

Fabio Baiocco and Antonio Scaramella are acknowledged 

for their crucial support in building the infrastructure of the 

EEE Catalogue. We wish also to acknowledge all the 

people who has participated to the implementation of the 

EEE Catalogue in the frame of the #0418 INQUA TERPRO 

Project. 

References 

Ali Z., Qaisar M., Mahmood T., Ali Shah M., Iqbal T.,, Serva 

L., Michetti A.M.. and Burton P.W. (2009) - The 

Muzaffarabad, Pakistan, earthquake of 8 October 2005: 

surface faulting, environmental effects and macroseismic 

intensity Geological Society, London, Special 

Publications 2009, 316:155-172; doi:10.1144/SP316.9. 

HERP (2009). Long-term assessment on seismic activity 

along Japan trench offshore Sanriku and offshore Boso 

Peninsular. 

http://www.jishin.go.jp/main/chousa/09mar_sanriku/index. 

htm [in Japanese with many figures]. 

Michetti, A. M., E. Esposito, L. Guerrieri, S. Porfido, L. 

Serva, R. Tatevossian, E. Vittori, F. Audemard, T. 

Azuma, J. Clague, V. Comerci, A. Gurpinar, J. McCalpin, 

B. Mohammadioun, N. A. Morner, Y. Ota, and E. 

Roghozin (2007). Intensity Scale ESI 2007. In Memorie 

Descrittive Carta Geologica d’Italia L. Guerrieri and E. 

Vittori (Editors), APAT, Servizio Geologico d’Italia— 

Dipartimento Difesa del Suolo, Roma, Italy, 74, 53 pp. 

Sawai Y., Shishikura M., Okamura Y., Takada K., Matsu’ura 

T., Tin Aung T., Komatsubara J., Fuji Y., Fujiwara O., 

Satake K., Kamataki T., and Sato N. (2007). A study on 

paleotsunami using handy geoslicer in Sendai Plain 

(Sendai, Natori, Iwanuma, Watari and Yamamoto), 

Miyagi, Japan. N. 7, p. 47-80, 2007 

http://unit.aist.go.jp/actfaulteq/seika/h18seika/pdf/sawai.pdf 

[In Japanese with 

English abstract and captions]. 

65


EARTHQUAKE 

ARCHAEOLOGY 



ACTIVE TECTONICS OF ALBANIA INFERRED FROM FLUVIAL TERRACES GEOMETRIES 

Guzman, Oswaldo (1,2), Jean-Louis Mugnier (1), Rexhep Koçi (3), Riccardo Vassallo (1), Julien Carcaillet (4), Francois Jouanne (1), 

Eric Fouache (5) 

(1) Institut de sciences de la Terre. Université de Savoie, Bâtiment des Belledonnes, Campus Scientifique, 73376 Le Bourget du 

Lac Cedex, France. E-mail : Oswaldo-Jose.Guzman-Guitierrez@etu.univ-savoie.fr, Jean-Louis.Mugnier@univ-savoie.fr, 

Riccardo.Vassallo@univ-savoie.fr, Francois.Jouanne@univ-savoie.fr 

(2) Dpto. Ciencias de La Tierra, Universidad Simon Bolivar, Edificio FE-II. Apartado 89000, Caracas 1081-A, Venezuela 

(3) Institut de Sciences de la Terre. Université Joseph Fourier. Maison des Géosciences, 1381 rue de la Piscine, 38400 Saint 

Matin d’Hères, France. E-mail : Julien.Carcaillet@ujf-Grenoble.fr 

(4) Institute of Seismology of the Academy of Science, Tirana, Albania. E-mail: rexhep.koci@yahoo.com 

(5) University of Paris 10, Nanterre, UMR 8591, EA435 Gecko, France. E-mail: eric.g.fouache@wanadoo.fr 

Abstract (Active Tectonics of Albania Inferred from Fluvial Terraces Geometries): We studied four rivers of the southern 

Albanides in order to quantify the geodynamic control on the incision of alluvial terraces. The Vjosa, Osum, Devoll and Shkumbin 

rivers flow from north-western Greece and southern Albania across the active Dinaro-Hellenic Alpine fold belt to the west. From 

mapping and dating of terrace abandonment, we reconstructed spatial and temporal variations of incision rates along those rivers. 

These reconstructions allow a quantification of the fluvial evolution at a higher resolution than a glacial-inter-glacial cycle, and 

therefore an estimation of the effect of geodynamic uplift on river incision. We identified ten preserved terraces units developed 

since the “MIS-6” up to historic time. Uplift seems to produce two distinct effects: an overall increase of the incision rate from west 

to east related to regional bulging and local pulses of increasing incision generated by activation of faults. 

Key words: Albania, Active Tectonics, Alluvial Terraces. 

INTRODUCTION 

Active tectonics, eustatic and climatic variation are 

the forcing that controls the formations of fluvial 

terraces. In a tectonically active setting characterised 

by long term uplift, Pazzaglia (submitted) suggests 

that the incision rate equals to the long-term uplift 

rate. Nonetheless, the morphology of the rivers in this 

context is strongly climate-dependent: a change in 

downstream base level could have a profound impact 

on long profile shape (Merrits et al., 1994), and 

hydrologically-driven changes in discharge and 

sediment load induce periodically lateral incision 

processes that widen the channel bottom producing 

strath fluvial terraces (Hancock & Anderson 2002) or 

aggradation leading to the development of filled 

terraces (Ouchi, 1985) 

The Albanide domain is rather small (less than 5. 10 4 

km 2 ). We consider that the climatic evolutions of the 

different Albanide catchments are rather similar. 

Therefore differences in the evolution of the studied 

catchments would not be explained by local climate 

changes. 

The active tectonics pattern in Albania is quite 

variable (Fig. 1), with a NE-SW shortening in the 

western part of the Albania, and a NW-SE extension 

in the inner part of the chain, it leads to spatially 

variable long term uplift (Aliaj, 2000). Therefore the 

comparison of the different catchments evolution 

would help to estimate the uplift rate in Albania and 

to outline the role of the variation of the tectonics rate 

in the sedimentation/erosion behaviour of the 

catchments. 


The Dinaro-Hellenic Alpine fold belt constitutes a 

segment of the wide Circum-Mediterranean Peri- 

Tethyan thrust belt. This belt as a result of the Dinaric 

subduction. The Albanian foothills have been thrust 

westwards over the Adriatic foredeep during the 

Alpine orogeny (Roure et al., 2004). 

The presenty-day geodynamic deformation results 

from the subduction of the Adriatic Sea floor beneath 

the foreland domain. This geodynamic setting 

produces contrasting relief with a flexural basin filled 

with Plio-Quaternary deposits and forming a flat 

costal plain in the foreland, a thin skinned fold and 

thrust belt in the midland, and a large basin and 

range zone in the hinterland, resulting from 

synorogenic Neogene-Quaternary grabens that 

crosscut the thrust system (Roure et al., 2004, 

Niewland, et al., 2001). 

The area (Fig.1) is still tectonically active and 

produces permanent microseismicity and frequent 

earthquakes reaching intensities of IX. GPS data 

suggest that the western Albania is being affected by 

southwestward motions relative to Adriatic 

microplate, illustrating the ongoing collision of 

external Albanides, whereas inner Albanides present 

southward motion, increasing from north to south, 

relative to both Adriatic and stable Eurasia (Jouanne 

et al., submitted). 

The neotectonics are controlled by NE-SW 

compressive stress in the western part connected to 

subduction of the Apulian lithosphere and extensional 

stress in the internal zones (Goldsworthy et al., 2002; 

Roure et al., 2004). The major extension direction 

varies from E-W close to the Ohrid and Prespa lakes 

66


EARTHQUAKE 

ARCHAEOLOGY 



to NNW-SSE at the southern end of the Ersekë 

basins (Jouanne et al., submitted). On the regional 

scale the Albanides are governed by long-term uplift 

(Aliaj, 2000). 

We concentrate our study on a zone located between 

two major transverse structures which cut the area: 

the Vlore-Fier-Elbasani-Dibra transfer zone in the 

41.5° 

Durres 

Erzen 

Tirane 

central part of Albania (Roure et al., 2004), and the 

prolongation of the Keffalinia strike-slip zone at the 

northwestern border of Greece (Baker et al., 1997). 

The Vjosa, Osum, Devoll and Skhumbin rivers 

constitute four of the six main rivers of Albania. They 

are close to each other; therefore, their climatic 

settings are comparable. 

N 

Librazdhi 

41° 

Karavatsas 

Lagoon 

C.p. 

S hkumbin 

Ar.A. 

Seman 

S.A. 

? 

? 

Lushnje 

Elbasan 

Berat 

E.G. 

Pa.D. 

T.T. 

Gramsh 

Progradec 

Devo l 

Ohrid 

Lake 

N.N.F. 

W.G.N.F.S. 

Prespa 

Lake 

Korçë 

L.T.T. 

40.5° 

Nartës 

Lagoon 

Ap.A. 

Vlorë 

O 

s u m 

W.E.N.F.S. 

Vithkuq 

Tepelenë 

V j 

o 

s a 

Ersekë 

Permët 

40° 

0 50 km 

42° 

41° 

40° 

19° 20° 21° 

Tirana 

Ionian 

Sea 

Reverse Fault 

Normal Fault 

39.5° 

19° 20° 

Sarandë 

Strike-slip Fault 

Anticline Fold 

K.N.F.S.N.N.F. 

Aristi 

Ioannina 

Borders 

Konitsa 

P.W.N.F. 

K.N.F. 

Kipi 

Scale 

0 20km 

Figure 1. Neo-tectonic map of the Southern Albania and North-western Greece (Modified from Aliaj et al., 2000 and 

Carcaillet, et al, 2009). Areas located within 200 m above Sea level are in light grey. Dark grey stars indicate published data 

(Lewin et al., 1991; Hamlin et al., 2000; Woodward et al., 2001; Carcaillet et al., 2009) and red stars indicate data computed 

in the present study. Fault and fold types are described in the picture caption and the overall current tectonic deformation is 

represented by black arrows (from Jouanne et al., submitted). (C.p.) Coastal pop-up; (Ar.A.) Ardenica Anticline; (Ap.A.) 

Apollonia Anticline; (S.A.) Shkumbin Anticline; (T.T.) Tomorrica Thrust; (L.T.T.) Lushnje - Tepelenë Thrust; (B.N.F.) Bulcar 

Normal Fault; (E.G.) Elbasan Graben; (N.N.F.) Nerotrivi Normal Fault; (K.N.F.S.) Konitsa Normal Fault Systems; (P.W.N.F.) 

Papingo West Normal Fault; (K.N.F.) Kipi Normal Fault; (W.G.N.F.S.) West Graben Normal Fault Systems; (W.E.N.F.S.) 

West Ersekë Normal Fault Systems; (PaD) PaleoDevoll. 

21° 

METHODS AND RESULTS 

The terraces has been mapped from field work, 

satellite image and 1/25000 topo sheets. Terrace 

thicknesses and elevations above the present-day 

river bed have been measured using a laser range 

distancemeter. Terrace age control was obtained 

from data previously published by Lewin et al. (1991), 

Hamlin et al. (2000), Woodward et al. (2001) and 

Carcaillet et al. (2009) as well as from in situ 

67


EARTHQUAKE 

ARCHAEOLOGY 



Incision rate (mm/yr) 

produced 

10 Be and radiocarbon ( 14 C) dating 

specifically performed for this study. 

We dated organic residues (thirteen 14 C ages) and 

quartz rich pebbles collected into layers developed 

during the last flooding event and considered the 

value as the age of abandonment of the alluvial 

terrace. This implies that the age of the terrace 

abandonment is contemporaneous to the age of 

deposition of the highest terrace layer. The samples 

for cosmonuclide (in situ produced 

10 Be) age 

determinations have been collected on siliceous rich 

pebbles along depth-profiles of two terraces of the 

lower Skhumbin and one on the top of the upper 

Osum. 

West 

5 

4.5 

4 

3.5 

3 

2.5 

2 

1.5 

1 

0.5 

0 

Post LGM 

Pre LGM 

Linear regression Post LGM 

by tectonic blocks 

120 145 170 

Distance from the Sea (km) 

Figure 3. Evolution of the incision rate 

along the river profile of the Vjosa. bed. In 

the river profile are located the Lushnje - 

Tepelenë Thrust (L.T.T.), Nerotrivi Normal 

fault (N.N.F.), Konitsa Normal Fault 

Systems (K.N.F.S.) Papingo West Normal 

Fault (P.W.N.F.). 

West 

T.T. 

28.6 m 

1.7 m/ka 

Incision 

Incision 

rate 

rate 

(mm/yr) 

(m/ka) 

0.8 

0.6 

0.4 

0.2 

0 

W.E.N.F.S. 

35 m 

2.1 m/ka 

195 220 

West 

1.2 

Post LGM 

1 Pre LGM 

Incision rate has been calculated by dividing the 

elevation of the terraces above the river by the age of 

the upper surface of the terraces; the vertical motions 

of the faults have been deduced from the local 

change of the incision rates. 

Our study has allowed estimating the vertical slip rate 

of actives structures in southern Albania ( Fig, 2, 3 

and 4, Table 1). These results confirm an important 

uplift rate, up to 2 mm/year (Carcaillet, et al. 2009), 

for the central Albania and the current activity of the 

Plio-Quaternary folds of external Albanides. They 

also show significant Quaternary displacement rates 

along normal faults of inner Albanides. Our results 

are in agreements with the E-W to N-S extension 

deduced from the GPS network. 

East 

1000 

500 

0 

Elevation above 

Sea level (m) 

3m 

0.2 m/ka 

Figure 2. Evolution of the incision rate along the 

river profile of the Osum. Left axis is the 

calculated incision rate; right axis is the elevation 

of the present-day river bed. Colour of the 

diamonds refers to the different terraces units. 

Into the lower box, appear the actual river profile 

where are located the actives faults. Bold lines 

represent normal faults belong of West Ersekë 

Normal Fault Systems (W.E.N.F.S.) producing 

surface displacements and the dashed lines 

represent the Tomorrica Thrust (T.T.) which 

would not influence in the evolution of the river 

profile. 

Konitsa Basin 

K.N.F.S. 

East 

L.T.T. 

N.N.F. P.W.N.F. 

0 

60 85 110 135 160 185 

>0.4m/ka 

1000 

500 


Sea level (m) 

East 

2.0 

Pre LGM 

Incision rate (mm/yr) 

Incision rate (m/ka) 

1.6 

1.2 

0.8 

0.4 

0 

Linear regression Pre LGM 

by tectonic blocks 

23 m 

0.9 m/ka 

5m 

0.2 m/ka 

Elbasan Basin 

>1 m/ka 

11 m 

0.4 m/ka 

9m 

0.3 m/ka 

? 

? 

> 0.8 m/ka 

Korça Basin 

L.T.T. 

Lower Shkumbin PaleoDevoll Devoll 

S.A. 

E.G. 

T.T. B.N.F. W.G.N.F.S. 

25 50 75 100 125 150 175 

Figure 4. Evolution of the incision rate along the combined river profile of the lower Shkumbin and Devoll. In the river profile are 

located the Lushnje - Tepelenë Thrust (L.T.T.), the Shkumbin Anticline (S.A.), Tomorrica Thrust (T.T.), the Burcal Normal Fault 

(B.N.F.) and the West Graben Normal Fault Systems (W.G.N.F.S.) producing surface displacements. 

800 

400 

0 


Sea level (m) 

68


EARTHQUAKE 

ARCHAEOLOGY 



Table 1. Kinematic of active structures active analyzed from the incision of the terraces levels of the Vjosa, Osum, Devoll and 

Shkumbin rivers. 

Name Structure River Vertical slip rate (mm/yr) 

Coastal pop - up PaleoSeman ~ 0.1 

Shkumbin Anticline Shkumbin ~ 0.2 

Elbasan Graben Shkumbin > 1 

Lushnje - Tepelenë Thrust Shkumbin ~ 0.9 

Lushnje - Tepelenë Thrust Vjosa ~ 0.2 

Tomorrica Thrust Devoll ~ 0.4 

Burcal Normal Fault Devoll ~ 0.3 

Nerotrivi Normal Fault Vjosa > 0.4 

Konitsa Normal Fault Systems Vjosa > 0.4 

Papingo West Normal Fault Vjosa ~1.8 between 25 and 17 Ka (Carcaillet, et al, 2009) 

West Graben Normal Fault Systems Devoll > 0.8 

West Ersekë Normal Fault Systems Osum 1.7 – 2.1 

SEISMICITY, PRESENT-DAY DEFORMATION AND 

ACTIVE TECTONICS 

Current tectonics of Albania is characterized by an 

important microseismicity, small and medium size 

earthquake and a few large events as shown by the 

occurrence of earthquakes with magnitudes 

exceeding 6 during the last century (Jouanne et al., 

submitted), (e.g. 1905, Shkodra earthquakes Ms=6.6, 

1911 Ohrid lake earthquake Ms = 6.7, 1920 Tepelena 

event Ms=6.4, 1926 Durres earthquake Ms=6., 1967 

Dibra earthquake Ms=6.6and 1979 Montenegro 

earthquake Ms=6.9). 

The strongest earthquakes occur in three well-defined 

seismic belts: a) The Ionian-Adriatic coastal 

earthquake belt at the eastern margin of the Adria 

microplate, which trends NW-SE represents the most 

seismically active zone in the country; b) The 

Peshkopia-Korça earthquake belt, which trends N-S, 

and c) The Elbasani-Dibra earthquake belt, which 

trends N-E, (Sulstarova et al. 2000; Aliaj, 2000). 

Focal mechanisms (Sulstarova et al., 2000, Aliaj, 

2000) and geodetic data (Jouanne et al., submitted), 

fit the geodynamics setting with a NE-SW 

compression across external Albanides and E-W to 

NNW-SSE extension in the internal Albanides. 

The comparison of our results and the deformation 

deduced from the GPS network (jouanne et al., 

submitted) indicates that: 

a) The horizontal shortening linked to the active thrust 

faults exceeds 2 mm/yr (assuming a mean dip of 30° 

for the thrusts) and is close to the deformation rates 

deduced from GPS ; b) the extension linked to normal 

fault system would be in the order of 1 mm/yr 

(assuming a dip of 60° for the normal faults) and is 

smaller than the deformation rates deduced from 

GPS; c) The normal faulting linked to the East 

Albanian graben systems seems more active than the 

normal faulting that affects the North West part of 

Greece. 

Acknowledgements: The authors thank the NATO SFP 

977993 and the Science for Peace team to have supported 

this work. 

References 

Aliaj, S. (2000). Neotectonics and seismicity in Albania, in: 

Geology of Albania, edited by: Meco, S., Aliaj, S., and 

Turku, I., Beitr. Regional. Geol. Erde, 28, 135–178. 

Baker, C., Hatzfeld, D., Lyon-Caen, H., Papadimitriou, E., 

Rigo, A. (1997). Earthquake mechanisms of the Adriatic 

Sea and Western Greece: implications for the oceanic 

subduction-continental collision transition. Geophysical 

Journal International 131, 559-594. 

Carcaillet, J., Mugnier, J.L., Koçi, R., Jouanne, F. (2009). 

Uplift and active tectonics of southern Albania inferred 

from incision of alluvial terraces. Quaternary Research 

71, 465-476. 

Goldsworthy, M., Jackson, J., Haines, J. (2002). The 

continuity of active fault systems in Greece. Geophysical 

Journal International 148(3), 596-618. 

Hancock, G., Anderson, R. (2002). Numerical modeling of 

fluvial strath-terrace formation in response to oscillating 

climate. Geological Society of America Bulletin 114(9): 

1131-1142 

Hamlin, R.,Woodward, J., Black, S., Macklin, M.G. (2000). 

Sediment fingerprinting as 580 a tool for interpreting longterm 

river activity: the Voidomatis basin, NWGreece. In: 

581 Foster, I.D.L. (Ed.), Tracers in Geomorphology. 

Wiley, Chichester, 473–501. 

Jouanne, F., Bushati, S., Mugnier, J.L., Shinko, I., Pasha, 

M., Koci, R. (Submitted). GPS constrains on current 

tectonics of Albania. Geophysical Research Letters. 

Lewin, J., Macklin, M.G., Woodward, J.C. (1991). Late 

Quaternary fluvial sedimentation in the Voidomatis basin, 

Epirus, Northwest Greece. Quaternary Research 35, 103- 

115. 

Merritts, D.J., Vincent, K.R., Wohl, E.E. (1994). Long river 

profiles, tectonism, and eustasy: a guide to interpreting 

fluvial terraces. Journal of Geophysical research, 99, 

14031-14050. 

Niewland, D.A., Oudemayer, B.C., Valbona, U. (2001). The 

tectonic development of 616 Albania: explanation and 

prediction of structural styles. Marine and Petroleum 617 

Geology 18, 161–177. 

Ouchi, S. (1985) Response of alluvial rivers to slow active 

tectonics movements, Geol. Soc. Am. Bull., 96, 504-515. 

Pazzaglia, F. J. River Terraces, in Wohl, E., ed., Treatise of 

Geomorphology, Elsevier. Submitted 

Sulstarova, E., Peçi, V., Shuteriqi, P. (2000). Vlora- 

Elbasani-Dibra (Albania) transversal fault zone and its 

seismic activity. Journal of Seimology 4, 117-131. 

Roure, F., Nazaj, S., Mushka, K., Fili, I., Cadet, J.P., 

Bonneau, M. (2004). Kinematic evolution and petroleum 

systems - An appraisal of the Outer Albanides. In: Mc 

Clay K.R. (Ed.), Thrust tectonics and hydrocarbon 

systems. AAPG memoir 82, 474-493. 

69


EARTHQUAKE 

ARCHAEOLOGY 



LECHAION, THE ANCIENT HARBOUR OF CORINTH (PELOPONNESE, GREECE) 

DESTROYED BY TSUNAMIGENIC IMPACT 

Hadler, Hanna (1, Andreas Vött (1), Benjamin Koster (2), Margret Mathes-Schmidt (2), Torsten Mattern (3), Konstantin 

Ntageretzis (1), Klaus Reicherter (2), Dimitris Sakellariou (4), Timo Willershäuser (1) 

(1) Institute for Geography, Johannes Gutenberg-Universität Mainz, Johann-Joachim-Becher-Weg 21, 55099 Mainz, GERMANY. 

E-mail: hadler@uni-mainz.de 

(2) Neotectonics and Natural Hazards Group, RWTH Aachen, Lochnerstr. 4-20, 52056 Aachen, GERMANY. 

(3) Faculty of Classical Archaeology, University of Trier, Universitätsring 15, 54296 Trier, GERMANY. 

(4) Institute of Oceanography, Hellenic Centre for Marine Research, 47 km Athens-Sounio Ave., 19013 Anavyssos, GREECE. 

Abstract: Lechaion, the harbour of ancient Corinth, is situated at the south-eastern extension of the Gulf of Corinth (Peloponnese, 

Greece). Due to extensive fault systems dominating the gulf, seismic activity is frequent and often related to landslides or 

submarine mass movements. Thus, the study area is highly exposed to tsunami hazard. By means of geo-scientific studies 

comprising geomorphological, sedimentological and geophysical methods, evidence of multiple palaeotsunami impact was 

encountered at the Lechaion harbour site and the surrounding coastal area. The detected tsunami signatures include 

allochthonous marine sediments intersecting quiescent harbour deposits, extensive units of tsunamigenic beachrock and geoarchaeological 

destruction layers. Our results suggest that the harbour at Lechaion was finally destroyed in the 6 th century AD by 

strong tsunami impact. 

Key words: Palaeotsunami, beachrock, ancient harbours, Lechaion 

INTRODUCTION 

Situated in the eastern Mediterranean, the Gulf of 

Corinth belongs to one of the seismically most active 

regions in the world. Due to ongoing continental 

rifting with high extension rates up to 16 mm/yr, this 

half-graben structure is surrounded by active onshore 

and submarine faults (Papazachos & Dimitriu 

1991; Sachpazi et al., 2003; Avallone et al., 2004). 

Strong earthquakes, often accompanied by 

coseismic displacement of the seafloor and/or 

inducing submarine slides are therefore common 

throughout the gulf. Water depths of maximum 900 

m, steep submarine slopes and a narrow shelf 

additionally enhance the potential of strong tsunamis 

(Hasiotis et al., 2002; Stefatos et al., 2006). 

Accordingly, numerous tsunami catalogues for the 

Mediterranean and especially for Greece report on 

historical tsunami events that occurred within the Gulf 

of Corinth (Soloviev et al., 2000; Papadopoulos, 

2003; Ambrasey & Synolakis, 2010). Solely for the 

20 th century, six tsunamis have been recorded; the 

most destructive occurred near Aeghio in 1963 

(Papadopoulos, 2003). 

The main objectives of our study were to (i) detect 

allochthonous high-energy event layers within the 

sedimentary record of the harbour basin at Lechaion 

and (ii) to reconstruct the late-Holocene coastal 

evolution of the study area. 

LECHAION – THE 

CORINTH 

ANCIENT HARBOUR OF 

Lechaion, the western harbour of ancient Corinth, is 

situated at the Isthmus of Corinth, which connects 

the Peloponnese with mainland Greece. The harbour 

was most probably founded around 600 BC when the 

Corinthians expanded their military and trading 

activities (Stiros et al., 1996). According to ancient 

sources, Lechaion served as a naval base from the 

4 th century BC and was connected to the city of 

Corinth by a harbour road protected by strong walls 

(Xen. Hell. 4.4.6-8 after Brownson, 1918). With the 

devastation of Corinth in 146 BC, Lechaion was 

abandoned but re-activated under Roman supremacy 

in 44 BC. In Roman times, the harbour underwent 

different phases of reconstruction and mainly served 

as trading base (Strab. Geogr. 8.6.20-23 after 

Hamilton & Falconer, 1903). The final abandonment 

of Roman Lechaion is reported to be associated with 

the destruction of ancient Corinth by a series of 

strong earthquakes in 521 or 551 AD. Though 

occasionally re-used in medieval times, the harbour 

never regained its former importance (Rothaus, 

1995). 

Today, certain harbour installations are still visible, 

including two outer moles, an entrance channel 

leading to an elongated inner harbour basin and the 

remains of an ancient quay wall (Paris, 1915). The 

present day topography is dominated by large 

mounds of sediments dredged from the harbour 

basin (Fig. 1). Between the inner harbour basin and 

the present beach the remains of an early Christian 

basilica, dating to the late 5 th century AD, were 

70


EARTHQUAKE 

ARCHAEOLOGY 



excavated. The basilica, 186 m long, represents the 

largest from this period and was completely 

destroyed by the 521 or 551 AD earthquake series 

(Krautheimer, 1989; Rothaus, 1995). 

Fig. 1: (a) Location of study area in Greece. (b) 

Overview of the Corinthia. Red box indicates harbour 

site of Lechaion, red dashes indicate beachrock. (c) 

Map of the Lechaion harbour basin and adjacent area. 

Vibracores LEC 1-3 are indicated by red dots, ERTmeasurements 

by red dashed lines and GRPmeasurements 

by white lines (maps modified after 

Google Earth images, 2009). 

TSUNAMI IMPACT AT THE LECHAION HARBOUR 

SITE 

Sedimentary evidence from the harbour basin 

As previous studies show, tsunami impact may be 

documented by layers of allochthonous coarsegrained 

sediment intersecting quiescent near-shore 

deposits (Vött et al. 2009). Low-energy environments 

such as harbour basins therefore represent excellent 

geological archives for palaeotsunami research. 

Sedimentological, geochemical and micropalaeontological 

studies were carried out at the Lechaion 

harbour site (see Fig. 1 for location of vibracores). 

The local stratigraphy comprises shallow marine 

(pre-harbour) deposits overlain by harbour-related, 

quiescent lagoonal sediments followed by limnic to 

terrestrial deposits accumulated when the harbour 

was already out of use (Fig. 2). The stratigraphical 

record of all vibracores is repeatedly interrupted by 

layers of allochthonous coarse-grained sediment and 

shell debris, partly characterized by fining upward 

sequences. These ex situ-layers extend up to 450 m 

inland. Erosional unconformities at the base (Fig. 2b) 

and immediately re-established quiescent conditions 

on top of the event layers indicate short term highenergy 

interference of the Lechaion harbour site. 

Fig. 2: (a) Vibracore LEC 2 (N 37° 55’ 58.6 E 22° 53’ 

03.5’’, -0.04 m a.s.l.) was drilled in the central 

Lechaion harbour basin. Here, two coarse-grained 

tsunami layers were found, intersecting the 

autochthonous, predominantly shallow-marine to 

lagoonal facies. Youngest tsunami deposit as 

encountered in core LEC 3 and covering early 

Christian basilica is dredged at site LEC 2 (see ). 

14 

C-samples are marked by *. (b) Detailed view of the 

younger event layer. The quiescent harbour 

environment was abruptly covered by allochthonous 

high-energy deposits (Photo by T. Willershäuser, 

2010). 

The event-stratigraphical correlation of all vibracore 

profiles reveals three distinct tsunami layers. The 

geochronological framework is based on radiocarbon 

dating and age determination of diagnostic ceramic 

fragments. Tsunamigenic impact could be timebracketed 

by radiocarbon dating to around 760 cal 

BC and 50 cal AD and to the 6 th century AD by 

geoarchaeological findings. The youngest event 

correlates to the geoarchaeological destruction layer 

presented below. 

Geoarchaeological destruction layers 

Large sediment mounds adjacent to the inner 

harbour basin are generally referred to as natural 

dunes or sediments obtained from dredging activity 

(Frazer, 1965; Rothaus, 1995; Stiros, 1996). They 

consist of sand and gravel, intermingled with 

numerous ceramic fragments and marine 

macrofossils. The sediment composition and grain 

size distribution thus exclude natural dune formation. 

However, gravel is also completely atypical for the 

siltation of a quiescent harbour basin environment 

(Marriner & Morhange 2007). According to their grain 

size, the sediments provide evidence of high-energy 

influence. Incorporating marine as well as 

terrigeneous material, they indicate strong 

tsunamigenic inundation and backflow affecting the 

coastal plain at Lechaion. Since associated to 

dredging activities, the mounds therefore document 

tsunami influence burying the Lechaion harbour 

basin. 

Widespread burial of the harbour site by high-energy 

influence must also be assumed regarding the ruins 

71


EARTHQUAKE 

ARCHAEOLOGY 



of the 5 th century AD Christian basilica. The entire 

foundation like the adjacent area is completely 

covered by an up to 2 m thick sediment layer. The 

surrounding topography as well as the sediment 

composition, however, exclude a colluvial, fluvial or 

mass denudative burial of the site. The deposits are 

also not related to the archaeological excavation of 

the site. A former door lintel used as a threshold after 

the destruction of the church attests that the entrance 

level was already elevated during historic times. 

Made up of a sandy matrix incorporating abundant 

gravel, numerous ceramic fragments and marine 

macrofossils, the high-energy sediment layer again 

reflects tsunami backflow. 

probable cause for the final destruction and 

abandonment of the harbour. 

Beachrock as indicator for tsunami impact 

Along the Corinthian coastline, extensive beachrock 

formations dominate the present coastal 

geomorphology. As recently described by Vött et al. 

(2010), beachrock deposits may be of tsunamigenic 

origin. At the Corinth Canal at a distance of 7 km 

from Lechaion, the beachrock complex extends up to 

300 m inland (Fig. 4a), while the maximum extension 

of the recent beach is less than 25 m. The deposits 

show a clearly laminated structure with multiple fining 

upward sequences reaching from gravel to fine sand. 

The inner structure reveals a landward orientated 

imbrication of the gravel components (Fig. 4b) 

documenting landward flow dynamics as may be 

induced during tsunami inundation. According to their 

sedimentary structure, these beachrock deposits 

must not automatically be regarded as lithified beach 

but may rather represent calcified tsunamigenic 

deposits (for further discussion see Vött et al. 2010). 

The “diolkos”, an ancient slipway across the Isthmus 

of Corinth, is partly covered by beachrock; used until 

the early 1 st century AD, the diolkos represents a 

terminus post quem for the tsunamigenic impact. 

Fig. 3: (a) Excavated remains of the 5 th century AD 

Christian basilica erected to the north of the former 

harbour basin. (b) The present day ground surface 

overtops the basilica’s original ground level by up to 2 

m (in photo: ca. 1.2 m). (c) The sedimentary cover 

consists of sand and gravel with abundant ceramic 

fragments and marine shell debris. 

Several earth resistivity transects were carried out 

around the ancient harbour basin (Fig. 1c). They 

revealed a sharp boundary between the allochthonous 

high-energy deposits and the underlying 

autochthonous fine-grained harbour sediments as 

well as a clear thinning landward of the event layer 

up to 450 m inland. Furthermore, ground penetrating 

radar profiles (Fig. 1c) revealed channel-like 

structures at their base. Orientated perpendicular 

towards the coastline, these incised channels 

indicate strong linear erosion in land-/seaward 

direction. Similar structures, created by strong 

backflow processes, were observed along the 

Chilean coast after the February 2010 Chile tsunami 

(Bahlburg & Spiske 2010). 

In a summary view, our results give evidence for the 

influence of tsunami impact on the Lechaion harbour. 

According to Rothaus (1995), only few ceramic 

fragments younger than late Roman to early 

Byzantine times were found in the vicinity of the 

harbour basin and basilica. Thus, we consider 

tsunami landfall in the 6 th century AD as the most 

Fig. 4: (a) Extensive beachrock deposits at the 

Corinth Canal. (b) The internal structure shows a 

fining upward sequence as well as imbricated pieces 

of gravel. 

CONCLUSION 

Based on our studies, multiple tsunami impact was 

determined for the Lechaion harbour site and 

adjacent coastal areas. At least three distinct event 

layers were identified, the youngest and obviously 

most destructive event dating to the 6 th century AD. 

The spatial distribution and geomorphological 

variability of the encountered tsunami deposits 

require an event-stratigraphical approach to 

understand and reconstruct the chronology of events. 

Based on the combination of historical accounts with 

geomorphological, sedimentological, geophysical and 

geoarchaeological data we conclude that the ancient 

harbour of Lechaion, though influenced by tsunamis 

before, was finally destroyed by tsunami impact, 

most probably triggered during the 521 or 551 AD 

earthquake series. The present day topography of 

the harbour site is due to a latter re-activation of the 

72


EARTHQUAKE 

ARCHAEOLOGY 



harbour in medieval or pre-modern times and thus 

only partly represents the ancient Corinthian harbour. 

Acknowledgement: We acknowledge funding of the 

project by the German Research Foundation, DFG (Bonn, 

Gz. VO 938/3-1). Sincere thanks are due to the 37 th 

Ephorate of Prehistoric and Classical Antiquities, Corinth, 

for support during field work and discussion. 

References 

Ambrasey, N. & C. Synolakis, (2010). Tsunami catalogs for 

the eastern Mediterranean, revisited. Journal of 

Earthquake Engineering 14, 309-330. 

Avallone, A., P. Briole, A.M. Agatza-Balodimou, H. Billiris, 

O. Charade, C. Mitsakaki, A. Nercessian, K. Papazissi, 

D. Paradissis & G. Veis, (2004). Analysis of eleven years 

of deformation measured by GPS in the Corinth Rift 

Laboratory area. Competes Rendus Geosciences 336 (4- 

5), 301-311. 

Bahlburg, H. & M. Spiske, (2010). The February 27, 2010 

Chile Tsunami. Sedimentology of runup and backflow 

deposits at Isla Mocha. American Geophysical Union, 

Abstract #OS42B-06. 

Brownson, C.L. (ed.), (1918). Xenophon, Hellenica. Harvard 

University Press. Cambridge. 

Frazer, J.G., (1965). Pausanias’s description of Greece. 

Biblo and Tannen. New York. 652 p. 

Hamilton, H.C. & W. Falconer, (1903). The Geography of 

Strabo. George Bell & Sons. London. 

Hasiotis, T., G. Papatheodorou, G. Bouckovalas, C. Corbau 

& G. Ferentinos, (2002). Earthquake-induced coastal 

sediment instabilities in the western Gulf of Corinth, 

Greece. Marine Geology 186, 319-335. 

Krautheimer, R., (1989). Early Christian and Byzantine 

Architecture. Penguin Books. London. 556 p. 

Marriner, N. & C. Morhange, (2007). Geoscience of ancient 

Mediterranean harbours. Eart-Science Reviews 80, 137- 

194. 

Papadopoulos, G.A., (2003). Tsunami hazard in the eastern 

Mediterranean: strong earthquakes and tsunamis in the 

Corinth Gulf, Central Greece. Natural Hazards 29, 437- 

464. 

Papazachos, B.C. & P.P. Dimitriu, (1991). Tsunamis in and 

near Greece and their relation to the earthquake focal 

mechanism. Natural Hazards 4, 161-170. 

Paris, J., (1915). Contributions à l’étude des portes antiques 

du mond grec. Notes sur Lechaion. Bulletin de 

Correspondance Hellénique 39 (1), 5-16. 

Rothaus, R., (1995). Lechaion, western port of Corinth: a 

preliminary archaeology and history. Oxford Journal of 

Archaeology 14 (3), 293-306. 

Sachpazi, M., C. Clément, M. Laigle, A. Hirn & N. Roussos, 

(2003). Rift structure, evolution, and earthquakes in the 

Gulf of corinth, from reflection seismic images. Earth and 

Planetary Science Letters 216, 243-257. 

Soloviev, S.L., O.N. Solovieva, C.N. Go, K.S. Kim & N.A. 

Shchetnikov, (2000). Tsunamis in the Mediterranean Sea 

2000 BC – 2000 AD. Kluwer Academic Publishers. 

Dordrecht. 237 p. 

Stefatos, A., M. Charalambakis, G. Papatheodorou & G. 

Ferentinos, (2006). Tsunamigenic sources in an antive 

European half-graben (Gulf of Corinth, Central Greece). 

Marine Geology 232, 35-47. 

Stiros, S., P. Pirazolli, R. Rothaus, S. Papageorgiou, J. 

Laborel & M. Arnold, (1996). On the date of construction 

of Lechaion, western harbor of Ancient Corinth, Greece. 

Geoarchaeology: An International Journal 11 (3), 251- 

263. 

Vött, A., H. Brückner, S.M. May, D. Sakellariou, O. Nelle, F. 

Lang, V. Kapsimalis, S. Jahns, R. Herd, M. Handl & I. 

Fountoulis, (2009). The Lake Voulkaria (Akarnania, NW 

Greece) palaeoenvironmental archive – a sediment trap 

for multiple tsunami impact since the mid-Holocene. 

Zeitschrift für Geomorphologie N.F., Suppl. Issue 53 (1), 

1-37. 

Vött, A., G. Bareth, H. Brückner, C. Curdt, I. Fountoulis, R. 

Grapmayer, H. Hadler, D. Hoffmeister, N. Klasen, F. 

Lang, P. Masberg, S.M. May, K. Ntageretzis, D. 

Sakellariou & T. Willershäuser, (2010). Beachrock-type 

calcarenitic tsunamites along the shores of the eastern 

Ionian Sea (western Greece) – case studies from 

Akarnania, the Ionian Islands and the western 

Peloponnese. Zeitschrift für Geomorphologie N.F., Suppl. 

Issue 54 (3), 1-50. 

73


EARTHQUAKE 

ARCHAEOLOGY 



STRUCTURAL CHARACTERISTICS AND EVOLUTION OF THE YANGSAN-ULSAN FAULT 

SYSTEM, SE KOREA 

S.-R. Han (1), M. Lee (1), J. Park (2), Y.-S., Kim (1*) 

(1) Dept. of Environmental Geosciences, Pukyong National University, Busan 608-737, Korea. *Email: ysk7909@pknu.ac.kr 

(2) GeoGeny Consultants Group Inc. 807-2 Bangbae-Dong Seocho-Gu, Seoul 137-831, Korea. 

Abstract (Structural characteristics and evolution of the YangSAN-Ulan fault system, SE Korea): The Yangsan-Ulsan Fault 

System (YUFS) is a major fault system in Korea. More than 40 Quaternary faults have been discovered around the YUFS. To 

understand the evolution and structural characteristics of the YUFS, we performed numerical modelling for the evolution of the 

YUFS and fault zone analysis on the eastern part of the Ulsan fault. The result of the modelling shows that the YUFS evolved into 

-fault, a low angle merging fault system. In a fault zone analysis, the hanging wall block of the fault shows relatively higher 

damage with higher fracture density and deformation structures compared with the footwall. This indicates that the hanging wall is 

more susceptible to deformation than the footwall during faulting events; this result is consistent with damage patterns in other 

Quaternary faults. Epicenters of recent earthquakes around the study area are also concentrated on the hanging wall of the Ulsan 

fault. This kind of study can help us to evaluate hazards resulting from future potential earthquakes in and around active fault 

systems. 

Key words: Yangsna-Ulsan fault system, 

-fault , Quaternary faults 

INTRODUCTION 

The Korean Peninsula is generally considered as a 

relatively safe region considering earthquake 

hazards, because it is located away from plate 

boundaries. However, many moderate and minor 

earthquakes occur around the Yangsan-Ulsan fault 

system (YUFS), in the southeastern part of the 

Korean Peninsula (Lee and Yang, 2006). Recently, 

more than 40 Quaternary faults were discovered near 

the YUFS (Jin et al., 2011). 

There are two nuclear power plant sites near the 

YUFS (Fig.1). To prevent nuclear catastrophe due to 

a potential large scale earthquake, it is necessary to 

understand the deformation characteristics and future 

evolution of the YUFS. For this purpose, a numerical 

modelling related to the evolution of the YUFS and a 

detailed field investigation on a reactivated fault were 

carried out in this study. 

STUDY AREA 

The study area is located in the Cretaceous 

Gyeongsang Basin, SE Korea (Fig. 1). The basement 

of the study area is composed of Cretaceous 

sedimentary rocks with later igneous and volcanic 

rocks. The Yangsan fault is a NNE-SSW trending 

right-lateral dominant strike-slip fault, and the NNW- 

SSE trending Ulsan fault is interpreted as a strike-slip 

fault overprinted by later reverse-slip. Recently, more 

than 40 Quaternary faults have been discovered near 

the Yangsan and Ulsan faults. Historical records 

show many earthquakes in this region, with an 

earthquake producing over 100 casualties recorded 

Fig. 1: Regional geologic map of Gyeongsang basin, 

the SE part of the Korean peninsula and Quaternary 

faults (circles) around the Yangsan and Ulsan faults. 

Two Nuclear Power plants are located near the YUFS 

(modified from Lee, 2000) 

in A.D. 779 (estimated M= 6.7), near the intersection 

of the two faults (Lee and Yang, 2006). 

74


EARTHQUAKE 

ARCHAEOLOGY 



 

Fig. 2: Comparison between coulomb stress changes, distribution of Quaternary faults and epicenters of earthquakes 

around the Yangsan-Ulsan fault system (Han et al., 2009). (a) Coulomb stress changes are concentrated in the eastern 

and southeastern region of the Yangsan fault and eastern region of the Ulsan fault in Quaternary. (b) Locations of 

Quaternary faults around the Yangsan-Ulsan fault system. (c) Earthquake distributions around the southeastern part of the 

Korean peninsula (1994-1998). Green circles indicate the epicenters of the recent earthquakes. Earthquakes are 

concentrated in the eastern part (hanging wall) of the Ulsan fault. 

EVOLUTION MODELLING OF THE YANGSAN- 

ULSAN FAULT SYSTEM 

Analysis of lineaments and geomorphology of the 

YUFS indicate that the YUFS is analogous to the 

characteristics of -fault. The concept of -fault was 

suggested by Du and Aydin (1995). This is the 

geometric relationship between two intersecting 

strike-slip faults. When a strike-slip fault propagates 

into a pre-existing strike-slip fault, the orientation of 

the propagation depends on the angle between the 

maximum principal stress and propagating fault. To 

understand the evolution of the fault system, a 

numerical modelling on the YUFS with regards to 

stress changes with time was performed, using the 

USGS open program Coulomb 3 

(http://earthquake.usgs.gov). This modelling was 

carried out based on a previously suggested tectonic 

evolution model for the YUFS (Park, 2004). The 

result of this numerical modelling indicates that the 

YUFS is a -fault. During the propagation of the 

Ulsan fault into the Yangsan fault in the late Miocene, 

coulomb stress increases at the north-western tip of 

the Ulsan fault. After the Ulsan fault connected to the 

Yangsan fault in Pliocene, a new segment was 

initiated at the northern part of the Ulsan fault due to 

stress changes. Since the penetration of the Ulsan 

fault into the Yangsan fault in Quaternary, coulomb 

stress increased in the north-eastern region of the 

Yangsan fault and the eastern region of the Ulsan 

fault. These regions are well consistent with the 

distribution of Quaternary faults and locations of 

epicenters of recent earthquakes around the YUFS 

(Fig. 2). 

FAULT GEOMETRY AND SLIP ANALYSIS OF A 

FAULT ZONE IN THE WEOLSEONG AREA 

In the SE part of Korea, two nuclear power plant 

(NPP) sites are located near the YUFS. In the new 

Weolseong NPP site located in the southeastern 

coast of Korea, an east dipping, N-S trending fault 

was discovered during construction. The main fault 

zone shows various different coloured fault gouge 

bands and shear fabrics indicating opposite slips (Fig. 

3). These features indicate that there were discrete 

stages of normal and reverse faulting events along 

this fault zone. The analyzed kinematic indicators, 

such as lineations, cleavages, and slickenlines in the 

fault zone suggest that the fault was initiated as a 

normal fault and was reactivated as a reverse fault 

under NW-SE compression. The hanging wall block 

of the fault shows relatively higher damage indicating 

higher deformation in the hanging wall than the 

footwall. 

CONCLUSION 

It is common that earthquakes occur by reactivation 

of pre-existing faults (e.g. Chen, 2002; Ota et al., 

2004; Ota et al., 2005). It is well known that most of 

the damage from earthquake is concentrated on the 

hanging wall block during the faulting (Ota et al., 

2005). A N-S trending, east dipping fault zone was 

discovered in the construction site of the new 

Weolseong NPP.Detailed analysis was performed on 

this fault zone and the result indicates multiple 

faulting events. According to the analysis, this fault 

zone initiated as a normal slip fault and it was 

75


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 3: Photomosaic (a), sketch (b) and sense indicators (c) of reactivated fault zone located at the new Weolseong nuclear 

power plant (Fig. 1). 

reactivated as a reverse fault. The damages in the 

wall rocks of this fault were concentrated on the 

hanging wall block rather than footwall block. This 

result indicates that the hanging wall block is subject 

to higher deformation than footwall block during 

faulting. This result is well consistent with the result 

from larger scale deformation characteristics around 

the YUFS producing intense distribution of 

Quaternary faults and earthquakes in the hanging 

wall of the Ulsan fault. Therefore, detailed analysis of 

fault zone and fault evolution modelling can 

contribute to understanding of the deformation 

characteristics of a fault zone and the evaluation of 

earthquake hazards. 

Therefore, the fault geometries and kinematics 

are important factors to evaluate conditions of 

important sites. Furthermore, fault damage zones 

(Kim et al., 2004) around faults are very important 

indications for the seismic hazard assessment 

due to secondary fractures and aftershocks (Kim 

and Sanderson, 2008). 

Acknowledgements: This work was funded by Korea 

Institute of Nuclear Safety under Grant M20702070001- 

08M0207-00110. 

References 

Lee, J.I. (2000) Provenance and thermal maturity of the 

lower Cretaceous Gyeongsang Supergroup, Korea. Ph.D. 

Thesis. Seoul National University. 

Lee, K. and Na, S.H. (1983) A study of microearthquake 

activity of the Yangsan fault. Journal of the Geological 

Society of Korea, 19, 127-135. 

Jin, K., Lee, M., Kim, Y.-S., Choi, J.-H. (2011). 

Archaeoseismological studies on historical heritage sites 

in the Gyeongju area, SE Korea. Quaternary International, 

doi:10.1016/j.quaint.2011.03.055. 

Lee, M. S., Kang, P. C. (1964) Geological map of Yangsan 

area. Geological Survey of Korea. ( In Korean with 

English abstract ) 

Lee, K., Yang, W.-S. (2006). Historical seismicity of Korea. 

Bulletin of the Seismo-logical Society of America 96, 

846e855.Stiros, S. 1988b. Archaeology, a tool to study 

active tectonics – The Aegean as a case study. Eos, 

Transactions, American Geophysical Union 13, 1636- 

1639. 

Ota, Y., Watanabe, M., Suzuki, Y., Sawa, H. (2004). 

Geomorphological identification of pre-exiting active 

Chelungpu Fault in central Taiwan, especially its relation 

to the location of the surface rupture by the 1999 Chichi 

earthquake. Quaternary International 115–166, 155– 166. 

Ota, Y., Chen, Y.-G., Chen, W.-S., (2005). Review of 

paleoseismological and active fault studies in Taiwan in 

the light of the Chichi earthquake of September 21, 1999. 


Park. J. (2004) The structural characteristics of the Tertiary 

basin in Gyeongju region, Korea. Master thesis. Seoul 

National University. ( In Korean with English abstract ) 

Du, Y. and Aydin, A. (1995) Shear fracture patterns and 

connectivity at geometric complexities along strike-slip 

faults. Journal of Geophygical Research, 100, 18093- 

18102. 

Han, S.-R., Park. J., Kim, Y.-S. (2009) Evolution modelling 

of the Yangsan-Ulsan fault system with stress changes. 

Journal of the Geological Society of Korea. 45, 361-377. ( 

In Korean with English abstract ) 

Chen, Y.-G., Chen, W.S., Wang, Y., Lo, P.W., Lee, J.C., Liu, 

T.K.(b) (2002) Review of paleoseismological and active 

fault studies in Taiwan in the light of the Chichi 

earthquake of September 21, 1999.Tectonophysics 508, 

63-77. 

76


EARTHQUAKE 

ARCHAEOLOGY 



WHERE LANDSLIDES REPRESENT THE MOST IMPORTANT 

EARTHQUAKE-RELATED HAZARDS: THE MOUNTAIN AREAS OF CENTRAL ASIA 

Havenith, Hans-Balder (1) 

(1) Georisks and Environment group, Department of Geology, University of Liege. B18. 4000 Liege. BELGIUM. Email: 

HB.Havenith@ulg.ac.be 

Abstract (Where landslides represent the most important earthquake-related hazards: the mountain areas of Central 

Asia): The goal of this paper is to show which areas are particularly prone to earthquake-induced slope failure. In this regard, the 

paper considers the short- and long-term effects of geological, tectonic, climatic and morphological conditions. Case histories 

related to five earthquake events in the Tien Shan and Pamir Mountains are outlined: the earthquakes of Kemin in 1911, Sarez in 

1911, Khait in 1949, Gissar in 1989 and Suusamyr in 1992. The potential impact of seismically triggered mass movement affecting 

loess deposits is further pointed out through comparison with the 1920 Haiyuan (China) earthquake. One conclusion is that in 

semi-arid mountain regions marked by a strong seismic activity, such as those in Central Asia, seismogenic landslides and related 

long-term effects may represent the most important geohazards. Further, the susceptibility to seismic slope instability is highest 

along active fault zones and on convex slopes made of soft or fractured materials. 

Key words: landslides, co- and post-seismic effects, dynamics 

INTRODUCTION: THE PERCEPTION OF SEISMIC 

LANDSLIDE HAZARDS 

During the last ten years, after a series of disastrous 

earthquake events in mountain regions in Taiwan 

(1999), El Salvador (2001), Pakistan (2005) and 

China (2008), increasing attention has been 

addressed to landslides triggered by earthquakes. 

Previously, landslides have been considered as 

minor effects of earthquakes compared to the impact 

of the ground shaking itself. Schuster and Highland 

(2001) partly attributed the perception of the relatively 

small impact of earthquake-triggered mass 

movements to the fact that many related losses are 

often referred to as direct consequences of the 

earthquake. However, for the earthquake that hit the 

Kashmir mountains on October 8, 2005, Petley et al. 

(2006) estimated that about 30% of the total number 

of killed people (officially 87350), i.e. 26500, had 

been victims of co-seismic landslides. Less than 

three years later, on May 12, 2008, the Wenchuan 

earthquake hit the Sichuan and neighbouring 

provinces of China and caused ‘more than 15000 

geohazards in the form of landslides, rockfalls, and 

debris flows, which resulted in about 20000 deaths’ 

(Yin et al. 2009). These casualties represent again 

almost 30 % of the official total number of fatalities of 

69197 of this event. 

In the remote areas of Central Asian mountains, M>6 

earthquakes generally do not cause catastrophes – 

but if they do, they do it through mass movements, 

like in the case of the M=8.2 Kemin, 1911, M=7.4 

Sarez, 1911, M=7.4 Khait, 1949, M=5.5 Gissar, 1989 

and M=7.3 Suusamyr, 1992 earthquakes (see 

location in Fig. 1). These case histories show that 

both rapid flows (mainly) in loess deposits and 

massive rock avalanches created the greatest 

disasters. To highlight the important issue related to 

landslides triggered in loess deposits, a comparison 

is made with one of the worldwide most catastrophic 

seismic events involving landslides: the M=7.8 

Haiyuan (or Gansu, 1920) earthquake in China. 

Special attention must also to be paid to post-seismic 

increase of landslide activity. In this regard, we have 

to take into consideration so-called ‘secondary or 

tertiary effects’ of earthquakes, such as landslide 

dams and related flooding impacts. The 2008 

Wenchuan earthquake clearly marked the 

importance of such effects – which could have killed 

another few thousands of people if efficient mitigation 

measures had not been taken by Chinese authorities. 

GEOLOGICAL, TOPOGRAPHIC AND CLIMATIC 

SETTING OF EARTHQUAKE-INDUCED SLOPE 

FAILURE 

Sassa (1996) observed that seismic landslide 

occurrence is strongly dependent on the proximity of 

the fault rupture. This observation is confirmed by 

many others made after the 1999 Taiwan, 2005 

Pakistan and 2008 Sichuan earthquakes. 

The geological factors have been extensively 

analysed by Keefer (1984) who concluded that any 

geological material with low geomechanical strength, 

be it soil or rock, may be susceptible to earthquakeinduced 

slope instability. The low strength may be 

related to weak cementation, intense weathering or 

fracturing, high water saturation or poor compaction. 

Concerning the structural elements (bedding, 

foliation, sediment-rock contact), no clear trend can 

be outlined. Especially for seismic rock slope failure, 

e.g., it is not clear if cross-bedding joints or bedding 

contacts support the development of a sliding surface 

77


EARTHQUAKE 

ARCHAEOLOGY 



or not. Here, clearly more research based on 

structural geology mapping is needed. 

Further, earthquake-induced landslides may be 

triggered from any surface morphology, in flat areas, 

such as lateral spreads, or on steep cliffs, such as 

rock falls. Still, several particularities can be outlined. 

From observations after the 2005 Pakistan 

earthquake, Sato et al. (2007) inferred that ‘there was 

a slight trend that large landslides occurred on 

vertically convex slopes rather than on concave 

slopes; furthermore, large landslides occurred on 

steeper (30° and more) slopes than on gentler 

slopes’. From my experience of earthquake-induced 

landslides in the Tien Shan (see below), it can be 

concluded that mainly the surface curvature has an 

influence on seismic slope stability at a global scale; 

particularly, hillcrests, higher parts of slopes and 

convex surface morphologies are prone to seismic 

slope failure. The influence of slope angle on seismic 

mountain regions marked by a high seismicity – such 

as the Tien Shan and the Pamir, but also the Hindu- 

Kush and others. Why semi-arid? If the climate is 

wet, the overall landslide hazards increase more than 

the specific hazard related to seismically triggered 

slope failure (e.g. Himalayas). On the contrary, if the 

climate is very dry – arid, the seismic hazards with 

direct impacts on building stability prevail (e.g. 

mountains in the Middle East, Southern Altai) since 

slope instability – also under seismic conditions - 

requires a certain amount of humidity in the top-soils 

to develop. In the Tien Shan, we observed that 

earthquakes, which occurred at the end of the dry 

summer season, caused relatively few slope failures 

(e.g., Ms=7.3 Suusamyr earthquake on August 19, 

1992; Ms=6.7 Nura earthquake on October 5, 2008); 

while, for instance, the Ms=7.4 Khait earthquake of 

July 10, 1949 – at the beginning of the dry highmountain 

summer - in northern Tajikistan triggered 

widespread slope instability. 

Fig. 1: Map of Tien Shan and Pamir Mountains in Central Asia with location of major faults and earthquakes (white filled 

circles show all recorded M>=7 earthquakes with the year of occurrence; the magnitude is indicated for analysed events) and 

related major mass movements (stars). 

slope stability is not clear; in some cases, especially 

in rocks, steeper slopes are more prone to instability; 

in others, especially in soft sediments, gentle slopes 

produce most of the mass movements, indicating that 

the combined effect of slope and geology has to be 

taken into consideration. 

The contribution of the climate to seismic slope 

failure has not yet been well investigated. According 

to our experience, the strongest contribution of 

earthquake-induced landslides to the total seismic 

and landslide hazards can be observed in semi-arid 

CASE HISTORIES FROM CENTRAL ASIA 

COMPARED WITH THE 1920 CHINA EVENT 

Nadim et al. (2006) assessed landslide and 

avalanche occurrence probabilities worldwide on the 

basis of morphological, geological, meteorological 

and seismological data. They clearly showed that all 

landslide hotspots are located in seismically active 

mountain ranges. For Central Asia, they estimate that 

global landslide hazard can be rated as medium to 

very high. They further noted that some areas in 

78


EARTHQUAKE 

ARCHAEOLOGY 



Tajikistan are marked by highest mortality risk due to 

landslides. 

The following case histories document the landslide 

hazard and risk triggered by earthquakes in the Tien 

Shan and the Pamir Mountains. A comparison will be 

made with the 1920 Haiyuan (China) event to outline 

the most important factors contributing to landslide 

hazard and risk related to the presence of thick loess 

deposits in Central Asia. 

The Kemin earthquake, 1911 

The Kemin Ms=8.2 earthquake of 1911 (January 3) is 

one of the strongest events ever recorded in the Tien 

Shan. The earthquake caused extensive landsliding 

along the activated fault segments over a length of 

200 km. The ‘Ananevo’ rockslide located in the north 

of lake Issyk Kul (at some 80 km east of the 

presumed epicentre) is one of the most prominent 

feature produced by the Kemin earthquake (Havenith 

et al., 2002, see Fig. 2). Failure took place at the 

southern end of a mountain ridge, just above the 

discontinuous Chon Aksu fault also activated by the 

1911 Kemin earthquake. This section of the Chon 

Aksu fault is a thrust gently dipping towards the 

northeast into the collapsed slope. Evidence of the 

presence of the fault is the related scarp with a height 

of almost 10 m at 12 km to the WNW. On the site 

itself, outcrops at the foot of the southwest-oriented 

slope show particularly disintegrated and weathered 

granitic rocks within a 100-200 m thick fault zone. 

thousands of inhabitants (Fig. 3) - the exact number 

of fatalities will never be known. This rock avalanche 

had been triggered from Borgulchak mountain at an 

altitude of about 2950 m and travelled more than 6 

km before reaching the inhabited valley at an altitude 

of 1550 m. The volume is about 40*10 6 m 3 . A 

significant part of the mass movement was made of 

loess, which probably contributed to the mobility of 

the initial rockslide. In the Yasman valley, opposite to 

the Khait rock avalanche, massive loess earth-flows 

are believed to have buried about 20 villages. In total, 

the Khait rock avalanche and loess earth-flows are 

likely to have killed more than 5000 people during the 

1949 event. 

Fig. 3: Khait rock avalanche; view towards the East from 

Yasman valley (unpublished photograph of 2005 provided 

by A. Ischuk). The length of the scarp is about 1 km. 

The Gissar earthquake, 1989 

South of Dushanbe, in Gissar, Tajikistan, a Ms=5.5 

earthquake on January 23, 1989 had triggered a 

series of earth-flows in loess. At least 200 people 

were killed and hundreds of houses were buried. 

According to Ishihara et al. (1990), those slides were 

all related to extensive liquefaction, which had 

developed for a horizontal acceleration of about 

0.15g. Ishihara et al. (1990) associated the 

liquefaction to the ‘collapsible nature’ of the highly 

porous loess material (a silt-sized deposit with an 

average content of clay of 15 % and a low plasticity). 

Fig. 2: Photograph of the Ananevo rockslide. The horsefarm 

building (lower left corner) is 60 by 60 m. 

The Sarez earthquake, 1911 

The Sarez earthquake, Ms=7.6, struck the central 

Pamir Mountains, Tajikistan, on February 18, 1911. 

Such an earthquake is likely to have triggered 

hundreds or thousands of mass movements, but only 

one is well documented: the giant Usoi rockslide, 

which fell from a 4500 m high mountain down to an 

elevation of 2700 m in the valley (Schuster and 

Alford, 2004). This rockslide has formed a dam with a 

volume of about 2*10 9 m 3 on Murgab River. 

The Khait earthquake, 1949 

According to Leonov (1960), the M=7.4 Khait 

earthquake that struck Northern Tajikistan on July 10, 

1949, produced a very destructive mass movement 

that had buried the villages of Khait and Kusurak with 

The Suusamyr earthquake, 1992 

The most recent large seismic event hitting Central 

Asian mountain regions was the Ms=7.3 Suusamyr 

earthquake on August 19, 1992, triggering various 

types of ground failures in the Northern-Central Tien 

Shan (Bogachkin et al., 1997). 

Most of the 50 people killed in the remote areas were 

victims of mass movements. Korjenkov et al. (2004) 

described a series of ground failures and also a great 

variety of gravitation cracks. Ground instability could 

be observed along the crest and southern slope of 

the Chet-Korumdy ridge – here, most landslides had 

developed from previously existing ground 

instabilities. The largest mass movement, a rock 

avalanche, had formed a landslide dam that partly 

failed in 1993, causing a long-runout debris flow and 

widespread flooding downstream. 

79


EARTHQUAKE 

ARCHAEOLOGY 



The Haiyuan or Gansu earthquake, 1920 

On December 16, 1920, a M=8.5 earthquake 

occurred near Ganyan Chi, Haiyuan County of the 

Ningxia Hui Autonomous Region in China (Zhang, 

1995). Several hundreds of thousands of houses 

collapsed and officially 234.117 people died. Zhang 

(1995) noticed that particularly high intensities were 

recorded over areas covered by thick loess deposits 

and that in those deposits ‘landslides were not only 

controlled by the intensity of the earthquake, but by 

the structure of the subsoil’. Zhang and Wang (2007) 

reported that about 100000 people were killed just by 

landslides in loess deposits. They observed that 

loess earth-flows triggered by the Haiyuan 

earthquake had developed on relatively gentle slopes 

compared to those triggered by rainfall in the same 

region. These observations highlight the particular 

susceptibility of loess areas to ground failure, such as 

it was clearly shown by Derbyshire et al. (2000) 

analysing geological hazards affecting the loess 

plateau of China. 

CONCLUSIONS 

A series of case histories of earthquake-induced 

landslides in Central Asia have been presented. 

These show that the most disastrous mass 

movements in Central Asia are long runout and rapid 

(>20 m/s) rock avalanches and loess earth-flows. 

While the giant rockslides are almost exclusively 

triggered by large magnitude seismic events (M7) in 

Central Asian mountain regions, loess earth-flows 

may also be triggered by smaller earthquakes – or 

even by climatic factors. Here, I presented some 

examples of fatal loess landslides triggered directly 

by a M=5.5 earthquake in Tajikistan. The comparison 

with the Haiyuan earthquake event of 1920 shows 

that such loess landslides can be very disastrous. 

The importance of mid- and long-term effects is 

outlined for both rockslides and loess landslides. 

Several case histories showed that one important – if 

not the most important – long-term consequence of 

massive rockslides can be the formation of a dam 

and the impoundment of a natural reservoir. Actually, 

the largest still existing rockslide dam on earth had 

formed in 1911 in the Pamir Mountains. 

I also wanted to show that similar earthquakes may 

not necessarily trigger the same number of 

landslides, due to different climatic conditions and 

groundwater level at the time of the earthquake. 

Finally, it is important to note that landslides are not 

only instantaneous effects of earthquakes – some 

had already developed before the seismic shock and 

some continued or started moving well after the 

shaking. To better assess the short- to long-term 

effects earthquakes on slopes, landslides need to be 

monitored by geophysical, seismological and 

geotechnical systems, coupled to multi-temporal 

satellite imagery and numerical modeling of multievent 

scenarios. In the frame of new projects on 

landslide problems in Central Asia, focus will be on 

the installation of such coupled monitoring–modeling 

systems. 

Acknowledgements: This study was supported by the 

NATO science for Peace and Security Project LADATSHA 

983289, 2009-2012. 

References 

Derbyshire, E., Meng, X. & T.A. Dijkstra, (2000). Landslide 

in the Thick Loess Terrain of North-West China. John 

Wiley, Chichester, U.K. 

Havenith, H.B., Jongmans, D., Faccioli, E., Abdrakhmatov, 

K. & P.Y. Bard, (2002). Site effects analysis around the 

seismically induced Ananevo rockslide, Kyrgyzstan. 

Bulletin of the Seismological Society of America, 92, 

3190-3209. 

Ishihara, K., Okusa, S., Oyagi, N. & A. Ischuk, (1990). 

Liquefaction-induced flow slide in the collapsible deposit 

in the Soviet Tajik. Soils and Foundations, 30: 73-89. 

Keefer, D.K., (1984). Landslides Caused by Earthquakes, 

Geological Society of America Bulletin, 95: 406–421. 

Korjenkov, A.M., Mamyrov, E., Omuraliev, M., Kovalenko, 

V.A. & S.F. Usmanov, 2004. Rock Avalanches and 

Landslides Formed in Result of Strong Suusamyr (1992, 

M=7,4) Earthquake in the Northern Tien Shan. Test 

Structures for Mapping of Paleoseismic Deformations by 

Satellite Images. In High Mountain Remote Sensing 

Cartography VII (HMRSC VII), M. F. Buchroithner (ed.), 

Institute for Cartography of the Dresden University of 

Technology. Kartographische Bausteine, Band 23, 

Dresden, 19 p. 

Leonov, N.N., (1960). The Khait, 1949 earthquake and 

geological conditions of its origin. In Proc. of Academy of 

Sciences of the USSR, Geophysical Series, 3: 409-424 

(in Russian). 

Nadim, F., Kjekstad, O., Peduzzi, P., Herold, C. & C. 

Jaedicke, (2006). Global landslide and avalanche 

hotspots. Landslides, 3: 159–173. 

Petley, D., Dunning, S., Rosser, N. & A.B. Kausar, (2006). 

Incipient Landslides in the Jhelum Valley, Pakistan 

Following the 8th October 2005 Earthquake. Disaster 

Mitigation of Debris Flows, Slope Failures and 

Landslides, Universal Academy Press Inc./Tokyo, Japan: 

47–55. 

Sassa, K., (1996). Prediction of earthquake induced 

landslides. In Proc. Landslides, Senneset (ed.), 115-131. 

Sato, H.P., Hasegawa, H., Fujiwara, S., Tobita, M., Koarai, 

M., Une, H. & J. IWAHASHI, (2007). Interpretation of 

landslide distribution triggered by the 2005 Northern 

Pakistan earthquake using SPOT5 imagery. Landslides, 

4:113–122. 

Schuster, R.L., L.M. Highland, (2001). Socioeconomic and 

environmental impacts of landslides in the western 

hemisphere. USGS Open-File Report 2001-0276. 

Schuster, R.L., D. ALFORD, (2004). Usoi Landslide Dam 

and Lake Sarez, Pamir Mountains, Tajikistan. Environmental 

& Engineering Geoscience, X, 2: 151–168. 

Yin, Y., Wang, F. & P. Sun, (2009). Landslide hazards 

triggered by the 2008 Wenchuan earthquake, Sichuan, 

China. Landslides. DOI 10.1007/s10346-009-0148-5. 

Zhang, Z., (1995). Geological Disasters in Loess Areas 

during the 1920 Haiyuan Earthquake, China. GeoJournal, 

36.2, 3: 269-274. 

Zhang, D., Wang, G., (2007). Study of the 1920 Haiyuan 

earthquake-induced landslides in loess (China). 

Engineering Geology, 94: 76–88. 

80


EARTHQUAKE 

ARCHAEOLOGY 



A CASE STUDY OF EARTHQUAKES AND ROCKFALL - 

INDUCED DAMAGE TO A ROMAN MAUSOLEUM IN 

PINARA, SW TURKEY 

Hinzen, Klaus-G. (1, Kehmeier, Helen (1), Schreiber, Stephan (1), Reamer, Sharon K. (1) 

(1) Earthquake Geology and Archaeoseismology Group, Cologne University, Vinzenz-Pallotti-Str. 26, 51429 Bergisch Gladbach. 

GERMANY. Email hinzen@uni-koeln.de 

Abstract (A Case Study of Earthquakes and Rockfall): A Roman mausoleum located in the ancient city of Pnara, southwest 

Turkey, shows clear signs of damages due to dynamic loading. Considering the seismotectonic potential of the area, earthquake 

ground motions are a possible cause of the damages. However, the building is located at the foot of a 90 m high cliff with a 

significant rock fall hazard. We present a 3D discrete element model of the mausoleum based on a 3D laser scan. The range of 

impact velocities of blocks of different size and form and the actual slope of the cliff have been incorporated into 2D model 

calculations. The deformations caused by simulated rock impacts are compared to the actual displacements of blocks quantified 

from the laser scan. In addition, analytic ground motions signals are used to study the principal reaction of the building. The 

second damage scenario using earthquake strong ground motions shows that the damage is more likely caused by an earthquake 

than impacting rocks. 

Key words: Quantitative Methods, Rockfall, Strong Ground Motion, Archaeoseismology. 

INTRODUCTION 

A Roman mausoleum (Fig. 1) in the southwestern 

Turkey in the ancient city Pnara shows clear signs of 

dynamic loading that deformed the building. Its 

simple block structure and the fact that most parts of 

the building are still standing makes it an excellent 

test case for a quantitative archaeoseismic analysis. 

While collapsed structures usually allow at most an 

estimate of a minimum ground motion threshold, 

deformed but standing structures offer a deeper 

insight into the causes of the deformations. 

(A) 

(C) 

(B) 

(D) 

The suitability of the ancient city of Pnara in SW 

Turkey for archaeoseismic studies has been proven 

in several previous studies (Sintubin et al., 2003; 

Yerli et al., 2010; Yerli et al. 2009). Yerli et al. (2011) 

and Hinzen et al. (2010) provide detailed information 

about the seismicity and tectonic setting of the site. 

The pronounced topography gives Pnara an unique 

character, but also introduces a rockfall risk for 

several exposed structures, including the Roman 

mausoleum. 

ROMAN MAUSOLEUM 

(E) 

(F) 

Laserscan Model 

The mausoleum was surveyed with a 3D phaselaserscanner 

(Fleischer et al. 2010; Schreiber et al., 

2010, Schreiber et al. 2009); nine individual scans 

(from in- and outside positions) were combined into a 

virtual model of 79 million 3D points (Fig. 1B-D). This 

model was used as the basis for a discrete element 

model (Fig. 1E and F) of the undamaged mausoleum 

and also to quantify the deformation of the still 

standing part of the structure. 

Fig. 1: (A) Photo with a view from SE of the Roman 

mausoleum in Pnara; (B) laserscan from the same 

perspective; (C) and (D) ortho-views of the scan of 

the west and south wall, respectively; (E) and (F) 

wireframe and rendered view of the discrete element 

model. 

81 

A clear increase of the amount of deformation from 

bottom to top rows exists. Some parts of the sill and 

the pediment have fallen down. This also applies to 

the columns that are no longer found in situ. A part of 

the ceiling of the colonnade fell down and broke a


EARTHQUAKE 

ARCHAEOLOGY 



massive block of the front platform (Fig. 1 A and B). 

No fallen blocks are found inside the structure. On 

the western (back-) side of the structure the fallen 

pediment blocks are located in close proximity to the 

wall. 

Discrete Element Model 

Descriptions by Vitruvius (1796) and intact examples 

of a Roman mausoleum were used to reconstruct the 

missing front section of the Pnara mausoleum 

(shown in dark colour in Fig. 1F). The size of each 

construction block was measured from the 3D scan 

and transferred into a model of discrete rigid blocks 

(Fig. 1E). The model contains 180 blocks and with a 

density of the local conglomerate of 2.87 Mgm -3 it 

has a total mass of 1.8x10 5 kg. 

Rockfall 

2D Cliff Model 

A 2D model of the cliff and the slope on which the 

mausoleum is built was used to estimate impact velocities 

of rockfall material of different size and form. 

Boulders currently resting on the slope with sizes of 

several cubic meters and fresh fracture faces on the 

cliff indicate the persisting rockfall hazard at the site. 

Figure 2 shows the distribution of impact velocities of 

falling material on the mausoleum summarizing the 

results from numerous simulation calculations. 

Impact to Mausoleum 

Scenarios with rocks measuring 0.5 to 2.0 m with 

impact velocities between 10 m/s and 35 m/s and 

impacting the mausoleum at different height levels 

were tested. 

Velocity (m/s) 

Elevation a.s.l. (m) 

50 

40 

30 

20 

10 

0 

600 

550 

500 

450 

400 

Blocksize (m) 

0 1 2 3 4 5 

350 

200 250 300 350 400 450 500 

Distance (m) 

160 

120 

80 

40 

0 

Velocity (km/h) 

rows of blocks; however the rest of the structure is 

not significantly deformed (Fig. 3). 

30 m/s 

G 

F 

E 

D 

C 

B 

A 

Z 

Y 

X 

W 

log 10 Displacement (m) 

1 

0 

-1 

-2 

-3 

-4 

-5 

X Z B D F 

Layer 

Fig. 3: (left) View of the mausoleum from south after 

the impact of a rock with 1 m edge length and 30 m/s 

impact velocity. Block layers of the structure are 

labelled with capital letters. (right): Boxplots of the 

distribution of the log of displacements of block corners 

in five layers. The upper boxes (hatched downwards 

and blue) show the displacements measured 

from the laserscan of the building; the lower boxes 

(hatched ascending and red) are the displacements 

after the rock impact. 

Fig. 4: Reaction of the Roman mausoleum to an 

analytic ground motion signal, a Morlet wavelet with 

1 Hz main frequency and 10 s duration. 

GROUND MOTION 

Analytic Signals 

Before using full 3D earthquake ground motions, a 

series of tests was carried out with analytic signals in 

the form of Morlet wavelets (Goupillaud et al., 1984). 

The colonnade of the structure is highly vulnerable to 

ground motion frequencies around 1 Hz (Fig. 4). At 

signals with main frequencies of 2 Hz and above, the 

typical corner expulsions and block shifts are 

observed. 

Synthetic Seismograms 

Site-specific Green’s functions and an arbitrary number 

of rectangular dislocation planes were used to 

calculate synthetic seismograms using the method 

described by Wang (1999). A model of the local active 

faults is based on the work by Yerli et al. (2011). 

Fig. 2: (top) Red dots give median values of impact 

velocities of differently sized blocks at the location of 

the Roman mausoleum. (bottom) 2D model of the 

slope with trajectories of rocks of different size and 

form falling from the top of the cliff; a photo of which is 

shown in the insert. 

Impacting rocks cause localized damage on the west 

wall of the structure. With rocks of 1 m size at velocities 

above 20 m/s blocks are being pushed inside 

the building. Above velocities of 30 m/s strong 

damage occurs if the rock impacts into the upper 

0 km 5 km 10 km 

S3 

S7 

30.5°N 

S6 

S8S9 

S2 

S5 

S10 

S4 

S1 

Fig. 5: Active faults in 

the vicinity of Pnara. 

Fault segments for 

different earthquake 

scenarios are labelled 

S1 to S10 (map 

from Yerli et al., 

2011). 

29.3°E 

82


EARTHQUAKE 

ARCHAEOLOGY 



Impact to Mausoleum 

An example of the deformation from a synthetic 

earthquake record is shown in Figure 6. In this 

scenario activation of segments S1, S10, S9, and S3 

was assumed. The M W 6.0 example earthquake did 

not topple the colonnade; however, the calculated 

displacements are very close to the actual ones with 

the exception of an underestimated F-layer. Test 

calculations with measured ground motions from the 

2009 L’Aquila earthquake did also destroy the 

colonnade. As part of the ongoing work we will 

1 

0 

Modeling of rock fall impacts and diverse earthquake 

scenarios indicate that the damages and displacements 

of building blocks of the Pnara mausoleum 

are more likely the result of an earthquake than being 

formed by rockfall. So far only a single earthquake 

has been used for the modeling. Further calculations 

are planned to test for the possibility of repeated 

earthquake action over the millennia since the 

construction of the mausoleum. 

Acknowledgements: We are grateful for the help during 

the field work by Barish Yerli and Claus Fleischer. Part of 

the work was financially supported by Deutsche Forschungsgemeinschaft 

(DFG HI660/2-1). 

G 

F 

E 

D 

C 

B 

A 

Z 

Y 

X 

W 

log 10 Displacement (m) 

-1 

-2 

-3 

-4 

-5 

0 4 8 12 16 20 

T m e s 

X Z B D F 

Layer 

Fig. 6: Same layout as Fig. 3 for deformations of a 

synthetic earthquake with the ground motions shown 

in the insert; upper trace EW component lower trace 

NS component of ground motion. 

further quantify the ground motion parameters which 

caused the deformations. 

CONCLUSIONS 

The mausoleum located above the Roman forum of 

the ancient city of Pnara is heavily damaged; 

however, most of its simple block structure is still 

standing. Deformations indicate that the building suffered 

from dynamic loading and the location suggests 

rockfall and earthquake ground motion as possible 

d ( m ) 

0.12 

0.08 

0.04 

-0.04 

0 

References 

Fleischer, C., K.-G. Hinzen & S. Schreiber (2010). Laser 

scanning eines roemischen Brunnens in der 

Archaeologischen Zone Koeln. Allgemeine Vermessungs-Nachrichten 

5/2010, 176-180. 

Goupillaud, P., A. Grossman & J. Morlet (1984). Cycle and 

related transforms in seismic signal analysis. 

Geoexploration 23, 85-102. 

Hinzen, K.-G., S. Schreiber & B.Yerli, (2010). The Lycian 

Sarcophagus of Arttumpara, Pnara (Turkey) -Testing 

Seismogenic and Anthropogenic Damage Scenarios. 

Bulletin of the Seismological Society of America 100 (6), 

3148-3164. 

Schreiber, S., K. G. Hinzen & C. Fleischer (2010). Closerange 

Laserscanning in the Archaeological Zone 

Cologne, Germany, Workshop on Remote Sensing 

Methods for Change Detection and Process Modelling, 

Cologne, November 18-19 2010. 

Schreiber, S., K. G. Hinzen & C. Fleischer (2009). An 

application of 3D laser scanning in archaeology and 

archaeoseismology: The medieval cesspit in the 

archaeological zone Cologne, Germany, 1st INQUA- 

IGCP-567 International Workshop on Earthquake 

Archaeology and Palaeoseismology, Baelo Claudia, 

Cadiz, Spain, 7-13 September 2009, 136-138. 

causes. Sintubin, M., P. Muchez, D. Similox-Tohon, G. Verhaert, E. 

Paulissen & M. Waelkens (2003). Seismic catastrophes 

Numeric modelling of rock fall impact indicates a 

different damage pattern from that one observed in 

the field. Impacting rocks tend to heavily deform the 

western wall facing the cliff; however, the rest of the 

building, especially block layers below the impact, 

are only insignificantly affected. Rocks with high 

impact velocities push blocks of the building to the 

inside; however, all fallen blocks from the top of the 

building were found outside of the structure. 

Tests of the dynamic reaction of the mausoleum with 

analytic ground motion signals show a high vulnerability 

of the colonnade to horizontal movements 

with frequencies around 1 Hz. Ground motions with 

frequencies of 2 Hz and above initiate rocking of 

blocks with an increasing tendency from bottom to 

top. Morlet wavelets with main frequencies of 2 Hz 

dislocate the blocks in the same pattern as it is 

observed today. Synthetic ground motions of local 

earthquakes with M W 5.6 to 6.0 also produce 

displacements similar to the observations. 

at the ancient city of Sagalassos (SW Turkey) and their 

implications for seismotectonics in the Burdur-Isparta 

area. Geological Journal 38 (3-4), 359-374.Vitruvius 

Pollio, M. & A. Rode (1796). Baukunst. Translated by 

August Rode, G. J. Goeschen (eds.), Leipzig, 10 Books. 

Wang, R. (1999). A simple orthonormalization method for 

stable and efficient computation of Green's functions. 

Bulletin of the Seismological Society of America 89, 733- 

741. 

Yerli, B., S. Schreiber & K.-G. Hinzen (2011). Seismotectonic 

background for archaeoseismologic studies in 

the Een Basin, SW Turkey. Natural Hazards, submitted. 

Yerli, B., J. Ten Veen, M. Sintubin, C. Karabacak, C. 

Yaliciner & E. Altunel (2010). Assessment of seismically 

induced damage using LIDAR: the ancient city of Pinara 

(SW Turkey) as a case study. In: Ancient Earthquakes. 

Geological Society of America, Special Paper 471, M. 

Sintubin, I. Stewart, T. M. Niemi and E. Altunel (eds.), 

157-170. 

Yerli, B., S. Schreiber, K. G. Hinzen, J. H. t. Veen, M. 

Sintubin & M. Sintubin (2009). Testing the Hypothesis of 

earthquake-related damage in structures in the Lycian 

ancient City of Pinara, SW Turkey, 1st INQUA-IGCP-567 

International Workshop on Earthquake Archaeology and 

Palaeoseismology, Baelo Claudia, Cadiz, Spain, 7-13 

September 2009, 173-176. 

83


EARTHQUAKE 

ARCHAEOLOGY 



NEOTECTONICS OF GRACIOSA ISLAND (AZORES) – UNCERTAINTY IN SEISMIC 

HAZARD ASSESSMENT IN A VOLCANIC AREA WITH VARIABLE SLIP-RATES 

Hipólito, Ana (1, José Madeira (2), Rita Carmo (1), João Luís Gaspar (1) 

(1) Centro de Vulcanologia e Avaliação de Riscos Geológicos, Universidade dos Açores, Complexo Científico, 3º Piso – Ala Sul, 

Rua da Mãe de Deus, 9500-321 Ponta Delgada, Açores, PORTUGAL. Email: ana.rc.hipolito@azores.gov.pt 

(2) Universidade de Lisboa, GeoFCUL, LATTEX/IDL-Laboratório Associado, Faculdade de Ciências das Universidade de 

Lisboa, Edifico C6, Campo Grande, 1749-016 Lisboa, PORTUGAL.Email: jmadeira@fc.ul.pt 

Abstract (Neotectonics of Graciosa Island (Azores) – uncertainty in seismic hazard assessment in a 

volcanic area with variable slip-rates): Graciosa is a mid-Pleistocene to Holocene volcanic island that lies on a 

complex plate boundary between the North American, Eurasian and Nubian plates. Large fault scarps displace the 

oldest volcanic units, but in the younger areas recent volcanism hides the surface expression of faulting, limiting 

neotectonic observations. Slip-rates deduced from neotectonic surveys are higher than those provided by kinematic 

plate motion models. This suggests a variability of deformation rates, alternating between high tectonic deformation 

periods, decreasing the recurrence interval of surface rupturing earthquakes, and phases of low slip-rate. 

Nevertheless, in historical time a few destructive earthquakes affected the island attesting for its seismic hazard. 

Key words: Neotectonics, Azores, Graciosa Island 

GEODYNAMIC AND VOLCANIC SETTING 

The Azores archipelago lies on a complex 

geodynamic setting: the Eurasian (Eu), North 

American (NA) and Nubian (Nu) triple junction 

(Azores Triple Junction - ATJ) (Fig.1). 

Graciosa Island is located on the west segment of 

the Eu-Nu boundary, a diffuse and complex 

deformation zone (Fig.1) sheared by a dextral 

transtensile regime (e.g. Madeira & Brum da Silveira, 

2003; Carmo, 2004; Hipólito, 2009). 

Fig. 1: Location of the Azores and main 

morphotectonic features of the region. The shaded 

area represents the sheared western segment of the 

Eu-Nu plate boundary; Plates: NA – North American; 

Eu – Eurasian; Nu – Nubian; Tectonic structures: 

MAR – Mid-Atlantic Ridge; EAFZ – East Azores 

Fracture Zone; GF – Gloria Fault; Islands: G – 

Graciosa Island. World topography and bathymetry 

from ESRI ® (2008), Azores bathymetry adapted from 

Lourenço et al. (1997). Datum: WGS 1984 (modified 

from Hipólito, 2009). 

84 

Fig. 2: Seismicity in and around Graciosa island from 

2003 to 2009, defining three alignments in West 

Graciosa Basin and one at the western border of the 

East Graciosa Basin. Red dots mark earthquake 

epicenters and magnitudes (M D ). Data from CIVISA 

(Centro de Informação e Vigilância Sismovulcânica 

dos Açores), 2009. 

This boundary acts as an ultra-slow oblique 

expansive center (Vogt and Jung, 2004) and a 

transfer zone accommodating the differential motion 

between Eu and Nu plates, due to the higher


EARTHQUAKE 

ARCHAEOLOGY 



spreading rates north of the Azores (e.g. DeMets, 

1994; Fernandes et al., 2003). 

Due to its geodynamic setting, the Azores 

archipelago presents frequent seismicity. Graciosa 

Island was affected by some significant earthquakes 

since its settlement in mid 15 th century. One of them 

(1837) had probably its epicenter on land (Madeira, 

1998; Silva, 2005). However, instrumental seismicity 

does not show significance seismic activity within the 

island (Fig.2). The epicenter distribution of the 

current activity reveals four NNW-SSE trending 

offshore alignments, one on the east flank of West 

Graciosa Basin, two crossing this basin floor to the 

west, and another alignment just east of the island 

(Fig. 2). 

STRUCTURAL DATA 

Geometric and kinematic fault analysis 

In Graciosa Island, neotectonic studies are limited by 

the absence of outcrops with well exposed fault 

planes. Tectonic features, corresponding to important 

fault scarps, several tens to hundreds of meters high, 

occur in the central region of the island. Unfortunately 

the faults producing these large features do not crop 

out and the small size of the island limits the 

observation of the full length of the faults. The thick 

and non-cohesive young volcanic fall deposits and 

the lava flows that mantle the topography were not 

tectonically displaced yet, hiding the trace of the 

faults in areas covered by recent volcanic units. 

Therefore, paleoseismological studies were not made 

due the sheer size of the fault scarps and the 

absence of recent surface faulting with favorable 

conditions for trenching. The observed faults crop out 

either in quarry walls exploring cinder cones or in 

inaccessible sea cliffs. In the first case the nature of 

the deposits (homogeneous, coarse size and low 

cohesion) does not allow the generation of 

slickensides, hindering kinematic analyses. 

Fig. 3: Digital Elevation Model of Graciosa and the main 

morphologic regions. Based on Carta Militar de Portugal, 

Sheet 21 – Stª Cruz da Graciosa, Instituto Geográfico do 

Exército (2001); UTM Projection; Datum: WGS 1984. 

Graciosa Island comprises one quiescent trachytic 

polygenetic volcano (the Caldera Volcano), an older 

volcanic complex (the Central-Southern Complex), 

resulting from the dismantling of an important central 

volcano, and several monogenetic eruptive centers 

included in the NW basaltic Platform that mantle the 

older units (Gaspar, 1996; Fig.3). The oldest 

volcanic-stratigraphic unit is around 620 ± 120 ka old 

(Féraud et al., 1980). The most recent volcanic event 

was a pre-settlement basaltic hawaiian-strombolian 

eruption at about 2 ka B.P. (Walker, unpublished 

data, in Gaspar, 1996). Currently the volcanic activity 

is just expressed by secondary manifestations, 

namely by thermal springs, fumarolic fields and 

diffuse degassing (e.g. Ferreira et al., 1993; Gaspar, 

1996). 

The island presents in general a smooth relief with 

maximum altitude of 402m (Fig.3). The central part is 

crossed by several NW-SE trending fault scarps 

parallel to the shape of the island (Fig.4). Those 

faults define a graben structure that is crossed in the 

SE by a NNE-SSW fault scarp which separates the 

older from the most recent volcanic units to the 

south. 

Fig. 4: Morphotectonic map of Graciosa Island. Based on 

Carta Militar de Portugal, Sheet 21 – Stª Cruz da Graciosa, 

Instituto Geográfico do Exército (2001); UTM Projection; 

Datum: WGS 1984. 

Generally, the mapped structures are normal faults or 

present normal component. Although, in most cases 

it was difficult to recognize a strike-slip component, 

these structures may also have strike-slip component 

(dextral or sinistral) typical of a tectonic transtensile 

regime. 

85


EARTHQUAKE 

ARCHAEOLOGY 



The tectonic map (Fig.4; Gaspar and Queiroz, 1995; 

Hipólito, 2009), the geometric (Fig.5) and kinematic 

(Fig.6) fault analysis (Hipólito, 2009) show the 

presence of two main fault systems: system A – 

composed of two sets of conjugated faults, one 

trending NW-SE, dipping to SW, presenting normaldextral 

or dextral-normal oblique slip, and another 

striking NNE-SSW, dipping to ESE, with oblique 

normal-left lateral or left lateral-normal slip (Fig. 6a); 

system B – includes NNE-SSW to NE-SW trending 

faults, dipping to WNW or NW, presenting normaldextral 

or dextral-normal oblique slip (Fig. 6b). A 

family of faults conjugated with these structures was 

not found. The strong fault dips (the 80º to 90º range 

dominate) also suggest oblique (normal/strike-slip) 

faulting (Fig.5). 

Slip-rates 

Only three tectonic structures allowed slip-rate 

estimation: the Saúde-Serra das Fontes-Hortelã, the 

South Serra das Fontes and the East Serra das 

Fontes faults (Fig. 4). However, in two of them only 

the normal slip component was determined, while in 

the other only the strike-slip could be measured 

(Table I). The Saúde-Serra das Fontes-Hortelã Fault 

trace is marked by a 5750m-long south–facing scarp 

trending N306º to N328º producing 49m of dextral 

strike-slip displacement. The South Serra das Fontes 

Fault has a 200m-high south-facing scarp 3500m 

long, trending N282º to N327º. The East Serra das 

Fontes Fault trace is marked by a 4750m-long and 

185m- maximum high east-southeast facing scarp, 

trending N20º. The total length of these faults is 

unknown because on-shore they are fossilized by 

younger units and there is no off-shore data allowing 

to trace them. 

The estimated values indicate slip-rates somewhat 

higher than those provided by plate motion kinematic 

models (Table II) and contrasting with the present 

low seismicity and neotectonic deformation in the 

island (Fig. 2). These slip-rates could be overestimated 

because the age of the used marker may 

not represent the whole deformation period. On the 

other hand, the real displacements may be greater 

than the measured ones due to infilling of tectonically 

depressed areas by younger deposits. 

Table I: Slip-rates of three faults in Graciosa Island. 

Fig. 5: Geometry of all analyzed faults: a) stereographic 

plot of fault planes (lower hemisphere; Schmidt net) – 

diagram; b) stereogram of fault poles density – 

diagram; c) Circular histogram of unweighted frequencies 

of fault plane directions and dip angles. TectonicsFP 

software (Ortner et al., 2002). 

Paleostress analysis suggests that the region is 

affected by two different stress fields that can 

alternate in time and/or in space; variations of the 

local stress field may occur, which are responsible for 

the generation of new faults or reactivation of preexisting 

structures (Hipólito, 2009). 

Table II: Relative velocities and azimuth directions of 

slip vector for Eu and Nu plates and for those plates 

relatively to NA plate in ATJ zone, according to the 

NUVEL-1A (DeMets et al., 1994), REVEL (Sella et al., 

2002) e DEOS2K (Fernandes et al., 2003) global 

kinematic models. 

Fig. 6: Stereographic plot of two main fault systems (lower 

hemisphere; Schmidt net – diagram). a: system A - NW- 

SE to NNW-SSE faults, with normal-dextral oblique slip, 

conjugate of NNE-SSW to NE-SW normal-left lateral 

structures; b: system B - NNE-SSW to NE-SW faults with 

normal-dextral oblique slip. Black arrows: strike-slip 

sense. TectonicsFP software (Ortner et al., 2002). 

DISCUSSION AND FINAL REMARKS 

As in other zones in the Azores (Hipólito et al., 2011), 

several limitations prevent a more detailed 

neotectonic survey of Graciosa Island. Recent 

volcanic deposits, that mantle the topography, hide 

86


EARTHQUAKE 

ARCHAEOLOGY 



the surface expression of major tectonic structures; 

basaltic lapilli deposits do not allow the generation of 

linear kinematic markers; there are no structures with 

scarp heights allowing paleoseismological studies, 

thus contributing to the assessment of seismic 

hazard. Nevertheless, the tectonic information from 

Graciosa Island is in agreement with the stress 

pattern proposed by other authors for the Azores 

region (e.g. Madeira, 1998, Lourenço et al., 1988; 

Carmo, 2004). 

The absence of seismic events producing surface 

rupture since settlement and the current low seismic 

activity in Graciosa Island, contrast with the youthful 

aspect of its tectonic morphology. The calculated 

slip-rates and the evident loss of geomorphic 

expression of fault scarps into the areas covered with 

the younger units (NW Platform and Caldeira 

Volcano) suggest that the occurrence of a period of 

important tectonic activity before 31 ka, with higher 

slip-rates than those observed in present times, 

responsible for the deformation of the central part of 

the island. That period was followed by a magmatic 

dominated phase, with the build-up of Caldeira 

Volcano and the installation of the fissural basaltic 

volcanism responsible for the formation of the NW 

Platform. During that period there was a reduction of 

fault slip-rates and consequent increase of the 

recurrence period of surface rupturing event. This is 

consistent with the occurrence of variations in 

deformation rates in the archipelago, with periods 

with slip-rates higher or lower than the average, as 

proposed by Madeira (1998). 

Thus, the current tectonics in Graciosa Island is not 

particularly active. Nevertheless, as it is located 

between two important seismogenic sources (West 

and East Graciosa Basins) and on a complex 

geodynamic setting, seismic hazard cannot be 

disregarded. 

Acknowledgements: Ana Hipólito and fieldwork were 

supported by CARIGE Project – Carta de Riscos 

Geológicos, Governo Regional da Região Autónoma dos 

Açores, Secretaria Regional da Habitação e Equipamentos. 

References 

Carmo, R. (2004). Geologia estrutural da região Povoação - 

Nordeste (ilha de S. Miguel, Açores). MSc thesis, Azores 

Univ., 133 p. 

DeMets, C., R. Gordon, D. Argus, S. Stein (1994). Effect of 

recent revisions to the geomagnetic reversal time scale 

on estimates of current plate motions. Geophys. Res. 

Lett. 21(20), pp. 2191-2194. 

Féraud, G., I. Kaneoka, C.J. Allegre (1980). K-Ar Ages and 

Stress Pattern in the Azores - Geodynamic Implications. 

Earth and Planet. Sci Lett. 46(2), pp. 275-286. 

Fernandes, R. M. S., B.A.C. Ambrosius, R. Noomen, L. 

Bastos, M.J.R. Wortel, W. Spakman, R. Govers (2003). 

The relative motion between Africa and Eurasia as 

derived from ITRF2000 and GPS data, Geophys. Res. 

Lett. 30 (16), 1828, pp. 1-1 - 1-5. doi: 

10.1029/2003GL017089. 

Ferreira, T., J. L. Gaspar, G. Queiroz (1993). 

Considerações sobre as emanações gasosas da Furna 

do Enxofre (Ilha Graciosa, Açores). Açoreana VII (4), pp. 

603-612. 

Gaspar, J.L.; G. Queiroz (1995). Carta Vulcanológica dos 

Açores, ilha Graciosa. Sheets A and B, scale 1:10000. 

Ed. Azores Univ. and Câmara Municipal de Santa Cruz 

da Graciosa. 

Gaspar, J.L. (1996). Ilha Graciosa (Açores): História 

Vulcanológica e Avaliação do Hazard. PhD thesis, 

Azores Univ., 361p. 

Hipólito, A. (2009). Geologia Estrutural da ilha Graciosa – 

Enquadramento no âmbito da Junção Tripla dos Açores. 

MSc thesis, Azores Univ., 155 p. 

Hipólito, A., F. Viveiros, R. Carmo, C. Silva, N. Cabral, J. 

Cabral, V. Alfama, T. Ferreira, J.L. Gaspar (2011). Soil 

degassing surveys as a tool to identify hidden faults in 

volcanic areas: preliminary results at the Ribeira Grande 

graben (Fogo Volcano, S. Miguel Island, Azores). 

Geophysical Research abstract. 13, EGU2011-8992. 

Lourenço, N., J. Miranda, J. Luís, A. Ribeiro, L. Mendes- 

Victor, J. Madeira, H. Needham (1998). Morpho-tectonic 

analysis of the Azores Volcanic Plateau from a new 

bathymetric compilation of the area. Marine Geophys. 

Res. 20, pp. 141-156. 

Luís, J.F., J. M. Miranda (2008). Reevaluation of magnetic 

chrons in the North Atlantic between 35ºN and 47ºN: 

Implications for the formation of yhe Azores Triple 

Junction and associated plateau. J. of Geophys. Res. 

113, B1o1o5. doi:10.1029/2007Jb005573. 

Madeira, J. (1998). Estudos de neotectónica nas ilhas do 

Faial, Pico e S. Jorge: uma contribuição para o 

conhecimento geodinâmico da junção tripla dos Açores. 

PhD thesis, Lisbon Univ., 428 p. 

Madeira, J., A. Brum da Silveira (2003). Active tectonics 

and first paleoseismological results in Faial, Pico and S. 

Jorge Islands (Azores, Portugal). Annals of Geophysics 

46 (5), pp. 733-761. 

Ortner, H., F. Reiter, P. Acs (2002). Easy handling of 

tectonic data: the programs TectonicVB for Mac and 

TectonicsFP for Windows. Computers and Geosciences 

28, 1193-1200. 

Sella, G., T. H. Dixon, A. Mao (2002). REVEL: a model for 

recent plate velocities from space geodesy. J. Geophys. 

Res. 107 (B4), 2081, pp. 11-1 – 11-12. doi: 

10.1029/2000JB000033. 

Silva, M. (2005). Caracterização da sismiciade histórica dos 

Açores com base na reinterpretação de dados de 

macrossísmica: contribuição para a avaliação do risco 

sísmico nas ilhas do Grupo Central. MSc thesis. Azores 

Univ., 163 p. 

Vogt, P.R., W.Y. Jung (2004). The Terceira Rift as hyperslow, 

hotspot dominated oblique spreading axis: A 

comparison with other slow-spreading plate boundaries. 

Earth Planet. Sci. Lett. 218, pp. 77–90. 

doi:10.10116/S0012-821X(03)00627-7. 

87


EARTHQUAKE 

ARCHAEOLOGY 



EVIDENCE FOR HOLOCENE TSUNAMI-IMPACT ALONG THE SHORELINE OF OMAN 

Gösta Hoffmann (1, Klaus Reicherter (2), Thomas Wiatr (2), Christoph Grützner (2) 

(1) Department of Applied Geosciences, German University of Technology in Oman, PO Box 1816, Athaibah, Muscat, PC 

130, Sultanate of Oman, email: goesta.hoffmann@gutech.edu.om 

(2) RWTH Aachen University, Inst. für Neotektonik und Georisiken, Aachen, 52056, Germany 

Abstract (Evidence for Holocene tsunami-impact along the shoreline of Oman): Three independent sets of evidence of past 

tsunami along the coastline of Oman are reported. The rocky coastline of the Sultanate of Oman between Fins and Sur is 

decorated by a number of large boulders and boulder accumulations forming ramparts. The boulders occur as individual blocks of 

almost 50 tons weight, as imbricated sets and “boulder trains”. The coast is made up of folded Tertiary limestones and beach rock 

of Quaternary age. The transport distance from the fractured cliff front of 6-10 m height above mean sea level varies between 

several meters up to 70 m inland. We found individual blocks of recent corals and overturned blocks with oysters and pools. T- 

LIDAR was used to analyse geomorphologic features and for volumetric estimates of boulder weight. Tropical cyclones such as 

Gonu in 2007 or Phet in 2010 as well as historical tsunamis are known to have affected Oman’s coastline in the past. Coastal 

changes by cyclones are known to have been negligible; therefore, we interpret the boulder ridges as tsunamigenic deposits. 

Additionally, fine grained lagoonal sediments were analyzed. A distinct shell layer with allochthonous species is documented. A 

tsunamigenic origin is most likely. Although no dating evidence of the observed boulder and lagoon deposits is available at the 

moment we conclude that the 1945 Makran tsunami affected Oman’s coastline. This conclusion is based on interviews with local 

people. 

Key words: Oman, tsunami, boulder deposits, T-LiDAR 

INTRODUCTION 

Recent tsunami events like the Indian Ocean tsunami 

on 26th December 2004 and the Tôhoku earthquake 

and tsunami on 11th March 2011 resulted in large 

numbers of casualties and immense damage to 

infrastructure. These events underline the need for 

tsunami hazard research and assessment for any 

potentially vulnerable region. In most cases this can 

only be done by studying past tsunami records. The 

coastlines of the Sultanate of Oman are prone to 

various natural hazards such as tropical cyclones, 

landslides and tsunamis. The devastating effects of 

the cyclone Gonu, caused by flash floods and 

landslides in June 2007 illustrated the need to 

investigate the recurrence intervals of such events in 

order to assess the vulnerability and to mitigate 

damages. So far no scientific research concerning 

recurrence intervals of natural hazards has been 

carried out. However, studies published by 

Heidarzadeh et al. (2008a, 2008b, 2009) and Jordan 

(2008) reveal past tsunami events in the Indian 

Ocean with possible effects on the coastline of Oman 

(Fig. 1). As the population of Oman and the 

neighboring countries is concentrated along the 

coastline and large infrastructure projects are 

planned or already completed a holistic scientific 

approach to decipher the geological record of past 

extreme events is overdue. On 27 th November 1945 

an earthquake occurred in the Makran Subduction 

Zone offshore Pakistan and triggered a tsunami 

(Jaiswal et al., 2009). Up to 4000 casualties were 

reported along the coastlines of NW India and 

Pakistan, including 5 m run-up along the coastlines of 

the Sultanate of Oman. Donato et al. (2008, 2009) 

analysed shallow sediment cores from the lagoon in 

Sur and recorded a 5-25 cm thick shell bed close to 

the surface. Based on the taphonomy and 

fragmentation a tsunamigenic origin is discussed as 

the most likely form of deposition related to the 1945 

tsunami. However, there are almost no historical 

documents available for Oman for this period as the 

country was isolated with no international contacts 

until the 1970s, living conditions were poor and no 

modern technology was in use. We report geological 

and historical evidence for the tsunami along Oman's 

coastline. These evidence are: (a) fine grained 

lagoon sediments, which show distinct layers with 

allochthonous, offshore species (mollusks and 

foraminifera); (b) boulder deposits encountered along 

cliff-coastlines and (c) eyewitness-reports of old 

people we interviewed. 

OBSERVATIONS 

The coastal area under investigation is situated in the 

eastern part of Oman between the cities Quariat and 

Sur. The area is sparsely populated as most of the 

country; small fishing villages are scattered along the 

coast. Only since 2008 there is a paved road 

connecting the cities Quariat in the north and Sur in 

the south. 

The geology of the area is dominated by Paleogene 

to Neogene limestone formations, which rise from the 

coast up to 1500 m to form the Selma Plateau. 

Geomorphologic evidence of Quaternary land-uplift is 

obvious along the entire coastline: coast-parallel, 

wave-cut terraces are encountered up to elevations 

of 300 m. Within the study area these terraces are 

88


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 1: Historical tsunamis in the Indian Ocean and working area in the Sultanate of Oman. 

cut into Paleocene-Early Eocene limestone 

formations. Quaternary deposits are either of fluvial 

origin or ancient to subrecent littoral deposits, usually 

preserved as beachrock. In most cases only 

erosional remnants of the beachrock are found and 

the underlying older strata dominate along the cliff 

coast. Several intertidal lagoons exist in the vicinity of 

Sur and Ras al Hadd. These lagoons serve as 

geological archives with a preservation potential for 

palaeo-tsunami and were investigated during several 

field campaigns in 2010. 

We collected seven sediment cores at various 

locations within Sur lagoon. The deepest core 

reaches 10 m below the present surface. The 

sequence is characterized by silty fine sand in the 

lower part (10-6 m) and fine-sand in the upper part (6 

– 0 m). Within the uppermost meter several distinct 

shell beds were identified. The shell and foraminifera 

assemblage contains allochthonous species living in 

the subtidal zone and offshore. Additionally, we 

collected 4 sediment cores in the lagoon of Ras al 

Hadd. The longest core is 3 m long. The base of the 

sequence is made up of sandy gravel partly 

Fig. 2: Study area along the east coast of the Sultanate of Oman. Inset shows bathymetric sections. 

89


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 3: Study area along the east coast of the Sultanate of Oman with the observed ramparts. 

cemented as beachrock, overlain by fine to medium 

sand. The mollusk and foraminifera assemblages in 

the upper 100 cm show a variety of species. 

Especially some bivalve species within this layer are 

allochthonous as their habitat is characterized as 

subtidal and offshore. 

In our 2011 survey we found boulder deposits south 

of the village of Fins (Fig. 2) and more distal finergrained 

deposits yielding shells and coral blocks on 

the cliff top and to approx. 60-70 m inland (Fig. 3). 

The blocks are partly imbricated and reach volumes 

of more than 20 m 3 (determined with terrestrial LiDAR 

scanning), corresponding to a weight of almost 50 

tons. Also, we found so called tsunami boulder trains, 

where blocks are aligned in a row (Fig. 3). Some 

blocks are toppled or upright with hit marks on the 

surface, erosional pot holes, Lithophaga borings and 

attached oysters (which provide dating material, 

dating is in progress). The boulders form ramparts 

and have a wavy, lobe-like pattern. Most blocks have 

a platy shape, which origins in layer thickness of 

reworked material, mainly beach rock and Tertiary 

limestones. We measured the long axis (a-axis) of 60 

boulders, a vast majority is oriented N30, possibly 

pointing towards the wave direction. The cliff tops are 

“cleaned”, however, drift wood of the tropical cyclone 

Phet in 2010 is found in height of approximately 6 m 

above mean sea level. Inland, boulders are found in 

a gravel/sand matrix with various fossil remains like 

shell and corals (Fig. 4). The finer-grained layers 

show fining-up cycles. We also interviewed old 

people living in the towns of Sur and Tiwi. An old 

man in Sur recalled an event that happened most 

probably during the 1940s: first the sea retreated, 

then, two waves washed onshore. The event took 

place at 02:00 am. 

Another old man of the village of Tiwi reported from 

hearsay, as he was born in 1946. He heard about an 

event that destroyed the local graveyard in the 

1940s. He described that the graveyard was located 

much further inland. Furthermore, he gave an 

account of fish (sardines) and mollusks (oysters) 

being washed into the Wadi Shab. The women who 

used to get freshwater from the wadi could not walk 

there anymore, but boats had to be used. The marine 

fishes lived in the wadi after the event. 

DISCUSSION 

The uppermost ~1m in the lagoon of Sur as well as in 

the lagoon of Ras al Hadd clearly indicates an eventlayer 

which can be either storm- or tsunamigenerated 

(see Kortekaas and Dawson 2007). As the 

lagoons are intertidal, reworking and bioturbation of 

the sediments is a common phenomenon that 

hampers a clear stratification. Boulders deposits 

along the east coast of the Sultanate of Oman form 

ramparts between Fins and Sur. Inland, boulder 

deposits are incorporated in finer-grained sediments. 

Observations of the coastline changes and 

modifications of the last two very strong tropical 

cyclones Gonu and Phet rather exclude tropical 

cyclones a “moving agent” for the large 50 tons 

boulders. This was also proven by comparing time 

series of Google Earth, where blocks are detectable, 

but remained in the position (before and after). 

90


EARTHQUAKE 

ARCHAEOLOGY 



The information gained from the interview in Sur is 

very helpful. The timing of an event resulting in seawater 

entering the house was given with 02:00 am 

which is 22.00 GMT (local time zone is GMT +4). 

This fits quite well with the time of the Makran 

earthquake on 27.11.1945 which is reported by 

Pendse (1948) as 21:57 GMT. The modeled travel 

time of the tsunami wave is 20-30 minutes 

(Heidarzadeh and Kijko. 2011). The description of a 

retreating sea fits general tsunami descriptions. The 

arrival of two separate waves fits observations in 

India (see Rajendran et al. 2008) where also two 

waves with disparity in arrival time are reported. A 

submarine landslide is assumed for the second 

wave. The descriptions given in the interview in Tiwi 

cannot unambiguously be related to any know event. 

However, the accounts are more likely to be the 

effect of a tsunami wave rather than wadi-flooding. 

Acknowledgements: This study was financially supported 

by the German Research Foundation (DFG-project 

Re 1361/14-1) and GUTech in Mascat. Our students Sina 

Rausch, Kathrin Wagner and Tobias Rausch are thanked 

for field and lab support, and Tobias also for graphic design. 

References 

Fig. 4: Sketch of the possible tsunamigenic 

deposits. 

Hence, we propose a tsunamigenic event being 

responsible for the rampart formation and the boulder 

deposits along this part of the coast. Dating of 

oysters, which have grown on the blocks and died 

during the deposition, is in progress. 

Donato, S. V., Reinhardt, E.G., Boyce, J.I., Rothaus, R., 

Vosmer, T. (2008). Identifying tsunami deposits using 

bivalve shell taphonomy, Geology, 36, 199-202. 

Donato, S. V., Reinhardt, E. G., Boyce, J. I., Pilarczyk, J. E., 

Jupp, B. P. (2009) Particle-size distribution of inferred 

tsunami deposits in Sur lagoon, Sultanate of Oman, Mar. 

Geol., 257, 54-64. 

Heidarzadeh, M., Pirooz, M., Zaker, N. H., Yalciner, A. C., 

Mokhtari, M., Esmaeily, A. (2008a) Historical tsunami in 

the Makran subduction zone off the southern coasts of 

Iran and Pakistan and results of numerical modeling, 

Ocean Eng., 35, 774-786. 

Heidarzadeh, M., Pirooz, M., Zaker, N. H., Synolakis, C. 

(2008b). Evaluating tsunami hazard in the northwestern 

Indian Ocean, Pure Appl. Geophys., 165, 2045-2058, 

Doi:10.1007/s00024-008-0415-8. 

Heidarzadeh, M., Pirooz, M., Zaker, N. H., Yalciner, A. C. 

(2009). Preliminary estimation of the tsunami hazards 

associated with the Makran subduction zone at the 

northwestern Indian Ocean, Nat. Hazards, 48, 229-243, 

Doi:10.1007/s11069-008-9259-x. 

Heidarzadeh, M., Kijko, A. (2011). A probabilistic tsunami 

hazard assessment for the Makran subduction zone at the 

northwestern Indian Ocean, Nat. Hazards 56, 577-593. 

Jaiswal, R., Singh, A., Rastogi, B. (2009). Simulation of the 

Arabian Sea tsunami propagation generated due to 1945 

Makran earthquake and its effect on western parts of 

Gujarat (India), Nat. Hazards, 48, 245-258, 

Doi:10.1007/s11069-008-9261-3. 

Jordan, B. R. (2008). Tsunamis of the Arabian Peninsula - a 

guide of historic events, Science of Tsunami Hazards, 27, 

31-46. 

Kortekaas, S., Dawson, A. G. (2007). Distinguishing 

tsunami and storm deposits: An example from Martinhal, 

SW Portugal, Sed. Geol. 200, 208-221. 

Pendse, C. G. (1948). The Makran earthquake of the 28th 

of November, 1945, Scientific Notes, Indian 

Meteorological Dept. 10, 141-145. 

Rajendran, C. P., Ramanamurthy, M.V., Reddy, N.T., 

Rajendran, K. (2008). Hazard implications of the late 

arrival of the 1945 Makran tsunami, Current Science 95, 

1739-1743. 

91


EARTHQUAKE 

ARCHAEOLOGY 



MONITORING COASTAL CHANGES ON THE IONIAN ISLANDS (NW-GREECE) 

BY MULTI-TEMPORAL TERRESTRIAL LASER SCANNING 

Hoffmeister, Dirk (1, Ntageretzis, Konstantin (2), Tilly, Nora (1), Curdt, Constanze (1), Bareth, Georg (1), Brückner, Helmut (1), 

Andreas Vött (2) 

(1) Institute for Geography, University of Cologne, Albertus-Magnus-Platz, 50923 Cologne. GERMANY. Email: 

dirk.hoffmeister@uni-koeln.de 

(2) Institute for Geography, Johannes Gutenberg-Universität Mainz. Johann-Joachim-Becher-Weg 21, 55099 Mainz. 

GERMANY. Email: konstantin.ntageretzis@uni-mainz.de 

Abstract: This study deals with the application of terrestrial laser scanning (TLS) for monitoring both gradual and abrupt coastal 

changes, the latter for example associated to seismic events. TLS is a widely-used method for accurate measurements and 

monitoring of surface changes over time. Several sites in northwestern Greece were surveyed during TLS campaigns in 2009 and 

2010. Our main objectives were (i) to measure the present configuration and geometry of a mushroom rock at Poros (Cefalonia 

Island), which repeatedly experienced co-seismic uplift and (ii) to monitor annual coastal changes of selected shoreline sections 

on Lefkada Island. At the Poros mushroom rock, up-to-date 3D data of different generations of notches, which were uplifted by 

palaeoseismic events offer a solid base for evaluating future event-related uplifts. At Lefkada, TLS datasets of different years were 

compared with each other, resulting in a net balance of elevation changes. Results are checked by photographs taken by a 

camera on top of the TLS. Comparing the years 2009 and 2010, clear differences in the grain size distribution on beach sediments 

can be observed as well as slight changes in the coastline configuration. However, shadowing effects of the complex surface and 

noise caused by sea water represent lead to considerable problems analysing the data. 

Key words: terrestrial laser scanning, coseismic uplift, coastal change, Ionian Islands 

INTRODUCTION 

Western Greece, especially the Ionian Islands, 

belong to the most active seismic regions in the 

Mediterranean Sea as they are directly exposed to 

the Hellenic Trench system and the Cefalonia 

transform fault (Cocard et al. 1999, Hollenstein et al. 

2008). Transform faulting, collision, and subduction 

can be found within less than 100 km distance 

(Sachpazi et al. 2000). 

ideal approach, easy to realize by multi-temporal 

surveys. Laser scanning is an active remote sensing 

technique, also known as Light Detection and 

Ranging (LIDAR). 

The tectonic setting of Cefalonia Island, located at 

the northwestern edge of the Hellenic Trench, 

consists of E dipping and NW/NNW-SE/SSE striking 

thrust sheets (Stiros et al. 1994). The vertical 

movement of Cefalonia Island is dominated by 

gradual subsidence interrupted by co-seismic uplifts 

(Hollenstein et al. 2008). 

In the southeastern part of Cefalonia, a mushroom 

rock (Fig. 1) in the harbour of Poros is well known for 

two uplifted Holocene notches at +0.6 m and +1.2 m 

above present sea level (a.s.l.), respectively 

(Pirazzoli et al. 1994, Stiros et al. 1994). The lower 

notch was co-seismically uplifted during the 1953 

earthquake, and the upper notch during a seismic 

event around 1.500 yr BP (Stiros et al. 1994, 

Pirazzoli et al. 1994). Due to the high seismic activity 

of the region, further uplift due to future earthquakes 

can be expected. 

For measuring the dimensions and for monitoring 

geomorphological features which are moved, due to 

co-seismic crustal movements, laser scanning is an 

Fig. 1: Mushroom rock with two elevated notches in the 

harbour of Poros (southeastern Cefalonia) (Photo: K. 

Ntageretzis 2009). 

Direct measurement of distances and angles 

between the sensor and reflecting targets provides 

highly accurate 3D point clouds. LIDAR can be 

applied from the ground surface as terrestrial laser 

scanning (TLS) (Heritage & Large 2009). The 

interpretation of 3D point clouds is used within the 

framework of various applications (Vosselmann & 

Maas 2010). For example, TLS is used to study 

erosion and denudation processes along cliffs (Lim et 

92


EARTHQUAKE 

ARCHAEOLOGY 



al. 2010) and hillslopes, such as landslides and rock 

fall (Abellan et al. 2011, Nguyen et al. 2011). Richter 

et al. (2011) quantified the erosion of a dune cliff and 

the change of beach width by multi-temporal airborne 

laser scanning. Rosser et al. (2005) used terrestrial 

laser scanning for dune cliff erosion monitoring. Also, 

the rapidly changing geomorphology of fluvial 

systems can be monitored by laser scanning 

approaches (Heritage & Milan 2009). 

In this study, we used TLS to retrieve high-resolution 

data of the introduced mushroom rock of Poros 

(Cefalonia Island), which provides a reliable base for 

long-term monitoring of uplifting movements by 

seismic events. Furthermore, a recent notch at 

Kaminia Beach (Lefkada Island) was scanned to 

monitor annual changes in the coastal sedimentary 

system by gradual sediment transport and the 

influence of storms. 

LOCATIONS 

Several sites were surveyed by TLS along the shores 

of the eastern Ionian Sea within the framework of an 

interdisciplinary research project on palaeo-tsunami 

impacts (Vött et al. 2010). TLS field campaigns were 

carried out in 2009 and 2010. In this paper, we focus 

on the sites of Poros (Cefalonia Island) and Kaminia 

Beach (Lefkada Island) (see Fig. 2). 

For the mushroom rock at Poros four scan positions 

were chosen to retrieve a good coverage. For the 

area of Kaminia Beach, two scan positions were 

selected, which cover the whole site. At one of these 

two scan positions, the scanner was tilted to obtain a 

view of the entire study area. The resolution for a 

detailed scan was around 0.8 cm at a distance of 

10m. 

A Topcon HiPer Pro DGPS instrument was used to 

measure the different scan positions with a relative 

accuracy of 1 cm. Positions were recorded in the 

WGS84 system, UTM Zone 34 N. Furthermore, 

positions of cylindrical reflectors on ranging poles 

were recorded. 

For the annual measurements, the base point of the 

local DGPS net was marked by a metal mark and 

measured 500 times to achieve a mean, enhanced 

position. All measurements in relation to this base 

point were within the stated accuracy. Additionally, in 

each year the similar scan positions were chosen. 

Point clouds were subsequently georeferenced by 

the DGPS points of the scan positions and the 

reflectors. Afterwards, the registration was enhanced 

by the ICP-algorithm. The mushroom rock at Poros 

was reconstructed using the Geomagic Studio 12 

software and textured according to the photographs. 

The results of the two campaigns at Kaminia Beach 

were clipped and high resolution digital elevation 

models (HRDEMs) for each year were established. In 

a first step, the results of the HRDEM comparison 

were checked visually by the pictures of mounted 

camera. 


The mushroom rock at Poros with the two 

generations of uplifted notches were reconstructed 

as an 3D-object and textured by photographs for a 

near-realistic model (Fig. 3). The data set of Fig. 3 is 

a valuable tool to precisely measure the present 

situation as well as the amount of uplift by 

palaeoseismic events. It also provides a reliable base 

for monitoring after future events. 

Fig. 2: Map of the study areas in the Ionian Sea (NW 

Greece). Poros lies on Cefalonia Island and Kaminia Beach 

on the Island of Lefkada. Map based on Modis and ASTER 

GDEM. 

METHODS 

We used a TLS LMS-Z420i Riegl instrument for this 

survey. The time-of-flight range measurements have 

an accuracy of 0.6 cm with a range between 2 m and 

1,000 m. A high-resolution digital Nikon D200 camera 

mounted on the head of the laser scanner took RGBphotos 

which were used to colourize the TLS point 

clouds and to control the results. For data acquisition 

and first post-processing steps, the RiSCAN PRO 

Riegl-software was applied. 

Fig. 3: Perspective view of the mushroom rock model at 

Poros (Cefalonia Island). 

93


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 4: Morphological net elevation changes of the shoreline at Kaminia Beach between 2009 and 2010 calculated from TLS data 

(center). Elevation changes go hand in hand with changes in the sedimentary structure along the shoreline (top and bottom 

photos); exemplary areas are marked by ellipses and discussed in the text. 

Comparing the point clouds from the years 2009 and 

2010 regarding the notch at Kaminia Beach no 

changes were observed. However, comparing the 

beach morphology of 2009 with the one of 2010 

considerable changes in the littoral zone can be 

observed. 

Results of the net elevation differences calculated 

from multi-temporal TLS data with according 

photographs are illustrated in Fig. 4. The sea weed, 

which in 2009 covered a large part of the beach has 

mostly disappeared in 2010 (yellow and blue 

ellipses). At the same time, a deposit of sandy 

sediments (orange ellipse) was considerably reduced 

up to 1 m in size. Another sand cover close to a wall 

(white ellipse) was replaced by coarse-grained 

material in 2010. The yellow ellipse marks an area 

close to the Kaminia notch where gravel was 

accumulated inland and the coast line was 

smoothened. This proves active littoral abrasion and 

documents that the present notch is formed by wave 

erosion and not by bio-erosion. The red ellipse shows 

an area where a sand cover was partly erroded and 

underlying gravel exhumed. However, some pieces 

of gravel with 30-50 cm diameter seem to have also 

been displaced. We assume that the observed 

changes in the sedimentary budget are mainly due to 

winter storm events. 

Post-processing of point clouds turned out to be 

difficult due to noise caused by water and shadow 

effects at places where dense gravel occurs (Fig. 5). 

A major problem, especially at Kaminia beach, is that 

measurements from the seaside are not possible. 

CONCLUSIONS AND OUTLOOK 

Our studies show that the application of TLS in 

different littoral settings is an appropriate tool for 

monitoring both abrupt and gradual coastal changes. 

94


EARTHQUAKE 

ARCHAEOLOGY 



With regard to palaeoseismic research, TLS based 

high-resolution 3D models of geomorphologic 

features moved by co-seismical uplift yield detailed 

and precise data, for instance at Poros (Cefalonia 

Island). Moreover, the 3D model may be used to 

quantify effects from future earthquakes. 

Furthermore, multi-temporal TLS datasets allows to 

detect and to monitor gradual coastal changes. 

Fig. 5: Two major methodological problems arose in 

coastal monitoring using TLS in littoral zones: noise caused 

by water and shadowing effects by coarse material in the 

upper littoral zone. 

The results in this paper represent an essential 

progress in coastal monitoring by calculating a net 

balance of sediment budget achieved by a 

comparison of HRDEMs from different years, 

generated by TLS data . 

Acknowledgements: We gratefully acknowledge financial 

support by the German Research Foundation (DFG, VO 

938/3-1). 

References 

Abellan, A., J. M. Vilaplana, J. Calvet, D. Garcia-Selles & E. 

Asensio (2011). Rockfall monitoring by Terrestrial Laser 

Scanning a case study of the basaltic rock face at 

Castellfollit de la Roca (Catalonia, Spain). Nat. Hazards 

Earth Syst. Sci. 11, 829-841. 

Cocard, M., H.-G. Kahle, Y. Peter, A. Geiger, G. Veis, S. 

Felekis, D. Paradissis & H. Billiris (1999). New 

constraints on the rapid crustal motion of the Aegean 

region: Recent results inferred from GPS measurements 

(1993-1998) across the West Hellenic Arc, Greece. Earth 

Planetary Science Letters 172, 39–47. 

Heritage, G. L. & A.R.G. Large (Eds.) (2009). Laser 

Scanning for the Environmental Sciences. 1. ed., Wiley- 

Blackwell, Chichester, UK. 278 p. 

Heritage, G. L. & D.J. Milan (2009). Terrestrial Laser 

Scanning of grain roughness in a gravel-bed river. 

Geomorphology 113 (1-2), 4-11. doi: 

10.1016/j.geomorph.2009.03.021 

Hollenstein, C., M.D. Müller, A. Geiger & H.G. KAHLE 

(2008). Crustal motion and deformation in Greece from a 

decade of GPS measurements, 1993-2003. 


Lim, M., N.J. Rosser, D.N. Petley & M. Keen (2011). 

Quantifying the Controls and Influence of Tide and Wave 

Impacts on Coastal Rock Cliff Erosion. Journal of Coastal 

Research 27 (1), 46-56. doi: 10.2112/jcoastres-d-09- 

00061.1 

Nguyen, H. T., T.M. Fernandez-Steeger, T. Wiatr, D. 

Rodrigues, D. & R. Azzam (2011). Use of terrestrial laser 

scanning for engineering geological applications on 

volcanic rock slopes - an example from Madeira island 

(Portugal). Nat. Hazards Earth Syst. Sci. 11 (3), 807-817. 

doi: 10.5194/nhess-11-807-2011 

Pirazzoli, P.A., S.C. Stiros, J. Laborel, F. Laborel-Deguen, 

M. Arnold, S. Papageorgiou & C. Morhange (1994). Late- 

Holocene shoreline changes related to paleoseismic 

events in the Ionian Islands, Greece. The Holocene 4 (4), 

397-405. 

Richter, A., D. Faust & H.G. Maas (2011, in press). Dune 

cliff erosion and beach width change at the northern and 

southern spits of Sylt detected with multi-temporal Lidar. 

Catena. 

Rosser, N. J., D. N. Petley, M. Lim, S. A. Dunning & R. J. 

Allison (2005). Terrestrial laser scanning for monitoring 

the process of hard rock coastal cliff erosion. Quarterly 

Journal of Engineering Geology and Hydrogeology 38(4), 

363-375. doi: 10.1144/1470-9236/05-008 

Sachpazi, A., C. Hirn, M. Clement, F. Laigle, E. Haslinger, 

P. Kissling, Y. Charvis, J.C. Hello, M. Lepine, M. Sapin, & 

J. Ansorge (2000). Western Hellenic subduction and 

Cephalonia transform: local earthquakes and plate 

transport and strain. Tectonophysics 319 (4), 301-319. 

Stiros, S.C., P. A. Pirazzoli, J. Laborel & F. Laborel-Deguen 

(1994): The 1953 earthquake in Cephalonia (Western 

Hellenic Arc): coastal uplift and halotectonic faulting. 

Geophysical Journal International 117, 834-849. 



Lang, P. Masberg, S. M. May, K. Ntageretzis, D. 

Sakellariou & T. Willershäuser (2010). Beachrock-type 



Akarnania the Ionian Islands and the western 

Peloponnese. Zeitschrift für Geomorphologie N.F., Suppl. 

Issue 54 (3), 1-50. 

Vosselmann, G. & H.-G. Maas (Eds.) (2010). Airborne and 

terrestrial laser scanning. 1st ed., Whittles Publishing, 

Dunbeath, UK. 311 p. 

95


EARTHQUAKE 

ARCHAEOLOGY 



PRELIMINARY REPORT ON THE VODICE FAULT ACTIVITY AND ITS POTENTIAL FOR 

SEISMIC HAZARD IN THE LJUBLJANA BASIN, SLOVENIA 

Jamšek, Petra (1), Lucilla Benedetti (2), Miloš Bavec (1), Jure Atanackov (1), Marko Vrabec (3), Andrej Gosar (4,3) 

(1) Geological Survey of Slovenia. Dimieva 14. SI-1000 Ljubljana. SLOVENIA. Email: petra.jamsek@geo-zs.si 

(2) Centre Européen de Recherche et d’Enseignement des Géosciences de l’Environnement. BP 80, Europôle Méditerranéen de 

l’Arbois. 13545 Aix-en-Provence Cedex 4. FRANCE. 

(3) University of Ljubljana, Faculty of Natural Sciences and Engineering, Department of Geology. Aškereva 12. SI-1000 

Ljubljana. SLOVENIA. 

(4) Slovenian Environment Agency, Seismology and Geology Office. Dunajska 47. SI-1000 Ljubljana. SLOVENIA. 

Abstract (Preliminary report on the Vodice fault activity and its potential for seismic hazard in the Ljubljana Basin, 

Slovenia): In Ljubljana Basin (Slovenia) the Vodice fault was investigated to decipher its recent activity and seismic hazard 

potential for the densely populated central part of Slovenia. Preliminary geomorphological analysis and field observations suggest 

it may have been recently reversely active. Moreover, using preliminary optical stimulation ages we estimated a slip-rate along the 

Vodice fault at about 0.2 – 0.4 mm/yr over the last 115 ± 32 ka. An earthquake of magnitude 6.2 – 6.3 could be expected on this 

fault which may have been the source of the Ljubljana 1895 magnitude 6.1 earthquake. Further analysis will follow to better 

constrain characteristics of this fault and seismic hazard of the area. 

Key words: active fault, seismic hazard, Ljubljana Basin, Slovenia 

INTRODUCTION 

The Ljubljana Basin (Fig. 1), the most densely 

populated and urbanized area of Slovenia, 

experiences constant seismic activity and has been 

the site of strong historical seismic events with 

magnitude as high as 6.1 (intensity VIII–IX EMS-98, 

Ljubljana earthquake 1895 (Ribari, 1982)). The 

basin is filled with Quaternary sediments reaching a 

thickness of up to 280 m in some parts, which may 

significantly enhance site effects and therefore 

increase the earthquake hazard in the area (Gosar et 

al., 2010). However, active faults capable of 

producing strong magnitude earthquakes in this area 

are poorly known. 

The Ljubljana Basin probably results from 

transpression with dextral strike-slip movement along 

NE-SW faults and thrusting along smaller scale E-W 

structures (Vrabec, 2001; Bavec et al., 2003; 

Benedetti et al., 2000). In this report we focus on the 

Vodice fault escarpment, located 15 km north of 

Ljubljana, the capital city located within the 

Quaternary basin infill. This escarpment is offsetting 

Quaternary surfaces for 5 to 25 m along a length of 

10 – 11 km. This feature was previously described as 

a terrace riser of the Sava river (Žlebnik, 1971), and 

later as a reverse fault (Bavec et al., 2004; Verbi, 

2006). To decipher whether the Vodice fault has 

been recently active and to asses its potential for 

seismic hazard we performed a detailed study of its 

morphology and topography. Herein, we present our 

preliminary interpretations. 

METHODS 

Fig. 1: The Ljubljana Basin map with main active 

structures (modified according to Buser, 2009), 

historical earthquake epicentres with magnitude above 

3.9 (according to Živi, 2009; note that earthquakes 

magnitudes are obtained from macroseismic data and 

that locations of historical earthquake epicentres are 

not well constrained), and isoseisms of the Ljubljana 

earthquake 1895 (according to Lapajne, 1989). 

We investigated its surface expression trough 

geomorphological analysis of topographical maps (1 : 

5.000 and 1 : 10.000 scale), digital elevation model 

(resolution 5 m), aerial images, and 2.5 m resolution 

SPOT images in stereo pairs. Each alluvial surface 

was carefully mapped and a series of topographical 

profiles were extracted across, and parallel to the 

scarp from 5 m resolution digital elevation model. 

96


EARTHQUAKE 

ARCHAEOLOGY 



Based on analysis of topographical profiles, detailed 

mapping and field observations we were able to 

evidence recent movement along the fault. Based on 

preliminary optical stimulation luminescence ages 

(Bavec et al., 2005) we evaluated the slip-rate along 

the Vodice fault, possible earthquake magnitude and 

average displacement per event. 

VODICE FAULT SURFACE EXPRESSION 

The flat Ljubljana Basin surface resulting from 

Quaternary deposition is perturbed in the area of 

Vodice by an unusual linear feature, ENE-WSW 

oriented, 10-11 km long scarp between Pšata river 

on the east and Sava river on the west (Fig. 2). East 

of Vodice, the scarp splays into two branches. A 

series of detailed topographic profiles levelled 

across, and parallel to the scarp show that the height 

of the scarp varies from 25 to 5 m for the southern 

branch and from 18 to 3 m for the northern branch 

(Figs. 2 and 3). Evidence of ongoing uplift is attested 

by the presence of abandoned streams across the 

structure and perched valleys (Fig. 3). Active streams 

strongly incise the northern, hanging wall 

compartment, this upper surface being clearly older 

as shown by its degradation mostly due to dolines. 

These observations lead us to interpret the Vodice 

scarp as the surface expression of an active reverse 

fault. The latter suggests that the scarp can not be 

interpreted as a terrace riser of a former course of 

the Sava river. 

Fig. 2: The Vodice fault surface expression. 

SUBSURFACE DEFORMATIONS 

Fig. 3: The series of topographic profiles across, and 

parallel to the Vodice fault scarp (red arrows = perched 

valeys). Note that the profile in the image is showing 

two fault branches because it is shalow, but in the 

depth the two branches are most probably joined into 

one fault. 

Further evidences of the fault activity are provided 

from subsurface deformations. Folding was reported 

in a clay pit at the eastern end of the scarp (Fig. 4, 

Drobne et al., 1960; Šifrer, 1961). Westward, where 

the fault cuts the N-S running Sava River, Quaternary 

conglomerates are also folded and offset (Fig. 5). In 

both cases the deformations are consistent with a 

reverse fault interpretation. 

FIRST EVALUATION OF SEISMIC HAZARD 

POTENTIAL 

Preliminary optical stimulation luminescence ages of 

deformed sediments, located at the eastern tip of the 

fault (Bavec et al., 2005), suggest an age of 115 ± 32 

ka for the upper alluvial surface (Fig. 2). Using this 

age and assuming a northward dip of 35 – 45°, we 

estimated a minimum slip-rate along the Vodice fault 

at about 0.2 – 0.4 mm/yr over the last 115 ± 32 ka. 

Fig. 4: Folded Quaternary sediments at the eastern end 

of the scarp reported by Šifrer (1961). 

On a 10 – 11 km long reverse fault, an earthquake of 

magnitude 6.2 – 6.3 could be expected. Such 

earthquake could trigger an average displacement of 

~ 0.5 m with an average recurrence time of 1200 – 

3000 yr (Wells & Coppersmith, 1994). 

97


EARTHQUAKE 

ARCHAEOLOGY 



Acknowledgements: This work is funded by the Slovenian 

Research Agency, through 1) The project L1–2383 

Seismotectonic model of the Ljubljana Basin (cofounded by 

the Slovenian Environment Agency), 2) The programme P1 

– 0011 Regional geology, and 3) The support for young 

researcher (contract 1000-09-310068). 

References 

Fig. 5: One of the Vodice fault outcrops. Fault plane (red 

line) dips towards NNW for 35°. Features that may also be 

interpreted as near surface expression of faulting are 

marked in black. 

DISCUSSION 

This preliminary geomorphological analysis and field 

observations suggest recent activity along E-W 

trending 10-11 km long Vodice reverse fault and 

correlates well with results of previous studies (Bavec 

et al., 2004; Verbi, 2006). However, to further prove 

Vodice scarp as the surface expression of a reverse 

fault the geophysical investigations are planned. 

Also, to better constrain the Vodice fault 

displacements and its slip-rate, further accurate 

quantitative analysis of the morphology are 

warranted (levelling survey using theodolite or 

diferential GPS) and Quaternary geochronology to 

date the offset surfaces. 

A first approximation suggests the Vodice fault could 

trigger magnitude 6.2 – 6.3 earthquake events. 

Considering its location and proximity to Ljubljana, 

the Vodice fault is a candidate to be the source of the 

Ljubljana earthquake 1895. To better assess the 

seismic hazard of this area further analysis will 

follow, such as geophysical investigations of the fault 

geometry (high-resolution seismic reflection and 

ground penetrating radar profiling) and 

paleoseismological investigations along the fault to 

define its seismic behaviour and decipher its seismic 

history. 

Recognition of seismic hazard is crucial in areas 

such as Ljubljana Basin, where destructive 

earthquakes can represent a huge danger for 

population and infrastructure. Investigations in active 

tectonics are the first step towards seismic hazard 

assessment, leading to protection of human lives as 

well as to a decrease of economical damage in case 

of a destructive seismic event. To determine the 

seismic hazard in the Ljubljana Basin we will also 

extend our investigations to other presumably active 

faults bounding the basin. 

Bavec, M., B. Celarc, M. Demšar, M. Poljak, D. Rajver, I. 

Rižnar, F. Preusser, M. Toman, J. Žibrat, M. Novak & K. 

Hribernik, (2005). Izdelava geoloških kart: letno poroilo 

za leto 2005. Geological Survey of Slovenia. Ljubljana. 

Bavec, M., B. Celarc, M. Poljak, M. Demšar, D. Rajver, D. 

Skaberne, T. Verbi, I. Rižnar, F. Preusser & M. Culiberg, 

(2004). Izdelava geoloških kart. Karta aktivnih prelomov v 

Sloveniji: letno poroilo za leto 2004. Geological Survey 

of Slovenia. Ljubljana. 

Bavec, M., M. Poljak, M. Demšar, D. Rajver, M. Komac, M. 

Toman, S. Stojanova, B. Muši, M. Vrabec, T. Verbi & I. 

Rižnar, (2003). Izdelava geoloških kart. Karta aktivnih 

prelomov v Sloveniji : letno poroilo za leto 2003. 1. del : 

Raziskave na obmoju Ljubljansko-kranjske kotline. 

Geological Survey of Slovenia. Ljubljana. 56 f. 

Benedetti, L., P. Tapponnier, G. C. P. King, B. Meyer & I. 

Manighetti, (2000). Growth folding and active thrusting in 

the Montello region, Veneto, northern Italy. Journal of 

Geophysical Research – Solid Earth 105, 739–766. 

Buser, S. (2009). Geological map of Slovenia 1:250.000. 

Geological Survey of Slovenia. Ljubljana. 

Drobne, F., R. Pavlovec & A. Šercelj, (1960). Some 

analyses and problems of Pleistocene sediments at 

Lokarji near Vodice. Kamniški zbornik 6, 163–194. 

Gosar, A., J. Rošer, B. Šket Motnikar, P. Zupani, (2010). 

Microtremor study of site effects and soil-structure 

resonance in the city of Ljubljana (central Slovenia). 

Bulletin of earthquake engineering 8 (3), 571–592. 

Lapajne, J. (1989). Veliki potresi na Slovenskem – III.: 

potres v Ljubljani leta 1895. Ujma 3, 55–61. 

Ribari, V. (1982). Seismicity of Slovenia. Catalogue of 

earthquakes (792 A.D.–1981). Seismological Survey of 

Slovenia. Ljubljana. 649 p. 

Šifrer, M. (1961). The basin of Kamniška Bistrica during the 

Pleistocene period. Slovenian Academy of Sciences and 

Arts. Ljubljana. 211 p. 

Verbi, T. (2006). Quaternary-active reverse faults between 

Ljubljana and Kranj, central Slovenia. Razprave IV. 

razreda SAZU 47 (2), 101–142. 

Vrabec, M. (2001). Structural analysis of the Sava Fault 

zone between Trstenik and Stahovica. PhD thesis. 

University of Ljubljana. Faculty of Natural Sciences and 

Engineering. Department of Geology. 94 f. 

Wells, D. L. & K. J. Coppersmith, (1994). New Empirical 

Relationships among Magnitude, Rupture Length, 

Rupture Width, Rupture Area, and Surface Displacement. 


974–1002. 

Živi, M. (2009). Catalogue of earthquakes in Slovenia. 

Internal documentation. ARSO – Slovenian Environment 

Agency. Ljubljana. 

Žlebnik, L. (1971). Pleistocene Deposits of the Kranj, Sora 

and Ljubljana Fields. Geologija 14, 5–51. 

98


EARTHQUAKE 

ARCHAEOLOGY 



DISTRIBUTION AND SEDIMENTARY CHARACTERISTICS OF TSUNAMI DEPOSITS ON 

PHRA THONG ISLAND, THAILAND 

Jankaew, Kruawun (1, Dominik Brill (2), Maria E. Martin (3), Yuki Sawai (4) 

(1) Department of Geology, Faculty of Science, Chulalongkorn University, Bangkok, 10330 THAILAND 

Email:kruawun.j@chula.ac.th 

(2) Department of Geography, University of Cologne, GERMANY 

(3) Department of Earth and Space Sciences, University of Washington, Seattle, USA 

(4) Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, JAPAN 

Abstract (Paleotsunami): Phra Thong Island is located along the western coast of Thailand directly facing the source area of the 

2004 Indian Ocean tsunami. Here we report a distribution of paleotsunami sand layers on Phra Thong Island, collected during four 

years of field work (2007-2011). Sedimentary characteristics of paleotsunami deposits are compared with those of the 2004 

tsunami. Thicknesses of the 2004 and paleotsunamis sand layers are controlled by micro-topography and subjected to postdepositional 

alteration. Sand layers of 2004 and paleotsunami layers contain fining upward sequence(s), which suggest deposition 

from suspension, with similar grain size distribution. The sedimentary structures of the 2004 tsunami, where present, are already 

obliterated six years after the event, making it difficult to interpret the flow direction and to differentiate between deposits from 

inflow and outflow of tsunami. This will affect accuracy of future work in estimating flow characteristics of past tsunamis based on 

the depositional record in the vegetated area such as Thailand. 

Key words: Paleotsunami, tsunami deposits, Indian Ocean, Thailand 

INTRODUCTION 

Phra Thong Island is located about 580 km 

southwest of Bangkok and about 120 km north of 

Phuket (Fig.1). It is about 15 km in length and 9 km in 

width, with the long axis in an approximate N-S 

direction and the western side directly facing the 

Andaman Sea. The island composes of beach-ridge 

plain on the western side and the mangrove-fringed 

tidal inlet to the east. The 2004 tsunami severely 

struck the island and completely wiped out one of the 

only three villages on the island. Paleotsunami 

deposits on Phra Thong Island were reported by 

Jankaew et al. (2008) and Fujino et al. (2009). 

o 

9 00’ N 

8 

o 

00’ N 

o 

98 00’ E 

20 km 

ANDAMAN 

SEA 

PT 

o 

98 30’ E 

PHANG NGA 

PHUKET 

o 

99 00’ E 

Fig. 1: Map showing location of Phra Thong Island 

(PT). 

Apart from the frequency, a magnitude of large 

tsunamis is important information needed to be 

established in certain coastal area in order have 

better mitigation measure in place for possible future 

catastrophic events. The magnitudes of tsunamis 

responsible for the sand layers at Phra Thong, can 

be assessed using numerical models with help from 

tsunami deposits to validate the model results. 

Because of its relatively flat terrain, Phra Thong 

offers a suitable site to study velocity and tsunami 

flow depth from characteristics of the deposits. 

Thickness and grain size distribution of tsunami 

deposits can be used in sediment transport models 

(Jaffe & Gelfenbaum, 2002; Jaffe & Gelfenbaum, 

2007) to derive estimates of flow velocity and water 

depth (Peters et al., 2007), which are important in 

building design and in planning evacuation route. 

We present distribution of paleotsunami deposits on 

Phra Thong Island and sedimentary characteristics of 

2004 and paleotsunamis. 

METHOD 

We dug shallow pits and trenches to observe 

sedimentary characteristics of tsunami sediments 

(both 2004 tsunami and paleotsunamis) and to 

collect samples for grain size analyses. In order to 

better observe and to be able to make a meaningful 

conclusion of the deposit thickness and grain size 

variation, observed in the 2004 and paleotsunami 

deposits, we collected samples along two transects 

parallel to the tsunami flow (Fig. 2). Sediment 

samples were collected to the end of the 2004 

99


EARTHQUAKE 

ARCHAEOLOGY 



tsunami sedimentation limit. Total length of the two 

transects are 1.5 km and 2 km, respectively. Along 

these transects wherever the paleotsunami sand 

layer is present, we also collected samples for grain 

size analyses. The samples were collected at 0.5 cm 

interval, and analysed using settling tube, measuring 

size range from 0-5 with 0.1 intervals. 

landward location and sediments remaining in 

suspension. The 2004 tsunami deposits contain up to 

o 

98 15’ E 

o 

98 21’ E 

9 

o 

09’ N 

Transect 1 

Transect 2 

Fig. 4: Thick 2004 tsunami deposit (almost 30 cm 

thick) in a swale. Sedimentary structures indicating 

flow direction are clearly visible in this picture. The 

sea is to the right of the picture. 

o 

9 03’ N 

RESULT 

5KM 

Fig. 2: Sediment samples were collected from two 

transect from the northern part of the island. 

The thickness of deposits of the 2004 tsunami and its 

predecessors varies across the beach-ridge plain 

and is primarily controlled by local topography. 

Thickness tends to be greatest in a low-lying swale 

(as much as 30 cm). On ridges and higher ground it 

tends to be thinner (few cm, 10 cm at most). Fig. 3 

shows a variation in thickness of the 2004 tsunami 

deposit along transect 2 (in Fig. 2). A thicker, and 

sometimes coarser, deposit in topographic lows is a 

response of the flow to the deeper water depth, 

which slows the flow and drive deposition (Apotsos et 

al., 2009). In the case of Phra Thong, if the swale is 

located on a path of back flow, which is often 

channelized, thickness is much higher. Back flow 

deposits contain both sediments eroded from 

4 fining-upward sequences. Fig. 4 is an example of 

2004 tsunami deposit in the swale located on a path 

of the back flow. In Fig. 4, the lowest fining-upward 

sequence was probably deposited out of the first 

tsunami wave. Above this sequence, a continuous 

dark gray layer of fine silt-clay which was probably 

deposited when the tsunami flow was still and 

between the two main waves. Above this silty clay 

layer there are 2-3 fining upward sequences, which 

were possibly deposited from back flow of the first or 

second wave. Sets of cross bedding in these upper 

sequences indicate a flow direction towards the sea. 

Revisiting the sites allow us to observe that dense 

vegetation in the tropics, especially in the low lying 

fresh water swales, already destroy most of these 

cross bedding sets. This will complicate the 

interpretation of flow depths and velocities based on 

the deposit thickness. 

On high ground of Phra Thong Island (about 4 m 

above msl.), with less vegetation cover, the 2004 

tsunami deposit lies above the sand ridge sediments 

with faint soil in between them. It is still possible to 

identify the 2004 tsunami deposits in the high ground 

6 years after the event, but allowing more time 

differentiating it from the sand ridge below will be 

problematic. At many locations, distinguishing the 

2004 tsunami laid sand and the underlying ridge 

sand was already difficult due to the lack of organic 

soil formation on top of the ridge sand. On the high 

ground immediately next to the swale, after the 

tsunami dumped a majority of sediments in the 

swale, the 2004 tsunami deposit composes mainly of 

silt which fell out of suspension (Fig. 5). The tsunami 

flow then picked up more sediment from the ridge 

area along its flow path. As a result, grain size of 

2004 tsunami on high ground is composed mainly of 

silt in contrast to sand and silt in the swale deposit. 

The thicknesses of the 2004 deposits at different 

locations are also shown. 

Fig. 3: Thicknesses of 2004 tsunami deposit along 

transect 2 (blue vertical line: VE x10). 

100


EARTHQUAKE 

ARCHAEOLOGY 



appear as a massive bed or faintly normally graded, 

more so than the 2004 deposits, but mean grain size 

at 0.5 cm thickness intervals show that they contain 

up to two fining-upward sequences 

DISCUSSION 

Fig. 5: Thin 2004 tsunami deposit (about 5 cm thick) 

on high ground. The deposit contains mainly silt and 

very fine sand with no sedimentary structure. The sea 

is to the left of the picture. 

Apart from variation in the thickness of the deposit, 

internal characteristics of 2004 deposit on Phra 

Thong also varies greatly from one place to another. 

Often the deposits show no internal structure or 

layering. Typically the deposits appear as a massive 

bed or faintly normally graded. In locations close to 

the sea, some deposits are coarsened upward. 

The 2004 tsunami sediments range in grain size from 

coarse sand to coarse silt, with distinctive bimodal 

distribution with the first mode in fine-sized range (2 

), and the second and highest mode in very finesize 

range (3.3 ). The coarser sediments are 

composed of coarse sands, derived from beach berm 

and possibly from offshore areas, and broken shell 

fragments, large foraminiferas and other 

macrofossils, whereas the finer sediments were 

possibly derived from the subtidal zone. The 2004 

sediment composes about 85% of clear to white 

quartz grains, 8% of shell fragments and microfossils, 

4% of muscovite and 3% of heavy minerals - mostly 

small grains of tin. Although typically 2004 tsunami 

deposits at Phra Thong appear as a massive bed or 

faintly normally graded, detailed grain size analyses 

show that they can contain 2-3 fining-upward 

sequences. The mean grain size of the 2004 tsunami 

is slightly bigger than that of the paleotsunami sand. 

The 2004 tsunami layer contains 2-3 fining-upward 

sequences while paleotsunami deposits contain 2 

fining-upward sequences. 

Paleotsunami deposits at Phra Thong Island are 

composed of grains ranging in size from coarse sand 

to coarse silt, with bimodal distribution. A small first 

mode in fine-sized range (2.3 phi), and second and 

the highest mode in very fine-size range (3.2 phi). 

Grains compose of about 91% of clear to white 

quartz grains, 7% of muscovite and 2% of heavy 

minerals - mostly small grains of tin. The shell 

fragments and microfossils are absent in the paleosand 

layer. Phra Thong paleotsunami deposits 

Deposition of tsunami sediment, although generally 

has a sheet-like geometry, is greatly controlled by the 

local topography with thicker deposit in the low-lying 

area. The thickness of tsunami deposit is further 

altered by post-depositional processes such as 

human and animal disturbances, erosion and 

bioturbation. However, the thick deposits in the lowlying 

area have a greater tendency to be preserved 

and remained as a geological record. Dense 

vegetation in a study area quickly destroy most of the 

sedimentary structures contain in the deposit which 

may complicate differentiation between inflow and 

outflow deposits, leading to over or under estimation 

of the thicknesses of tsunami deposited by each 

wave. Sedimentation limit of the paleo-event is also 

difficult to define, especially in the case of Phra 

Thong Island where deposits are mainly preserved in 

the swales and are subjected to bioturbation. 

CONCLUSION 

Future work in estimating hydrodynamic properties of 

past tsunamis from sediment deposits in the tropics 

should not be dependent on thickness of the 

deposits. 

Acknowledgements: This work was supported by Japan 

Society for the Promotion of Science (to YS and KJ), DFG 

grant numbers BR 877/27 and KE 190/26 (to DB and KJ) 

and NRCT grant number GE3/2554 (to KJ). 

References 

Apotsos, A.; B. Jaffe, G. Gelfenbaum, and E. Elias, (2009). 

Modeling time-varying tsunami sediment deposition. 

Proceedings of Coastal Dynamics 2009, 1-15. 

Fujino, S., H. Naruse, D. Matsumoto, T. Jarupongsakul, A. 

Sphawajruksakul, and N. Sakakura, (2009). Stratigraphic 

evidence for pre-2004 tsunamis in southwestern 

Thailand. Marine Geology 262, 25-28. 

Jaffe, B. E. and G. Gelfenbaum, (2002). Using tsunami 

deposits to improve assessment of tsunami risk. 

Solutions to Coastal Disasters’02, Conference 

Proceedings, ASCE, 836-847. 

Jaffe, B. E. and G. Gelfenbaum, (2007). A simple model for 

calculating tsunami flow speed from tsunami deposits, 

Sedimentary Geology 200, 347-361. 

Jankaew, K., B.F. Atwater, Y. Sawai, M. Choowong, T. 

Charoentitirat, M.E. Martin, and A. Prendergast, (2008). 

Medieval forewarning of the 2004 Indian Ocean tsunami 

in Thailand. Nature 455, 1228–1231. 

Peters, R.; Jaffe, B. & Gelfenbaum, G. (2007). Distribution 

and sedimentary characteristics of tsunami deposits 

along the Cascadia margin of western North America, 


101


EARTHQUAKE 

ARCHAEOLOGY 



ARCHAEOSEISMOLOGY OF THE AD 1545 EARTHQUAKE 

IN CHIANG MAI, NORTHERN THAILAND 

Kázmér, M. (1, Kamol Sanittham (2), Punya Charusiri (3), Santi Pailoplee (3) 

(1) Department of Palaeontology, Eötvös University, Pázmány sétány 1/c, Hungary. Email: mkazmer@gmail.com 

(2). Department of Mathematics and Statistics, Chiangmai Rajabhat University, Chiang Mai, Thailand 

(3) Department of Geology, Chulalongkorn University, Bangkok, Thailand 

Abstract (Archaeoseismology of the A.D. 1545 earthquake in Chiang Mai, northern Thailand): The A.D. 1545 Chiang Mai 

earthquake in northern Thailand was studied by historical and archaeological sources.The temple Wat Chedi Luang has lost about 

half of the original 80-metres height due to southward-directed collapse. Twenty-one temple sites – out of 74 visited – has tilted 

pagodas, up to 5° in various directions, dominated by a SE trend. All damaged temples were built before the 1545 earthquake. We 

suggest that a city-wide liquefaction event caused tilting. The responsible earthquake possibly occurred along the Doi Suthep 

Fault within city limits. Possible activity of distant faults is assessed. 

Key words: palaeoseismology, Thailand, liquefaction, Wat Chedi Luang 

INTRODUCTION 

An important obstacle to the assessment of 

earthquake hazard at present is the lack of 

information about old earthquakes (Ambraseys, 

2009: xii). The locations of larger historical 

earthquakes have been found to be known well 

enough to guide field studies for further in situ 

investigations. Properly run field studies provide 

reliable observations for the assessment of damage, 

intensity, and its distribution, ground effects and 

surface faulting. Field studies of old earthquakes are 

time-consuming and often present subtle problems 

but they are essential (Ambraseys 2009: 16). Here 

we provide a brief description of traces of a 

significant earthquake in Northern Thailand, and 

provide assessment of seismic parameters of the 

event. 

Fig. 2: Archaeological reconstruction of the pre-earthquake 

dimensions of Wat Chedi Luang as seen from the east. 

Total height was approx. 80 m. The portion above the heavy 

line is the art historian’s vision about its looks. 

Fig. 1: Wat Chedi Luang in Chiang Mai, Thailand, seen from 

the southeast. Damaged during the AD 1546 earthquake, 

the upper half of the stupa fell to the south. 

Both historical documents and archaeological data 

are available describing the A.D. 1545 earthquake in 

Northern Thailand. We studied the Buddhist temples 

in and around the old city of Chiang Mai (Kázmér & 

102


EARTHQUAKE 

ARCHAEOLOGY 



Sanittham, 2011) to identify possible earthquakeinduced 

damages preserved in the buildings’ 

structure and orientation. 

Currenty earthquake activity in northern Thailand is 

interpreted within the framework of Thoen Fault 

(Chiang Saen, May 2007, M L = 6.3), Mae Tha and 

Pha Youv fault zones tens of kilometres away 

(Pailoplee et al., 2010). Since recurrence time of 

major earthquakes seems to be longer than the 

instrumental period of 50 years, archaeoseismology 

is a necessary tool to extend the observation period 

to centuries. 

high was built in 1479-1481. The base was enlarged 

and strengthened in 1512 (Podjarawaraporn, 2547). 

On 28 July 1545 there was a huge rainstorm and an 

earthquake, which caused the chedi to topple, 

leaving only half if its structure to stand (Fig. 2). The 

power and richness of the Medieval Lanna Empire 

already in the decline, no funds have ever been 

available to restore the damaged building to its 

former glory. The chedi was left in this damaged 

condition for more than four centuries. A cosmetic 

restoration in 1992 completed the strengthening by a 

60 x 60 m base, . 

METHODS 

Historical, archaeological, and geologicalgeophysical 

data are combined to understand the 

Chiang Mai earthquake of AD 1545. The published 

historical description was cross-checked with 

archaeological data of the site of Wat Chedi Luang 

and elsewhere. We visited 74 temples of Chiang Mai 

city. While recording earthquake archaeological 

effects (Rodriguez-Pascua et al. 2011), we measured 

the angle and direction of tilt of the chedi (stupa) by a 

stonemason’s tiltmeter and a compass, respectively. 

Coordinates of chedi location were taken from the 

digital map of Northern Thailland (ThinkNet 2010). 

Construction ages were drawn from Thai-language 

publications. When no printed source was available, 

we accepted the dating of tourist information tablets 

in the monasteries. 

TILTED BUILDINGS CITYWIDE 

In addition to the famous damaged chedi, numerous 

religious and secular monuments in and around the 

old city bear evidence for some kind of earthquake 

damage. The most obvious evidence is tilting of 

chedis: the pointed top part of the monument clearly 

deviates from the vertical by a few degrees (Fig. 3). 

(The lightweight metal decoration at the very top is 

almost always heavily tilted; we did not take these 

into account, only the brick portion below.) Historical 

data on construction time of the chedis indicate that 

all of them were built in the 14-15 th century AD, 

before the A.D. 1545 earthquake (Fig. 4). Locations 

and tilt directions are mapped on Fig. 5. 

Tilt directions are dominated by a conspicuous SE 

trend (Fig. 6). 

There is no official English transliteration system for 

the Thai language. English spelling of Thai names is 

inconsistent to the extent that one’s own name is 

written differently on subsequent occasions. In this 

paper we use names as found on the electronic map 

of ThinkNet (2010), which is neither official, nor better 

than any other spelling. 

HISTORICAL DATA 

There was a damaging earthquake in Chiang Mai city 

(Northern Thailand) on 28 July 1545 in the afternoon 

hours between 4.30 and 6.00 pm. „The earth 

trembled and shook, groaned and moaned, very 

intensely. The finials, (top parts) yòt, of the Jedi 

Luang and of the jedi in Wat Phra Sing broke off and 

fell down, and also the finials of many other jedis”, 

recorded the contemporary Chiang Mai Chronicle in 

Lanna language (translated by Penth, 2006). 

WAT CHEDI LUANG IN CHIANG MAI 

The largest chedi (stupa, pagoda) ever built in what 

is Thailand today is the Wat Chedi Luang, standing in 

the monastery of the same name in the centre of old 

Chiang Mai city (Fig. 1). Built in 1391, it has been 

reconstructed and enlarged several times, A huge 

chedi, 56 x 56 m rectangular basement, approx. 80 m 

Fig. 3: Tilted buildings citywide 

103


EARTHQUAKE 

ARCHAEOLOGY 



The Doi Suthep fault, the master fault of the halfgraben 

of the Chiang Mai Basin is 3 km away. It is 

not known to be active: a minimum M 5.5 seismic 

event here could cause liquefaction. 

Fig. 4: All tilted chedis (stupas) were built before the AD 

1545 earthquake. 4 – Wat Chiang Yeun, 8 – Wat Hua 

Khuang, 12 – Wat Lok Mo Li, 17 – Wat Phra SIngh, 23 – 

Wat Chai Prakiat, 28 – Wat Phuak Hong, 34 – Wat Chet 

Rim, 42 – Wat Umong, 46 – Wat Chiang Man, 50 – Wat 

Srisupan, 51 – Wat Nantharam, 56 – Wat Daowadueng (no 

tilting was observed), 67 – Wat Bupharam, 72 – Wat 

Chomphu. 

The SW-NE trending, left-lateral Mae Kuang Fault 32 

km to the NE is possibly inactive since the Tertiary, 

althought the fault trace is particularly conspicuous in 

the landscape (Rhodes et al. 2004). A minimum M 

6.3 seismic even would have been sufficient to cause 

liquefaction in Chiang Mai. 

SUBSOIL 

The tilted chedis are all on the alluvial plain of the 

Ping River, extending over at least 4 km 2 . 

Groundwater lever was high during our survey in 

August 2010, about 70 to 100 cm below ground, as 

seen in several wells within the temple compounds. 

Historical data indicate a rainy summer season for 

1545, too. 

We suggest that a city-wide liquefaction event, 

caused uneven settlement and subsidence of the 

buildings in the saturated soil. The dominant SE-ward 

tilt direction possibly reflects strong motion 

directionality. 

INTENSITY 

While modified Mercalli intensity VII is the damage 

threshold for many archaeological sites (Kovach and 

Nur, 2006), we assume that damages to Wat Chedi 

Luang related to the 1545 earthquake require a 

larger intensity due to the especially compact 

construction of the building. The pagoda, built like a 

pyramid, is certainly a more earthquake-resistant 

structure than any ordinary city house, even palace. 

Intensity IX or higher (good masonry damaged 

seriously, in areas of loose sediment, sand, mud, and 

water ejected – Rapp, 1986) seems more probable. 

Fig. 5: Tilted chedis (Fig. 4) (black dots) on the alluvial plain 

of Ping River (Margane & Tatong, 1999). Ticks towards 

direction of tilting. Untilted chedis are marked with empty 

circles. Rectangle indicates walled city of old Chiang Mai. 

The Lampang-Thoen fault zone 120 km to the SE is 

active (Chiang Saen, May 2007, M L = 6.3). The 

segments are long enough to produce M 7 

earthquakes (Pailoplee et al., 2009). A minimum M 7 

seismic event is needed to cause liquefaction in 

Chiang Mai city. 

Intensity VIII to IX (heavily damaging to destructive) 

is assumed on the ESI 2007 environmental intensity 

scale (Michetti et al., 2007): liquefaction with 

settlement up to 30 cm or more. The total affected 

area was in the order of 1000 to 5000 km 2 

(Reicherter et al., 2009), i.e. all of the Chiang Mai- 

Lamphun Basin. 

EPICENTER AND MAGNITUDE 

Known and possibly active faults were assessed for 

source of the earthquake, and the minimum 

magnitude for liquefaction calculated after Obermeier 

(1996, Fig. 42). 

Fig. 6:Tilt directions plotted in polar bar chart. Horizontal 

axis – number of pagodas tilted in a certain direction. Note 

the prominent southeastward tilting of several chedis. 

104


EARTHQUAKE 

ARCHAEOLOGY 



The Sagaing Fault in Myanmar, forming the boundary 

between the Sunda and Burma plates is 200 km to 

the W. It regularly produces M > 7 earthquakes (M 

7.0-7.4) (Hurukawa & Maung, 2011). However, the M 

7.5 earthquake on December 3, 1930, did not cause 

any liquefaction event in Chiang Mai we are aware 

of. A lack of proper attenuation model for Thailand 

(Chintanapakdee et al., 2008) prevents formulating a 

suitable explanation. There is local model developed 

for Chiang Mai (Kannika & Takada, 2009), although 

for rock sites, not for alluvium. We suggest that 

earthquake intensities display a strong directionality 

along the right-lateral Sagaing Fault: higher 

intensities occurring parallel and lower intensities 

perpendicular to the fault, thus protecting Chiang Mai 

from major plate-boundary events. 

Whether any of the above or another fault is 

resonsible for the AD 1545 earthquake is an open 

question as yet. Studies on strong motion direction 

causing the major damages (see, for example. 

Korjenkov & Mazor, 1999, 2003; Kázmér & Major, 

2010; Hinzen, 2008, 2009) may help tor resolve 

some of the open questions. 

Acknowledgements: Our thanks are due to the staff of 

Mahamakat Buddhist University, Lanna Campus library for 

their help in accessing relevant literature in the Thai 

language. Financial help from Hungarian National Science 

Foundation (OTKA) grant K 67.583 and and an IGCP 567 

travel grant are sincerely acknowledged here. Raúl Pérez 

López is thanked for his comments, which improved this 

manuscript. 

References 

Publication dates of Thai-published books in Thai only are 

given as of the Thai calendar. Deduce 543 for conversion to 

western, Gregorian years. 

Ambraseys, N.N. (2009). Earthquakes in the Mediterranean 

and Middle East. A Multidisciplinary Study of Seismicity 

up to 1900. Cambridge, Cambridge University Press. 968 

p. 

Chintanapakdee, C., Naguit, M.E. & Charoenyuth, M. 

(2008). Suitable attenuation model for Thailand. – 14th 

World Conference on Earthquake Engineering: 

Innovation Practice Safety, Beijing, 2008. [8 p.] 

Download: http://www.14wcee.org/Proceedings/files/02- 

0088.PDF (accessed 25 June 2011) 

Hinzen, K.-G. (2008). Can ruins indicate a back azimuth?: 

Seismological Research Letters, 79 (2), 290. 

Hinzen, K.-G. (2009). Simulation of toppling columns in 

archaeoseismology. Bulletin of the Seismological Society 

of America 99 (5), 2855–2875. 

Hurukawa, N. & Maung, P.M. (2011). Two seismic gaps on 

the Sagaing Fault, Myanmar, derived from relocation of 

historical earthquakes since 1918. – Geophysical 

Research Letters 38, L01310, 5 PP., 

doi:10.1029/2010GL046099 

Kannika, P. & Takada, T. (2009). Updating framework for 

site-specific attenuation relation of seismic ground motion 

in Thailand. MSc thesis, Building Research Institute, 

Tsukuba, Japan, 70 p. 

Kázmér, M., & Major, B. (2010). Distinguishing damages of 

two earthquakes – archeoseismology of a Crusader 

castle (Al-Marqab citadel, Syria). In: Ancient 

Earthquakes. (Stewart, I., Sintubin, M., Niemi, T., Altunel, 

E. eds). Geological Society of America Special Paper, 

471, 186 –199. 

Kázmér, M. & Sanittham, K. (2011). Archeoseismology of 

the A.D. 1545 earthquake in Chiang Mai, northern 

Thailand. In: International Symposium on the 2001 Bhuj 

Earthquake and Advances in Earthquake Science, AES 

2011. 22-24 January 2011, Institute of Seismological 

Research, Raisan, Gandhinagar, India. Abstract Volume 

AES 2011, 117-118. 

Korjenkov, A.M. & Mazor, E. (1999). Seismogenic origin of 

the ancient Avdat Ruins, Negev Desert, Israel. Natural 

Hazards 18, 193–226. 

Korjenkov, A.M. c Mazor, E. (2003). Archaeoseismology in 

Mamshit (southern Israel): Cracking a millennia-old code 

of earthquakes preserved in ancient ruins: 

Archäologischer Anzeiger, 2003 (2), 51–82. 

Kovach, R.L & Nur, A. (2006). Earthquakes and archeology: 

Neocatastrophism or science?: Eos, Transactions, 

American Geophysical Union 87 (32), 317. 

Margane, A. & Tatong, T. (1999): Aspects of the 

hydrogeology of the Chiang Mai-Lamphun Basin, 

Thailand that are important for the groundwater 

management. – Zeitschrift für angewandte Geologie 45, 

188-197. 

Michetti, A.M., Esposito, E. et al. (2007). Environmental 

Seismic Intensity Scale 2007 – ESI 2007. In: Vittori, E:, 

Guerreri, L. eds. Memorie Descrittive della Carta 

Geologica d’Italia. LXXIV. Servizio Geologico d’Italia, 

Roma, 7–54. 

Obermeier, S.F. (1996): Use of liquefaction-induced 

features for paleoseismic analysis. – Engineering 

Geology 44, 1-76. 

Pailoplee, S., Takashima, I., Kosuwan, S. & Charusiri, P. 

(2009). Earthquake activities along the Lampang-Theon 

fault zone, northern Thailand: evidence from 

paleoseismological and seismicity data. – Journal of 

Applied Science Research 5 (2), 168-180. 

Pailoplee, S., Sugiyama, Y. & Charusiri, P. (2010). 

Probabilistic seismic hazard analysis in Thailand and 

adjacent areas by using regional seismic source zones. – 

Terestrial, Atmospheric and Ocean Sciences 21, 757– 

766. 

Penth, H. (2006). Earthquakes in Old Lan Na: part of natural 

catastrophes. Chiang Mai University CMU Journal 5 (2), 

255-265. 

Podjarawaraporn (2547). Pra wad Wat Chedi Luang wor ra 

wi harn. [History of Wat Chedi Luang.] Text by Pra Bhuda 

Podjanawaraporn (Chan Kusato), pictures by Pra Kro 

Soponkaweewat (Thanachan Suramanee). Published by 

Wat Chedi Luang, Chiang Mai, Thailand, 160 p. ISBN 

974-92159-8-2 (In Thai) 

Rapp, G., Jr. (1986). Assessing archaeological evidence for 

seismic catastrophes. Geoarchaeology 1, 365–379. 

Reicherter, K., Michetti, A.M. & Silva Barroso P.G. (2009). 

Palaeoseismology: historical and prehistorical records of 

earthquake ground effects for seismic hazard 

assessment. Geological Society, London, Special 

Publication 316, 1-10. 

Rhodes, B.P., Perez, R., Lamjuan, A. & S. Kosuwan, S. 

(2004). Kinematics and tectonic implications of the Mae 

Kuang Fault, northern Thailand. Journal of Asian Earth 

Sciences 24 (1): 79-89. 

Rodríguez-Pascua M.A., R. Pérez-López, J.L. Giner- 

Robles, P.G. Silva, V.H. Garduño-Monroy and K. 

Reicherter (2011). A Comprehensive Classification of 

Earthquake Archaeological Effects (EAE) in 

Archaeoseismology: application to ancient remains of 

Roman and Mexican cultures. Quaternary International. 

DOI: 10.1016/j.quaint.2011.04.044. 

ThinkNet (2010). Map of 14 northern provinces of Thailand 

+ CD. Bilingual Mapping Software. ThinkNet Co., Ltd, 

Bangkok, Thailand. (In Thai and English) 

105


EARTHQUAKE 

ARCHAEOLOGY 



OUTLINE OF THE 3.11 TOHOKU EARTHQUAKE IN JAPAN 

Yoshihiro Kinugasa (1 

(1) Association for the Development of Earthquake Prediction, Tokyo, Japan. Email: king@8f.adep.or.jp 

Abstract (Outline of the 3.11 Tohoku earthquake in Japan): Author will discuss the following subjects on the 3.11 Tohoku 

Earthquake (Mw=9.0), based on the published data by the time of the Corinth2011 workshop. 

1. Seismological Characteristics of the Earthquake; Location, Magnitudes, Focal Mechanism, Intensity Distribution, etc. 

2. Tsunami 

3. Ground Deformations 

4. Re-activation of Normal Fault 

5. Paleoseismology along the Tohoku Coast 

6. Impacts on the Fukushima NPPs. 

Key words: Tohoku Earthquake, Tsunami, Paleoseismology, Fukushima NPPs 

106


EARTHQUAKE 

ARCHAEOLOGY 



THE EVIDENCE OF TSUNAMIGENIC DEPOSITS IN THE GULF OF CORINTH (GREECE) 

WITH GEOPHYSICAL METHODS FOR SPATIAL DISTRIBUTION 

Koster, Benjamin (1), Klaus Reicherter (1), Andreas Vött (2), Christoph Grützner (1) 

(1) Neotectonics and Natural Hazards Group, RWTH Aachen University, Lochnerstr. 4 - 20, 52056 Aachen, Germany. Email: 

b.koster@nug.rwth-aachen.de 

(2) Natural Hazard Research and Geoarchaeology, Institute for Geography, Johannes Gutenberg-Universität Mainz, Johann- 

Joachim-Becher-Weg 21, 55099 Mainz, Germany 

The evidence of tsunamigenic deposits in the Gulf of Corinth (Greece) with geophysical methods for spatial distribution: 

Drill core sampling in coastal areas in the Mediterranean proved evidence for tsunamis. Sedimentary analyses were conducted to 

identify tsunamigenic deposits, but did not reveal larger scale sedimentary structures or spatial distribution of tsunamites in a 

regional scale. We used ground penetrating radar (GPR) in combination with electrical resistivity tomography (ERT) 

measurements and sedimentological research methods in different areas. The combination of these three methods allows us to 

generate 3D visualizations, which give clues for tsunamite distribution and sediment architecture. GPR data indicate 

unconformable thicknesses of tsunamigenic layers, channel-like structures of backwash deposits, in some extent non-planar 

erosion basement, as well as abrasion-scours in various places, and boulder accumulation inside the deposits. 

Key words: tsunami, GPR, ERT, Greece 

INTRODUCTION 

Former studies of various authors on tsunamis 

mainly focused on a hydromechanical analysis of 

specific tsunami events (e.g., Bondevik et al., 2005) 

or sedimentary analyses of drill cores (e.g., 

Reicherter et al., 2010; Shiki et al., 2008; Vött et al., 

2009). The latter method encompasses sieve curves 

as well as magnetic susceptibility measurements and 

micropaleontology to prove tsunamigenic features. 

Characteristics of the sediments, such as finingupward 

sequences, coarse shell debris, upward 

rising magnetic susceptibility and marine foraminifera 

in sandy sediments, give amongst others evidence 

for tsunami events. X-ray fluorescence spectroscopy 

measurements (XRF) were performed in some 

cases. OSL and 14 C-dating can be used for dating. 

All of these studies defined characteristics of the 

deposits (e.g., Bryant et al., 2005; Shiki et al., 2008), 

but do not show the spatial distribution of an event or 

the larger scale sediment structures of tsunami 

deposits. With the knowledge of spatial distribution 

and extent of erosion due to tsunamis it would be 

easier to understand processes during a tsunami 

event and to estimate the possible damage by a 

future tsunami. Since drilling is time-intensive and 

expensive (depending on extend), this method can 

by far not cover an entire coastal area. As the 

distribution and preservation of tsunamigenic 

deposits is highly variable according to several 

studies (e.g., Dawson & Stewart, 2007), there is a 

strong interest in a low-cost and easy to use imaging 

technique. 

Only one published study dealt with GPR for 

detecting tsunami deposits. Switzer et al. (2006) 

investigated a wash-over fan, but did not validate the 

data by other geophysical methods. Therefore, it is 

still not clear whether or not the detected sediments 

were deposited by a tsunami. Furthermore, a limiting 

factor of GPR measurements is a wet environment. A 

shallow ground water table or even sea water 

intrusions close to the ocean can reduce data quality 

significantly. Relative dielectric permittivity r and the 

conductivity of tsunamigenic deposits are 

unknown. However, we can show that GPR has the 

ability to distinguish between tsunami deposits made 

up of marine sands, boulders and shells and clayey 

background sediments although this is an ambitious 

challenge. 

STUDY AREA 

Our study area is located near Lechaion, one of the 

ancient harbors of Corinth (Fig. 1). It was probably 

the most important harbor of this type in antiquity, 

and one of the most important harbors in Greece for 

more than one millennium (Rothaus, 1995). Today it 

is partially buried by up to 2 meters of sediment. 

Fig. 1: Study area in Greece, Lechaion close to Corinth, 

brown areas illustrate topographic elevation; red box 

indicates area of GPR measurements (see Fig. 2 for 

details); green arrow displays possible tsunami propagation 

Lechaion may very well have been affected by a 

series of seismic events and possible tsunami in late 

fourth century after Christ. Reconstruction of the 

107


EARTHQUAKE 

ARCHAEOLOGY 



harbor in AD 353 - 358 (Stiros et al., 1996) 

contingently supports this idea. 

We collect GPR data in combination with drill cores 

and electrical resistivity tomography (ERT) in order to 

test our method in an extraordinary environment, an 

ancient harbor which could have been affected by a 

tsunami (Soloviev, 1990). 

400 MHz antenna promised the best compromise. 

Data processing included static correction, 

background removal, gain adjustment and velocity 

adaption for depth calculations based on a hyperbola 

analysis. Boulders were detected due to hyperbolic 

features in the data. Results of sedimentary drill 

cores (Fig. 4) and ERT profiles in the study area give 

evidence for three tsunami events (Hadler et al., this 

volume). 

VISUALIZATION & RESULTS 

Three GPR profiles were taken parallel to the coast, 

one profile was recorded perpendicular to the 

shoreline (Fig. 2). Three drill cores (Fig. 4) were 

taken between 50 and 150 meters away from GPR 

profiles in the ancient inner harbor. All GPR 

measurements took place on the top of the ancient 

harbor facility, which is buried under a possible third 

tsunami event layer. The base and inner structures of 

the possible tsunami deposits could be imaged in all 

the profiles. 

Fig. 2: Map of the study area with locations of drill points 

and GPR measurements 

METHODS 

GPR measurements were performed in patterns 

directly adjacent to drilling locations and ERT 

profiles. We used the GSSI 400 MHz antenna with a 

survey wheel, the SIR-3000 unit, and a handheld 

GPS (Fig. 3). 

Fig. 3: GPR with 400 MHz antenna, survey wheel, SIR-3000 

unit and GPS 

Trace increment was set to 0.02 m for detailed 

investigation, the range was set to 120 ns TWT and 

the sample rate to 512. From drillings and field 

observations a target depths up to 3.50 m could be 

assumed. The thickness of the assumed tsunami 

sediment layer reaches up to 2.00 m, so the 

Fig. 4: Correlation of drill cores in the study area of 

Lechaion; two possible tsunami layers were detected (red 

boxes with red dashed lines for correlation); these layers 

include fining-upward sequences as well as erosive bases 

108


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 5: GPR profile 276 taken parallel to the coastline at the ancient harbor Lechaion (Greece) in combination with ERT profile 

LEC_ERT_3: A) processed GPR data, B) processed GPR and ERT data; C) processed and combined analyzed data; yellow 

colored areas show tsunamigenic deposits; the base is illustrated by orange line; dashed orange lines refers to inner structures 

Fig. 6: GPR profiles 275, 276 & 277 taken parallel to the coastline at the ancient harbor Lechaion (Greece); direction of profiles is 

plotted in the circle on top left (red arrow is north; both black linings show x- and y-direction of ground surface); the yellow plain is 

a visualized inner structure of the possible upper tsunamigenic deposit; orange lines on the GPR data underline the reflected 

boundaries of layers or inner structures (dashed orange line); big channel structures within the possible tsunamite are clearly 

visible with accumulation of boulders in depressions; as well some smaller channels could be construed as abrasion-scours 

The combination of ERT and GPR measurements in 

the study area suggests that there are bigger channel 

structures with erosive bases and boulder deposits 

inside these channels (Fig. 5 & 6). They point toward 

the ocean and are not part of the buried harbor. The 

channels can be part of a flow-system during the 

backwash processes after a tsunami 

(Dawson & Stewart, 2007). In some cases, channel- 

109


EARTHQUAKE 

ARCHAEOLOGY 



like structures could also be interpreted as abrasionscours, 

which can originate by backwash processes 

with high backflow velocities. Some kind of crossbedding 

is visible in the GPR data as well (Fig. 5). 

The tsunamigenic deposits reaches depths up to 

2.00 m. The sedimentary evidence from the drill 

cores could not be verified due to high attenuation in 

lower depths (>2.50 m). The inner layer-structures of 

the tsunamites (maybe due to multiple waves) show 

an unconformable thickness (Fig. 5). Boulders in the 

sediments appear as hyperbolas (v = 0.12 m/ns). 

Boulders with diameters larger than the resolution 

limit of the 400 MHz GPR antenna are located inside 

the deposits and could be detected by GPR. ERT 

profiles show as well electrical resistivity contrasts at 

the boundary between the tsunamigenic deposits and 

underlying harbor sediments. For the other GPR 

profiles the correlation with ERT data has been done 

similarly, if ERT profiles were available. 

CONCLUSIONS 

Due to highly variable sedimentation processes and 

materials (gravels, sand or silt/clay to some extend 

with boulders) in the Mediterranean and worldwide in 

the context of a tsunami event, deposited sediments 

differ extremely. Therefore, no specific values for 

relative dielectric permittivity r or the conductivity 

can be declared for these variable sediments. Drill 

cores or outcrops are always necessary to prove 

tsunami characteristics and to correlate these results 

with the GPR data. Distinctive contrast changes in 

the GPR data help proving the spatial distribution of 

tsunami deposition interfaces. Only the combination 

of the presented methods is the key for conclusions 

on detailed spatial distribution of tsunamites. 

The main result is the visualization of channelized 

structures in the tsunamigenic horizon. The 

structures most likely originate from backwash 

processes. Our data lead us to conclude that the 

topography of an effected area plays an important 

role for the expansion of the channels, since we 

observed different channel types as well e. g., in 

Spain. It is possible to detect the upper and lower 

boundary of the tsunamite in some cases, depending 

on the grain size of the tsunamigenic material in 

contrast to the background sedimentation. 

With the 400 MHz GPR antenna it is also possible to 

detect bigger structures like abrasion-scours, 

unconformable thicknesses of tsunamigenic 

beddings, in some extent non-planar erosion 

basement and boulder accumulations inside the 

deposits. Typical thinning-inland structures (Dawson, 

1994) could not be detected in this case within the 

GPR profile perpendicular to the coast. A GPR with 

higher resolution should be useful to detect further 

sedimentary structures in tsunamigenic deposits in 

the future. 

OUTLOOK 

In the future, GPR and other shallow geophysical 

methods will be used to detect run-up distances and 

for creating large-scale models considering 

topography to detect sediment thickness and volume. 

With these data it would be possible to calculate the 

physical power and the possible damage of the 

tsunami wave and typical sediment structures. Due to 

account on spatial distribution information it could be 

possible the reconstruct the topography of the 

landscape before and after a paleo-tsunami. 

Another aim is to get detailed information of the 

deposits by trenching and using methods like LiDAR, 

multispectrometry, magnetic susceptibility and the 

documentation commonly used in archeological 

excavations.3-dimensional data or block-plots can be 

generated based on these methods to evaluate new 

features and characteristics of tsunamigenic 

deposits. 

References 

Bondevik, S., F. Løvholt, C. Harbitz, J. Mangerud, A. 

Dawson & J.I. Svendsen, (2005). The Storegga Slide 

tsunami – comparing field observations with numerical 

simulations. Marine and Petroleum Geology 22, 195-208. 

Bryant E., R.L. Wiegel, F.P. Shepard & D. Myles, (2005). 

Natural hazards. Cambridge University Press, 2nd 

edition, 214-225. 

Dawson, A.G. & I. Stewart, (2007). Tsunami deposits in the 

geological record. Sedimentary Geology (200), 166-183. 

Dawson, A. G. (1994). Geomorphological effects of tsunami 

run-up and backwash. Geomorphology 10, 83-94. 

Hadler, H., A. Vött, B. Koster, M. Mathes-Schmidt, T. 

Mattern, K. Ntageretzis, K. Reicherter, D. Sakellariou, T. 

Willershäuser, (2011). Lechaion, the ancient harbour of 

Corinth (Peloponnese, Greece) destroyed by 

tsunamigenic impact. 2nd INQUA-IGCP-567 International 

Workshop on Active Tectonics, Earthquake Geology, 

Archaeology and Engineering, Corinth, Greece. This 

volume. 

Reicherter, K., D. Vonberg, B. Koster, T. Fernández- 

Steeger, C. Grützner & M. Mathes-Schmidt, (2010). The 

sedimentary inventory of the 1755 Lisbon tsunami along 

the southern Gulf of Cádiz (southwestern Spain). 


147-173. 

Rothaus, R., (1995). Lechaion, western port of Corinth: a 

preliminary archaeology and history. Oxford Journal of 

Archaeology 14 (3), 293-306. 

Shiki, T., T. Tachibana, O. Fujiwara, K. Goto, F. Nanayama 

& T. Yamazaki, (2008). Characteristic features of 

tsunamiites. In: Tsunamiites – Features and Implications 

(T. Shiki, Y. Tsuji, T. Yamakazi & K. Minoura eds). 

Elsevier Scientific Publishing Company, 319-336. 

Soloviev, S.L., (1990). Tsunamigenic Zones in the 

Mediterranean Sea. Natural Hazards (3), 183-202. 

Stiros, S., P. Pirazzoli, R. Rothaus, S. Papageorgiou, J. 

Laborel & M. Arnold, (1996). On the Date of Construction 

of Lechaion, Western Harbor of Ancient Corinth, Greece. 

Geoarchaeology 11 (3), 251-263. 

Switzer A.D., C.S. Bristow & B.G. Jones, (2006). 

Investigation of large-scale washover of a small barrier 

system on the southeast Australian coast using ground 

penetrating radar. Sedimentary Geology 183, 145-156. 



Fountoulis, (2009). The Lake Voulkaria (Akarnania, NW 

Greece) palaeoenvironmental archive - a sediment trap 



1-3. 

110


EARTHQUAKE 

ARCHAEOLOGY 



COSEISMIC SURFACE RUPTURING IN THE EPICENTRAL AREA OF GERMANY’S 

STRONGEST HISTORICAL EARTHQUAKE 

S. Kübler(1), A. M. Friedrich(1), M. R. Strecker(2) 

(1) Ludwig-Maximilians-University Munich, Department of Earth and Environmental Sciences, Luisenstr. 37, 80333 Munich, 

Germany, Email: kuebler@iaag.geo.uni-muenchen.de 

(2) University of Potsdam, Institute of Geosciences, Karl-Liebknecht-Str. 24, 14476 Potsdam, Germany 

Abstract (Coseismic surface rupturing in the epicentral area of Germany’s strongest historical earthquake): The Lower 

Rhine Embayment is one of the most seismically active regions in intraplate Europe. A trenching investigation carried out south of 

the city of Düren near the estimated epicenter of Germany's largest historical event (1756 AD, M L 6.2 ± 0.2) revealed evidence of 

significant coseismic deformation of the earth's surface. Deformation is expressed by co-planar sets of fractured and rotated 

pebbles as well as liquefaction of fine-sand deposits. Two event horizons indicate an older rupturing event that occurred 

presumably in the Early to Mid Holocene and a younger rupturing event that occurred in the Latest Holocene, and may correspond 

to the 1756 Düren earthquake. 

Key words: paleoseismology, historical earthquake, coseismic rupture, Germany 

DISCUSSION 

Based on historical documents and instrumental data 

the Lower Rhine Embayment is currently one of the 

most seismically active regions in intraplate Europe. 

At least 21 M > 5 instrumental and historical 

earthquake events have been recorded and 

documented, respectively, for western Germany, 

eastern Belgium and southern Netherlands 

(Leydecker, 2002). One of the largest of these events 

was the earthquake of February 18 th 1756 near the 

city of Düren at the western border of the LRE. 

Damage related to this event included triggered 

landslides and destroyed buildings and castles, but 

the occurrence of surface ruptures has not been 

reported for this or any other historic event in this 

region. 

The LRE is characterized by NW-SE striking normal 

faults. Fault plane solutions indicate an extensional 

normal faulting regime (Hinzen, 2003). Several 

paleoseismic studies in the LRE indicate that strong, 

surface-rupturing events may have repeatedly 

occurred in this region since the Late Pleistocene 

(e.g., Camelbeeck & Meghraoui, 1998; Vanneste et 

al., 2001; Vanneste & Verbeeck, 2001), but there has 

been (Ahorner, 1996) and continues to be (Houtgast 

et al., 2003, 2005) controversy about whether faults 

in the LRE mainly move by creep, or whether 

earthquakes that are large enough to break the 

surface commonly occur in this region. The 1756 

Düren earthquake had a local magnitude of 6.2 ± 0.2 

based on empirical studies (Meidow, 1994), the total 

moment for this event was estimated at 1.6 x 10 17 

Nm (M W 6.1; Hinzen & Reamer, 2007). Events of this 

magnitude are commonly associated with surface 

ruptures, but there is still great uncertainty in the 

reliability of historic documents for estimating the 

correct magnitude of historic events. Therefore, the 

identification of a surface rupture related to the Düren 

event would be important to better understand the 

mechanical behaviour of faulting in this tectonically 

active region. 

In contrast to paleoseismic and geomorphic studies 

in arid regions, where fault scarps are exposed for 

many kilometers along strike and their preservation 

potential is excellent, the recognition and 

characterization of potentially active faults is much 

more difficult in the moderately humid Lower Rhine 

area. The dense vegetation cover and intense 

agricultural landuse hamper the recognition of 

seismogenic surface ruptures in the densely 

populated region. Low displacement rates on 

individual faults, and hence a sparse earthquake 

record due to long recurrence intervals of large 

seismic events, additionally aggravate seismic 

hazard evaluation. Thus, multidisciplinary 

paleoseismic studies are important to identify 

seismically active fault segments in the LRE. 

In order to search for a potential surface rupture 

related to the 1756 Düren event, we carried out 

multidisciplinary reconnaissance studies including 

geomorphic mapping, shallow drilling and shallow 

geophysical prospecting in the vicinity of Düren, 

wherever we could find late Holocene fluvial deposits 

covering the projection of potentially active faults. At 

one location, we identified two gentle 0.5 and 0.6 m 

high surface scarp in late Holocene deposits in 

prolongation of the E-dipping, 16-km long Schafberg 

fault. The trends of the two scarps differ: the south- 

111


EARTHQUAKE 

ARCHAEOLOGY 



western step strikes 125° and is slightly curved, 

whereas the north-eastern step is perfectly straight 

and strikes 155°. The study site is situated a few 

kilometres W of Untermaubach, where the Schafberg 

fault crosses the Rur river valley. Geophysical 

prospecting identified two anomalous zones 

characterized by resistivity minima, high near-surface 

conductivity, and offset radar reflectors that coincide 

with the two 0.5 and 0.6 m high surface scarps 

(Streich, 2003). 

We excavated the fault along an up to 5 m deep and 

85 m long trench approximately 200 m west of the 

current Rur river course. At the trench site, the 

Schafberg fault is covered by < 5 m thick, scarcely to 

moderately well layered Holocene sandy gravel 

deposits and fine-grained flood deposits overlaying 

Lower Devonian shale and sandstone. The trench 

exposes two asymmetric channels that coincide with 

the position of the observed surface scarps. The 

channel fill consists of sandy silt and clay-rich layers. 

At the base of both channels, we detected a high 

concentration of organic material including trees and 

branches. We mapped various types of soft-sediment 

deformation in the gravel deposits including tilted and 

rotated clasts as well as fractured clasts that show 

offsets on a millimeter to centimeter scale. We further 

detected liquefaction features in sand and silt 

deposits. 

Our mapping revealed a narrow zone of localized 

deformation below the north-eastern scarp expressed 

by abundant fractures with aligned and broken clasts 

extending vertically throughout the entire gravel unit. 

We carried out detailed gravel analysis including 

mapping of 237 fractured clasts and the long-axis 

orientation of ~ 10.000 clasts. Results define a ~ 10 

m wide fault zone that coincides with the surface 

offset and an almost 40 m wide deformation zone 

where gravel deformation is still prominent. In 

contrast, gravel below the south-western scarp 

exhibits no indicators of coseismic deformation and is 

therefore most likely a fluvial channel. Liquefaction 

occurs 20 m east and 50 m west of the suspected 

rupture zone. As the overlaying flood deposits do not 

show signs for seismogenic deformation, the contact 

between the gravel and overlaying silt and clay 

appears to mark the most recent event horizon. An 

older event horizon – a paleosoil with high amounts 

of organic material – is preserved at a depth of 2.5 m 

within the gravel deposits. Here, underlying gravel 

deposits show clear deformation features while 

gravel above this marker horizon is undisturbed. The 

maximum vertical displacement across the fault 

appears to range between 0.8 and 1.2 m based on 

two offset marker horizons. This is in accordance 

with estimates derived from borehole data, 

geophysical prospecting and morphometric analysis. 

We interpret the observed deformation features, 

especially the co-planar sets of rotated and fractured 

clasts, as the result of coseismic deformation at the 

near-surface end of the rupture, and we rule out slow 

deformation due to aseismic creep as governing 

process to cause rupturing of pebbles this close to 

the surface. Preliminary radiocarbon data bracket the 

younger event horizon to Latest Holocene age, which 

rules out periglacial processes as cause for the 

observed sediment deformation. Age results also 

imply that this fault may be a possible source for the 

1756 earthquake. Further analyses are in progress. 

We identified coseismic deformation at the trench 

site, because special conditions produced a number 

of features not normally observed in other fault 

exposures. The thin sedimentary cover (< 5 m) 

above basement rocks and the high water table may 

have played an important role in producing this 

unusual deformation pattern. However, this newly 

investigated trench site yields the first evidence for 

Late Holocene seismogenic surface rupturing in the 

German part of the LRE, and thus confirms the 

importance of paleoseismic studies for seismic 

hazard analysis in humid low-strain intraplate 

regions. 

Acknowledgements: Trench excavation and geological 

field work was financed by the DFG, research grant 

awarded to Friedrich and Strecker (AF 6513). 

References 

Ahorner, L., 1996, How reliable are speculations about large 

paleo-earthquakes at the western border fault of the Roer 

Valley Graben near Bree, In: Bonatz, M., ed., Comptes- 

Rendus des 81ièmes Journées Luxembourgeoises de 

Géodynamique, Walferdange, Grand Duchy of 

Luxemburg, p. 39-57. 

Camelbeeck, T. and M. Meghraoui (1998). Geological and 

geophysical evidence for large paleoearthquakes with 

surface faulting in the Roer Graben (northwest Europe). 

Geophys. J. Int., 132, 347 - 362. 

Hinzen, K. G. (2003). Stress field in the Northern Rhine 

area, Central Europe, from earthquake fault plane 

solutions, Tectonophysics, 377, 325 - 356. 

Hinzen, K. G., and S. K. Reamer (2007). Seismicity, 

seismotectonics, and seismic hazard in the Northern 

Rhine Area, In: Intraplate Seismicity (Stein, S. and 

Mazzotti, S. eds). GSA Books, 225-243. 

Houtgast, R.F., R. T. van Balen, K. Kasse, and J. 

Vandenberghe (2003). Late Quaternary tectonic evolution 

and postseismic near surface fault displacements along 

the Geleen fault (Feldbiss fault zone-Roer valley rift 

system, the Netherlands), based on trenching, Geologie 

en Mijnbouw, 82(2), 177-196. 

Houtgast, R.F., R. T. van Balen, and C. Kasse (2005). Late 

Quaternary evolution of the Feldbiss fault (Roer valley rift 

system, the Netherlands) based on trenching, and its 

potential relation to glacial unloading, Quaternary 

Science Reviews, 24(3-4), 489-508. 

Leydecker, G. (2002). Earthquake Catalogue for the 

Federal Republic of Germany and Adjacent Areas for the 

Years 800 - 2001. Datafile, Federal Institute for 

Geosciences and Natural Resources, Hannover, 

Germany. 

Meidow, H. (1994). Comparison of the macroseismic field of 

the 1992 Roermond earthquake, the Netherlands, with 

those of large historical earthquakes in the Lower Rhine 

Embayment and its vicinity, Geologie en Mijnbouw, 73(2- 

4), 282-289. 

Streich, R (2003). Geophysical prospecting of suspected 

Holocene fault activity in the Lower Rhine Embayment, 

Germany, MSc Thesis. University of Potsdam, Germany. 

Vanneste, K. and Verbeeck, K. (2001). Paleoseismological 

analysis of the Rurrand fault near Jülich, Roer Valley 

graben, Germany: coseismic or aseismic faulting history? 

Geologie enMijnbouw, 80, 155-169. 

112


EARTHQUAKE 

ARCHAEOLOGY 



Vanneste, K., Verbeeck, K., Camelbeeck, T., Paulissen, E., 

Meghraoui, M., Renardy, F., Jongmans, D. and Frechen, 

M. (2001). Surface-rupturing history of the Bree fault 

scarp, Roer Valley graben: evidence for six events since 

late Pleistocene. Jour. Seism., 5, 329-359. 

113


EARTHQUAKE 

ARCHAEOLOGY 



SEISMOTECTONIC AND SEISMIC HAZARD MAPS OF LITHUANIA (BALTIC REGION) – 

RECENT IMPLICATIONS OF INTRACRATONIC SEISMICITY 

Lazauskiene, Jurga (1 and Pacesa, Andrius (2) 

(1) Lithuanian Geological Survey under the Ministry of Environment. S. Konarskio 35; Vilnius University, M.K. iurlionio 21/27, 

(1) LT-03123 Vilnius, Lithuania. Email: jurga.lazauskiene@lgt.lt, 

(2) Lithuanian Geological Survey under the Ministry of Environment. S. Konarskio 35, LT-03123 Vilnius, Lithuania. Institute of 

(2) Geology and Geography, T.Ševenkos, 13, LT-03123 Vilnius, Lithuania. Email: andrius.pacesa@lgt.lt, 

Abstract (Seismotectonic and seismic hazard maps of Lithuania (Baltic region) – recent implications of intracratonic 

seismicity): Lithuania, situated in the western part of the East European Craton, is regarded as an intracratonic area of low 

seismicity. Several dozens of earthquakes of intensity up to VII (MSK-64) were recorded since 1616 implying the possible 

occurrence of stronger earthquakes. The northern part of in the Baltic region is seismically more active than the southern, but the 

Kaliningrad earthquakes of 2004 showed the necessity to re-assess the seismicity of the region. The identification of seismogenic 

faults in the Baltic region is rather complicated due to the small scale of tectonic structures and significant errors of location of 

seismic events and even faults. Still, recently the seismic hazard and seismotectonic maps of Lithuania were compiled implying the 

highest seismic hazard of 32,6 cm/s 2 PGA in eastern and 25-30 cm/s 2 in northern Lithuania. The majority of the territory is 

described by PGAs of 10-20 cm/s 2 . 

Key words: intracratonic seismicity, Baltic region, seismic hazard, seismotectonic framework, 

INTRODUCTION 

The territory of Lithuania comprises a part of the 

Baltic sedimentary basin situated in the western part 

of the East European Craton that is characterised by 

low seismic activity - the historical sources since 

1616 to 1964 record only a few tens of weak or 

moderate earthquakes (Pasa et al., 2005). The 

historical seismic activity in the eastern Baltic region 

is significantly lower comparing with seismicity of the 

Fennoscandian shield. High seismic activity of the 

Fennoscandian shield and adjacent Baltic Sea 

territories during the Late Glacial and Holocene (last 

13 000 years) is also well documented by numerous 

paleoseismic investigations and corresponding 

publications. The earthquakes caused landslides in 

glacial till, seismically-induced soft sediment 

deformation structures, “seismites”, are common in 

trench exposures in the vicinity of the faults in 

northern Sweden and even with tsunami events 

reported in the Baltic Sea (Mörner, 2005, 2008). 

Still, the eastern Baltic region is more seismically 

active comparing with the more “inland” aseismic 

territories of the craton. Several tens (~ 40) of 

earthquakes of intensities of VI-VII (MSK-64 scale) 

and local magnitudes up to M L = 5 are recorded in the 

Baltic region and neighbouring since 1616 (fig. 1) 

The strongest instrumentally registered earthquakes 

are Ossmussare (Estonia) earthquake of 1976 (with 

maximal magnitude up to M L = 4,75) and the 

Kaliningrad (Russia) earthquakes of 2004 of 

magnitudes, respectively, M L = 4.75 and M L = 5.0 (M w 

= 5.2, table 1). The other more significant 

earthquakes in the region are: 

Fig. 1: Main tectonic faults and seismic events in the Baltic 

region (after Sharov (ed.) et al., 2007; Pasa et al., 2005 ). 

1–3 faults: 1 – superregional, 2 – regional, 3 – subregional; 

5–8 – epicentres of earthquakes with local magnitudes: 

5 – M L = 1-2, 6 – M L = 2-3, 7 –M L = 3-4, 8 – M L = 4-5. 

February 22, 1821, Kokneses (Estonia), M L = 4.5; 

December 28, 1908, Gudogai (Belarus), M L = 4.5; 

114


EARTHQUAKE 

ARCHAEOLOGY 



December 29, 1908, Madona (Latvia); M L = 4.5 

(Boborikin et al., 1993). All the mentioned 

earthquakes within the same Baltic sedimentary 

basin in the same tectonic setting show the recent 

seismic activity of the Baltic region. 

INTRACRATONIC SEISMICITY OF THE BALTIC 

REGION 

The seismic activity in the Baltic region has an 

irregular distribution - the northern part of the region 

is more seismically active than the southern one (fig. 

1). The maximum activity is recorded in Latvia that is 

characterized by most intense faulting of the 

sediment layers. Seismic activity is slightly lower in 

Estonia, while the territory of Lithuania seems to be 

the most quite. The boundary between two areas of 

different seismicity approximately coincides with the 

northern state border of Lithuania. A quite similar 

boundary between northern and southern parts of the 

region was established in previous global seismic 

hazard assessment studies e.g. WSHAP (Giardini, 

1999) and European–Mediterranean Seismic Hazard 

Map (Jimenez et al., 2003). 

No earthquakes are registered instrumentally or 

reliably recorded historically in the territory of 

Lithuania. Peter of Duisburg, a 14th Century 

chronicler of the Teutonian Knights (Chronicon 

Terrae Prussiae) reported “a ground shaking which 

was felt in Skirsnemune castle” (south-western part 

of Lithuania) in 1328; after the event the castle was 

abandoned. However, this historical record causes 

some doubts and is considered controversially: 

A. Nikonov (personal communication) includes this 

seismic event of 1328 (and one of 1303 in Prussia; 

(Chronicon Terrae Prussiae)) into revised 

seismological catalogue of the South-East Baltic 

region, while Grunthal and Riedel (2007) strongly 

deny the existence of these events. Thus, this event 

was not included to the complete catalogue of 

seismic events of Eastern Baltic Region. There are 

also three records about the ground shaking during 

the years 1908-1909 in western and central part of 

Lithuania, but the primary sources rises some doubts 

and could not be considered as reliable. The first 

local seismic event with magnitude of M = 2.1 

presumably of tectonic origin was instrumentally 

registered 4-th of September, 2001 by one of short 

period seismic station, located in the NE part of 

Lithuania. The seismic signal of this seismic event 

was the typical one for tectonic events (Pasa et al., 

2002). Still, as it was registered only in one station, 

no accurate location of this event was possible and 

the epicentre of the earthquake could have been 

located in a distance of ~80 km from the station. The 

passive seismic experiment PASSEQ was carried out 

in the eastern part of Europe, including territory of 

Lithuania in 2006–2007. The preliminary data 

received from the project indicated some possibility 

of tectonic seismic event in the middle Lithuania 4-th 

July, 2007 (coordinates - 55,053; 24,264) and one 

more event in Kaliningrad offshore of the Baltic Sea, 

14-th June, 2007 (coordinates - 54,796; 19,232) 

(Kozlovskaya et al., 2010). The coordinates of the 

epicentre of the onshore Lithuania event were 

identified near Kaunas town, located within the 

Middle Lithuanian Suture zone (the zone between 

two different Precambrian domains) with possibly 

increased tectonic activity. Both onshore and the 

Baltic Sea offshore events have occurred at the 

night-time. The seismic signals have the 

characteristics of tectonic events, but the quality of 

data is too poor to define magnitudes and depths of 

epicentres of these seismic events and, respectively, 

too poor for the reliable conclusions. Also, it must be 

noted, that the network of local seismic stations in the 

eastern Baltic region, especially Lithuania, is rather 

sparse and the instrumental seismological 

information is quite poor. 

No doubts, the seismic events in Kaliningrad area 

showed the necessity to review the understanding of 

the seismicity of the Baltic region - earlier it was 

considered that the maximum magnitude of the 

earthquakes in this part of region might be M L = 4.8; 

but the magnitude of the Kaliningrad earthquake was 

M L = 5.0 (M w = 5.2; table 1). Thus, taking into 

consideration the accepted margin of 0.5 (based on 

common agreement), the maximum magnitude could 

be assumed to be as high as M w = 5.7 that implies 

the possibility of occurrence of some stronger 

earthquakes in the Baltic region. 

Date Time Lat Long Magn Depth 

10/25/1976 8:39:00 59.2 23.58 4.7 M L 15 - 18 

9/21/2004 11:05:04 54.908 20.029 5.0 M w 16 ± 9 

9/21/2004 13:32:31 54.849 20.088 5.2 M w 20 ± 10 

Table 1. The strongest seismic events of Eastern Baltic 

region with magnitude M>4 (compiled by A. Pasa: the first 

one - Ossmussare (Estonia) earthquake, the other two – 

Kaliningrad (Russia) earthquakes. Magnitude types: M L – 

local magnitude, M w – moment magnitude. Coordinates of 

epicentres are provided in geographical system of 

coordinates: Lat - latitude (North), Long – longitude (East); 

time – in GMT; Magn – magnitude; Depth - in km. After: 

Gregersen et al., 2007. 

The seismotectonic framework of the study areas has 

been outlined several times during the last decades 

(Sharov et al., 2007; Suveizdis, 2003). A number of 

faults and fault zones have been distinguished in the 

Baltic region and adjacent territories based on 

geological and geophysical data. Still, different 

authors provide quite different tectonic and 

seismotectonic maps of the eastern Baltic region and 

there is no single commonly accepted tectonic map of 

this region currently; location, orientation, length or the 

other parameters of the same fault might be interpreted 

differently. No doubts, summarizing different maps one 

can infer some dominating fault zones, directions and 

spacing (fig. 1). 

DISCUSSIONS 

As it was mentioned, majority of seismic events in the 

eastern Baltic region are historical ones. The primary 

sources of information do not provide any evaluations 

of errors of epicentre locations for historical events, 

but, most likely, the errors vary from ten to several 

115


EARTHQUAKE 

ARCHAEOLOGY 



tens of kilometres. Even the epicentres for the two 

strongest Kaliningrad earthquakes were scattered in 

the area with a diameter of about 30 km (Pasa et 

al., 2005). It is very hard to associate single 

earthquakes with some certain faults unambiguously 

due to significant errors of location of seismic events 

and even location of faults. Additionally, the 

identification of the seismogenic faults in the region is 

rather complicated due to the small scale of tectonic 

structures located within an intracratonic area. It must 

also be pointed out that not all the earthquakes in the 

Baltic region are related to fault zones. Moreover, the 

majority of the previous global seismogenic and 

seismic hazard studies (Grünthal et al., 1999; 

Jimenez et al., 2003) that included also the territory 

of the Baltic region, strongly implied the local 

seismogenic sources with diffused seismicity. Thus, it 

was rather complicated to understand the 

geodynamic control on the seismicity of the relatively 

seismotectonically non-active cratonic area. 

Accordingly, until recently, no any seismic hazard 

map has been compiled neither for the territory of 

Lithuania, nor for the wider Baltic region. Finally, in 

year 2011, on request of the Lithuanian Geological 

Survey, the seismic hazard and seismotectonic maps 

(Šliaupa, 2011) have been compiled for the territory 

of Lithuania. The complex analysis of the geological, 

geophysical, geodetic, structural, seismic and 

geodynamic data allowed to distinguish 5 active 

seismogenic zones and 5 potentially active 

seismogenic zones in the territory of Lithuania. The 

maximal seismic potential has been implied for the 

W-E trending Silute-Polock and Northern Prieglius– 

Birstonas fault zones (Northern Prieglius has 

„hosted“ Kaliningrad earthquakes), transecting the 

central and southern part of Lithuania, with a 

maximum magnitude up to M = 5,5 of the possible 

earthquake predicted. It is implied that possible 

seismic activity of the faults is related to the regional 

stress field that affects the lithosphere of the Baltic 

region. No uniform stress pattern can be found for 

Baltic countries. Two stress provinces are suggested 

in Lithuania: NW-SE horizontal compression in the 

western part, whereas the main horizontal stress in 

the eastern part is NE-SW oriented. The western 

zone is attributed to the North Atlantic stress 

province, while the eastern is possibly a part of the 

Mediterranean stress province (Šliaupa, 2011; fig. 2). 

Therefore, as the seismic stresses are rather variable 

within the territory of Lithuania this might influence 

the character of the fault activation. 

The newly compiled seismic hazard map of Lithuania 

(Šliaupa, 2011) shows that the highest seismic 

hazard with peak ground acceleration equal to 32,6 

cm/s 2 (0,025 g) with 10% probability of exceedance 

within 50 years could be expected in the eastern part 

of Lithuania; also, the increased seismic hazard 

(peak ground acceleration being equal to 25-30 

cm/s 2 (0,030 g) with 10% probability of exceedance 

within 50 years) also is estimated in the northern part 

of Lithuania. Therefore, the majority of the territory of 

Lithuania is described by rather law seismic hazard - 

peak ground acceleration values varies in a range of 

10-20 cm/s 2 with 10% probability of exceedance 

within 50 years. 

Fig. 2. Directions of maximum dilatation in the territory of 

Lithuania (Sliaupa, 2011). Grey areas roughly indicate N-S 

maximum dilatation, blank – NW-SE maximum dilatation; 

areas subjected to bi-axial compression are bounded by 

dotted countures.Triangules indicates GPS points of the 

zero class; squares – GPS points if the first class; numbers 

-the numbers of GPS stations; dotted lines – maximal 

dilatation. 

The results of the new seismic hazard assessment 

show good coincidence with the results of previous 

global seismic hazard assessment - according to the 

results of Global Seismic Hazard Assessment 

Program (Grünthal et al., 1999) seismic level in the 

territory of Lithuania was estimated in the range of 

10–30 cm/s 2 for the standard seismic level of civil 

engineering (10% probability of exceedance within 50 

years or for a return period of 475 years) and 

according to the European map of seismic hazard 

(ESC-SESAME project, (Jimenez et al., 2003) the 

standard seismic level of civil engineering (return 

period of 475 years) was estimated as high as 

20cm/s 2 . 

References 

Boborikin, A..M., Avotinia, I.Y., Yemelianov, A. P., Sildvee, 

N. N., Suveizdis, P. (1993). Catalogue of historical 

earthquakes of Belarus and the Baltic Region. 

Seismological report of seismic stations of Minsk– 

Pleshchenitsi and Naroch for 1988. Minsk. 126–137. 

Kozlovskaya, E., Budraitis, M., Janutyte, I., Motuza, G., 

Lazauskiene, J. and PASSEQ-Working Group, (2010). 3- 

D crustal velocity model for Lithuania and its application 

to local event studies. EGU General Assembly 2010, 

Vienna, Austria, 02-07 May, 2010, Abstract and 

programme book, Abstract No. EGU2010-10625. 

Giardini, D. (Ed.). 1999. The Global Seismic Hazard 

Assessment Program 1992–1999. Annali Geofis. 42 (6) 

Special Issue. 

Gregersen S., P. Wiejacz, W. Debski, B. Domanski, B. 

Assinovskaya, B. Gutterch, P. Mantiniemi, V.G. Nikulin, 

A. Pacesa, V. Puura, A.G. Aronov, T.I.Aronova, G. 

Grunthal, E.S. Husebye, S. Sliaupa. (2007). The 

exceptional earthquakes in Kaliningrad district, Russia on 

September 21, 2004. Physics of the Earth and planetatry 

interiors 164, 63-74. 

Grünthal, G. and GSHAP Region 3 Working Group. (1999). 

Seismic Hazard Assessment for Central, North and 

Northwest Europe: GSHAP Region 3. Annali di Geofisica 

42 (6), 999–1011. 

116


EARTHQUAKE 

ARCHAEOLOGY 



Grünthal, G., Riedel, P. (2007). Zwei angebliche Erdbeben 

in den Jahren 1303 und 1328 im heutigen Raum 

Kaliningrad. Zeitschrift für Geologische Wissenschaften. 

Berlin, 35 (3). 157–163 (in Germany with English 

summary). 

Jiménez, M. J., Giardini, D. and Grünthal, G. (2003). The 

ESC-SESAME Unified Hazard Model for the European– 

Mediterranean Region. EMSC/CSEM Newsletter. 19, 2– 

4. 

Mörner, N.A. (2005). An interpretation and catalogue of 

Paleoseismicity in Sweden. Tectonophysics 408, 265- 

307. 

Mörner, N.A. (2008). Tsunami events within Baltic. Polish 

Geological Institute Special Papers, 23. 71-76. 

Pasa, A., Šliaupa, S., Satknas J. (2005). Recent 

earthquake activity in the Baltic region and seismological 

monitoring in Lithuania. Geologija 50, 8-18. (in 

Lithuanian). 

Pasa, A., Narbuntas, J., Nemcovas, G. (2002). The 

annual bulletin of the seismological monitoring in 

Lithuania. 2001. Vilnius. 40 p. (internal report in 

Lithuanian with English summary). 

Sharov N. V., Malovichko A. A., Schukin Ju. K. (eds.). 

(2007). Earthquakes and microseismicity of Eats 

European Platform – recent geodynamic approach. Book 

1 Earthquakes. Petrozavodsk: KarNC RAN, 381 p. (in 

Russian). 

Suveizdis (ed.). (2003). Tectonic structure of Lithuania. 

Institute of Geology and Geography, Vilnius. 160 p. (in 

Lithuanian). 

Šliaupa, S. (2011). Assessment of the seismic observations 

in the territory of Lithuania. Book 2 - Seismic hazard map 

of Lithuania. 136 p. (internal report, in Lithuanian). 

117


EARTHQUAKE 

ARCHAEOLOGY 



PRELIMINARY STUDY ON DAMAGED STONE MONUMENTS IN GYEONGJU, SE KOREA 

Minjung Lee (1), Young-Seog Kim (1,*) 

(1) GSGR, Dept. of Earth Environmental Sciences, Pukyong National University, Daeyeon3-dong, Busan 608-737, Korea. 

*Email: ysk7909@pknu.ac.kr 

Abstract (Preliminary study on damaged stone monuments in Gyeongju, SE Korea): Concerns over the possibility of future 

earthquakes are high in Eastern Asia, since the catastrophic earthquake in north-eastern Japan on the 11 th of March, 2011. The 

Korean peninsula is regarded as a relatively safe area from earthquakes because it is located within the Eurasian intracontinental 

region. According to historical records, the Gyeongju area was struck by many large earthquakes. Recently a long crack was 

discovered on the east facing exterior surface of Seokgatap, a famous pagoda in Gyeongju. The crack length is 132 cm and its 

width is 5 mm. During the investigation into the cause of the damage, we performed kinematic analysis of the cracks on the 

pagoda. We cannot find any consistency between the damage of the pagoda and earthquakes, recently cracks were found within 

the pagoda, which was most probably developed by a prolonged continuous force. Other probable earthquake damage can, 

however, be seen in other historical heritage sites within this city and its surrounding regions. Some of the damage at this and 

other sites are also recorded in historical references. Therefore, more careful studies are necessary to identify and distinguish the 

origins of the damage. 

Key words: stone heritage, physical damage, Gyeongju, paleoseismology. 

INTRODUCTION 

The Korean peninsula is located within the relatively 

safe Eurasian intracontinental region. In some 

neighbouring countries around Korea such as Japan 

and Taiwan, big earthquakes occur frequently. Over 

forty Quaternary faults were recently discovered 

along the Yangsan and Ulsan faults which are major 

tectonic features in SE Korea (Fig. 1). According to 

Korean historical records (e.g. Samguksagi, 

Mukseojipyeon), several relatively strong seismic 

events, greater than intensity VII, have affected the 

Korean peninsula in the past (Lee and Yang, 2006). 

Gyeongju is a good place to study damage of stone 

heritages, because it has many stone artefacts at its 

heritage sites and it was the capital city of the Silla 

Dynasty for almost 1000 years from 57 BC to 935 AD. 

During this time, many large seismic events (over 

intensity VII) affected the Gyeongju and Ulsan areas 

destroying many heritage sites. In AD 779, a huge 

seismic event was recorded and resulted in about 

100 casualties. A number of buildings were 

destroyed due to this earthquake. It is inferred to be a 

destructive earthquake with a magnitude of 6.7 (Lee 

& Yang, 1983, 2006). The purpose of this study is to 

examine stone heritage sites and to distinguish and 

interpret the affects of recorded paleoseismological 

events at cultural heritage sites in the Gyeongju area. 

GENERAL GEOLOGY 

The basement of the study area consists of 

Cretaceous sedimentary rocks, the Taegu Formation, 

which forms part of the Gyeongsang Basin. This 

formation is discordantly intruded by Cretaceous and 

Tertiary igneous rocks (Fig. 1). In recent years, more 

than forty Quaternary faults have been reported near 

the Yangsan and Ulsan fault systems, which are the 

major fault systems in and around the Gyeongsang 

Basin (Lee and Jin, 1991; Kyung & Okada, 1995; 

Chang, 2001; Kim and Jin, 2006). The geometry of 

the intersection zone between the Yangsan and 

Ulsan faults is similar to simulated -faults (a low 

angle merging fault system; Du & Aydin, 1995, Kim et 

al., 2000). 

Fig. 1: Location map and geological map of the study area 

(modified from Lee et al., 1995). 

EARTHQUAKE RECORDS 

Instrumental recording of earthquakes in Korea 

began in 1905, and about one thousand earthquakes 

of mostly small magnitude (


EARTHQUAKE 

ARCHAEOLOGY 



between 2 to 1904 AD (Jin et al., in press, Lee & 

Yang, 2006). 

Based on the compiled earthquake catalogue (Fig. 2), 

more reliable records were reported after the Koryeo 

Dynasty. We have identified 3 clusters and their 

respective recurrence intervals. We observed that 

after the first cluster there was a recurrence interval 

of 368 years. After this recurrence interval, 

earthquakes occurred frequently over a period of 193 

years. This was followed by another recurrence 

interval of 53 years which was in turn followed by a 

110 year period of repeated earthquakes. We are 

now 200 years into a period without earthquake 

events. 

DESCRIPTION OF THE SEOKGATAP 

Bulguksa temple was built in AD 751 and is one of 

the most famous temples in Korea. There are two 

main pagodas in the temple site. The three-story 

Seokgatap which stands at 8.2 m is a traditional 

Korean-style granite pagoda with simple lines and 

minimal detailing. Seokgatap is over 13 centuries old. 

The other one - Dabotap - is 10.4 m tall. In contrast 

to Seokgatap, Dabotap is famous for its highly ornate 

structure. 

The temple was renovated during the Koryeo 

Dynasty (AD 918 – 1392) and the early Joseon 

Dynasty (AD 1392 – 1910). Historical records say 

that the Seokgatap was destroyed twice, in 1024 and 

in 1036 by big earthquakes (National Museum of 

Korea, 1997). After 1604, reconstruction and the 

expansion of Bulguksa started, this was followed by 

about 40 renovations leading up to 1805. During the 

Colonization of Korea between 1910 and 1945, the 

Japanese conducted restoration, but there were no 

records of the work done, and many known treasures 

disappeared during this time. After World War and 

the Korean War, partial restoration was conducted 

and completed in 1966. Upon an expansive 

archeological investigation, major restoration was 

conducted between 1969 and 1973, bringing 

Bulguksa to its current form. These stone pagodas 

are now preserved in their original Silla style. 

To interpret the factors related to fracturing and the 

instability of the pagoda, we performed a kinematic 

analysis on the developed cracks. Cracks on the 

structure demonstrate the factors that control crack 

creation and spreading. We can use crosscutting 

relationship, kinematic indicators, inferring stress and 

damage zone theory to interpret the conditions that 

the pagoda has been subjected to. 

According to an engineering survey for the 

Seokgatap, the following problems of structural 

stability are observed. The pagoda is tilled 0.9 to the 

Northwest. Foundation stones are dislocated with 

respect to each other by about 4 cm (Fig. 3). The 

center of the pagoda has subsided by about 3 cm. 

Also, we observed that there are cracks in the flat 

stones that are laid on other stones ranging in length 

between 1 and 6.5 cm. Some blocks have a gap 

between each other ranging from 3 to 4.5 cm. Also, a 

long crack of 132 cm in length and width of 5 mm 

have recently been detected on the eastern side of 

the pagoda. By comparing these results with 

previous state, the structural stability of the upper 

part of the pagoda is not safe and damage currently 

occurring. 

Fig. 3: (a) Photograph of western part of Seokgatap (b) 

sketch of western part of Seokgatap. 

Fig. 2: Histogram of historic earthquake catalogue from 2 AD to 1810 AD for >ML = 4. It shows three earthquake clusters from 

1013 AD and various recurrence intervals in the study area (modified from Jin et al., in press). 

119


EARTHQUAKE 

ARCHAEOLOGY 



DISCUSSION 

Quaternary faults have mainly been focused on the 

subject of paleoseismological studies in Korea. 

Recently, an archaeoseismological approach was used 

to infer the causes of the falling of a Buddha statue in 

Gyeongju area (Jin et al., in press). 

Archaeoseismological research in Gyeongju is 

important because deformed man-made structures 

with known age and original state offer supplementary 

information on past seismic events. Studying damaged 

monuments such as Seokgatap, Cheomseongdae and 

Sukbinggo suggest different major factors contributed 

to the damage at each site. The Cheomseongdae was 

built as an observatory and leans to the north by about 

4 and it shows a horizontal shift in its large ashlars 

(Fig. 4), which may have been caused by episodic 

forces. If we carry out a similar engineering survey and 

a kinematic analysis (slip sense, opening and tilting 

etc.) in order to know how the cracks and gaps are 

developed in the pagoda at Cheomseongdae as we did 

at Seokgatap, we could be able to estimate the reason 

of the damage. 

Fig. 4: (a) Overview of the damaged Cheomseongdae 

observatory in Gyeongju. (b) Sketch of southern part of 

Cheomseongdae. 

Acknowledgements: This work was funded by the Korea 

Meteorological Administration Research and Development 

Program under Grant CATER 2008-5502 

References 

Chang, T.W. (2001). Quaternary tectonic activity at the 

Eastern Block of the Ulsan Fault. Journal of the Geological 

Society of Korea, 37, 431-444 (in Korean with English 

abstract). 

Du, Y and Aydin, A., (1995). Shear fracture patterns and 

connectivity at geometric complexities along strike-slip 

faults. Journal of Geophysical Research, 100, 18093-18102. 

Jin, K., Lee, M., Kim, Y.-S., Choi, J.-H. (2011). 

Archaeoseismological studies on historical heritage sites in 

the Gyeongju area, SE Korea. Quaternary International, 

doi:10.1016/j.quaint.2011.03.055. 

Kim, Y.-S., Andrews, J.R., Sanderson, D.J. (2000). Damage 

zones around strike-slip fault systems and strike-slip fault 

evolution, Cracking Haven, southwest England. 

Geoscience Journal, 4, 53-72. 

Kim, Y.-S., Jin, K. (2006). Estimated earthquake magnitude 

from the Yugye Fault displacement on a trench section in 

Pohang, SE Korea. Journal of the Geological Society of 

Korea, 42, 79-94 (in Korean with English abstract). 

Kyung, J.G., Okada, A. (1995). Liquefaction phenomena due 

to the occurrences of great earthquakes; some cases in 

central Japan and Korea. Journal of the Geological Society 

of Korea, 31, 237-250. 

Lee, K. and Na, S. H. (1983). A study of microearthquake 

activity of the Yangsan fault. Journal of the Geological 

Society of Korea, 19, 127-135. 

Lee, K., Jin, Y.G. (1991). Segmentation of the Yangsan fault 

system: geophysical studies on major faults in the 

Kyeongsang basin. Journal of the Geological Society of 

Korea, 27, 434-449. 

Lee, K. (1998). Historical earthquake data of Korean, Journal 

of the Korea Geophysical Society, 1, 3-22. 

Lee, K., Yang, W.-S. (2006). Historical seismicity of Korea. 

Bulletin of the Seismo-logical Society of America 96, 

846e855.Stiros, S. 1988b. Archaeology, a tool to study 

active tectonics – The Aegean as a case study. Eos, 

Transactions, American Geophysical Union 13, 1636-1639. 

National Museum of Korea (1997). News. 

Kim, B.-S. (1145) Samguksagi. 

CONCLUSIONS 

Gyeongju is located around the junction between the 

Yangsan and Ulsan faults. Recently many Quaternary 

faults have been reported along the Yangsan and 

Ulsan faults. This has encouraged research on fault 

activities such as Quaternary faults and related 

paleoseismicities. According to historical records, the 

study area was significantly affected by earthquakes. 

Large seismic events produced heavy casualties and 

destroyed historical property. To identify the reason for 

the damage, we have performed kinematic analysis on 

the Seokgatap, in Gyeongju. Comparing previous 

report of structural stability with its present state, it is 

being destroyed by anonymous forces and movements 

such as foundation subsidence. Further detailed 

research on other destroyed stone monuments in 

Gyeongju would be helpful to clarifying the relationship 

between damage and historical earthquakes. So this 

study can give some valuable information to heritage 

preservation and earthquake hazard study in Korea. 

120


EARTHQUAKE 

ARCHAEOLOGY 



EARTHQUAKE GEOLOGY AND RELATED HAZARD IN KACHCHH, GUJARAT, 

WESTERN INDIA 

Malik, Javed N.(1), Michio Morino (2), Mahendra S. Gadhavi (3,4), Khalid Ansari (1), Chiranjeeb Banerjee (1), 

B. K. Rastogi (3),Fumio Kaneko (2), Falguni Bhattacharjee (3), Ashok. K. Singhvi (5) 

(1) Department of Civil Engineering, IIT Kanpur. Kanpur 208016. UP. INDIA. Email: javed@iitk.ac.in 

(2) OYO International Corporation Yushima 1-Chome Bldg. 4F, 1-6-3 Yushima, Bunkyo-ku, Tokyo 113-0034, JAPAN. 

Email: morino@oyointer.com; kaneko@oyointer.com 

(3) Institute of Seismological Research, Gandhinagar 382018, Gujarat, INDIA. Email: mahendrasinh@gmail.com; 

brastogi@yahoo.com; falguni.geo@gmail.com 

(4) L. D. College of Engineering, Ahmedabad, Gujarat. INDIA 

(5) Physical Research Laboratory, Ahmedabad. INDIA. Email: aksprl@yahoo.com 

Abstract: The Kachchh falls under seismic zone V outside the Himalaya. It is marked by several E-W striking longitudinal faults 

viz. the Allah Bund Fault (ABF), Island Belt Fault (IBF), South Wagdh Fault (SWF), Kachchh Mainland Fault (KMF), and Katrol Hill 

Fault (KHF). Several large to moderate magnitude earthquakes have struck this area during last 300 years, viz. 893 AD, 1668 

Indus Delta (M7); 1819 Allah Bund (M7.8), 1956 Anjar earthquake (Ms6.1), and the recent 2001 Bhuj earthquake (Mw7.6). The 

rupture of 2001 remained concealed, suggestive of occurrence on blind fault. We carried out active fault identification, mapping 

and paleoseismic studies along ABF, KMF and KHF. CORONA satellite photos were used for identification of active fault traces. 

Ground Penetrating Radar (GPR) profiling helped us in locating appropriate site for trenching across KMF and KHF. Our study 

suggests that all three fault are active and were ruptured during recent historic past. 

Key words: Kachchh, active faults, paleoseismic studies, Western India. 

INTRODUCTION 

The state of Gujarat has two fold hazard posed by 

earthquakes, one on-land along the active faults in 

Kachchh region and its neighbourhood, and another 

offshore along the Makran Subduction Zone (MSZ) 

located in the west (Figure 1). The entire coastline is 

highly vulnerable from the tsunami triggered by an 

earthquake occurring along the MSZ. Two such 

tsunami events have been recorded during historic 

and recent times by earthquakes along MSZ, i.e., 

during 325 BC and 1945 AD. Apart from this the 

Kachchh region has been struck by several large to 

moderate magnitude earthquakes during last 300 

years, viz. 893 AD, 1668 Indus Delta (M7); 1819 

Allah Bund (M7.8), 1956 Anjar earthquake (Ms6.1), 

and the recent 2001 Bhuj earthquake (Mw7.6) (Malik 

et al., 1999a; Bilham, 1999). From these events, only 

1819 Allah Bund earthquake has been reported to 

have accompanied with 80-90 km long surface 

rupture and uplift resulting into formation of about 4-6 

m high scarp (Quittmeyer and Jacob, 1979; Johnstan 

and Kanter, 1990). The recent 2001 Bhuj earthqukae 

with magnitude Mw7.6, the rupture remained 

concealed below the ground at a depth of 7-10 km, 

suggestive of occurrence on blind fault (Mandal and 

Horton, 2007). It is a matter of concern that if 

movements on a blind fault are capable of producing 

large magnitude earthquakes, than having 

earthquakes of similar magnitude or larger on active 

faults with surface rupture cannot be ruled out (Malik 

et al., 2008; Morino et al., 2008). Active faults are 

considered to be the source for large magnitude 

earthquakes in seismically active regions. Their 

proper identification and distribution significantly help 

in knowing the seismic potential and associated 

hazard in the region. The landscape of Kachchh is 

marked by several E-W striking longitudinal faults viz. 

the Allah Bund Fault (ABF), Island Belt Fault (IBF), 

South Wagdh Fault (SWF), Kachchh Mainland Fault 

(KMF), and Katrol Hill Fault (KHF). Under the project 

sponsored by Gujarat State Disaster Management 

Authority (GSDMA) on Seismic Microzonation of 

Gandhidham, Kachchh, we carried out active fault 

mapping and paleoseismic investigations along ABF, 

KMF and KHF. In this paper we highlight in brief our 

findings based on detailed satellite data interpretation 

for identification of active faults and related 

geomorphic features as well as paleo-earthquake 

signatures preserved in sediment succession 

(Figures 1). 

Active fault and paleoseismic investigations: 

Katrol Hill Fault: 

Several new active fault traces were identified along 

Katrol Hill Fault (KHF) (Figure 1). These fault traces 

were identified based on satellite photo interpretation 

and field survey. Trenches were excavated to identify 

the paleoseismic events, pattern of faulting and the 

nature of deformation. Active fault traces were 

recognized about 1km north of the topographic 

boundary between the Katrol Hill and the plain area. 

121


EARTHQUAKE 

ARCHAEOLOGY 



Figure 1. Generalized structural map of Kachchh region (after Biswas and Deshpande, 1970). Inset at the left top show DEM of 

India highlight the location of Kachchh peninsula and Makran Subduction Zone (MSZ). Inset at the lower right shows major 

geomorphic zones of Kachchh. Yellow box marks the study area along Kachchh Mainland Fault and Katrol Hill Fault, and along 

Allah Bund Fault. NHR- Northern Hill Range, KHR- Katrol Hill Range, KMF Kachchh Mainland Fault and KHF- Katrol Hill Fault. 

The fault exposure along the left bank of Khari River 

with 10m wide shear zone in the Mesozoic rocks and 

showing displacement of the overlying Quaternary 

deposits is indicative of continued tectonic activity 

along the ancient fault. The E-W trending active fault 

traces along the KHF in the western part changes to 

NE-SW or ENE-WSW near Wandhay village. 

Trenching survey across a low scarp near Wandhay 

village reveals three major fault strands F1, F2, and 

F3 (Figures 2a & b). These fault strands displaced 

the older terrace deposits comprising Sand, Silt and 

Gravel units along with overlying younger deposits 

from units 1 to 5 made of gravel, sand and silt. 

Stratigraphic relationship indicates at least three 

large magnitude earthquakes along KHF during Late 

Holocene or recent historic past. 

Kachchh Mainland Fault: 

Two trenches were dug along the KMF, one near 

Jhura and another near Lodai villages. We reported 

first identified active fault exposure from Kachchh 

region along the Kachchh Mainland Fault (KMF) 

other than the 1819 Allah Bund earthquake. The 

active fault scarps striking E-W were identified near 

Lodai village along KMF. North facing scarps with 

height from 10-15 m are the manifestation of the 

displaced alluvial fan surface along this fault. 

Occurrence of discontinuous linear mound ranging in 

height from 3-5 m aligned along the strike about 100 

m north of the main scarp are suggestive of younger 

tectonic movement and progressive shift of tectonic 

activity towards north along new imbricated fault 

(Figure 3). Three low to high angle reverse fault 

strands (F1, F2 and F3) displacing young Quaternary 

deposits (late Pleistocene-Holocene?) classified as A 

to F units comprising gravel and sand-silt facies were 

identified in a trench excavated at the base of the 

linear mound along KMF (Figure 3). Our preliminary 

observations revealed occurrence of at least two 

large magnitude earthquakes along the F3 fault, and 

may be older events along the F1 and F2. Latest 

event (Event-I) occurred along F3 after the deposition 

of unit B registering the displacement of ~33 cm, 

penultimate event (Event-II) occurred after the 

deposition of unit C with ~40 cm of displacement. 

The maximum displacement of about 73 cm along F3 

indicates cumulative displacement accommodated 

during more than one event. The total displacement 

of ~98 cm along F2 strand displacing the E and F 

units is the result of more than one event, and since 

the F2 probably displaced the unit C suggests that 

the movements occurred during penultimate (Event 

II) and during the Event III, older than penultimate. 

Displacement of Mesozoic succession during older 

events and unit B during the latest Event I along F1 

suggests repetitive movement along this fault. The 

fragile nature of ~3-4 m wide shear zone formed in 

Mesozoic rocks (shale+sandstone) also point 

towards repetitive tectonic movement along KMF. 

122


EARTHQUAKE 

ARCHAEOLOGY 



Figure 2a. Photo-mosaic of the eastern wall of Wandhay trench exposed across the Katrol Hill Fault near Wandhay Dam. The 

trench was excavated across a very low fault scarp (refer Figure 2, for location). Three strands of thrust faults (F1, F2 and F3) 

dipping towards south were identified registering the latest event along F1, the penultimate as well as latest event along F2 and 

the older event along F3. 2b.Trench log of eastern wall of Wandhay trench. Terrace deposits are composed of Sand, Gravel-1, 

Silt, and Gravel- 2. These terrace deposits are covered by units 1 to 5 (after Morino et al., 2008). 

Figure 3. East wall view of trench excavated across active fault scarp near Lodai village along Kachchh Mainland Fault (KMF). 

Three main fault strands F1, F2 and F3 shows reverse faulting with variable dip ranging from 10°-55° towards south. 

To confirm further active faulting along the KMF, 

paleoseismic investigation near Jhura Village about 

30 km west of Lodai revealed an active fault 

displacing overbank deposits of Kaila River (Figure 2 

and 6). Two fault strands F1 and F2 were identified in 

the trench. The northern F1 shows a low-angle 

reverse fault with inclination of 15° towards the south. 

At least two faulting events were inferred on the basis 

of upward fault termination with clear angular 

unconformity. The net-slip during a single faulting 

event considering deformation on the hanging wall of 

F1 fault is over 5 m, suggestive of a large magnitude 

event during late Holocene period. 

Allah Bund Fault in Great Rann of Kachchh: 

The Kachchh region is not only well known for the 

occurrence of large magnitude earthquakes, but also 

for having a major Harappan (4000-4500 year) and 

historical sites. One of such major sites was 

Dholavira located on Khadir Island (Figure 1). Few 

sites in Great Rann of Kachchh (GRK), probably 

flourished until 1819 Allah Bund earthquake (?). Till 

date it is not fully understood as whether these sites 

were affected by the major seismic events in the past 

and also the presently evolved landscape was 

influenced by tectonic movements. The geologists, 

archaeologists, and scholars of ancient Indian history 

have mentioned the existence of numerous mighty 

southwest flowing rivers viz. the Sindhu (Indus), 

Shatadru or Nara (Sutlej) and Sarasvati, during Pre- 

Vedic and Vedic times (~4000 yr). These rivers 

flowed into then existing Arabian Sea, presently the 

GRK. 

We excavated 6-8 trenches in Allah Bund region 

GRK. Study reveals occurrence of at least 3 major 

events during recent past, which were probably 

responsible for the disruption of major channels (?), 

changing the landscape and destruction of the 

settlements (Figure 4). Trenches on the hanging wall 

of ABF shows thick massive yellowish medium-fine 

sand overlain by 1-1.5 m thick laminated sequence of 

123


EARTHQUAKE 

ARCHAEOLOGY 



silty-sand and clay. This suggests change in 

depositional environment from fluvial to fluvial-marine 

or tidal environment (high sea-level during 4000-6000 

yr?). Trench at Vigukot revealed prominent sandsheets 

at three levels indicative of 3 major 

liquefaction events, triggered by near source 

earthquakes (Figure 4), the latest event probably be 

the 1819 Allah Bund. Preliminary OSL ages of the 

sediments dated from the sand blow, soft sediment 

deformational structures, faulted sedimentary units 

from the trenches excavated on the hanging wall and 

across the Allah Bund Fault suggests occurrence of 

at least 2-3 events during 2.0-3.0 ka, with the most 

recent event during 2.0 ka. 

(a) 

(b) 

(c) 

Figure 4. (a) Vigukot Fort in GRK, fortification area is marked by broken while line. VT1, VT2, VT3 and VT4 are the locations of 

trenches excavated in and around Vigukot Fort, (b) field-photo showing location of trench VT1 and VT3. Location of VT3 is marked 

by yellow-box. Oriented bricks on the surface represents remnant of old foundation and (c) east wall view of 1.2 m deep and 3.5 m 

long trench excavated across a 80 cm thick wall (for location refer figure 4b). Exposed succession exhibits massive medium to fine 

sand at the base – suggestive of fluvial environment, capped by 1 m thick laminated succession – indicative of tidal environment. 

Preliminary interpretations suggests occurrence of at least three major event marked by prominent liquefied sand embedded in 

form of sand-sheet. CL1, CL2 and CL3 mark three cultural or occupation levels, the foundation in the centre of the trench belongs 

to the latest cultural level (CL3), which got severely affected during latest event, most likely the 1819 Allah Bund event(?). 

Acknowledgements: The authors are thankful to GSDMA, 

Gujarat, for funding this project. We are grateful to Prof. S. 

K. Jain, IIT Kanpur, now at IIT Gandhinagar and Dr. Alpa 

Sheth, VMS Consultants Private Limited, for their constant 

encouragement, support and discussion during this work. 

References 

Bilham, R. (1999). Slip parameters for the Rann of 

Kachchh, India, 16 June 1819 earthquake quantified 

from contemporary ccounts. Coastal Tectonics. 

Stewart, I. S., and Vita-Finzi, C. (eds.). Geological 

Society of London 146, 295-318. 

Biswas, S. K. (1980): Structure of Kutch-Kathiawar Region, 

W. India.- Proc. 3rd Ind. Geol. Congr., Pune, 255-272. 

Biswas, S. K., and Deshpande, S. V. (1970): Geological 

and tectonic maps of Kutch. - In: , 7, 115-116 

Johnston, A. C. and Kanter, L. R. (1990). Earthquakes in 

stable continental crust. Scientific American, 68-75. 

Malik, J. N., Sohoni, P. S., Karanth, R. V. and Merh, S. 

S. (1999a). Modern and Historic seismicity of Kachchh 

Malik, J. N., Morino, M., Mishra, P., Bhuiyan, C., and 

Kaneko, F. (2008). First active fault exposure identified 

along Kachchh Mainland Fault: Evidence from trench 

excavation near Lodai village, Gujarat, Western India. 

Journal Geological Society of India, 71, 201-208. 

Morino, M., Malik, J. N., Gadhavi, M. S., Ansari, K., Mishra, 

P., Bhuiyan, C., and Kaneko, F. (2008). Active Low- 

Angle Reverse Fault and Wide Quaternary Deformation 

Identified in Jhura Trench across Kachchh Mainland 

Fault, Kachchh, Gujarat, India. Journal of Active Fault 

Research, Japan, 29, 71-79. 

Morino, M., Malik, J. N., Mishra, P., Bhuiyan, C., and 

Kaneko, F. (2008). Active fault traces along Bhuj Fault 

and Katrol Hill Fault, and trenching survey at Wandhay, 

Kachchh, Gujarat, India. Journal of Earth System 

Sciences, 117(3),181–188. 

Quittmeyer, R. C. and Jacob, K. H. (1979). Historical and 

modern seismicity of Pakistan, Afghanistan, 

Northwestern India and South-eastern Iran. Bull. Seism. 

Soc. of America, 69, 773-823. 

124


EARTHQUAKE 

ARCHAEOLOGY 



SEISMOGENIC SLUMPS IN PALAEO-DEAD SEA SEDIMENTS 

Marco, Shmuel (1) and G. Ian Alsop (2) 

(1) Department of Geophysics and Planetary Sciences, Tel Aviv Univesity. Ramat-Aviv, Tel-Aviv 69978, ISRAEL 

(shmulikm@tau.ac.il) 

(2) Department of Geology and Petroleum Geology, School of Geosciences, University of Aberdeen, Aberdeen AB24 3UE, UK 

(ian.alsop@abdn.ac.uk) 

Abstract (Seismogenic slumps in palaeo-Dead Sea sediments): We analyze a series of slumps in lake sediments overlying the 

Dead Sea Fault. The slumps are interpreted as seismites that have been triggered by earthquakes, thus providing a palaeoseismic 

record for the DSF. The direction of slumping inferred from the geometry and orientations of folds and thrusts varies systematically 

along the entire ~100 km length of the western Dead Sea Basin. They are interpreted to form part of a large-scale radial slump 

system directed towards the depocentre of the precursor to the Dead Sea. The recognition that slumps may be reworked by 

younger seismically-triggered events suggests that in some cases the seismic recurrence interval may be shorter than previously 

anticipated. 

Key words: paleoseismology; seismites; slumps; Dead Sea Fault 

DISCUSSION 

Seismites found in sediments of lakes that straddled 

the Dead Sea Fault have provided a 70-kyr-long 

palaeoseismic record, one of the longest on Earth 

(e.g., Ferry et al., 2011; Kagan et al., 2011; Ken-Tor 

et al., 2001; Marco et al., 1996; Migowski et al., 

2004). Still, questions concerning the significance of 

their detailed shapes and the physical process that 

governed the formation of these seismites remain 

open. Understanding the seismites evolution and the 

underlying physics are tools for reconstructing the 

properties of past earthquakes on the basis of the 

observed deformations. These could also be useful in 

detecting and understanding off-fault earthquake 

indicators in other submarine environments. 

The Rayleigh-Taylor instability can explain 

mushroom-like symmetric deformation of inverseddensity 

stratification, where heavy strata overlay 

lighter strata (Figure 1a). However, stable 

stratification and asymmetric folds predominate the 

seismites in the Lisan Formation, which was 

deposited during the last glacial period in Lake Lisan 

(the precursor to the Dead Sea). Seismites in stably 

stratified lacustrine marls favor a mechanism of 

earthquake-triggered shear known as the “Kelvin- 

Helmholtz Instability” (Heifetz et al., 2005). Field 

observations and numerical simulations show that 

the deformation begins as moderate wave-like folds 

due to shear at the water-sediment interface (Wetzler 

et al., 2010). It evolves into asymmetric folds, then 

reclining folds, and in cases of backwash and/or 

vertical relaxation we observe backward-oriented 

folds (Fig. 1). If the flow becomes turbulent the layers 

are fragmented, re-suspended, and ultimately redeposited 

as breccia. During past earthquakes this 

process stopped at different stages, depending on 

the strength and duration of the shaking. 

Our detailed analysis of the seismites (Alsop and 

Marco, 2011b) reveals that some of them were 

affected by later events. Failure to recognize 

reworked seismites would reduce the number of 

earthquakes inferred from simple counting. In cases 

of incomplete reworking careful scrutiny can 

distinguish two earthquakes represented in one 

seismite. For example, we observe breccia layers 

that are folded as part of slumps, which we interpret 

as multiple event seismites (Fig. 2). 

We recognize that the direction of slumping inferred 

from the folds and thrusts varies systematically along 

the entire ~100 km length of the western Dead Sea 

Basin (Fig. 3). Hence the asymmetry of the folds 

represents the sense of shear, which was determined 

by very subtle slope of less than 1°. In the case of 

Lake Lisan coherent pattern of the fold vergence 

indicates that the location of the depocenter dictated 

the flow (Alsop and Marco, 2011a). 

Acknowledgments: We thank Mr. John Levy, together with 

the Carnegie Trust and the Royal Society of Edinburgh for 

travel grants to IA, and the Israel Science Foundation for 

grant 1539/08 to SM. 

References 

Alsop, G. I., and Marco, S., 2011a, A large-scale radial 

pattern of seismogenic slumping towards the Dead Sea 

Basin: Journal of the Geological Society. Accepted. 

Alsop, G. I., and Marco, S., 2011b, Soft-sediment 

deformation within seismogenic slumps of the Dead Sea 

Basin: Journal of Structural Geology, v. 33, no. 4, p. 433- 

457. 

El-Isa, Z. H., and Mustafa, H., 1986, Earthquake 

deformations in the Lisan deposits and seismotectonic 

implications: Geophys. J. R. Astron. Soc., v. 86, p. 413- 

424. 

Ferry, M., Meghraoui, M., Abou Karaki, N., Al-Taj, M., and 

Khalil, L., 2011, Episodic Behavior of the Jordan Valley 

125


EARTHQUAKE 

ARCHAEOLOGY 



Section of the Dead Sea Fault Inferred from a 14-ka-Long 

Integrated Catalog of Large Earthquakes: Bulletin of the 

Seismological Society of America, v. 101, no. 1, p. 39-67. 

Hall, J. K., 1996, Digital topography and bathymetry of the 

area of the Dead Sea Depression: Teconophysics, v. 

266, p. 177-185. 

Heifetz, E., Agnon, A., and Marco, S., 2005, Soft sediment 

deformation by Kelvin Helmholtz Instability: A case from 

Dead Sea earthquakes: Earth Planet. Sci. Lett., v. 236, p. 

497-504. 

Kagan, E., Stein, M., Agnon, A., and Neumann, F., 2011, 

Intrabasin paleoearthquake and quiescence correlation of 

the late Holocene Dead Sea: J. Geophys. Res., v. 116, 

no. B4, p. B04311. 

Ken-Tor, R., Agnon, A., Enzel, Y., Marco, S., Negendank, J. 

F. W., and Stein, M., 2001, High-resolution geological 

record of historic earthquakes in the Dead Sea basin: J. 

Geophys. Res., v. 106, no. B2, p. 2221-2234. 

Marco, S., Stein, M., Agnon, A., and Ron, H., 1996, Long 

term earthquake clustering: a 50,000 year paleoseismic 

record in the Dead Sea Graben: J. Geophys. Res., v. 

101, no. B3, p. 6179-6192. 

Migowski, C., Agnon, A., Bookman, R., Negendank, J. F. 

W., and Stein, M., 2004, Recurrence pattern of Holocene 

earthquakes along the Dead Sea transform revealed by 

varve-counting and radiocarbon dating of lacustrine 

sediments: Earth and Planetary Science Letters, v. 222, 

no. 1, p. 301-314. 

Wetzler, N., Marco, S., and Heifetz, E., 2010, Quantitative 

analysis of seismogenic shear-induced turbulence in lake 

sediments: Geology, v. 38, no. 4, p. 303-306. 

Figure 1. Schematic line 

drawings and photographic 

examples of structures 

generated during slump 

sheet initiation (a), 

translation (b, c), cessation 

(d), relaxation (e) and 

compaction (f). Folded beds 

are shown in yellow, while 

axial planes (blue) and 

thrusts (red) are also 

highlighted together with 

possible deformation styles. 

Not all slump sheets will 

show the full range and 

evolution of structures 

depicted here. Northeast is 

on the right of photographs 

(Figure from Alsop and 

Marco, 2011b). 

126


EARTHQUAKE 

ARCHAEOLOGY 



Figure 2. Seismites 1 and 2 that are 

recognized as breccia layers are 

included in a seismogenic slump. Note 

that the upper part of the slump is also 

brecciated. We therefore interpret the 

sequence as containing evidence for 

three seismic events. Coin diameter is 

20 mm. 

Figure 3. Schematic illustrations of slump structures showing radial vergence directions. We show the mean vergence directions of 

over 350 measured folds on the west and slump directions from El-Isa and Mustafa (1986) on the east, suggesting the control of 

the depocenter on the slump transport (Alsop and Marco, 2011a). Map from Hall (1996). 

127


EARTHQUAKE 

ARCHAEOLOGY 



MAPPING AND MEASURING HOLOCENE FAULT SCARPS IN DENSE FORESTS WITH 

LIDAR 

McCalpin, James P. 

GEO-HAZ Consulting, Inc., P.O. Box 837, Crestone, Colorado 81131 USA. Email: mccalpin@geohaz.com 

Abstract (Mapping and Measuring Holocene Fault Scarps in Dense Forests with Lidar): In the past decade new fault scarps 

have been discovered in the forests of North America by the use of Lidar DEMs. Most of the surveyed regions had been examined 

previously with aerial photographs, but no scarps were seen through the tree canopy. Lidar in those same areas shows obvious 

fault scarps. The scarp heights and slope angles can be accurately measured directly from the DEM, and compare favourably to 

field measurements at the same sites. We describe several case histories in the USA where Lidar-detected scarps have been 

trenched and analyzed for their seismic source characteristics. 

Key words: fault scarps, Holocene, Lidar 

INTRODUCTION: THE PROBLEM OF FINDING 

FAULT SCARPS IN DENSE FORESTS 

The identification and mapping of Quaternary fault 

scarps is well advanced in arid and semi-arid regions 

that contain little vegetation. Sub-humid and humid 

regions, in contrast, are often heavily forested and 

the ground surface cannot be seen in aerial 

photographs. Until recently the only way to locate 

and map fault scarps in forested areas was to enter 

the forest on foot and begin searching, but without a 

previously-identified target, a search would be 

random and extremely time-consuming. As a result, 

the density of fault scarps mapped to date in forested 

areas is a small fraction of that mapped in open 

areas. The scarcity of mapped Quaternary faults in 

forested areas has, in turn, translated into lower 

predicted seismic hazards, particularly in areas with 

low historic seismicity. This low predicted seismic 

hazard may be incorrect for many regions, resulting 

mainly from the lack of a reliable technique for finding 

Quaternary faults in forests. 

The advent of light detection and ranging (LiDAR) in 

the 1990s has changed the situation dramatically. In 

this paper I describe four case histories where new 

fault scarps have been discovered in dense forests of 

the western USA and Alaska. 

LiDAR Basics 

LiDAR is also known by the more descriptive phrase 

“Laser altimeter terrain mapping.” An airborne 

scanning laser rangefinder collects millions of 

distance measurements between the airplane and 

the ground, as many as 30,000 points per second at 

~15 cm accuracy. The plane’s position is measured 

by differential GPS and an inertial navigation system. 

Each laser pulse measures multiple returns (distance 

measurements) along a single beam, with the first 

return from the top of local vegetation, and the last 

return from the ground surface. The last returns are 

amalgamated into a “bare earth” digital elevation 

model (DEM) which shows the shape of the ground 

surface beneath the forest canopy (virtual 

deforestation). The current cost of a LiDAR survey 

and bare-earth DEM is ca. $150–$400/km2. 

Fig. 1: Schematic diagram of LiDAR data acquisition 

The Puget Sound area in Washington State: The 

Seattle Fault and its Backthrusts 

The first LiDAR survey of the Puget Sound area in 

2000 revealed two previously unsuspected late 

Quaternary faults, the Toe Jam and Waterman Point 

faults. The Toe Jam fault scarp on Bainbridge Island 

was trenched by US Geological Survey in 1998 and 

1999 (Nelson et al., 2003). Trenching confirmed that 

the scarp records at least one, and probably more, 

large earthquakes since the latest glacial maximum 

(LGM, ca. 18 ka). A fossil beach terrace that 

surrounds this part of Bainbridge Island was uplifted 

about 7 meters in a single large earthquake about 

1,100 years ago. Likewise, trenching in August 2001 

confirmed that the Waterman Point scarp, like the 

Toe Jam scarp, follows a south-verging thrust fault 

that has moved since the Latest Glacial Maximum. 

128


EARTHQUAKE 

ARCHAEOLOGY 



blanketed by a dense coastal rain forest, which has 

defeated past attempts to unravel its neotectonics. 

Fig. 2. LiDAR image that revealed the scarp of the 

Toe Jam fault on Bainbridge Island, west of 

downtown Seattle, Washington, USA. The fault is an 

east-west-trending, north-dipping reverse fault. 

Western USA; The Upper Rio Grande Rift, Colorado 

The 1000 km-long Rio Grande rift traverses the 

desert state of New Mexico and much of the state of 

Colorado. In the desert Quaternary fault scarps are 

unobscured by vegetation, but as the rift floor 

gradually rises northward to 2500 m and ultimately 

3000 m in central Colorado, the rift margins become 

densely forested. Fault scarps higher than about 10 

m can be seen on aerial photographs (i.e., about half 

as high as the average tree height of 20-25 m). 

Smaller scarps cannot be seen on airphotos, but can 

be seen on LiDAR. 

The Williams Fork normal fault was discovered in 

2002 in a dense pine forest at the foot of the Williams 

Fork Mountains in central Colorado. Even though the 

aerial photographs did not show any fault scarps in 

the dense forest, Kirkham (2004) walked to the range 

front and discovered multiple-event Quaternary fault 

scarps there. That fault is now the northernmost 

known Quaternary fault associated with the Rio 

Grande rift zone, where scarps are clearly late 

Quaternary in age, and trenches show displacement 

of late Quaternary strata. 

This discovery encouraged the collection of 2400 

km2 of additional LiDAR of the rift margins in 2010 by 

the Colorado Geological Survey and USGS. The new 

LiDAR DEMs show numerous previously unknown 

fault scarps on the forested rift margins. These fault 

scarps fill in large gaps in the maps of the Quaternary 

Fault and Fold Database of the USA. 

Alaska: The Yakutat Microplate, Reverse Faults and 

Sackungen 

The Yakutat microplate lies in south coastal Alaska, 

comprising a small plate fragment caught between 

the North American plate and the Pacific Plate (Fig. 

3). The northward plate convergence of 55 mm/yr 

has created a spectacular barrier of coastal hills and 

mountains. These mountains intercept Pacific storms 

and lead to an annual precipitation of about 2000 

mm/yr. Due to the high precipitation the terrain is 

Fig. 3. Location map of the Yakutat Terrane 

(microplate) in south coastal Alaska. 

Re-folding of foreland fold and thrust belt structures 

has created Holocene fault scarps in mountain blocks 

throughout the western part of the microplate. The 

morphology of the fold belt is atypical of classical 

foreland fold belts, where valleys and ridges are 

typically elongated parallel to fold axes. The Yakutat 

landscape is composed by irregular to elongated, 

north to northeast trending mountain blocks 

separated by flat-floored valleys filled with alluvial 

and glacial deposits. This morphology reflects 

Quaternary re-folding of foreland folds about plunging 

hinge-lines and vigorous erosion by glaciers and 

rivers. The mountain blocks are covered with swarms 

of discontinuous scarps (Fig. 4), mainly formed by 

displacement along bedding during second-phase 

folding of Tertiary strata. 

Fig. 4. Photograph of flexural-slip scarps on a typical 

mountain ridge in the Yakutat microplate. 

Flexural-slip scarps occur in a variety of orientations 

with respect to hill slopes and ridge crests; parallel to 

ridge crests, cutting across ridge crests, and 

extending across alpine valleys. Scarps forming by 

flexural-slip folding are tens to hundreds of meters in 

length, have vertical offsets up to several meters, 

most face 'up-slope', and some have a dominant 

lateral slip component (Li et al., 2010). Ratios of fault 

scarp displacement to length (D/L) range between 

0.1 to 0.001, overlapping with, but generally larger 

than, D/L for discrete tectonic faults. Distances 

between scarps measured normal to bedding are 

tens of meters. The fundamental controls on 

129


EARTHQUAKE 

ARCHAEOLOGY 



localization of bedding parallel scarps are the 

mechanical competence of bedding and mechanical 

anisotropy imparted by the first-phase structures (Li 

et al., 2010). Other factors may include complex 

stress distributions within steep-sided mountain 

blocks, and stress transients generated by ground 

motion during large to great magnitude earthquakes. 

Fig. 5. Faults (red) and gravitational scarps (green) shown by LiDAR around the Martin Lake detailed study site, in the 

western Yakutat microplate. Black dashed line with hachures shows the LGM glacial trimline. From McCalpin et al., in 

press. 

Alaska: The Foothills Fault Zone 

The Foothills fault zone is a zone of north-verging 

thrusts and interconnecting strike-slip faults on the 

north side of the Alaska Range in central Alaska. 

Some alternative routes for the proposed Alaska- 

Canada Gas Pipeline cross the fault zone. Recent 

neotectonic studies in the crossing areas (Carver et 

al., 2008; 2010) have discovered fault scarps along 

the densely-forested range front, visible on LiDAR 

imagery but not detectable on aerial photographs. 

DISCUSSION 

LiDAR DEMs promise to change the way 

neotectonic-paleoseismic investigations are 

performed in forested regions. LiDAR should be the 

standard tool used for the critical step of identifying 

and mapping Quaternary fault scarps (Fig. 6). The 

success of this step determines the success of the 

remaining steps in the investigation. 

Acknowledgements: I thank the St. Elias Erosion- 

Tectonics Project (STEEP) and the National Science 

Foundation for funding the neotectonic work in the Yakutat 

microplate. 

References 

Carver, G.A., Bemis, S.P., Solie, D.N., and Obermiller, K.E. 

(2008). Active and potentially active faults in or near the 

Alaska Highway corridor, Delta Junction to Dot Lake, 

Alaska: Preliminary Interpretive Report 2008-3d, Alaska 

Division of Geological and Geophysical Surveys, 

Fairbanks, AK, 32 p 

Carver, G.A., Bemis, S.P., Solie, D.N., Castonguay, S. and 

Obermiller, K.E. (2010). Active and potentially active 

faults in or near the Alaska Highway corridor, Dot Lake to 

Tetlin Junction, Alaska: Preliminary Interpretive Report 

2010-1, Alaska Division of Geological and Geophysical 

Surveys, Fairbanks, AK, 41 p. 

Kirkham, R.K. (2004) Quaternary faulting in the Williams 

Fork Valley graben, north-central Colorado, and 

comparison with late Quaternary deformation near 

Spinney Mountain, central Colorado. Unpublished Final 

Technical Report submitted to the U.S. Geological 

Survey, NEHRP Program, Grant 02HQGR0102, revised 

09-JAN-2004, 50 p. 

Li, Z., Bruhn, R.L., Pavlis, T.L., Vorkink, M. and Zeng, Z. 

(2010). Origin of sackung uphill-facing scarps in the Saint 

Elias orogen, Alaska: LIDAR data visualization and stress 

modeling. Geol. Soc. Amer. Bull., 122 (9-10), 1585-1599. 

McCalpin, J.P. (ed). (2009). Paleosesmology, 2 nd 

Edition: 

Academic Press-Elsevier, 848 p. 

McCalpin, J.P., Bruhn, R.L., Pavlis, T.L., Gutierrez, F., 

Guerrero, J and Lucha, P., in press, Antislope scarps, 

gravitational spreading, and tectonic faulting in the 

western Yakutat microplate, south coastal Alaska. 

Geosphere., 2011. 

Nelson, A. R., Johnson, S. Y., Kelsey, H. M., Wells, R. E., 

Sherrod, B. L., Pezzopane, S. K., Bradley, L. A., Koehler, 

R. D., and Bucknam, R. C. (2003). Late Holocene 

earthquakes on the Toe Jam fault, Seattle fault zone, 

Bainbridge Island, Washington. Geol. Soc. Am. Bull. 

115(11), 1388–1403. 

Pavlis, T. L. and Bruhn, R. L. (2011). Application of LIDAR 

to resolving bedrock structure in areas of poor exposure: 

An example from the STEEP study area, southern 

Alaska, Geological Society of America Bulletin, v. 123, p. 

206-217, doi: 10.1130/B30132.1. 

130


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 6. Flow chart of a typical paleoseismic investigation that includes fault trenching, showing the early need for 

LiDAR surveys in forested areas. You cannot trench a fault scarp if you never find it. Adapted from McCalpin (2009). 

131


EARTHQUAKE 

ARCHAEOLOGY 



CHANGES ON THE GEOMORPHIC SETTINGS OF 

SAND-POOR ENVIRONMENT COAST OF BANDA ACEH, INDONESIA 

SUBJECT TO TECTONIC AND TSUNAMI EVENTS 

Meilianda, Ella (1), Ben Maathuis (2), Marjolein Dohmen-Janssen (3) 

(1) INTERCOAST, University of Bremen, MARUM - Leobener Strasse D-28359 Bremen, GERMANY. Email: 

emeilianda@marum.de 

(2) Water Resource Department, Faculty of GeoInformation Science and Earth Observation (ITC), The Netherlands.+31 53 

487 4391; Email: maathuis@itc.nl 

(3) Water Engineering & Management (WEM), University of Twente, PO Box 217, 7500 AE Enschede, NETHERLANDS, email: 

c.m.dohmen-janssen@utwente.nl 

Abstract: Geomorphic settings before and after the December 2004 tsunami of the Banda Aceh coast, Sumatra, Indonesia were 

investigated in this study. The coast of Banda Aceh is a sand-poor environment contains thin layer of mobile sand perched on top 

of a consolidated Holocene prograding delta. Changes on the coastal shoreface morphology as the response to the tsunami 

waves were varying subject to the difference of geomorphic settings. The seawater inundated to the coastal plain as far inland as 

the shoreline position of 0.6 ky BP, during which a comparable magnitude of tsunami confirmed to have occurred in this region. 

This study demonstrates that such huge tsunami event occurred abruptly, but led to changes of the geomorphology developed in 

the Holocene. 

Key words: tsunami, tectonic, geomorphic setting, Holocene 

INTRODUCTION 

Despite the growth of tsunami-related studies since 

December 2004’s tsunami event, there was little 

discussion on the geomorphological adjustment and 

development of the affected coast. Given the fact that 

such a powerful earthquake leading to an 

extraordinary magnitude of tsunami is a rare event in 

human-life time history, there are ample knowledge 

gaps that hinder thorough investigations towards this 

level. In addition, even before the tsunami of 

December 2004 occurred, that the topic of 

paleotsunami requires inevitable sedimentological 

and geomorphological research, since such extreme 

event effects on sedimentary transport or 

considerable alterations of the coastal configuration. 

In spite of this, only 5% of the existing tsunami 

literature is related to such issues (Scheffers & 

Kelletat, 2003). 

The geomorphological state and development of the 

coast in pre-tsunami time is essential to investigate 

the extent of the environmental damage of the coast 

after being hit by such a huge catastrophe. In the 

absence of this knowledge for the coastal 

management practice, it is impossible, to forecast or 

to set-up future scenarios of the geomorphological 

development of the affected coasts. 

In terms of sedimentary transport and 

geomorphological research, our main challenge is 

the lack of readily available data of the pre-existing 

geomorphology of the affected coast, particularly in 

the less-investigated study location of Banda Aceh, 

Sumatra Island, Indonesia. Pragmatically, the 

investigation on the coastal geomorphological 

development using various types of available 

information such as literature, historical maps, 

satellite images, bathymetric charts and in-situ 

sample data are tailored to construct the 

geomorphological interpretation. The objective of the 

present study is therefore to qualitatively measure 

the impact of the December 2004 tsunami on the 

coastal geomorphology based on the reconstruction 

of geomorphology of the Banda Aceh coast in the 

past. 

The work in this study rests on the notion that each 

older geomorphological unit provides a boundary for 

the more recent units and therefore co-determines 

more recent geomorphological developments. Some 

preliminary studies after the tsunami event revealed 

that the extent of the tsunami inundation on the 

Banda Aceh coastal plain reached as far as about 5 

km away from the coastline (Dohmen-Janssen et al., 

2006). 

Given the lack of geological data to support the 

quantification of the coastal destruction, we use data 

on the Holocene sea-level fluctuation of the region as 

a proxy and combined it with a Digital Elevation 

Model (DEM) of the coastal plain to interpret the 

geomorphological development of the coastal area. 

Additionally, this study also uses the generic 

knowledge of geomorphology and sediment 

stratification from bore hole samples to support the 

geomorphological interpretation. The geomorphic 

settings of the coastal area are also given by 

synthesizing the various but very limited geological 

and geomorphological studies of the coastal region. 

Overall, the study was aimed to answer the following 

research questions: 1) Which geomorphological units 

132


EARTHQUAKE 

ARCHAEOLOGY 



can we distinguish at the Banda Aceh coast and how 

do they respond to the earthquake and tsunami?; 2) 

To what extent have the earthquake and tsunami of 

26 December 2004 affected this coastal system? 

STUDY SITE, DATA SETS AND METHODS 

Banda Aceh is located at the northwestern tip of 

Sumatra Island, Indonesia (Fig. 1). The coastal plain 

has elevations from -0.5 m to +11 m relative to 

present sea level and occupies 125 square km of the 

northwest valley of the Barisan mountain range, 

which is the backbone of Sumatra Island. Krueng 

Aceh River is the main river crossing this low-lying 

coastal plain. In the 1990s, the middle reach of the 

river was bifurcated by a floodway channel, as a 

means to divert the surplus discharge of the river 

which caused annual flooding of the city. The 

coastline is dissected by parallel lagoon inlets in 

between some of the beach ridges sections. 

This study interpreted the geomorphology of the 

Banda Aceh coast from various types of maps and 

satellite images covering the past century and from 

literature covering the geological and 

geomorphological evolution of the surrounding 

region. All maps, images and charts were georeferenced 

to a master map (i.e. ortho-rectified aerial 

photo of 2005) which was processed in ArcGIS TM . 

The nautical charts, bathymetric data and a posttsunami’s 

topographic map of the coastal plain were 

transformed into Digital Elevation Models (DEMs) 

using Triangulation Irregular Network (TIN). 

topographic maps, bathymetric charts and bore hole 

samples as the proxy to the non-existing data (e.g. 

from core sampling, etc.) to reconstruct the 

development of the geomorphic settings of Banda 

Aceh. 

GEOMORPHIC SETTINGS OF BANDA ACEH 

Similar to the low latitude regions in the world, Banda 

Aceh (5° N) experienced a sea-level fall during the 

late Holocene (e.g. since 6.0 ky BP). Sea-level raised 

in Malacca Strait from -13 m in 8.0 ky BP to about +5 

m around 5.5 ky BP and then declined towards the 

present level (Geyh et al., 1979). Similar phenomena 

were also observed in Australia (Woodroffe and 

Horton, 2005), Singapore (Hesp et al. 1998) and 

Thailand (Sinsakul, 1990) for the same period. 

Tjia & Fujii (1989) suggested that at the west coast of 

Malaysian Peninsula the sea level reached its 

maximum of about 5 metres above present sea-level 

at approximately 5.0 ky BP. They used the carbon 

dating technique to investigate Holocene shoreline of 

the coasts of Malaysian Peninsula using abrasional 

and biogenic indicators. The sampled location at the 

west coast of Langkawi Island which is situated in the 

same tectonic region as to Banda Aceh and both 

locations are situated in the solitary Andaman Sea 

environment. In the present study, the sea-level 

fluctuation indicated by Tjia and Fujii (1989) was 

compared with the elevations as well as the units of 

morphology identified from the DEM of Banda Aceh 

coastal plain, which was derived from the 0.5-mcontour-interval 

topographic map. In this way, the 

shorelines associated with different phase of 

transgression and regression of sea-level during the 

Holocene were extracted. At the same time, the predominant 

process leads to the coastal morphology 

can be interpreted. 

Geomorphic settings of Banda Aceh coast 

Fig. 1: Topographic map of Banda Aceh coastal plain, 

Sumatra, Indonesia derived from DEM (RSTM TM 

2003 for 

the land part and nautical chart 1978 by BAKOSURTANAL, 

Indonesia). 

The Holocene sea-level fluctuation curves produced 

by Tjia & Fujii (1989) were used in our study to 

reconstruct the Holocene shoreline changes and the 

development of the Banda Aceh coastal plain 

towards the modern shoreline position. Overall, we 

use the results of these various studies in 

combination with the analysis of satellite images, 

Banda Aceh coastal plain and shoreface region 

consists of two distinctive geomorphic settings. The 

northeastern part is a tilted coastal zone (during the 

Pleistocene) which developments during the 

Holocene was influenced by the marine regression 

process. During this period, broad parallel coastal 

ridges and swales were developed over the modern 

coastal plain, and the shoreface has a concave - 

shape profile with sandy surface, indicating the predominant 

influence of the marine regression. On the 

other hand, the southwestern part consists of a broad 

alluvial flat-plain over a depression zone associated 

with the Sumatran Fault zone. The shoreface 

stratigraphic layers mainly consisted of silty-clay 

structures which profile is convex-shape, indicating 

an alluvial progradation which was developed during 

the Holocene (Meilianda, 2009). 

From the shoreline reconstruction using this method 

we analyse that the shoreline response to sea level 

fluctuations at Banda Aceh during the Holocene was 

highly influenced by the mechanisms of repeated 

changes of river mouths position and intermittent 

beach-ridge formation caused by periods of above 

133




EARTHQUAKE 

ARCHAEOLOGY 

average onshore winds; a similar climate condition of 

Jakarta Bay in Java Island according to Verstappen 

(1973). 

Modern coastal geomorphic settings and changes 

after December 2004’s tsunami 

Parallel to the completion of this study, several new 

findings about the recurrence of tsunami events in 

the Indian Ocean region have emerged. Monecke, et 

al. (2008) suggested that the three sediment coring 

samples at Meulaboh (west coast of Aceh), Simeulue 

(offshore west coast of Aceh) and Phra Thong (west 

coast of Thailand) show sediment deposit layers 

unconformity which age range are correlated with the 

historic tsunami occurrences in this region. 

Two sample units in Meulaboh and Phra Trong 

suggested the ranges of age of the deposit of AD 

1290 to 1400 (or 660 to 550 y BP) and AD 1300 to 

1450 (or 650 to 500 y BP), respectively (Monecke et 

al., 2008). In their report, they used the age 

estimates of coastal terraces in the Andaman Islands 

region by Rajendran, et al. (2008) to confirm their 

interpretation as evidence for subduction 

earthquakes. The problem with estimates made by 

Rajendran, et al. (2008) was that the range of age 

they provide was considerably large. The marine 

terrace of Andaman Islands was suggested to age of 

AD 1170-1600 and AD 550-1330 (or equal to 750- 

350 y BP and 1400-620 y BP). In this study, we 

produced an alternative proxy to that provides 

comparable confirmation to the analysis of great 

tsunami recurrence studied by Monecke, et al. 

(2008). 

In the 2004 tsunami, the inundation by seawater 

reached as far as 5 km inland, crossing the entire 

coastal plain of Banda Aceh that was already 

established since 3.5 ky BP (compare the tsunami 

inundation with the shoreline position of 2.8 ky BP in 

Fig. 1). It left deposits of various thickness and 

texture among the ruins and debris. The outer belt of 

beach ridge was mostly breached and even 

completely disappeared. After four days, the 

inundation reduced to about 50% from its initial 

extent (see also Dohmen-Janssen et al., 2006), 

which changed the low-lying housing areas, wetland 

and lagoon system behind the barrier islands into 

submerged areas. These areas are somehow 

associated with the one that geomorphologically 

modified by the last marine transgression (i.e. 

tsunami) in 0.6 ky BP. This shows that a huge 

tsunami event occurs only in a very short time-scale, 

but it leads to changes in geomorphology that has 

developed in a long-term scale, i.e. in centuries to 

millennia. 

The Banda Aceh coast in the recent pre-tsunami time 

consisted of a narrow sandy beach perched on top of 

the toe of the beach ridges from the previous coastal 

development (Pleistocene and Holocene 

morphological units). The preceding development 

has made the modern coast of Banda Aceh to be 

highly dissected within a 25-km coastal stretch. 

Figure 2 shows the five representative cross-shore 

shoreface profiles at Banda Aceh (P1, P2, P3, P4 

and P5 in Fig. 2). Each profile shows the elevation 

changes of three points in time within the recent 

century, i.e., before the tsunami (1893 and 1924), 

and after the tsunami (2006). 

elevation (m) 

(a) 

elevation (m) 

-50 

(b) 

elevation (m) 

(c) 

(d) 

elevation (m) 

(e) 

elevation (m) 

0 

-10 

-20 

-30 

-40 

0 

-10 

-20 

-30 

-40 

-50 

0 

-10 

-20 

-30 

-40 

-50 

0 

-10 

-20 

-30 

-40 

-50 

0 

-10 

-20 

-30 

-40 

-50 

1893 1924 2006 

1893 1924 2006 

1893 1924 2006 

1893 1924 2006 

1893 1924 2006 

cross profile P1 

0 300 600 900 1200 1500 1800 2100 2400 2700 3000 3300 3600 3900 4200 4500 4800 

distance from shoreline (m) 


0 300 600 900 1200 1500 1800 2100 2400 2700 3000 3300 3600 3900 4200 4500 4800 



0 300 600 900 1200 1500 1800 2100 2400 2700 3000 3300 3600 3900 4200 4500 4800 



0 300 600 900 1200 1500 1800 2100 2400 2700 3000 3300 3600 3900 4200 4500 4800 



0 300 600 900 1200 1500 1800 2100 2400 2700 3000 3300 3600 3900 4200 4500 4800 


silty clay (soft) 

Sandy silty clay 

(medium stiff) 

Poorly graded sand 

(very loose) 

Sandy clayey silt 

(stiff) 


(stiff) 

Silty sand (loose) 


(very soft) 

Silty sand (loose) 

Silty clay (very soft) 

Sandy silt (very 

soft) 

Poorly graded sand 

(loose) 

Sandy silty clay 

(very soft) 

Thickness scale : = 1 meter deep 

Fig. 2: Shoreface profile development in the past century 

and the associated bore hole samples. The downward 

arrows point at the sampling location at a depth ca. 10 m 

from mean sea level using rotational machine bore drilling 

log technique. The right panels of each profile display the 

stratification of the samples. Profile locations are in Fig. 1. 

From the borehole samples shown in Fig. 2 it can be 

observed that most of the intercalated layers 

underneath the top layers of each profile on the 

southwestern part of the coast (P1 and P2) consist of 

stiff deposits. These layers were associated with the 

alluvial delta progradation during the last Holocene. 

Furthermore, we suppose that the intercalating layers 

found near the Sumatran Fault zone (e.g. P2) were 

influenced by some tectonic-related event (e.g. land 

subsidence or tsunami). Mobile fine sediments were 

subsequently perched on top of these Holocene 

deposits. The deposits associated with the 

prograding delta process consist of very fine 

sediments (e.g. silty clay or clayey silt) which 

eventually become stiff structure of deposit layers in 

134


EARTHQUAKE 

ARCHAEOLOGY 



a long time and becomes the basement of the 

subsequent development (marine regression). 

The shoreface profile of Banda Aceh in response to 

the tsunami can be seen in cross shore profiles in 

1924 and 2006 shown in Fig. 2. The amount of 

eroded shoreface sediments varied from one profile 

to another. Interestingly, each shoreface profile 

maintained its convexity or concavity upon being 

eroded by the tsunami. The upper layer of the 

shoreface consists of young loose sediments 

underlain by the more consolidated morphological 

units (see bore hole data in Fig. 2) which indicate the 

combination between alluvial and regression 

processes during the Holocene. As a result, the 

upper layer were easily eroded during the tsunami 

(e.g. soft clay or silty layer deposit at the upper 

shoreface of profile P1 and the entire profile P3), 

while the shoreface composed by older deposits 

were less eroded (e.g. profiles center part of profile 

P1, profile P2 and P4). The shoreface erosion has 

modified the shoreface profiles of the entire coast 

and therefore has eventually modified the 

hydrodynamics of the coastal system. 

DISCUSSION 

The shoreface profile changes in Figure 2 show that 

the convex shoreface was influenced by the predominant 

alluvial progradation during sea-level 

regression, whilst the concave shoreface was 

influenced by sea-level regression process during the 

Holocene. In addition, the resistance of the shoreface 

against severe erosion by the tsunami waves may 

depend on the degree of consolidation or the rate of 

weathering of the shoreface deposits. On the other 

hand tectonic intensity must be influential in the 

consolidation processes (e.g. liquefaction or 

compaction) on a specific coastal section preceding 

the December 2004 tsunami. In particular, tectonics 

and extreme waves such as tsunami have left their 

signature in the present geomorphology. 

Regarding the complexity of the processes involved 

and the resulting geomorphology, proper attention 

should therefore be given to the boundary conditions 

which are imposed by the development of older 

geomorphological units. The complex 

geomorphology underlying the Banda Aceh coastal 

plain resulted from various processes in history and 

ultimately determines the shoreface geomorphic 

settings and the morphological development, 

composition and texture of beach sediments, and the 

rates of shoreline position changes. This case 

demonstrated that it is essential to understand the 

geomorphic settings of a coast before attempting to 

model the large-scale behaviour of these types of 

coastal systems. 

This study was aimed to establish the general 

framework for future research on this specific study 

location which deserves attention to gain knowledge 

on the tectono-tsunami process and its impact on the 

coastal system. Borehole samples and reviews from 

other studies about this study area incorporated in 

the analysis in the present study fairly ascertain our 

interpretations on the geomorphic settings inherited 

from various morphological developments in history. 

Further study on the stratification of sediment deposit 

layers, carbon-dating, and side-looking radar may be 

utilized in future to support the overall interpretations 

of the present study. 

Acknowledgements: This study was conducted in 

the Water Engineering and Management Group of 

University of Twente, Enschede, The Netherlands, in 

collaboration with the International Institute for Geo- 

Information Science and Earth Observation (ITC), 

Enschede, The Netherlands. The project is co-funded 

by the Syiah Kuala University of Banda Aceh, 

Indonesia and the University of Twente, Enschede, 

The Netherlands. Data sets was provided by BRR- 

Nias/NAD Banda Aceh. 

References 

Dohmen-Janssen, C.M., Meilianda, E., Maathuis, B.H.P. & 

Wong, P.P. (2006). State of Banda Aceh beach before 

and after the tsunami. Proceeding of 30 th International 

Conference on Coastal Engineering, ASCE, Sand 

Diego, California. 

Geyh, M.A., Kudrass, H.R. & Streif, H., (1979). Sea-level 

changes during the late Pleistocene and Holocene in 

the Strait of Malacca. Nature 278(5703), 441 – 443. 

Hesp, P.A., Hung, C.C., Hilton, M. , Ming, C.L., Turner, I.M., 

1998. A first tentative Holocene sea-level curve for 

Singapore. Journal of Coastal Research 14(1), 308-314. 

Meilianda, E., (2009). Past, present and future 

morphological development of a tsunami-affected coast: 

A case study of Banda Aceh. Ph D Thesis, University of 

Twente, Enschede, The Netherlands. 

Monecke, K., Finger, W., Klarer, D., Kongko, W., McAdoo, 

B.G., Moore, A.L. & Sudrajat, S.U., (2008). A 1000-year 

sediment record of tsunami recurrence in northern 

Sumatra. Nature (Letter) 455, 1232-1234. 

Rajendran, K., Rajendran, C.P., Earnest, A., Ravi Prasad, 

G.V., Dutta, K., Ray, D.K. & Anu, R., (2008). Age 

estimates of coastal terraces in the Andaman and 

Nicobar islands and their tectonic implications. 


Scheffers, A. & Kelletat, D., (2003). Sedimentologic and 

geomorphologic tsunami imprints worldwide – a review. 

Earth-Science Reviews 63, 83-92. 

Sinsakul, S., 1990. Evidence of sea-level changes in the 

coastal area of Thailand: A review. In: Proceedings 

IEM/ICE Joint Conference on Coastal Engineering in 

National Development (1991), pp. B1-B30. 

Tjia, H.D., (1992). Holocene sea-level changes in the 

Malay-Thai Peninsula, a tectonically stable 

environment. Geology Society of Malaysia 31, 157-176. 

Tjia, H. D. & Fujii, S., (1989). Late Quaternary shorelines in 

Peninsular Malaysia. In: Spafa Final Report Seminar in 

Prehistory of Southeast Asia 1987, 239–257. 

Verstappen, H. Th., (1973). A geomorphological 

reconnaissance of Sumatra and adjacent islands. 

Wolters-Noordhof Publication, Groningen. 

Woodroffe, S.A. & Horton, B.P., (2005). Holocene sea-level 

changes in the Indo-Pacific. Journal of Asian Earth 

Science 25, 29-43. 

135


EARTHQUAKE 

ARCHAEOLOGY 



SEISMIC ANALYSIS OF LIQUID STORAGE TANKS 

Konstantin Meskouris, Britta Holtschoppen, Christoph Butenweg, Julia Rosin (1) 

(1) Chair of Structural Statics and Dynamics, Mies-van-der-Rohe Strasse 1, 52074 Aachen, GERMANY. 

Email: meskouris@rwth-aachen.de 

Abstract (Tank design for seismic loading): In the event of strong earthquakes, it is important that the structural integrity of 

tanks containing liquids is maintained in order to not jeopardize the population's supply with essential goods. The continuous 

operation of tanks after strong earthquakes requires safe and effective design rules. The overall seismic behaviour of tanks is, 

however, quite complex, since the dynamic interaction effects between tank wall and liquid must be considered. The interaction 

can be simplified with the concept of generalized single-degree-of-systems representing the convective, rigid impulsive and flexible 

impulsive vibration modes of tank and liquid. This concept is well accepted for anchored tanks with a fix connection to a rigid 

foundation. This paper presents the state of the art of tank design with special focus on the practicability of the available design 

rules. Analytical and numerical calculation approaches are compared on the example of a typical tank geometry, taken the 

relevant interaction effects into account. 

Key words: liquid filled tank, seismic design, impulsive vibration mode, critical facilities 

INTRODUCTION 

Recent earthquake events showed that heavy 

seismic damages of tanks may lead to environmental 

hazards, fire following earthquakes and temporary 

loss of essential facilities. The vibration of liquid filled 

tanks subject to seismic loading depends on the 

inertia of the liquid and on the interaction effects 

between the liquid and the tank shell. Different 

calculation methods are available for describing the 

vibration behaviour and the earthquake loads. These 

methods are either quite simple (Housner, 1963) or 

very accurate but complex (Fischer et al., 1991). 

Therefore a well comprehensive and feasible method 

is needed, that provides realistic results for the 

seismic behaviour of tanks with an acceptable 

computing time. The following considerations apply 

to cylindrical, anchored tanks with a fix connection to 

a rigid foundation. 

SEISMICALLY INDUCED LOAD COMPONENTS 

OF LIQUID FILLED TANKS 

The seismic loads acting on wall and bottom of 

cylindrical tanks (Figure 1) can be divided into the 

following components (Meskouris et al., 2010): 

- the convective load component; the fluid 

vibration in the rigid tank (sloshing), 

- the impulsive rigid load component; caused by 

the inertia of the liquid, if the rigid tank moves 

together with the foundation, 

- the impulsive flexible load component; 

representing the combined vibration of the 

flexibile tank shell (e.g. steel tanks) with the 

liquid. 

Fig. 1: Cylindrical tank 

Convective pressure 

Figure 2 shows the mode of vibration and the 

pressure distribution corresponding to the convective 

pressure component. 

The pressure distribution is defined as: 

∞ 

p k 

(ξ, ζ, θ, t) = 2∙R∙ρ L 

(λ 2 J 1(λ n ∙ ξ) 

n -1) J 1 (λ n ) cosh(λ n∙ γ ∙ ζ) 

 

cosh(λ n ∙ γ) 

n=1 (1) 

⋅ [cos(θ)] [a kn (t) ∙ Γ kn ] 

with 

p k convective pressure component due to 

horizontal excitation 

n summation index; number of considered 

sloshing modes (here: n = 1) 

R inner tank radius 

ρ L liquid density 

136


EARTHQUAKE 

ARCHAEOLOGY 



J 1 

first order Bessel function: 

∞ 

J 1 (λ n ∙ ξ) = 

(-1)k 

2k+1 

k! ∙ Γ(1 + k + 1) ∙ λ n ∙ ξ 

2 

k=0 

λ n null derivation of Bessel function: 

λ 1 = 1,841, λ 2 = 5,331, λ 3 = 8,536 

ξ dimensionless radius: ξ = r/R 

ζ dimensionless height: ζ = z/H 

θ angle of circumference 

γ tank slenderness: γ = H/R 

a kn(t) horizontal acceleration-time history as a result 

of an equivalent single-degree-of-freedom 

system with a period T kn for the n th -eigenmode 

of the sloshing wave. By using the response 

spectrum analysis the spectral accelerations 

corresponding to the natural periods T kn 

should be calculated based on the elastic 

response spectrum. 

Γ kn participation factor for the convective pressure 

component for the n th -eigenmode. 

Fig. 2: Convective pressure - mode of vibration and 

pressure distribution 

Taking into account the first sloshing eigenmode 

(n = 1) and the pressure distribution of the tank shell 

(ξ = 1), equation (1) can be simplified to: 

cosh(1,841∙ γ ∙ ζ) 

p k 

(ξ = 1, ζ, θ, t) = R ∙ ρ L 

0,837 ∙ 

cosh(1,841∙ γ) (2) 

⋅ [cos(θ)][a k1 (t) ∙ Γ k1 ] 

ν n ν n = 2n + 1 

π 

2 

modified first order Bessel function: 

I 1 

I 1 

’ 

I 1 ν n 

γ ∙ ξ = J 1 i∙ ν n 

γ 

∙ ξ 

i n 

∞ 

ν 

2k+1 

nγ 

1 

∙ ξ 

= 

∙ 

k! ∙ Γ(1 + k + 1) 2 

k=0 

Derivation of the modified Bessel function 

regarding to DIN EN 1998-4 (2007) 

I ν n 

γ ∙ξ = I 0 ν n 

γ ∙ξ - I 1 ν n 

γ 

∙ξ 

ν n 

γ 

∙ ξ 

ν 

2k+0 

∞ 

n 

I 1 

∙ξ γ 

= 

k! ∙ Γ(0 + k + 1) ∙ 2 

k=0 

∑ 

∞ 

k=0 

ν 

2k+1 

nγ 

1 

∙ ξ 

∙ 

k! ∙ Γ(1+k+1) 2 

- 

ν n 

∙ ξ γ 

a is,h(t) horizontal acceleration-time history. By using 

the response spectrum analysis a is,h(t) should 

be replaced by the spectral acceleration S a 

corresponding to T = 0 s. 

a gR reference peak ground acceleration on type A 

ground 

S soil factor 

γ I importance factor according to DIN EN 1998- 

1 (2010) or DIN EN 1998-4 (2007) 

Γ is,h participation factor for the rigid impulsive 

pressure component: Γ is,h = 1,0, because the 

rigid tank is moving together with the 

foundation. 

The natural period T kn for the n th -eigenmode of the 

sloshing wave is calculated with: 

2π 

T kn = 

g ∙ λ n ∙ tanh (λ n∙ γ) 

(3) 

R 

Rigid impulsive pressure 


pressure distribution corresponding to the rigid 

impulsive pressure component. 

The pressure distribution is given by the expression: 

∞ 

p is,h 

(ξ, ζ, θ, t) = 2 ∙ R∙ γ∙ ρ L ∙(-1)n 

with: 

p is,h 

I 1 ν n 

∙ ξ γ 

 

 

γ 

ν2 

n I ν n 

n=0 (4) 

⋅ [cos(ν n ∙ ζ)][cos(θ)]a is,h (t) ∙ Γ is,h 

rigid impulsive pressure component due to 


Fig. 3: Rigid impulsive pressure - Mode of vibration 

and pressure distribution 

Taking into account the pressure distribution of the 

tank shell (ξ = 1), equation (4) is simplified to: 

p is,h 

(ξ = 1, ζ, θ, t)= R∙ρ L 

2 ∙ γ ∙ (-1)n 

∞ 

n=0 

ν n 

2 

I 1 ν n 

γ 

 

I ν n 

γ 

cos(ν n∙ζ) 

⋅ [cos(θ)]a is,h (t)∙Γ is,h (5) 

The corresponding natural period is T = 0. 

137


EARTHQUAKE 

ARCHAEOLOGY 



Flexible impulsive pressure 


pressure distribution corresponding to the flexible 

impulsive pressure component. 

The flexible impulsive pressure component is 

calculated in an iterative procedure using the addedmass-model 

according to DIN EN 1998-4 (2007), 

Annex A. Within the framework of the procedure the 

tank wall is loaded with iterative calculated additional 

mass portions of the activated fluid. The pressure 

distribution is given by the expression: 

p if,h 

(ξ, ζ, θ, t) = 

∞ 

I 1 ν n 

∙ ξ γ 

2R ρ L 

ν n 

∙ I γ ν n cos(ν n∙ ζ) f(ζ)∙cos(ν n ∙ ζ) dζ 

n=0 

γ 0 

with: 

p if,h 

f(ζ) 

1 

⋅ [cos(θ)] a if,h (t) ∙ Γ if,h 

flexible impulsive pressure component due to 


deflection curve of the first (anti-symmetric) 

mode of oscillation of the tank-fluid 

combination 

a if,h(t) horizontal acceleration-time history as a result 

of an equivalent single-degree-of-freedom 

system. By using the response spectrum 

analysis the spectral accelerations 

corresponding to the first natural period T if,h,1 

should be used. 

Γ if,h 

(6) 

participation factor for the flexible impulsive 

pressure component due to horizontal 

excitation 

F(y) correction factor: F(γ) = 0,157 ∙ γ 2 + γ + 1,49 

s(ζ = 1/3) wall thickness of the tank at 1/3 filling 

height 

Fig. 4: Flexible impulsive pressure - Mode of vibration 

and pressure distribution. 

The function curve of f(ζ) in (8) is generally not 

known. It depends on the impact of the liquid onto 

the tank, in other words the aforementioned pressure 

function p if.h. Thus the joint bending form must be 

correctly determined iteratively. In DIN EN 1998-4 

(2007), Annex A the "added-mass concept" is 

proposed. According to this the resonating fluid, 

activated with the first bending shape, is added to the 

tank wall density. Then with the new "dry" tank 

model, the more accurate bending form is 

determined (Fischer et. al., 1991). 

The participation factor Γ if,h for the flexible impulsive 

pressure component is calculated as follows: 

Γ if,h = 

∫ 1 

0 

p if,h (ζ)dζ 

1 

∫ f(ζ) ∙ p if,h 

(ζ)dζ 

0 

, s(ζ) = const. (7) 

with: 

p if,h (ζ) pressure function of the flexible impulsive 

pressure component as a function of the filling 

height 

s(ζ) wall thickness of the tank 

Taking into account the pressure distribution of the 

tank shell (ξ = 1), equation (6) can be simplified to: 

p if,h 

(ξ = 1, ζ, θ, t) = 

∞ 

I 1 ν n 

γ 

= R∙ρ L 

2∙ ν nγ I ν n ∙cos(ν n∙ζ) f(ζ)∙cos(ν n ∙ζ) dζ 

n=0 γ 0 

1 

[cos(θ)]a if,h (t) ∙ Γ if,h 

(8) 

Fig. 5: Iteration process (Holtschoppen et. al., 2011) 

The first natural period T if,h,1 is calculated as follows: 

W L 

T if,h,1 = 2 ∙ F(γ) 

π ∙ g ∙ E ∙ s(ζ = 1/3) 

with: 

W L 

H ∙ρ 

= 2 ∙ R ∙ F(γ) 

L 

E ∙ s(ζ = 1/3) 

fluid weight: W L = π ∙ R 2 ∙ H ∙ ρ L 

∙ g 

(9) 

However as Figure 5 shows, even this iterative 

process is impractical because the calculation is 

based on a complex mathematical pressure function 

(6) which requires a coupling of a mathematical 

software tool with a finite element program. 

Though, with comprehensive parameter studies it is 

shown that the bending form of any tank can be 

described by using a parameterized sine wave which 

138


EARTHQUAKE 

ARCHAEOLOGY 



can be mapped correctly to the natural frequencies 

for all common geometric and material configurations 

(Cornelissen, 2010): 

f(ζ)=a⋅ sin π ⋅(ζ-b)⋅c+d (10) 

2 

The bending form of the combined vibration – and 

with this the parameters a, b, c, d – depends on 

- the tank slenderness (γ = H/R), 

- the Poisson’s ratio ν, 

- the ratio fluid mass to tank mass and 

- a changing wall thickness along the tank heigh. 

However, the last named influences are 

comparatively small. By using the sine function (10) it 

is possible to determine the flexible impulsive 

pressure without iteration. For practical use and to 

represent the pressure the equation 

p if,h 

(ξ,ζ,θ,t)=R⋅ρ L 

⋅ cos(θ) ⋅Γ if,h ⋅a if,h (t)⋅C if,h (ξ,ζ) (11) 

is suitable. C if,h(ξ,ζ) corresponds to the normalized 

pressure at θ=0. Figure 6 shows the variation of the 

factor C if,h(ξ=1,ζ) for the tank shell (ξ = 1) for different 

tank slendernesses. 

For simplicity, the formulations of all pressure 

functions are also provided in a standardized, 

tabulated form (Meskouris et. al., 2011). 

Fig. 6: Normalized flexible impulsive pressure C if,h. 

The participation factor for the flexible impulsive 

pressure Γ if,h, which is specified in equation (7), can 

be tabulated as well, assuming that the tank wall is 

constant and the tank mass is insignificant. 

CONCLUSION 

This paper provides guidance for the implementation 

of normative demands for the seismic design of liquid 

filled tanks. With the tabulation of different factors the 

cumbersome mathematical formulas for calculating 

the load components are avoidable, which allows an 

easy load generation for finite element analysis. 

References 

Cornelissen, P., (2010). Erarbeitung eines vereinfachten 

impulsiv-flexiblen Lastansatzes für die Berechnung von 

Tankbauwerken unter Erdbebenlast. Diplomarbeit, 

Fakultät für Bauingenieurwesen, RWTH Aachen 

DIN EN 1998-1, (2010). Auslegung von Bauwerken gegen 

Erdbeben – Teil 1: Grundlagen, Erdbebeneinwirkungen 

und Regeln für Hochbauten, Deutsche Fassung EN 

1998-1:2004+AC:2009, Deutsches Institut für Normung 

(DIN), Berlin 

DIN EN 1998-4, (2007). Auslegung von Bauwerken gegen 

Erdbeben – Teil 4: Silos, Tanks und Pipelines, Deutsche 

Fassung EN 1998-4:2006, Deutsches Institut für 

Normung (DIN), Berlin 

Fischer, F.D., Rammerstorfer, F.G., Scharf, K., (1991). 

Earthquake Resistant Design of Anchored and 

Unanchored, Liquid Storage Tanks under Three- 

Dimensional Earthquake Excitation. In: Structural 

Dynamics (G.I.). Springer Verlag, 317-371, ISBN 3-540- 

53593-4 

Holtschoppen, B., Cornelissen P., Butenweg, C., Meskouris, 

K., (2011). Vereinfachtes Berechnungsverfahren 

zur Berücksichtigung der Interaktionsschwingung bei 

flüssigkeitsgefüllten Tankbauwerken unter seismischer 

Belastung. Baustatik Baupraxis 11, Institut für Baustatik, 

Technische Universität Graz, ISBN 978-3-85125-115-9 

Housner, G.W., (1963). The dynamic behaviour of water 

tanks, Bulletin of the Seismological Society of America, 

Vol. 53, 381-387 

Meskouris, K., Holtschoppen B., Butenweg, C., Park, J., 

(2010). Seismische Auslegung von Silo- und 

Tankbauwerken. Festschrift für Prof. Dr.-Ing. Mangering, 

Universität der Bundeswehr München 

Meskouris, K., Hintzen, K.-G., Butenweg, C., Mistler, M., 

(2011). Bauwerke und Erdbeben. Vieweg und Teubner 

Verlag 

139


EARTHQUAKE 

ARCHAEOLOGY 



GEOLOGICAL CRITERIA FOR EVALUATING SEISMICITY: LESSONS LEARNED FROM 

THE PO PLAIN, NORTHERN ITALY 

Michetti, Alessandro M. (1, Leonello Serva (1), Andrea Berlusconi (1, Livio Bonadeo (1, Fabio Brunamonte (1, Francesca 

Ferrario (1Gianfranco Fioraso (2), Franz Livio (1, Giancanio Sileo (1), Eutizio Vittori (3) 

(1) Dipartimento di Scienze Chimiche e Ambientali, Università dell’Insubria, Via Valleggio 11, 22100, Como. ITALY. Email: 

alessandro.michetti@uninsubria.it 

(2). CNR, Istituto di Geoscienze e Georisorse, Via Valperga Caluso, 35,10125, Torino. ITALY. E-mail: g.fioraso@csg.to.cnr.it 

(3). Dipartimento Difesa del Suolo, Servizio Geologico d’Italia, ISPRA, Via V. Brancati 48, 00144 Roma, ITALY. Email: 

eutizio.vittori@isprambiente.it 

Abstract (Geological criteria for evaluating seismicity: lessons learned from the Po Plain, N Italy): We compare available 

literature and new geological and paleoseismological data, in particular from the Monte Netto - Brescia site, to maintain that A) 

strong earthquakes similar to the intensity IX to X MCS - M6.5 to 7 - historical events occurred in 1117 and 1222, though very rare, 

must be considered unlikely but credible events along virtually all the Quaternary faults throughout the whole Po Plain Foredeep, 

at least until we learn from geological and geophysical studies how/if the Quaternary tectonic structures near Lake Garda are truly 

different from those in other areas such as, for example, the Turin Hills, Monferrato, Insubria or around Piacenza, B) the late 

Quaternary (and especially Holocene) history of deformation and the local seismic landscape are far more valuable tools in 

estimating seismic hazard than has generally been appreciated until now in Italy – as well as in most parts of the world. The lack of 

significant earthquakes for 8 centuries in the Garda region and for longer elsewhere must prompt an effort to much deeper 

understanding of the actual seismic potential in one of the most populated and economically developed areas of Europe. 

Key words: Paleoseismology, late Quaternary tectonics, Po Plain, Seismic Hazard 

INTRODUCTION 

In the field of paleoseismology and earthquake 

hazard assessment, the year 2011 has been 

characterized by two “unexpected” destructive 

events, the Christchurch eq. in New Zealand and the 

Tohoku eq. in Japan. In particular, the devastating 

impact from the March 11, 2011, M9 Tohoku 

coseismic environmental effects – and of course in 

particular the massive tsunami – is definitely bound 

to deeply influence the standards for seismic hazard 

assessment worldwide for decades. 

In fact, these events have clearly shown that “the 

most important contribution to the understanding of 

long term seismicity, which is critical to the siting and 

design of safe structures and to the establishment of 

realistic building codes, is to learn more – region by 

region – of the late Quaternary history of 

deformation”. This of course includes the evidence 

for paleoseismicity. This is a quote from Allen (1975), 

a paper that has inspired also the title of this note. A 

retrospective look to the origin of the science of 

Paleoseismology is surprisingly instructive today, 

while we are still shocked by the effects of the 2011 

catastrophic seismic crises. 

The purpose of the present note is to argue that the 

approach discussed by Allen (1975), and further 

developed – among others – by Serva (1996) and 

Michetti et al. (2005) with the introduction of the 

notion of seismic landscape, is definitely validated 

not only by the investigations conducted before and 

after the 2011 events in New Zealand and Japan, 

which are seismically very active countries, but also 

by the lessons learned during the 4 decades of 

seismotectonic studies in the Po Plain, in Northern 

Italy, one of the most populated and developed areas 

of Europe, hosting a substantial portion of the Italian 

industrial production, many infrastructures and a 

number of high risk plants. 

From historical seismicity studies we know that the 

Po Plain, near Lake Garda, has been hit by two 

strong earthquakes in the XII – XIII centuries, the 

January 13, 1117, Verona, and Christmas 1222, 

Brescia, events, both in the range of intensity IX-X 

MCS, equivalent to M6.5 to 7 (Fig. 1). No other 

comparable earthquakes have occurred since than in 

the Central-Western Po Plain region. 

We use available literature and new geological and 

paleoseismological data to maintain that A) strong 

earthquakes similar to the Verona and Brescia must 

be considered credible events – though unlikely – 

along virtually all the Quaternary faults throughout 

the Po Plain Foredeep, at least until we understand 

from geological and geophysical studies if/how the 

Quaternary tectonic structures near Lake Garda are 

truly different from those in other areas such as in the 

Turin Hills, Monferrato, Insubria or around Piacenza 

(Figs. 1 and 2), B) the late Quaternary (and 

especially Holocene) history of deformation and the 

local seismic landscape are far more valuable tools in 

estimating seismicity and associated seismic hazard 

than has generally been appreciated until now in Italy 

- like in most parts of the world, and C) the “absence 

of evidence” for historical or instrumental strong/large 

140


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 1: Historical and instrumental earthquakes catalogue (modified from CPTI, 2004) and map of Quaternary capable faults in the 

Po Plain (after Livio et al., 2009). SP, Superga; MF, Monferrato; SV, Spina Verde; SL, San Colombano al Lambro; MN, Monte 

Netto; CV, Ciliverghe; CS, Castenedolo; MR, Mirandola; SO, Soncino. 

seismic events within a region should never be 

construed as “evidence of absence” of seismic 

hazard, unless systematic and specific paleoseismic 

analyses have been conducted in that region. 

Indeed, in the past decade our knowledge about the 

magnitude and rates of the key geomorphic agents 

which control the late Quaternary landscape 

evolution of the Po Plain and surrounding piedmont 

belts (along the margins of the Apennines to the S, 

and the Southern Alps to the N), has greatly 

improved. In particular, due to the large amount of 

novel information on the active tectonics and seismic 

potential of this region, it has become increasingly 

clear the role played by strong seismic events in this 

process. 

As already pointed out in the literature (e.g., Desio, 

1965; Carraro et al., 1995), both the Apennines and 

the Southern Alps piedmont belts are characterized 

by Quaternary tectonic features, i.e. drainage 

anomalies, isolated hills (e.g., Superga, Spina Verde, 

Fig. 2: Digital Elevation Model of the Superga and Monferrato Hills (vertical exaggeration 5x). 

141


EARTHQUAKE 

ARCHAEOLOGY 



San Colombano, Figs. 1 and 2), and buried structural 

highs (e.g., Soncino, Mirandola), evidence of the 

Quaternary growth of the two mountain belts beneath 

the Po Plain. However, only in the past decade, 

thanks to the integrated research conducted by the 

University of Insubria and partners (ISPRA- 

Geological Survey of Italy, ENI E&P, UCL London, 

Colorado University-Boulder, Innsbruck University, 

Università Statale di Milano, Università di Brescia, 

INGV, CNR Torino, Regione Lombardia) it has been 

possible to study the first paleoseismic site showing 

evidence for repeated latest Pleistocene to Holocene 

compressional surface faulting earthquakes along the 

Monte Netto Backthrust, Brescia (Livio et al., 2009; 

Figs. 3 and 4). At this site, 3 paleoseismic events 

have been recorded in the past ca. 40 kyr B.P. Based 

on trench data and geomorphic investigations, the 

magnitude estimated for these events is consistent 

with the earthquake size of the Christmas 1222 

Brescia event (M6.2 to 6.8; e.g., Serva, 1990; 

Guidoboni & Comastri, 2005). 

It is important to note that this result has been 

possible only due to the availability of extensive oil 

exploration subsurface data (seismic reflection and 

deep stratigraphic boreholes, courtesy of ENI E&P); 

and large quarry excavations at Monte Netto site, 

which provided outstanding exposures of late 

Quaternary growth anticlines, bending moment 

surface faulting, and coseismic liquefaction (data and 

site access courtesy of Fornaci Laterizi Danesi SpA). 

In fact, the Monte Netto structure is a unique case 

study in the Po Plain. Which is logic in a sense, due 

to i) the moderate seismicity level of this tectonic 

province, ii) the local high erosional and depositional 

rates compared to slip rates of capable faults, iii) the 

need for expensive investigations (seismic reflection 

profiles, boreholes, trenching, dating) due to the 

reverse style of faulting, and a geomorphic and 

human environment unfavorable to paleoseismic 

analyses. 

Literature data and our ongoing research strongly 

suggest that several Quaternary structures in the Po 

Plain share the same tectonic and geomorphic 

features of the Monte Netto Backthrust. However, 

virtually no comparable paleoseismic information and 

detailed structural characterization is available for 

these structures. 

DISCUSSION 

We are systematically investigating the structural 

features and evidence for capability along the 

Quaternary faults in Figure 1; to be scientifically 

prudent, at present we cannot rule out that many of 

these faults are able to generate earthquakes similar 

to the Verona and Brescia Middle Age events. So, in 

view of the difficulties in interpreting the historic 

record from Middle Age events, and the lack of 

detailed paleoseismic information for most of the 

structures showing some evidence for capability, we 

stress the need to be exceedingly conservative in 

estimating the probability of major damaging 

earthquakes in the Po Plain. “We have been 

surprised too often in the past, and we cannot afford 

to be surprised too many times in the future” (Allen, 

1975); this must be the case for the Po Plain, point 

made even more clear after the Japan and New 

Zealand experience in 2011, where, it must be said, 

the basic knowledge was seemingly much better. 

The Po Plain hosts one of the most industrially 

developed areas in the world, and like in many other 

“megacities” located in seismically active regions, 

every year the seismic risk keeps growing up, with 

more and more serious societal effects to be 

expected. The Po Plain has not been the subject of 

an intensive seismotectonic research effort since the 

end of the Italian Nuclear Program more than 20 

years ago (e.g., Serva, 1990). The state of the art in 

paleoseismology, and the seismic events of 2011, 

demonstrate that our conservatism is scientifically 

sound. The lack of events in the last 8 centuries in 

Fig. 3: Decametric Late Quaternary growth anticline at the Monte Netto site. Note the gravity graben due to due to bending 

moment normal faulting on the fold crest; new exploratory trenching across this graben revealed 3 paleoseismic surface faulting 

events in the past 40 kyr B.P. The coseismic liquefaction shown in Figure 4 is located near the core of the anticline, indicated by 

the person for scale. 

142


EARTHQUAKE 

ARCHAEOLOGY 



the Garda region and longer (how much?) elsewhere 

must not indulge us to optimism, but, on the contrary, 

warn us to get prepared, how unlikely a M6-7 seismic 

event can be. As a matter of fact, siting and design of 

safe structures, and realistic building codes for the 

Po Plain should be based on a much better 

understanding of the long term seismicity, that is on 

the detailed knowledge of the late Quaternary history 

of deformation and Holocene paleoseismicity. This is 

the kind of knowledge revealed by the new data 

collected at the Monte Netto site, as of yet basically 

the only paleoseismic site available in the region. 

Fig. 4: Monet Netto site, liquefaction feature near the 

core of the anticline described in Figure 3. 

Acknowledgements: This work has been partly funded by 

Project S1 INGV-DPC 2007-2009; and by grants from the 

Operational Programme Cross Border Cooperation IT / CH 

2007-2013 – project “SITINET: census, networking and 

development of geological and archaeological sites” ID 

7621984. 

References 

Allen, C., (1975), Geological criteria for evaluating 

seismicity. Geological Society of America Bulletin; 86 (8), 

1041-1057; DOI: 10.1130/0016-7606 

Desio, A., (1965), I rilievi isolati della Pianura Lombarda ed i 

movimenti tettonici del Quaternario. Rendiconti Istituto 

Lombardo Accademia Scienze e Lettere, Sez. A, 99, 881- 

894. 

Carraro, F., G. Collo, M.G. Forno, M. Giardino, F. Maraga, 

A. Perotto & D. Tropeano, (1995), L’evoluzione del 

reticolato idrografico del Piemonte centrale in relazione 

alla mobilità quaternaria. In: R. Polino & R. Sacchi, Eds., 

Atti del Convegno Rapporti Alpi-Appennino e guide alle 

escursioni, Peveragno (CN), 31 maggio - 1 giugno 1994, 

Accademia Nazionale delle Scienze detta dei 40, 1995 - 

XII, 441-465. 

Guidoboni, E. & A. Comastri, (2005), Catalogue of 

earthquakes and tsunamis in the Mediterrean area from 

the 11th to the 15th century. 1037 pp., SGA Storia 

Geofisica Ambiente srl, Bologna. 

Livio, F., A. Berlusconi, A.M. Michetti, G. Sileo, A. Zerboni, 

L. Trombino, M. Cremaschi, K. Mueller, E. Vittori, C. 

Carcano, & S. Rogledi, (2009), Active fault-related folding 

in the epicentral area of the December 25, 1222 (Io = IX 

MCS) Brescia earthquake (Northern Italy): 

seismotectonic implications. Tectonophysics, 476 (1-2), 

320-335, doi:10.1016/j.tecto.2009.03.019 

Michetti A.M., F. Audemard & S. Marco, (2005), Future 

trends in paleoseismology: Integrated study of the 

seismic landscape as a vital tool in seismic hazard 

analyses. Tectonophysics, 408 (1-4), 3-21. 

Serva, L., (1990). Il ruolo delle Scienze della Terra nelle 

analisi di sicurezza di un sito per alcune tipologie di 

impianti ad alto rischio. Il terremoto di riferimento per il 

sito di Viadana (MN). Bollettino Società Geologica 

Italiana, 109 (3), 375-411. 

Serva, L., (1996), Criteri geologici per la valutazione della 

sismicità: considerazioni e proposte. Atti dei Convegni 

Lincei 122, “Terremoti in Italia”, Accademia Nazionale dei 

Lincei, 1-2 Dicembre 1994, Roma, pp. 103-116. 

143


EARTHQUAKE 

ARCHAEOLOGY 



ANCIENT SEISMITES AS GEODYNAMICAL INDICATOR: APPROACH TO CONSTRUCT A 

REACTIVATION EVENT ON THE MAIN BOUNDARY THRUST IN THE HIMALAYAN 

REGION 

Mishra, Anurag (1), D.C. Srivastava (2), Jyoti Shah (3) 

(1) Department of Earth Sciences, Indian Institute of Technology, Roorkee, Haridwar, Uttarakhand-247667. 

Email: anurag2009iitr@gmail.com 

Abstract: Diversified soft-sediment deformation structures exposed in the nearest vicinity of the Main Boundary Thrust in the 

South-eastern Kumaun Himalaya, show progressive increase in abundance and complexity towards the thrust. After establishing 

liquidization as dominant deformation mechanism, a whole to part study has been done to identify most possible triggering 

process, excluding all the other probable processes, integrating criteria based approaches in combination. The structures show 

their development due to seismic origin, hence called as seismites, define a clear relationship with the Main Boundary Thrust. 

Magnetostratigraphic dates of the associated sediments show 4-5 Ma as the age of development of the structures. Hence these 

structures, acting as geodynamical indicator, show paleoseismicity of the area, and dates back a reactivation event on the Main 

Boundary Thrust in the Kumaun Himalaya around 4-5 Ma, when established age for the formation of this thrust is around 11 Ma. 

Key words: Soft-sediment deformation structures, seismites, Main Boundary Thrust 

INTRODUCTION 

The Alpine-Himalayan orogeny originating in the late 

period of the Mesozoic era led to development of a 

series of orogen scale detachment boundaries, 

named as the South Tibet Detachment Zone, the 

Main Central Thrust, the Main Boundary Thrust and 

the Main Frontal Thrust (Fig. 1a) from North to South 

(Jackson and Bilham, 1994). 

The Main Boundary Thrust gains importance due to 

recent active seismicity of the thrust causing 

catastrophic earthquakes and accommodation of 

large scale crustal shortening along the thrust 

(Nakata, 1989). But the dating of the Main Boundary 

Thrust has not been given much consideration in 

previous works in comparison to all the other 

boundary thrusts, perhaps due to poor exposures 

and lack of cross-cutting relationships. The available 

dates for the formation of thrust do not converge 

perhaps due to ambiguous relation between the 

activity of the thrust and applied methods, such as 

one of the date suggest it’s development before 11 

Ma whereas another younger than 5 Ma (Yin, 2006). 

Most of the available dates for the activity of the Main 

Boundary Thrust are largely based on study of 

change in depositional pattern in the Himalayan 

Foreland and study of exhumation based on rate of 

subsidence of the Himalayan Foreland. It is 

presumed that exhumation is directly incorporated 

with the uplift in the hanging wall of the Main 

Boundary Thrust. But the Main Boundary Thrust is 

Southerly of the Thrusts cutting through the 

Himalaya, and exhumation can be related with any of 

the Northerly thrust cutting through the rocks. 

The soft-sediment deformation structures of seismic 

origin are called as seismites (Seilacher, 1969). They 

have been used for understanding geodynamic 

activity of causative seismogenic faults and 

paleoseismicity of an area. 

Fig. 1: (a) Litho-tectonic units and demarcating boundaries 

in the Himalayan terrain. (b) Satellite imagery showing 

Himalayan terrain, location of Kumaun Himalaya and our 

study area. 

We use the soft-sediment deformation structures 

exposed in the vicinity of the Main Boundary Thrust 

for constraining one of the movement events along 

the thrust, in case they are seismites and show an 

undisputed relationship with the Main Boundary 

144


EARTHQUAKE 

ARCHAEOLOGY 



Thrust. We follow a 3 fold-process: To classify these 

structures, to understand their origin in terms of 

deformation mechanism, driving force system and 

triggering agent, and finally finding out possible 

relationship with the Main Boundary Thrust. 

OBSERVATIONS 

The convolute laminations developed in medium to 

fine grained sandstone lithology (Fig. 3). Folded 

laminations, thicker at the crest and thinner at the 

trough show that sediments became hydro-plastic in 

nature during development of these structures (Fig. 

4). At the crest, the laminations are blurred and show 

homogenization of sediments (Fig. 4). 

A large variety of the soft-sediment deformation 

structures are abundantly exposed in the vicinity of 

the Main Boundary Thrust in the southeastern 

Kumaun Himalaya (Fig. 1b). The structures identified 

in the area are the deformed cross-stratifications, the 

liquefaction pockets, the slump folds, the mushroom 

structures, the convolute laminations, the sand 

dykes, the synsedimentary faults, the fluid-escape 

structures and the flame structures. 

Fig.4: Convolute lamination in relatively medium grained 

sandstone. Arrows show blurring of laminations and 

homogenization of sediments. 

Fig. 2: Recumbent cross-stratification in sandstone. Arrows 

point to minor faults cutting through the liquefied 

laminations. Note the fluid-escape structure in the rightupper 

part originating from the liquefied bed below. 

Fig. 5. Liquefaction pocket in sandstone. Arrows mark the 

offset in laminations. 

Liquefaction pockets are developed due to complete 

homogenization of sediments and obliteration of 

primary structures (Fig. 5) by liquefaction within the 

local pockets of few centimeters (Lowe, 1975). 

Laminations are offset along these pockets (Fig. 5). 

Fig. 3: Typical convolute laminations in sandstone 

The recumbent cross-stratifications show well 

preserved laminations and thickness of individual 

laminae varies throughout the structure (Fig. 2). Most 

plausible explanation for their development is 

deformation of liquefied sediments due to current 

drag, as foresets are folded towards paleocurrent 

direction (Allen and Banks, 1972, Allen, 1982). The 

minor faults are developed due to cohesive nature of 

sediments, when grain to grain packing established 

after liquefaction (Owen, 1987, 1996). 

Sand dykes developed due to intrusion of coarsegrained 

sandstone upward, originated by drag force 

of fluids (Fig 6 and Fig. 7). Their origin is well 

established and related to fluidization. 

Mushroom structures developed due to fluidization in 

the upward regime, called as elutriation, by which a 

plume type (Fig. 8) of forceful injection took place, 

aligning clay minerals in the neck connected to the 

clay enriched dome shaped main body (Lowe, 1975, 

Allen, 1982). 

145


EARTHQUAKE 

ARCHAEOLOGY 



body, and these structures are confined within 

undeformed horizons. 

Fluid escape structures (Fig. 9) are developed due to 

expulsion of fluids to surrounding beds of same 

comparable lithology due to any of the processes that 

causes extraction of fluids from the beds such as 

liquefaction, fluidization or compaction of sediments. 

Fig. 6: Sand dyke in Sandstone. Note the alignment of 

primary laminations towards the direction of injection. 

Fig. 9. Fluid escape structures in sandstone 

Fig. 7. Multiple sand dyke system in medium-grained 

sandstone. Note the reorientation of parallel laminations 

due to drag of upward sand injection. A synsedimentary 

fault is cutting through the parallel laminated sandstone and 

sand dyke. 

Fig. 10. Syn-sedimentary faults in sandstone. Notice the 

geometry of faults. 

Several variably oriented planar or listric 

synsedimentary faults (Fig. 10) developed that cut 

through the soft-sedimentary deformation structures. 

Compaction may be one of the plausible 

mechanisms for the development of these structures, 

but observing the insufficient thickness of 

overburden, large scale offset of many faults and 

superposition on earlier phases of fluidized 

structures, seismic activity seems to be most 

possible cause. 

Fig. 8. Two classes of mushroom structures in medium 

grained sandstone. In left photograph, the upper domal part 

is detached from the main body, whereas in the right one it 

is connected through a narrow neck. Notice the darkening 

of structure in compared to surrounding lithology and 

warping and folding of surrounding laminations (shown by 

arrow) 

Their development indicates requirement of instant 

force, so that elutriation and sudden expulsion could 

take place. The best possible originating force for 

these structures is seismic shaking, as there is no 

difference in grain size between injecting and injected 

DISCUSSION 

Liquefaction and fluidization are the most dominant 

deformation mechanisms for development of these 

structures. Triggers such as groundwater movement, 

rapid sediment deposition, storm waves, meteoritic 

impact and earthquake shocks can be responsible for 

the development of the soft-sediment deformation 

structures (Van Loon and Brodzikowski, 1987). We 

follow the criteria given by Sims (1973), Owen and 

Moretti (2011) and Owen et al. (2011) for identifying 

the trigger of these structures. These structures are 

confined within rocks of varied lithology such as, in 

several varieties of sandstone and in mudstonemarlstone 

beds. This indicates the probability of out 

146


EARTHQUAKE 

ARCHAEOLOGY 



of system, allogenic trigger for the development of 

these structures. Rapid sedimentation cannot be a 

cause of development as magnetostratigraphic 

dating reveals same rate of sedimentation throughout 

the area (Kotalia et al., 2001). Storm waves can be 

neglected due to established fluviatile origin of these 

sediments (Tandon, 1976). We suggest seismic 

origin, after excluding all the probable triggers. Our 

inference gets more strength as it fits on criteria 

based approaches and gives several evidences for 

approval, given in the following lines: 

1) The structures confined within the same 

stratigraphic horizons, follow these beds for 

a sufficient lateral extent of about more than 

2 km in linear tract in the nearest vicinity of 

the Main Boundary Thrust. 

2) These are confined within cohesionless 

loose and friable sediments that are 

potentially liquefiable. 

3) These structures are comparable with the 

structures developed in seismic shaking 

experiments. 

4) The structures are exposed in the vicinity of 

the Main Boundary Thrust that is one of the 

most seismically active thrusts in the 

Himalayan Terrain. 

5) Several small scale structures such as 

minor faults in the recumbent-cross 

stratification are developed. 

6) Mushroom structures are very suggestive 

for seismic trigger. 

7) Frequent occurrence of faults with liquefied 

structures indicates need of instant shaking 

as soon as the grain to grain contact 

develops that suggest possible seismic 

origin. 

It is generally agreed that seismic activity of not less 

than 5.5 magnitude on Richter scale can possibly be 

recorded in sediments, as seismites (Ambraseys, 

1988). Hence this event shows an earthquake event 

of at least 5.5 magnitude on Richter scale. Two 

phases of seismic trigger has been established on 

the basis of overprinting of structures. The sand dyke 

representing the fluidization of one phase must has 

been developed earlier than synsedimentary fault of 

second phase cutting through it, that shows brittle 

nature of origin (Fig. 10). 

Confined within the 4-5 Ma horizon (Kotalia et al., 

2001), the soft-sedimentary deformation structures 

increase in abundance and, towards the Main 

Boundary Thrust. As the development of the Main 

Boundary Thrust occurred c. 10-11 Ma. (Miegs et al., 

1995). Hence, it is evident that the ancient seismites, 

discussed in this article, are related to the 4-5 Ma old 

reactivation event on the Main Boundary Thrust. 

Acknowledgements: I express my thanks to conference 

organizers for giving me chance to present my work. 

References 

Allen, J.R.L., Banks, N.L., 1972. An interpretation and 

analysis of recumbent-folded deformed cross-bedding. 

Sedimentology 19, 257-283. 

Allen, J.R.L., 1982. Sedimentary Structures: their character 

and physical basis. Development in Sedimentology. Vol. 

30. Elsevier, Amsterdem. 663 pp. 

Ambraseys, N.N., 1988. Engineering seismology. 

Earthquake Engineering Structural Dynamics 17, 1–105. 

Jackson, B., Bilham, R., 1994. Constraints on Himalayan 

Deformation Inferred from the vertical velocity fields in the 

Nepal and Tibet. Journal of Geophysical Research 99 

(B7), 13897-13,912. 

Kotalia, B.S., Nakayama, K., Bhalla, M.S., Phartiyal, B., T. 

Kosake, 2001. Lithology and magnetic stratigraphy of 

Lower-Middle Siwalik succession between Kathgodam 

and Ranibagh, Kumaun Himalaya. Journal Geological 

Society of India 58, 411-423. 

Lowe, D.R., 1975. Water escape structures in coarsegrained 

sediments. Sedimentology 22, 157–204. 

Miegs, A. J., Burbank, D. W., Beck, R. A., 1995. Middle late- 

Miocene (>10 ma) formation of the main boundary thrust 

in the western Himalaya. Geology. 23, 423-426. 

Nakata, T., 1989. Active faults of the Himalaya of India and 

Nepal. Geological Society of America. Special paper. 

232, 243-264. 

Owen, G., 1987. Deformation processes in unconsolidated 

sands. In: Jones, M.E., Preston, R.M.F. (Eds.), 

Deformation of Sediments and Sedimentary Rocks, 

Geological Society of (London) Special Publication. No. 

29, pp. 11–24. 

Owen, G., 1996. Experimental soft-sediment deformation 

structures formed by the liquefaction of unconsolidated 

sands and some ancient examples. Sedimentology 43, 

279–293. 

Owen, G., Moretti, M., 2011. Identifying triggers for 

liquefaction-induced soft-sediment deformation in sands. 


Owen, G., Moretti, M., Alfaro, P., 2011. Recognising triggers 

for soft-sediment deformation: current understanding and 

future directions. Sedimentary Geology 235, 133-140. 

Seilacher, A., 1969. Fault-graded beds interpreted as 

seismites. Sedimentology 13, 155–159. 

Sims, J.D., 1973. Earthquake-induced structures in 

sediments in Van Norman Lake, San Fernando, 

California. Science 182, 161–163. 

Tandon, S.K., 1976. Siwalik sedimentation in a part of 

Kumaun Himalaya. Sedimentary Geology 16, 131–154. 

Van Loon, A.J., Brodzikowski, K., 1987. Problems and 

progress in the research on soft sediment deformations. 

Sedimentary Geology 50, 167–193. 

Yin, A., 2006. Cenozoic evolution of the Himalayan orogen 

as constrained by along-strike variation of structural 

geometry, exhumation history, foreland sedimentation. 

Earth Sciences Reviews 76, 1-131. 

147


EARTHQUAKE 

ARCHAEOLOGY 



SAMPLING BIASES IN THE PALEOSEISMOLOGICAL DATA 

Mouslopoulou Vasiliki (1,2), Andrew Nicol (3), John J. Walsh (1), 

John G. Begg (3), Dougal B. Townsend (3), Dionissios T. Hristopulos (2) 

(1) Fault Analysis Group, University College Dublin, Dublin 4, Ireland. Email: vasiliki@mred.tuc.gr, john@fag.ucd.ie 

(2) Dpt. of Mineral Resources Engineering, Technical University of Crete, 73100, Greece. Email: dionisi@mred.tuc.gr 

(3) GNS Science, PO Box 30368, New Zealand. Email: a.nicol@gns.cri.nz ; j.begg@gns.cri.nz ; d.townsend@gns.cri.nz 

Abstract (Sampling biases in the paleoseismological data): The recent earthquakes in Christchurch, New Zealand, show that 

active faults, capable of generating large-magnitude earthquakes, can be hidden beneath the Earth’s surface. Here we combine 

near-surface paleoseismic data with deep (


EARTHQUAKE 

ARCHAEOLOGY 



Pliocene (3-4 Ma) and has produced kilometre-scale 

cumulative displacements on normal faults (Nicol et 

al., 2007; Giba et al., 2010; Mouslopoulou et al., In 

review), active faulting at the ground surface 

(Townsend et al., 2010; Mouslopoulou et al., in 

review) and historical seismicity (Sherburn & White, 

2006). 

Crustal extension is also accompanied by 

widespread Late Quaternary ( 

on the Taranaki Peninsula. The most recent phase of 

volcanic activity started about 100 ka (last eruption 

ca. 250 years ago) and resulted in the formation of 

the impressive volcanic cone of Mt. Taranaki. 

Although Mt. Taranaki rises to 2518 m above sealevel 

(a.s.l.) and dominates the landscape, studies 

show no direct link between the timing of prehistoric 

large-magnitude earthquakes in the rift and volcanic 

eruptions or episodes of cone collapse (Townsend et 

al., 2010). 

Here we combine seismic-reflection and trench data 

for individual faults to chart their growth in 

displacement through time. Two-dimensional (2D) 

seismic-reflection lines occupying the southwest of 

Mt. Taranaki, extend across the entire onshore width 

of the rift and in a rift-parallel direction for about 20 

km (Fig. 1). The 2D seismic survey comprises a total 

of 21 fault-perpendicular and 17 fault-parallel lines 

spaced at 0.5 to 5 km. These seismic lines image all 

of the known active faults and provide estimates of 

 

sites. Seismic reflectors extending to depths of up to 

ca. 5 km have been tied to five wells (Figs 1 and 2a) 

in which the stratigraphy was dated using 

micropaleontology (King & Thrasher, 1996). The six 

youngest of these seismic reflectors (ca. 0.5, 2, 3, 

3.6, 5.5 and 10 Ma) were traced along and across 

the faults around the grid of seismic lines and provide 

information on the accumulation of fault 

displacements during the Plio-Pleistocene. For 

individual faults in the area of study fault vertical 

displacements are up to ca. 1.5 km and decrease 

with reduction in horizon age from 3.6 Ma (Fig. 3a). 

The paleoearthquake history of the six active faults 

with resolvable surface traces is defined by 

constraints from 10 excavated trenches. Associated 

trench data permit identification of a total of 23 large 

paleoearthquakes (M>5.5) that ruptured the ground 

surface since 27 ka (Fig. 2b). The earthquake record 

is complete since 5 ka on all six active faults and for 

27 kyr on two of these faults. Vertical displacements 

measured in the trenches range from 0.1 m to 4 m 

(e.g., Townsend et al., 2010; Fig. 2b) and accrued 

from earthquakes with single-event displacements of 

up to 1.2 m. These displacements, which represent 

point measurements on fault traces (accrued during 

one or more earthquakes), are supplemented by 

scarp heights recorded along the lengths of each 

active trace. Paleoearthquake ages are primarily 

constrained by 14 C dating of near-surface (< 6 m 

deep) stratigraphy displaced by active faults (Fig. 

2b). Further details of the trenching are documented 

by Townsend et al. (2010). 

SAMPLING BIASES 

Seismic-reflection lines record many more faults than 

can be inferred from field mapping of active traces. 

Thirteen faults displace the 0.5 Ma horizon and 

clearly extend upwards towards the ground surface 

(Fig. 2a). Field mapping and interpretation of aerial 

ortho-photographs suggest that six of these faults 

have surface scarps up to 4 m high and displace a 

diachronous landscape mainly ranging in age from 

~8 to 27 ka (Fig. 1) (Townsend et al., 2010; 

Mouslopoulou et al., in review). Trench data confirm 

that these faults are active and have ruptured the 

ground surface during large-magnitude prehistoric 

earthquakes (Figs. 2b) (Townsend et al., 2010). The 

remaining seven faults that displace the 0.5 Ma 

horizon generally have large cumulative 

displacements (>1000 m throw) and also displace 

near-surface horizons (


EARTHQUAKE 

ARCHAEOLOGY 



lengths are generally between 20 and 30 km. 

Analysis of individual faults suggests that the subsurface 

fault lengths are about 2-8 times larger than 

their equivalent surface traces. Some of this 

discrepancy arises because fault scarps


EARTHQUAKE 

ARCHAEOLOGY 



often through impressive fault scarps (Nicol et al., 

2009). These faults are likely to be those which have 

recently accommodated successive earthquakes 

(e.g. during the Holocene). This view is supported by 

seismic reflection lines which show that the number 

of surface-rupturing faults in the rift is significantly 

less than the total number of faults in the rift (Figs 1 

and 2a). Examination of the buried faults indicates 

that, in addition to the six ground-rupturing faults, 

there are at least a further seven faults, including the 

larger faults in the system, that reach within ~50 m of 

the modern ground surface and are probably active 

(i.e. still accruing displacement and capable of 

producing future ground-rupturing earthquakes). For 

these additional seven active faults, long-term 

displacement rates are greater than Holocene rates 

by up to a factor of five (Fig. 3a). Thus, faults with 

increased and decreased displacement rates in the 

Holocene could be present in approximately equal 

proportions. As a result of these temporal rate 

changes, the sums of long-term rates on all faults in 

the system and Holocene displacement rates on the 

13 active faults identified are similar (i.e. long-term 

3.4±0.5 mm/yr and Holocene 3.1±0.5 mm/yr). This 

similarity in the sum of the rates over each time 

interval supports the view that regional rates of 

extension may not have changed, and that the 

temporal clustering of earthquakes reflects migration 

of the locus of fault activity from one fault to another. 

An important consequence of this migration is that 

faults which are currently in a relatively quiescent 

phase of earthquake activity may in the next 10 kyr, 

for example, become relatively active. Earthquake 

hazard analysis must take account of these relatively 

quiescent faults. 

CONCLUSIONS 

In the Taranaki Rift displacement rates vary 

temporally on individual faults by in excess of an 

order of magnitude over timescales of thousands to 

millions of years. These changes are attributed to 

fault interactions rather than to changes in regional 

strain rates. During the Holocene fault displacement 

rates were both faster (~50%) and slower (~50%) 

than their million-year averages. Faults moving faster 

than the long-term average can be identified in the 

landscape, while those moving slower cannot. The 

numbers of active faults and their active trace lengths 

are underestimated by at least 50% using 

geomorphic mapping of the scarps. Therefore, the 

number of potential earthquake sources may be 

significantly higher than it is represented in seismic 

hazard models. As a result, some future earthquakes 

will occur on faults that were not previously known to 

be active. Integration of seismic-reflection with 

paleoseismic data provides a basis for identifying 

active faults not observed at the ground surface, 

estimating maximum fault-rupture lengths and 

improving earthquake hazard assessment. 

References 

Coppersmith, K.J., (1989). On spatial and temporal 

clustering of paleoseismic events. Seismological 

Research Letters 59, 299-230. 

Giba, M., A. Nicol & J.J. Walsh, (2010). Evolution of faulting 

and volcanism in a back-arc basin and its implications for 

subduction processes. Tectonics 29, 

doi:10.1029/2009TC002634. 

King, P.R. & G.P. Thrasher, (1996). Cretaceous and 

Cenozoic geology and petroleum systems of the Taranaki 

Basin, New Zealand. Institute of Geological & Nuclear 

Sciences Monograph 13, 243 p. 

Marco, S., M. Stein & A. Agnon, (1996). Long-term 

earthquake clustering: a 50 000-year paleoseismic record 

in Dead Sea Graben. Journal of Geophysical Research 

101, 6179-6191. 

Mouslopoulou, V., J.J. Walsh & A. Nicol, (2009). Fault 

displacement rates on a range of timescales. Earth and 

Planetary Science Letters 278, 186-197. 

Mouslopoulou, V., A. Nicol, J.J. Walsh, J.G. Begg, D.B. 

Townsend & D.T. Hristopulos, (in review). Relations 

between paleoearthquakes and million year fault growth 

in an active rift. Journal of Structural Geology, 2011. 

Nicol, A., J.J. Walsh, K. Berryman, P. Villamor, (2006). 

Interdependence of fault displacement rates and 

paleoearthquakes in an active rift. Geology 34, 865-868. 

Nicol, A., C. Mazengarb, F. Chanier, G. Rait, C. Uruski & L. 

Wallace, (2007). Tectonic evolution of the Hikurangi 

subduction margin, New Zealand, since the Oligocene. 

Tectonics 26(4), TC4002. doi:10.1029/2006TC002090. 

Nicol, A., J.J. Walsh, V. Mouslopoulou, & P. Villamor, 

(2009). Earthquake histories and Holocene acceleration 

of fault displacement rates. Geology 37, 911–914. 

Palumbo, L., L. Benedetti, D. Bourles, A. Cinque & R. 

Finkel, (2004). Slip history of the magnolia fault 

(Apennines, Central Italy) from 36Cl surface exposure 

dating: evidence for strong earthquakes in the Holocene. 

Earth and Planetary Science Letters 225, 163–176. 

Schwartz, D.P. & K.J. Coppersmith, (1984). Fault behavior 

and characteristic earthquakes: examples from the 

Wasatch and San Andreas fault zones. Journal of 

Geophysical Research 89, 5681–5698. 

Sherburn, S. & R.S. White, (2006). Tectonics of the 

Taranaki region, New Zealand: earthquake focal 

mechanisms and stress axes. New Zealand Journal of 

Geology and Geophysics 49, 269–279. 

Sieh, K., M. Stuiver & D. Brillinger, (1989). A more precise 

chronology of earthquakes produced by the San Andreas 

Fault in Southern California. Journal of Geophysical 

Research 94, 603-623. 

Townsend, D., A. Nicol, V. Mouslopoulou, J.G. Begg, R.D. 

Beetham, D. Clark, M. Giba, D. Heron, B. Lukovic, A. 

McPherson, H. Seebeck & J.J. Walsh, (2010). 

Paleoearthquake histories across a normal fault system 

in the southwestern Taranaki Peninsula, New Zealand. 

New Zealand Journal of Geology and Geophysics 53, 1- 

20. 

Weldon, R., K. Scharer, T. Fumal, G. Biasi, (2004). 

Wrightwood and the earthquake cycle: what a long 

recurrence record tells us about how faults work. GSA 

Today 14, 9, 4-10. 

Wells, D.L. & K.J. Coppersmith, (1994). New empirical 


width, rupture area and surface displacement. Bulletin of 

the Seismological Society of America 8, 974-1002. 

Acknowledgements: This research was funded by an IIF 

Marie Curie Fellowship of the European Community’s 7th 

Framework Program (contract no. PIIF-GA-2009-235931) 

and an IRCSET (Irish) Postdoctoral Fellowship. 

151


EARTHQUAKE 

ARCHAEOLOGY 



EARTHQUAKES IN AQABA, JORDAN OVER THE PAST 2,000 YEARS: 

EVIDENCE FROM HISTORICAL, GEOLOGICAL, AND ARCHAEOLOGICAL DATA 

Tina M. Niemi (1) 

(1) Department of Geosciences, University of Missouri-Kansas City, 5100 Rockhill Road, Flarsheim Hall 420, Kansas City, MO 

64110 U.S.A. Email: niemit@umkc.edu 

Abstract (Earthquakes in Aqaba, Jordan Over the Past 2,000 Years: Evidence from Historical, Geological, 

and Archaeological Data): Aqaba lies at the boundary between the Gulf of Aqaba and Wadi ‘Arabah segments of 

the Dead Sea Fault Zone (DSFZ) and is therefore situated to sustain earthquake damage from rupture of either 

segment. The rupture history of these fault segments remains rather enigmatic because of the low population 

density around them throughout history. This paper presents a synthesis of paleoseismic trenching, archaeological 

excavations, and historicl data and provides a model for the recurrence of earthquakes along the southern DSFZ. 

Significant periods of active seismicity in the 4 th , 7 th -8 th , 11 th -13 th , and 15-16 th centuries suggest a three- to fivecentury 

recurrence rate of faulting of the Gulf of Aqaba and Wadi ‘Arabah segments of the DSFZ. It is interesting to 

note that earthquakes have been coincident with major political transitions that have occurred in this region and thus 

are likely to have played a significant role in these cultural shifts. 

Key words: Archaeoseismology, Dead Sea Transform, Jordan, Earthquakes 

INTRODUCTION 

The Dead Sea fault zone (DSFZ) is an left-lateral, 

strike-slip transform plate boundary between the 

Arabian and Sinai plates (Fig. 1). Earthquakes in the 

southern ‘Arabah and Gulf of Aqaba of Jordan and 

Israel are created by motion along the DST. The Mw 

7.2 Nuweiba earthquake of November 22, 1995 that 

ruptured a submarine fault in the Gulf of Aqaba (Gulf 

of Eilat) was the largest earthquake in the modern 

instrumented era (Hofstetter et al., 2003) along the 

Dead Sea fault. Most of the significant damage was 

concentrated in cities in the Sinai Peninsula near the 

epicenter, but damage was also reported from the 

Saudi Arabian coastline and the cities of Aqaba, 

Jordan and Eilat, Israel; both about 70 km north of 

the epicenter. 

Historical earthquake data reported in recent 

catalogues (Guidoboni, 1994; Guidoboni and 

Comastri, 2005; Ambraseys, 2009) suggest that the 

seismic events in A.D. 110-114?, 363, 749, 1068, 

1212, 1458, and 1588 likely caused damage in the 

region of southern Jordan and Aqaba. It is unclear 

whether events recorded in the Dead Sea region or 

Jerusalem, such as the earthquakes of A.D. 419, 

597, 634, 659, 1293, and 1546, could have also 

caused damage in Aqaba. Data from earlier 

catalogues have largely been superseded by more 

recent catalogue compilations. 

The city of Aqaba is situated at the northern end of 

Gulf of Aqaba along the southern part of the Dead 

Sea Transform fault system that separates the Sinai 

and Arabian tectonic plates. Furthermore, Aqaba lies 

Fig. 1: Regional setting o the Dead Sea fault system showing 

recent location of recent earthquake foci. Map after Hartman 

(2011). 

152


EARTHQUAKE 

ARCHAEOLOGY 



along the transition from the marine to the continental 

Eilat/Aqaba sedimentary basins. Faults controlling 

the structurally dynamics of sedimentation as well as 

the seismic activity lie both onshore and offshore. In 

order to understand the history of earthquakes in the 

region, an understanding of the seismogenic faults in 

both the marine and continental environments is thus 

essential. 

1.3 m across the older Qf1 and Qf2 surfaces. These 

data indicate that scarp heights reflect cumulative slip 

events. The most recent scarp-forming event fault 

occurred after A.D. 1045-1278 based on a corrected, 

calibrated radiocarbon age from charcoal collected 

from a buried campfire at the base of the scarp in 

Trench T-1. This likely represents fault motion in one 

of the historical earthquakes affecting southern 

Jordan (e.g. 1068, 1212, 1458, or 1588). 

ACTIVE FAULTING 

Offshore geophysical surveys have recently identified 

four submarine fault zones in the northern Gulf of 

Aqaba (Tibor et al., 2010). Two fault zones flank the 

margins of the gulf (the Eilat fault and Aqaba fault) 

and continue onland as faults that truncate the distal 

portions of alluvial fan systems. Both the onshore 

and offshore Eilat marginal faults have normal fault 

displacement (e.g., Ben-Avraham 1985; Bowman & 

Gerson, 1986). Along the eastern margin of the gulf 

lie the Aqaba and west Aqaba fault zones. The very 

steep bathymetric escarpment with granitic bedrock 

truncated at the eastern shoreline indicates that a 

significant amount of vertical offset is accommodated 

on faults of the Aqaba fault zone. The zone is wide 

with three offshore strands. Hartman (2011) mapped 

strands of a “West Aqaba” fault that may also 

accommodate strike-slip motion. In the middle of the 

basin, the Ayla fault zone that appears to bound 

subsidence across a localized basin was apparently 

active in the early Holocene. The main offshore 

strike-slip fault continues northward as the Evrona 

fault in Israel. This fault extends into Jordan where it 

crosses the Wadi Muhtadi alluvial fan and continues 

northward to the Dead Sea. The fault in Jordan is 

called the Wadi ‘Arabah fault. Faulting in the 1068 

earthquake has been documented on this fault in the 

Evrona sabkha (Amit et al., 1999, 2002; Zilberman et 

al. 2005). 

In Aqaba a concerted effort has been made to map 

the onshore continuation of the Aqaba fault zone. 

Due to the dense urbanization, mapping the 

northward continuation of faults has been difficult. Air 

photo interpretation of the Aqaba regional surficial 

geology suggests that the Aqaba fault emerges from 

the gulf and that slip is transferred to northwesttrending 

cross faults (Niemi and Smith 1999; Slater 

and Niemi 2003; Mansoor et al. 2004). Because the 

cross faults are linear and not offset, this geometry 

constrains the location of the Aqaba fault to lie south 

and/or east of the cross faults and at the toe of the 

eastern alluvial fan surfaces. 

Geological trenches (T-1 through T-5) were 

excavated across four NW-trending cross-faults (Fig. 

2) that produce active tectonic subsidence at the 

head of the Gulf (Mansoor, 2002; Slater and Niemi, 

2003). Mapping of alluvial fan and buried soil 

horizons in the trenches reveal multiple fault ruptures 

on the highest scarps and fewer distinct ruptures on 

the lowest scarp (Mansoor, 2002). The scarp heights 

range from 25 cm across the youngest Qf3 surface to 

153 

Fig. 2: Map of the city of Aqaba showing the location of 

major archaeological sites. Active cross faults (CF) mapped 

from aerial photos and in the archaeological excavations of 

J-East are also shown (Thomas et al., 2007). 

ARCHAEOSEISMOLOGY 

Aqaba has a rich cultural history. The earliest 

evidence of sedentary occupation found are the Late 

Chalcolithic sites of Tell Magass, and Tell Hujayrat 

al-Ghuzlan located 3-4 km northeast of the gulf 

(Khalil & Schmidt, 2009). An Iron through Hellenistic 

period (8 th - 4 th centuries B.C.) site called Tell el- 

Kheleifeh (Pratico, 1993) is located along the Jordan- 

Israel border. Along the coast and extending 

underneath the modern city, the remains of the early 

Roman-Byzantine (1 st century B.C.-7 th century A.D.), 

Early Islamic (7 th -12 th century), and Mamluk through 

Ottoman (13 th -early 20 th century) cities of Aila (or al 

‘Aqabah) have been partially excavated (e.g. 

Whitcomb, 1994; Parker, 2007; De Meulemeester & 

al-Shqour, 2008). 

Excavation of the Aila ruins from the 1 st to the 8 th 

centuries in Area J-East of the Roman Aqaba Project 

exposed a monumental mudbrick structure heavily 

damaged by successive earthquakes. Nine faults 

were mapped across the site (Thomas et al., 2007). 

Based on subsidence across the fault locations,


EARTHQUAKE 

ARCHAEOLOGY 



changes in floor elevations, and layers of collapsed 

mudbrick, the archaeological data suggest that the 

site was ruptured in an early 2nd century earthquake, 

an early 4th century earthquake, and the 363 A.D. 

earthquake. The monumental use of the structure 

was converted to domestic use in the late 4th to early 

5th century. 

We also have evidence for primary ground rupture for 

at least four post-date 363 A.D. earthquakes that 

fault the ruins in the J-East area of Aila. Primary fault 

rupture is documented in stratigraphic sections and 

plan maps of walls of various construction ages. Two 

earthquakes occurred during the Byzantine to 

Umayyad period (sixth to eighth centuries). There is 

a hiatus of deposition at this location between the 

Umayyad and the modern age. The two most recent 

earthquakes, with 42 and 35 cm of dip slip, occurred 

some time after the 8 th century and likely correlate to 

the historical earthquakes after the 11th century. No 

stratified materials were found at this site that could 

be used to further refine the timing of these seismic 

events. 

The early Islamic (8 th -11 th centuries A.D) site of Ayla 

was excavated in 1986-1995. Whitcomb 

hypothesized that the drainage running through the 

ancient site originated in erosion along the structural 

weakness of a fault and placed such a fault on the 

site plan map (Whitcomb, 1994). However, 

excavations by Rucker and Niemi (2005) of the NE 

corner tower of the walled city in the wadi and 

interpretation of 1918, 1945, and 1953 air photos 

indicate the wadi is man-made. There is evidence at 

Islamic Ayla for damage as a result of the 749 (or 

746 or 757? See Ambraseys 2005; 2009) earthquake 

followed by extensive reconstruction at the beginning 

of the Abbasid period. Major damage occurred in the 

town of Ayla in the March 18, 1068 earthquake. One 

contemporary source living in Baghdad, Ibn al- 

Banna, wrote “As for Aila, its inhabitants all perished 

except for 12 persons who had gone fishing at sea, 

thus escaping death.” (Guidoboni and Comastri, 

2005, p. 53). The site of Islamic Ayla was apparently 

never reoccupied to any significant degree after the 

earthquake. 

The site of Early Islamic Ayla was not rebuilt, but a 

new castle or caravan station was built about 1 km to 

the southeast. Excavations in and around the Aqaba 

castle from 2000-2008 have revealed three different 

phases in the “khan” or castle from the late 12 th to 

16 th centuries (De Meulemeester and Al-Shqour, 

2008). The extant castle was built in 1515 and rebuilt 

in 1587/8, probably after the Gulf of Aqaba 

earthquake of January 4, 1588 which, based on 

historical accounts, was felt in NW Arabia, Aqaba, 

and Sinai (Guidoboni and Comastri, 2005; 

Ambraseys, 2009). The archaeological data from the 

Aqaba castle (De Meulemeester and Al-Shqour, 

2008) also appear to support rupture of the Gulf of 

Aqaba fault segment in the earthquake of 1212 and 

possibly of the Wadi ‘Arabah fault segment in 1458. 

Our data suggest significant periods of active 

seismicity in the 4 th , 7 th -8 th , 11 th -13 th , and 15-16 th 

Centuries. These data suggest a recurrence rate of 

faulting and damaging earthquakes either from the 

Gulf of Aqaba segment or the Wadi ‘Arabah segment 

of the Dead Sea fault system of about every three to 

four centuries. It is interesting to note that 

earthquakes have been coincident with major political 

transitions that occurred in the region. Seismic 

activity is thus likely to have played a significant role 

in these cultural shifts. 

Acknowledgements: The Wadi ‘Arabah Earthquake 

Project has been supported by the Roman Aqaba Project, 

the American School of Oriental Research, a UMKC Faculty 

Research grant, the University of Missouri Research Board, 

and two grants from the Committee on Research and 

Exploration of the National Geographic Society. I am 

grateful for the assistance of Dr. Fawwaz al Khraysheh, 

former General Director, and Dr. Sawsan Fakhri, Aqaba 

Region director of the Department of Antiquities of Jordan, 

who granted permission to work in Aqaba. 

References 

Ambraseys, N.N. (2005). The seismic activity in Syria and 

Palestine during the middle of the 8 th 

century; an 

amalgamation of historical earthquakes. Journal of 

Seismology 9, 115-125. 

Ambraseys, N. (2009). Earthquakes in the Mediterranean 

and Middle East: A Multidisciplinary Study of Seismicity 

up to 1900. Cambridge University Press. Cambridge, 

U.K. 947 p. 

Amit, R., E. Zilberman, N. Porat, N. & Y. Enzel, (1999). 

Relief inversion in the Avrona playa as evidence of largemagnitude 

historical earthquakes, southern Arava valley, 

Dead Sea rift. Quaternary Research 52, 76-91 

Amit, R., E. Zilberman, Y. Enzel & N. Porat, (2002). 

Paleoseismic evidence for time dependency of seismic 

response on a fault system in the southern Arava Valley, 

Dead Sea rift, Israel. Geological Society of America 

Bulletin 114 (2), 192-206. 

Ben-Avraham, Z. (1985). Structural framework of the Gulf of 

Eilat (Aqaba), Northern Red Sea. Journal of Geophysical 

Research 90, 703-726. 

Bowman, D. & R. Gerson, (1986). Morphology of the Latest 

Quaternary Surface Faulting in the Gulf of Elat Region, 

Eastern Sinai. Tectonophysics 128, 97-119. 

De Meulemeester, J. & Al-Shqour, R. (2008). The Islamic 

Aqaba Project 2008: Preliminary Unpublished Report. 

Ghent University, Belgium, 48 p. 

Guidoboni, E. (1994). Catalogue of Ancient Earthquakes 

and Tsunamis in the Mediterranean Area from the 11 th to 

the 15 th Century. Istituto Nazionale di Geofisica. Roma, 

Italy. 504 p. 

Guidoboni, E., & A. Comastri, (2005). Catalogue of Ancient 

Earthquakes in the Mediterranean Area up to the 10 th 

Century. Istituto Nazionale di Geofisica. Roma, Italy. 

1037 p. 

Hartman, G. (2011). Quaternary Evolution of a Transform 

Basin: The Northern Gulf of Eilat/Aqaba. Dissertation. Tel 

Aviv University, Israel. 100 p. 

Hofstetter, R., Y. Gitterman, V. Pinksy, N. Kraeva, & L. 

Feldman, (2008). Seismological observations of the 

northern Dead Sea basin earthquake on 11 February 

2004 and its associated activity. Israel Journal of Earth 

Sciences 57 (2), 101-124. 

Khalil, L., & Schmidt, K. (2009). Prehistoric ‘Aqaba I. 

Orient-Archäologie Band 23, Rahden/Westf., Germany, 

Verlag Marie Leidorf. 419 p. 

154


EARTHQUAKE 

ARCHAEOLOGY 



SWATH BATHYMETRY AND MORPHOLOGICAL SLOPE ANALYSIS OF THE 

CORINTH GULF 

Nomikou, P. (1), Alexandri M. (2), Lykousis V. (2), Sakellariou D. (2), Ballas D. (2) 

(1) University of Athens, Department of Geology and Geoenvironment, Panepistimioupoli Zografou, 15784 Athens, Greece, 

evi@ath.hcmr.gr 

(2) Institute of Oceanography, Hellenic Centre for Marine Research, POBox 712, 19013 Anavyssos, Greece, 

matina@ath.hcmr.gr, vlikou@ath.hcmr.gr, sakell@ath.hcmr.gr, dballas@ath.hcmr.gr 

Abstract (Swath Bathymetry and Morphological Slope Analysis of Corinth Gulf): The swath bathymetric survey of the Gulf of 

Corinth has been conducted during multiple campaigns of HCMR’s vessel R/V "AEGAEO" between March 2001 and November 

2004, using the 20 kHz, SEABEAM 2120 system. The bathymetric map was produced using a 20 meter grid interval and plotted 

with a Mercator projection at 1:300.000 scale with 20 m contours. The major part of the central Gulf is a 800-870 m deep, 40km 

long, 9-12km wide, WNW-ESE elongate flat area. West of Aigio, the basin is shrinking towards its central run off axis, and finally 

diminishes at Rio-Antirrio straight. Eastwards the central basin terminates at the Alkyonides islets, east of which a sub-basin with 

maximum depth of 350 m is formed. The steep (30–40%) southern margin of the basin between Kiato to Aigio, is incised by 

numerous small canyons. The northern margin dips gently till the 400m isobaths and becomes steeper between 400-800m. The 

very detailed illustration of the Gulf’s bathymetry and morphology reflects the offshore active tectonics and faulting of the seafloor 

and the deformation during Quaternary. 

Key words: Swath Bathymetry, Sea Bed Morphology, Corinth Gulf 

INTRODUCTION 

The Corinth Rift, which separates the Peloponese from 

continental Greece, is a N100E oriented elongate 

graben, 105 km long, bounded by systems of very recent 

normal faults (less than 2 my old). The Corinth Gulf 

reflects the modification of the geomorphology and 

crustal structure formed during the alpine orogenesis by 

the extension tectonics and the resulted deformation of 

the Aegean region during Miocene-to-Quaternary. The rift 

transects obliquely the Pindos mountain chain resulted 

from the alpine nap tectonics. 

This structure is the most seismically active zone in 

Europe, and the fastest opening area of continental 

break-up, with up to 1.5 cm yr-1 of north-south extension 

rate (Clarke et al, 1998), and more than 1 mm yr-1 of 

uplift rate of the southern margin and 0.7-1 mm yr-1 

subsidence of the northern margin (Lykousis et al., 2009). 

The high rates of tectonic faulting and uplift lead to the 

outcropping of very recent fault planes with large offsets. 

SURVEY AND SYSTEM DEPLOYMENT 

The multibeam bathymetric surveys were carried 

out by the R/V AEGAEO of the Hellenic Centre 

for Marine Research, using a SEABEAM 2120 

swath system during 2001-2004 within the frame 

of various EU and National projects like ASSEM 

(Lykousis et. al., 2006) (Fig.1). The SEABEAM 

2120 is a hull-mounted swath system operating 

at 20 kHz in water depths not exceeding 6000 m. 

It has an angular coverage sector of 150 0 with 

149 beams, covering a swath width from 7.5 to 

11.5 times the water depth for depths from 20 m 

to 5 km. The maximum swath coverage can 

reach 9 km at maximum depth and gives 

satisfactory data quality at speeds up to 11 

knots. During some cruises we perform high 

resolution 3.5kHz and Air Gun profiling, side 

scan sonar imaging, gravity and box coring, 

deployment of a current-meter and CTD 

measurements. 

Offshore active faulting on the seafloor of the Gulf of 

Corinth has been studied in detail during dozens of 

offshore campaigns led by Greek and international 

research teams (Sakellariou et al, 2001; 2004; 2007; 

Stefatos et al, 2002; Moreti et al, 2004; Zelt et al, 2004; 

McNeill et al, 2005; Lykousis et al, 2006; 2008; 2007a,b; 

2009; Bell et al, 2008; Taylor et al, 2011). 

The tectonic graben of the Corinth Gulf has been filled up 

by gravitational deposits (turbidity flows, debris flows, 

mudflows etc.), while fun delta prograding deposits during 

Pleistocene sea-level changes, coupled by extensive 

slumping are the predominant sedimentary processes in 

the steep flanks of the gulf (Heezen et al. 1966; 

Ferentinos et al. 1988; Papatheodorou & Ferentinos 

1997; Lykousis et al. 1998; Perisoratis et al. 2000; 

Lykousis et al. 2007a). 

155 

Fig.1: Bathymetric Track lines during 2001-2004 

oceanographic cruises with R/V AEGAEO.


EARTHQUAKE 

ARCHAEOLOGY 



Fig.2: Multibeam Bathymetric Map of Corinth Gulf using 20m isobaths 

SWATH BATHYMETRY 

The multibeam data have been extensively processed by 

means of data editing, cleaning of erroneous beams, 

filtering of noise, processing of navigation data and 

interpolation of missing beams. The resulting bathymetric 

map of Corinth Gulf was originally compiled at 1/300.000 

scale, which was greatly reduced for publication with 9 

different colors corresponding to 100 m depth intervals 

and with additional isobaths of 20 m (Fig.2). This map 

permits the first detailed description of the overall 

topography of the sea floor as well as the mapping of the 

major morphotectonic structures within this area. 

shows that the distribution of slope values within 

the studied area can be subdivided into five 

categories: (1) areas of mean morphological 

slope 0-5%, (2) areas of 5-15%, (3) areas of 15- 

25%, (4) areas of 15-25% and (5) areas of 35- 

50%. 

Most of the central part of the Gulf from 800 up to 870 m 

depth is a very wide flat area forming an extensive WNW- 

ESE elongated basin of 40km length and a width of 9 km 

at the west up to 12 km at the east (Alexandri et al., 

2003) (Fig. 2). 

West of Aigio, the basin is shrinking towards its central 

run off axis, and finally diminishes at Rio-Antirio straight. 

Eastwards the central basin is reaching the Alkyonides 

islets where a sub-basin with maximum depth of 350 m is 

formed. The steep southern margin, that reaches the 

800m isobath is scored, from Kiato to Aigio by numerous 

small canyons trending NE-SW transversal to the main 

direction of the gulf (Fig. 3). In contrast, the northern 

margin spans up to 400m depth where abruptly drops to 

800 m creating the northern border of the central basin. 

Fig. 3: 3D view of Corinth Gulf combined offshore and 

onshore data. 

MORPHOLOGICAL ANALYSIS 

The bathymetric map of the area was analyzed from the 

perspective of slope distribution and the results are 

presented as a slope distribution map (Fig.5). This map 

Fig. 4: 3D view of fan delta failures (debris flows) along 

the southern flanks of Western Gulf of Corinth (imaged 

from ENE) (Lykousis et al., 2009). 

156


EARTHQUAKE 

ARCHAEOLOGY 



Fig.5: Slope distribution Map of Corinth Gulf 

Fig.6: Detailed swath bathymetry map of the Western Corinth Gulf using 10m isobaths 

This classification of the slope magnitudes clearly 

illustrates the zones where there is an abrupt change of 

slope, which usually reflects the position of active tectonic 

structures or steep slopes. 

The highest slope values (35-50%) are observed in the 

southern margin from Kiato to Aigio due to numerous 

small canyons. The northern margin of the gulf if 

characterized mostly by morphological slopes up to 25% 

showing a smooth transition to the basin. On the contrary, 

the central part of the gulf is characterized by a 

flat bottom of the basin with low morphological 

slopes from 0-5%. 

WEST CORINTH GULF 

The western Gulf of Corinth is the currently most 

active part of the Gulf and has been the site of 

the Aegion M:6 earthquake in 1995 and many 

other strong tremors during the last century. 

157


EARTHQUAKE 

ARCHAEOLOGY 



The seabed morphology of the western Gulf of Corinth is 

dominated by slope canyon systems feeding an axial 

channel which is only present within this part of the gulf 

(Fig. 6). Canyons prevail the southern margin but are rare 

on the northern margin of the gulf. The basin floor 

reaches a maximum depth of 800m towards the east. 

Slope failures are abundant in the southern margin and 

are related to prodelta sediment instabilities and to a 

minor degree in the northern margin (Eratini area, etc.). 

couple of sites were related to historical tsunarnis and are 

expected to be the potential tsunarnigenic sites the W. 

Corinth Gulf in the future (Lykousis et al., 2007a; 

Lykousis et al., 2009). Infinite slope stability analysis 

indicated that slopes steeper than 2-3 degrees are 

capable to induce slope failures taken account the 

expected (seismic) ground accelerations for the next 50 

years (Lykousis et al., 2009). The northern margin is 

dominated by a basement horst uplifted by the North (N) 

and South (S) Eratini faults (Stefatos et al., 2002). The 

horst is most prominent near the Psaromita Peninsula 

and is progressively buried and reduced in topography to 

the east and west, with faults as long as 15 km (McNeil et 

al., 2005). 

DISCUSSION 

The swath bathymetric survey of the Gulf of Corinth 

provides insight into morphological expression of the 

active seafloor processes, like faulting and submarine 

sliding, erosions and sediment deposition and enables 

detailed mapping of the offshore structural elements. 

The major part of the central Gulf is a 800-870 m deep, 

40km long, 9-12km wide, WNW-ESE elongate flat area. 

West of Aigio, the basin is shrinking towards its central 

run off axis, and finally diminishes at Rio-Antirio straight. 

Eastwards, the central basin terminates at the Alkyonides 

islets, east of which a sub-basin with maximum depth of 

350 m is formed. Both, the steep (30–40%) southern 

margin and the gentler dipping north margin of the basin 

are fault controlled. The steep margins are incised by 

numerous small canyons. 

References 

Alexandri M., Nomikou P., Ballas D., Lykousis V. & Sakellariou 

D. (2003): Swath bathymetry map of Gulf of Corinth. Geoph. 

Res. Abstracts, Vol. 5, 14268, EGS 2003. 

Bell, R., McNeill, L.C., Bull, J.M., and Henstock, T.J., 2008. 

Evolution of the western Gulf of Corinth continental rift, 

Greece. Geological Society of America Bulletin, 120, 156-178. 

Clarke, P.J., Davies, R.R., England, P.C., Parson, B., Billiris, H., 

Paradissis, D., Veis, G., Cross, P.A., Denys, P.H., Ashkenazi, 

V., Bingley, R., Kahle, H.-G., Muller, M.-V., and Briole, P., 

1998, Crustal strain in central Greece from repeated GPS 

measurements in the interval 1989-1997: Geophysical Journal 

International, 135, 195-214. 

Ferentinos, G., Papatheodorou, G. & Collins,M.B., 1988. 

Sediment transport processes on an active submarine fault 

escarpment: Gulf of Corinth, Greece, Mar. Geol., 83, 43–61. 

Heezen, B.C., Ewing, M. & Johnson, G.L., 1966. The Gulf of 

Corinth floor, Deep-Sea Res., 13, 381–411. 

Lykousis V., Sakellariou D., Blandin J., Person R., Etiope G., 

Alexandri M., Nomikou P., Rousakis G., 2006, The ASSEM 

sea bed observatory for long term multihazard monitoring: Gulf 

of Corinth experiment. 8 . . . & , 

. 151-154. 

158 

Lykousis, V., Sakellariou, D., Papanikolaou, D., 1998. 

Sequence stratigraphy in the northern margin of the 

Gulf of Corinth: implications to upper Quaternary 

basin evolution, Bull. Geol. Soc. Greece, 32, 157– 

164. 

Lykousis V., Sakellariou D., Rousakis G., Alexandri S., 

Kaberi H., Nomikou P., Georgiou P., Balas D. 

,2007a. Sediment failure processes in active 

grabens: the Western Gulf of Corinth (Greece). In: 

Lykousis V., Sakellariou D., Locat J. (eds.), 

Submarine Mass Movements and their 

Consequences, 297-305, Springer. 

Lykousis V., Sakellariou D., Moretti, I., Kaberi H. 

,2007b. Late Quaternary basin evolution of the Gulf 

of Corinth: Sequence stratigraphy, sedimentation, 

fault–slip and subsidence rates. Tectoniphysics, 440, 

29-51. 

Lykousis, V., G. Roussakis , D. Sakellariou, 2009. 

Slope failures and stability analysis of shallow water 

prodeltas in the active margins of Western Greece, 

northeastern Mediterranean Sea. J. Earth Sciences, 

98, 807–822. 

McNeill, L., Cotterill, C., Stefatos, A., Henstock, T., 

Bull, J., Collier, R., Papatheoderou, G., Ferentinos 

G., and Hicks, S. 2005, Active faulting within the 

offshore western Gulf of Corinth, Greece: 

implications for models of continental rift deformation: 

Geology, 33, 241-244. 

Moretti, I., Lykousis, V., Sakellariou, D., Reynaud, J.- 

Y., Benziane, B., and Prinzhoffer, A., 2004, 

Sedimentation and subsidence rate in the Gulf of 

Corinth: what we learn from Marion Dufresne’s longpiston 

coring: Comptes Rendus Geoscience, 336, 

291-299. 

Papatheodorou, G. & Ferentinos, G., 1997. Submarine 

and coastal sediment failure triggered by the 1995, 

Ms = 6.1 Aegion earthquake, Gulf of Corinth, 

Greece, Mar. Geol., 137, 287–304. 

Perissoratis C., Piper D.J.W. and Lykousis V., 2000. 

Alternating marine and lacustrine sedimentation 

during late Quaternary in the Gulf of Corinth rift 

basin, central Greece. Mar. Geol. 167: 391-411. 

Sakellariou, D., Lykousis,V. & Papanikolaou, D. (2001) 

Active faulting in the Gulf of Corinth, Greece. In: 

36th CIESM Congress Proceedings, 36 pp. 43. 

Sakellariou, D.,Kaberi, H. & Lykousis,V. (2004) 

Infuence of active tectonics on the recent 

sedimentation of the Gulf of Corinth basin. In: 10th 

International Congress of Greek Geological Society, 

Abstracts 15-17 April, p. 232-233, Thessaloniki. 

Sakellariou D., Lykousis V., Alexandri S., Kaberi H., 

Rousakis G., Nomikou P., Georgiou, P., Ballas D., 

(2007): Faulting, seismic-stratigraphic architecture 

and Late Quaternary evolution of the Gulf of 

Alkyonides basin – East Gulf of Corinth, Central 

Greece. Basin Research, 19/2, p. 273-295. 

Stefatos, A., Papatheoderou, G., Ferentinos, G., 

Leeder, M., and Collier, R., 2002, Active offshore 

faults in the Gulf of Corinth, Greece: Their 

seismotectonic significance: Basin Research, 14, 

487-502. 

Taylor, B., Weiss, J.R., Goodliffe, A. M., Sachpazi, M., 

Laigle, M. & Hirn, A., 2011. The structures, 

stratigraphy and evolution of the Gulf of Corinth rift, 

Greece Geophys. J. Int. (2011) 185, 1189–1219 

Zelt, B.C., Taylor, B., Weiss, J.R., Goodliffe, A.M., 

Sachpazi, M., and Hirn, A., 2004. Streamer 

tomography velocity models for the Gulf of Corinth 

and Gulf of Itea, Greece. Geophys. J. Int., 159, 333- 

346.


EARTHQUAKE 

ARCHAEOLOGY 



IS THE DECADENCE OF LEPTIS MAGNA (LYBIA) THE CONSEQUENCE OF A 

DESTRUCTIVE EARTHQUAKE? 

Pantosti, Daniela (1), Stefano Pucci (1), Paolo Marco De Martini (1), Alessandra Smedile (1 

(1) Istituto Nazionale di Geofisica e Vulcanologia. Via di Vigna Murata, 605. 00143- Roma. ITALY. 

Email: daniela.pantosti@ingv.it, stefano.pucci@ingv.it, paolomarco.demartini@ingv.it, alessandra.smedile@ingv.it 

Abstract (Is the decadence of Leptis Magna (Lybia) the consequence of a destructive earthquake?): The relationships 

between human modification of the environment and natural events in the Roman city of Leptis Magna are analyzed. Historical 

and archaeological sources indicate extreme natural events as the cause of the town’s decline: earthquakes, flooding, and 

tsunamis. Stratigraphical and geomorphological surveys investigated history and dynamics of the depositional and erosional 

systems of the settlement area by integrating independent constraints as archaeological and absolute radiocarbon dating. The 

results highlighted that once the Romans society could no longer guarantee the maintenance of defensive structures, destructive 

floods affected the town. Conversely, large earthquakes or tsunami have been discarded as primary cause of the town decline. 

Key words: Natural Hazards, Flooding, Earthquakes, AD 365 Crete earthquake and tsunami 

The Roman city of Leptis Magna (UNESCO world 

heritage) in western Libya was a magnificent, 

flourishing, strategic town in Tripolitania reaching its 

maximum expansion during the Empire of Septimius 

Severus (193-211 A.D.). The town is built on a wide 

alluvial fan fed by wadi (creek in Arabic) Lebda 

whose outlet served as a harbor since Phoenician 

times (Figure 1). An ingenious hydraulic system 

composed by a dam and an artificial channel 

diverting the wadi water was built by the Romans to 

prevent the town from flooding. The decline of Leptis 

Magna irreversibly started in the 4th Century. 

Historical and archaeological sources suggest that 

the town decadence is related to the fact that the 

harbor became inefficient and was abandoned due to 

its complete sediment infill. Several causes are taken 

in consideration to explain this infilling, these are: i) 

violent flooding following the collapse of a dam built 

to regulate the course of the wadi because of the 

large 365 A.D. Crete earthquake (Salza Prina Ricotti, 

1995; Di Vita, 1990); ii) lack of maintenance due to 

the decline of the settlement induced by severe 

damage after the 365 A.D. earthquake (Di Vita, 1990 

and 1995), or other local seismic sources (Guidoboni 

et al., 1994; Stiros, 2001); iii) inundation of a tsunami 

wave caused by the 365 A.D. earthquake that left a 

huge amount of debris and modified the local coastal 

morphology (Guidoboni et al., 1994; Ambraseys et 

al., 1994; Lorito et al., 2007, Shaw et al., 2008); iv) 

bad orientation and geometry of the harbor structures 

with respect to the local marine currents that were 

bringing sediments inside (Salza Prina Ricotti, 1972- 

1973). Most of these hypotheses are connected to 

the M8.5, 365 A.D. Crete event that, being the largest 

historical earthquake and tsunami ever occurred in 

the Mediterranean, has certainly stimulated the 

formulation of many hypothesis that put in connection 

any critical situation recognized in the mid of the 4th 

Century. with the occurrence of this catastrophe. 

159 

We studied in detail the geomorphology and the 

stratigraphy of Leptis Magna to understand if the 

causes of the decline of the town can be deciphered in 

the natural events recorded in the local geology (Pucci 

et al., 2011). To build up a consistent chronological 

record of significant natural events and their interaction 

with human history, geomorphological analyses (the 

study of cores and natural/artificial exposures), and the 

reconstruction of the different depositional 

environments based on the palaeontological content, 

were integrated to radiometric dating and 

archaeological knowledge. 

Given the strategic role of the harbor in the 

flourishing/decline of the town, this is certainly a key 

site to study to understand the history and men-nature 

interactions but, to fully understand the functioning of 

the natural and human system and equilibrium besides 

the harbor, we investigated the wadi, along with its 

hydraulic system built to control flooding (dam and 

aegere), the town, and the surrounding coastal area. 

At the same time we have surveyed the buildings in 

town to check for evidence of seismically induced 

collapses. 

The integration of all these observations allowed us to 

reconstruct the following history for the site (Pucci et 

al., 2011 - Figure 2): 

1) Building of hydraulic protections (dam and a 

channel diverting the wadi along a defensive aegere) 

to prevent flooding and have at disposal a water 

reserve, 32-131 A.D. No deposition in the lower reach 

of wadi (north of the dam), in town and in the harbor (B 

in figure 2). 

2) Functioning of the hydraulic system, expansion of 

the town toward the lower reach into the dry wadi bed 

and on its lower terraces. Enlargement of the harbor 

into the Severian monumental structure allowed by the 

lack of sediment/water discharge at the wadi outlet. 

Efficiency at least up to the time of Septimius Severus


EARTHQUAKE 

ARCHAEOLOGY 



(193-211 A.D.). Deposition only in the upper reach of 

the wadi (south of the dam), siltation effects cleared 

by artificial sweeping of the dam basin (C in figure 2). 

3) Maintenance (sweeping) of the hydraulic 

protections start to fail during the 3rd Century; onset 

of the dam siltation and beginning of the decadence 

of the town. In about 100 yrs the reservoir is 

completely filled; the wadi starts to spill out from the 

dam with the first alluvial deposits passing again the 

lower wadi reach and harbor area and with beginning 

of erosion at the base of the right shoulder of the dam 

(D in figure 2). 

4) Collapse of the right shoulder of the dam, 

entrenchment of the siltation body, restoration of the 

natural equilibrium of the wadi longitudinal profile by 

remobilization of a large amount of deposits that were 

previously stored in the dam basin. At 320-440 A.D. 

the wadi floods again the alluvial plain, begins to bury 

the town with alluvial sediments and to discharge large 

amounts of sediments into the harbor (E in figure 2). 

Figure 1. Geological and geomorphological map of the area of Leptis Magna. Locations S1-S7 are those studied in detail through 

stratigraphic and laboratory analyses. Inset in the lower left locates the study area on the north African coast (modified from Pucci 

et al., 2011). 

160


EARTHQUAKE 

ARCHAEOLOGY 



5) Harbor is completely infilled by wadi sediments at 

the 6th Century (F in figure 2). 

6) No evidence for important seismic collapses or 

damage was found by inspecting the buildings, as 

they appear today and the dam. Unfortunately, there 

was no possibility to check the original documents 

related to the early archaeological excavations to 

confirm the lack of major shaking effects. A wall in a 

recent excavation (Prof. F. Tommasello, pers. 

comm.), collapsed on top of 1 m-thick alluvium 

burying already the town, may be evidence for an 

earthquake. If seismic, this collapse occurred well 

after that repeated flooding draped the town with a 

thick layer of alluvium and thus post-dates 320-440 

A.D. 

Figure 2. Conceptual sketch of the evolution of the wadi Lebda longitudinal profile, from the dam to the harbor, and its depositional 

history in relation with the different studied locations (upper wadi, lower wadi, harbor, town - modified from Pucci et al., 2011). 

161


EARTHQUAKE 

ARCHAEOLOGY 



The data collected, although not definitive to 

characterize the true cause for the decline of Leptis 

Magna, allow evaluating the initial hypotheses (Pucci 

et al., 2011). 

1) Flooding due to the collapse of the dam because 

of the 365 A.D. earthquake is discarded. This is 

because no traces for earthquake effects were found 

in the existing structures (including the dam), but also 

because no water was present in the dam basin at 

the time of the dam shoulder collapse. In fact, as it is 

today, the basin should have been completely filled 

with sediments due to the siltation process incepted 

by lack of dam sweeping. Should the dam shoulder 

have collapsed when the basin was still containing 

water (to produce the violent flooding) no siltation 

would have ever occurred as the dam was no more 

efficient and the wadi would have re-gained its 

course and profile. 

2) Lack of maintenance due to severe damage to 

the town after the 365 A.D. This hypothesis is partly 

correct. In fact, lack of maintenance is the primary 

cause for siltation in the dam and for all the 

consequent effects (flooding in town and sediments 

infilling in the harbor) that made the initial 

(economic?) decline irreversible. However, the 

reason for this lack of maintenance does not seem 

related to the damage from an earthquake (and 

certainly not the 365 A.D. Crete event) as no serious 

evidence for seismically induced damage was found. 

3) 365 A.D. tsunami harbor inundation is discarded 

as well. Coring in the harbor we reconstructed the 

whole history of the sedimentation in that area, from 

marine to continental. No evidence for such a violent 

event was found. A possible layer representing the 

deposit from a paleotsunami was found in the upper 

part of harbor stratigraphy, when alluvium was 

already deposited in the harbor after the 6th Century. 

4) Bad orientation and geometry of the harbor as the 

cause of debris deposition from marine currents is 

clearly discarded because the reconstruction of the 

depositional environments in the harbor has shown 

initial marine conditions (up to the 3rd Century) that 

gradually had a continental input to become purely 

alluvial (5th Century). If the initial hypothesis was 

correct, the harbor infill would have only a marine 

origin. 

In conclusion, we can rule out the hypothesis that an 

earthquake, and particularly the Crete 365 A.D., was 

the cause for the decadence of Leptis Magna. On the 

contrary, from the analyses of the recent geological 

history of the site it is likely that social and economic 

criticalities (possibly related to the continuous 

robberies and attacks from the Austurians) are the 

primary cause for this decadence that made the 

Lepticians unable to cope with natural hazards and 

preserve their “risk management plans”. The fact that 

nature took up the place again, causing widespread 

flooding and huge sediments remobilization made it 

too difficult for the Lepticians to recover and the initial 

decline of the town became irreversible. 

Acknowledgements: This work was funded by Istituto 

Nazionale di Geofisica e Geofisica e Vulcanologia — INGV 

with contribution of the EU Transfer project. We are thankful 

to the Italian Archaeological Mission leaded by L. Musso for 

hosting us and for the fruitful discussions and collaboration. 

References 

Ambraseys, N.N., Melville, C.P., Adams, R.D., (1994). The 

Seismicity of Egypt, Arabia and the Red Sea: a 

Historical Review. Cambridge Univ. Press, Cambridge. 

Di Vita, A., (1990). Sismi, urbanistica e cronologia assoluta. 

Terremoti e urbanistica nella città di Tripolitania tra il I 

secolo a.C. ed il IV d.C. In: L’Afrique dans L’Occident 

Romani (Ier Siècle Av. J.-C.- IVe Siècle Ap. J.- C.), 

Actes du Colloque Organisé par l’Èrcole Français de 

Rome sous le Patronage de l’Institut National 

d’Archéologie et d’Art de Tunis (Rome, 3-5 décembre 

1987), Roma, pp. 425e494. 

Di Vita, A., (1995). Archaeologists and earthquakes: the 

case of 365 AD. Annali di Geofisica 38, 971-976. 

Guidoboni, E., Comastri, A., Traina, G., (1994). Catalogue 

of Ancient Earthquakes in the Mediterranean Area up to 

the 10th Century. ING SGA, Rome. 

Lorito, S., Tiberti, M.M., Basili, R., Piatanesi, A., Valensise, 

G., (2007). Earthquake generated tsunamis in the 

Mediterranean Sea: scenarios of potential threats to 

Southern Italy. Journal of Geophysical Research, 

doi:10.1029/2007JB004943. 

Pucci, S., D. Pantosti, P.M. De Martini, A. Smedile, M. 

Munzi, E. Cirelli, M. Pentiricci, L. Musso, (2011). 

Environment-man relationships in historical times: The 

balance between urban development and natural forces 

at Leptis Magna (Libya), Quaternary International, 

doi:10.1016/j.quaint.2011.03.050 

Salza Prina Ricotti, E., (1972-1973). I porti della zona di 

Leptis Magna. Rendiconti della Pontificia Accademia di 

Archeologia 45, 75-103. 

Salza Prina Ricotti, E., (1995). Leptis Magna la città delle 

ombre bianche. Archeo 9 (127), 50-91. 

Shaw, B., Ambraseys, N.N., England, P.C., Floyd, M.A., 

Gorman, G.J., Higham, T.F.G., Jackson, J.A., Nocquet, 

J.M., Pain, C.C., Piggott, M.D., (2008). Eastern 

Mediterranean tectonics and tsunami hazard inferred 

from the AD 365 earthquake.Nature Geoscience 1, 268- 

276. doi:10.1038/ngeo151. 

Stiros, S.C., (2001). The AD 365 Crete earthquake and 

possible seismic clusteringduring the fourth to sixth 

centuries AD in the Eastern Mediterranean: a reviewof 

historical and archaeological data. Journal of Structural 

Geology 23, 545-562. 

162


EARTHQUAKE 

ARCHAEOLOGY 



ON-SHORE PROLONGATION OF BATHYMETRICALLY RECOGNIZED FAULT ZONES 

BASED ON GEODETIC GPS OBSERVATIONS ALONG SANTORINI VOLCANO 

Papageorgiou, Elena (1, Nomikou, Paraskevi (2) 

(1) University of Athens, Department of Geophysics and Geothermy, Zografou 15784, Greece, Email: epapageo@geol.uoa.gr 

(2) University of Athens, Department of Dynamic, Tectonic and Applied Geology, Zografou 15784, Greece, 

Email: evi@ath.hcmr.gr 

Abstract (On-shore prolongation of bathymetrically recognized fault zones based on geodetic GPS observations along 

Santorini volcano): Ground deformation monitoring on Santorini Volcanic Complex (SVC) revealed different horizontal kinematics 

in several parts of the volcano. However, in the current reposed state of SVC, ground deformation is primarily the result of tectonic 

activity. This work aims to combine and interpret the results of both GPS measurements and bathymetric survey conducted at 

SVC. The GPS data were intermittently collected during numerous campaigns between 1994 and 2005, while bathymetric survey 

was completed by the Hellenic vessel R/V ‘AEGAEO’ in November 2001, using the 20 kHz, SEABEAM 2120 swath system and 

mapping a total area of 2480 km². In this attempt a correlation between the morpho-tectonic features exposed on the seabed by 

the bathymetric survey, and the surface displacements observed by the GPS measurements could be finally achieved. An 

interpretation will be held, on whether these submarine tectonic features continue and intersect the volcanic island, especially in 

the areas where the surface displacements enable the plausible existence of contemporary tectonics. 

Key words: Tectonics, Bathymetry data, GPS Displacements, Santorini Volcano 

INTRODUCTION 

Santorini Volcanic Complex (SVC) lies in an area of 

complex extensional and subduction-related 

tectonics in a continental environment (Le Pichon & 

Angelier, 1979; Fytikas et al., 1984; Papanikolaou, 

1993; Jackson, 1994). The development of the 

volcanic field has been strongly influenced by 

regional faults.The current geodynamic state of SVC 

is controlled by two major volcanoes, the Nea 

Kammeni Volcano built up in the central caldera and 

the submarine Columbo Volcano, NE of Cape 

Columbo. Two major faulting zones Kammeni and 

Columbo (KFZ, CFZ, respectively) cross the 

volcanoes in a NE-SW direction, for several km, 

forming vents through which the magma ascend to 

the surface (Fouqué 1879; Reck 1936). Generally, 

the major structural features in Santorini are striking 

in a NE direction (Heiken & McCoy 1984; Druitt et al., 

1989, 1999; Jackson 1994; Mountrakis et al., 1998; 

Pe-Piper & Piper, 2005). 

Seismic data (Bohnhoff et al., 2006; Dimitriadis et al. 

2009) indicate that the main volcanic centers of the 

Santorini volcanic field (Christiana, Santorini, 

Columbo and minor volcanic cones) are aligned 

along the NE-SW trending Santorini-Columbo 

volcano-tectonic line, a deep-seated, strike-slip 

feature (Sakellariou et al., 2010; Nomikou et al., 

2011a) which served as conduit for the rising 

magma. 

Herein, we present the joint data interpretation of 

GPS measurements and bathymetry survey of the 

surrounding area of Santorini, in an attempt to relate 

the submarine rupture system, with its onshore 

163 

continuation as deduced from surface displacement 

variations on either side of the possible prolonged 

faults. 

GPS OBSERVATIONS 

DGPS has proven to be an important tool for 

intermediate and long-term monitoring of active 

volcanoes, as it provides high resolution 3-D 

information about the deformation field (e.g. Dvorak 

& Dzurisin, 1997; Fernandez et al., 1999 and many 

others). At the SVC, DGPS measurements 

conducted during the period 1994-2005 have 

detected non-trivial amounts of ground displacement, 

in both the horizontal and vertical axes (e.g. 

Papageorgiou et al., 2007). This enabled the 

identification of several domains with different 

kinematic characteristics. The interested reader may 

find details about the field measurement and data 

reduction procedures in Papageorgiou et al. (2007, 

2010). 

In particular, the GPS analysis presents several 

distinct domains with different horizontal kinematics, 

which could be defined most possibly by the major 

tectonic fault zones of the area. Thus, areas with 

uniform GPS displacements both in rate and 

orientation that characterize homogeneous regions 

were separated among others with considerable 

changes in their kinematic pattern. The observed 

differential deformation between these deforming 

domains is thought primarily to be the result of 

differential motion across faulting structures located 

at the boundaries. For the purpose of this paper, a 

synthetic map was constructed showing the above 

mentioned homogeneous regions, based merely on


EARTHQUAKE 

ARCHAEOLOGY 



their kinematics, as well as the areas/zones which 

may correspond to the major regional faults, as they 

still affect the tectonic–volcanic evolution of the SVC, 

as has already been observed in the past (Fig. 1). 

Fig. 1: Horizontal displacements relative to MTPI (NE Thera) for the period 1994-2005 (Papageorgiou et al., 2010). 

In Northern Thera, in between the GPS stations 33 

and 43 a different kinematic pattern is observed, 

indicating the possible existence of an area/zone 

(case I) of kinematic change (Fig. 1). Similarly, in 

central caldera, at Nea Kammeni Volcano, a second 

kinematic differentiation is observed, between the 

northern GPS stations 15 and 22 and the central 

GPS station 45 (case II). In both cases (I and II), fault 

zones are implicated to consist the boundary of the 

different homogeneous deforming domains, where in 

either side of which the horizontal GPS 

displacements change both in rate and orientation 

(Fig. 1). 

This is a very complex pattern and certainly difficult 

to interpret. Nevertheless, the observed deformation 

pattern enables the drafting of a qualitative model of 

contemporary tectonics, which is also presented and 

discussed in detail in Papageorgiou et al. (2010). 

However, the most important differences of surface 

displacements which are observed for the time 

interval 1994-2005 are thought to be explained with 

the existence of tectonic features. As it appears, 

these features coincide with the major fault zones in 

the area, the Kammeni, and Columbo Fault Zones 

(Fig. 2). 

BATHYMETRIC SURVEY 

The multibeam bathymetric survey was completed by 

the Hellenic vessel R/V ‘AEGAEO’ in November 

2001, using the 20 kHz, SEABEAM 2120 swath 

system and mapping a total area of 2480 km². The 

164 

bathymetric map was produced using a 50 meter grid 

interval and plotted with a Mercator projection at a 

scale of 1:100.000 with 10m contours (Alexandri et 

al., 2003; Nomikou et al., 2011a). What comes out 

from the bathymetric map (Fig. 2) is the identification 

of several morphological discontinuities that finally 

correspond to fault zones as indicated from seismic 

profiling data (Fig. 3) (Sakellariou et al., 2010; 

Nomikou et al., 2011). The volcanic cones in the 

regional area of the SVC presented in Fig. 3 show 

linear distribution which is controlled by strike-slip 

faults that run parallel to the long axis of the 

Anydhros basin. 

Furthermore, the CFZ is an active, possibly rightlateral, 

40 km long, strike-slip fault-zone, which has 

enabled the upward migration of magmatic fluids, the 

creation of dikes and the development of submarine 

volcanic cones, with Columbo submarine volcano as 

the most productive among them (Sakellariou et al., 

2010). The KFZ in addition, passing through Nea 

Kammeni volcano, in the centre of the caldera is 

directly related to CFZ and displays similar 

characteristics. 

A particularly noteworthy feature that should be 

mentioned is a hydrothermal vent field at the northern 

caldera that is located in line with the normal fault 

system of the Columbo rift, and also near the margin 

of a shallow intrusion that occurs within the 

sediments of the north caldera (Sigurdsoon et al., 

2006; Nomikou et al., 2011).


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 2: Multibeam Bathymetric Map of Santorini Volcanic Field using 20m isobaths (Nomikou et al., 2011). 

Fig. 3: Tectonic map of Santorini-Columbo volcanic field (Sakellariou et al., 2010). 

165


EARTHQUAKE 

ARCHAEOLOGY 



DISCUSSION 

The results of GPS measurements agree with the 

swath bathymetry data, as the observed submarine 

tectonic fault zones seem to prolongate onshore 

crosscutting the whole of the volcanic edifice. This 

phenomenon is more obvious in Northern Thera with 

the continuation of the Columbus Fault Zone up to 

the northern part of the Santorini caldera. The GPS 

horizontal displacements vectors show evidently 

different kinematic pattern on both sides of the CFZ, 

agreeing for the distinct existence of the CFZ and its 

continuation on land. However, a consideration 

should be taken into account on whether the GPS 

data reflect the existence of two different tectonic 

blocks on both sides of the CFZ, or whether it may 

indicate an updoming due to ascending magma 

along the fault zone which at the same time is used 

by the volcanic dikes. 

This is also evident in other parts of the SVC. The 

displacement field in Nea Kammeni volcano, is 

differentiating in each side of the Kammeni Fault 

Zone (KFZ), indicating primarily two kinematic blocks 

bounded by the KFZ, and secondly its onshore 

prolongation at Nea Kammeni. 

However, the GPS results show a more complex 

kinematic model of the SVC and further numerical 

modelling should be considered in order to get a 

more representative tectonic regime. Yet, this study 

enabled the combination of two different methods, as 

is the GPS observations and bathymetric data, in 

order to successfully certify the existence of major 

fault zones on the surface. 

Acknowledgements: Part of the GPS raw data was given 

by Prof. E. Lagios. Part of the collection and analysis of 

GPS data was financed by (i) The European Union (75%), 

(ii) The General Secretariat for Research & Technology 

(25%), and (iii) Terramentor E.E.I.G. Funding for offshore 

work at Santorini Volcanic Field was provided by 

GEOWARN project. The European Commission is 

acknowledged for their financial contribution to the project. 

The officers and the crew of the R/V AEGAEO are gratefully 

acknowledged for their important and effective contribution 

to the field work. 

References 

Alexandri, M., D. Papanikolaou & P. Nomikou (2003). 

Santorini volcanic field - new insights based on swath 

bathymetry. Abstracts IUGG, 30 June-11July 2003, 

Sapporo, Japan. 

Bohnhoff, M., M. Rische, Th. Meier, D. Becker, G. 

Stavrakakis & H-P. Harjes (2006). Microseismic activity in 

the Hellenic Volcanic Arc, Greece, with emphasis on the 

seismotectonic setting of the Santorini–Amorgos zone. 

Tectonophysics, 423, 17–33. 

Dimitriadis, I., E. Karagianni, D. Panagiotopoulos, C. 

Papazachos, P. Hatzidimitriou, M. Bohnhoff, M. Rische, & 

T. Meier (2009). Seismicity and active tectonics at 

Coloumbo Reef (Aegean Sea, Greece): Monitoring an 

active volcano at Santorini Volcanic Center using a 

temporary seismic network. Tectonophysics, 465, 136- 

149. 

Druitt, T.H., R.A. Mellors, D.M. Pyle, & R.S.J. Sparks 

(1989). Explosive volcanism on Santorini, Greece. Geol. 

Mag., 126, 95-126. 

Druitt, T.H., L. Edwards, R.A. Mellors, D.M. Pyle, R.S.J. 

Sparks, M. Lanphere, M. Davies, & B. Barreirio (1999). 

Santorini Volcano. Geol. Soc. Mem., 19, 165. 

Dvorak, J.J., & D. Dzurisin (1997). Volcano geodesy; the 

search for magma reservoirs and the formation of 

eruptive event. Rev. Geophys., 35, 343–384. 

Fernández, J., J.M. Carrasco, J.B. Rundle, & V. Arána 

(1999). Geodetic methods for detecting volcanic unrest: A 

theoretical approach. Bull. Volcanol., 60, 534–544. 

Fouqué, F., 1879. Santorin et ses eruptions. Masson et 

Cie., Paris, 440p. 

Fytikas, M., F. Innocenti, P. Manetti, R. Mazuoli, A. 

Peccerilo & L. Vilari (1984). Tertiary to Quaternary 

evolution of the volcanism in the Aegean Region. In 

Dixon J.E. & Robertson A.H.F. (eds.). The Geological 

Evolution of the Eastern Mediterranean. Geol. Soc. 

London, Spec. Pub., 17, 687-699. 

Heiken, G., & F. McCoy (1984). Caldera development 

during the Minoan eruption, Thera, Cyclades, Greece. J. 

Geophys. Res., 89, 8441-8462. 

Jackson, J.A. (1994). Active tectonics of the Aegean region. 

Annual Reviews of Earth and Planetary Sciences, 22, 

239-271. 

Le Pichon, X., & J. Angelier (1979). The Hellenic Arc and 

trench system: a key to the Neotectonic evolution of the 

Eastern Mediterranean area. Tectonophysics, 60, 1-42. 

Mountrakis, D., S. Pavlides, A. Chatzipetros, S. Meletidis, 

M. Tranos, G. Vougioukalakis & A. Kilias (1998). Active 

deformation in Santorini. In: Casale R., Fytikas M., 

Sigvaldarsson, G., & G. Vougioukalakis (eds): The 

European laboratory volcanoes. European Commission, 

EUR 18161, 13-22. 

Nomikou, P., S. N. Carey, D. Papanikolaou, K.L.C Bell, D. 

Sakellariou, M. Alexandri, & K. Bejelou (2011a). 

Exploration of the submarine cones in the Kolumbo 

Submarine Volcanic Zone of the Hellenic Arc (Aegean 

Sea, Greece). Journal of Global and Planetary Change, 

In review. 

Nomikou, P., D. Papanikolaou, M. Alexandri,& D. 

Sakellariou (2011b). Submarine Volcanoes along the 

Agean Volcanic Arc. Tectonophysics, In review. 

Papanikolaou, D. (1993). Geotectonic evolution of the 

Aegean. Bull. Geol. Soc. Greece, XXVII, 33-48. 

Papageorgiou, E., E. Lagios, S. Vassilopoulou & V. Sakkas, 

(2007). Vertical & Horizontal Ground Deformation of 

Santorini Island deduced by DGPS measurements. 

Proceedings of the 11th International Conference Geol. 

Soc. Greece, Athens, Greece. Bull. Geol. Soc. Greece, 

40 (3), 1219-1225. 

Papageorgiou E., A. Tzanis, P. Sotiropoulos, & E. Lagios 

(2010). DGPS and magnetotelluric constraints on the 

contemporary tectonics of Thera Island, Greece. Bull. 

Geol. Soc. Greece, XLIII, 1,344-356. 

Pe-Piper, G. & D.J.W. Piper (2005). The South Aegean 

active volcanic arc: relationships between magmatism 

and tectonics. Develop. in Volc., 7, 113-133. 

Sakellariou D., H. Sigurdsson, M. Alexandri, S. Carey, G. 

Rousakis, P. Nomikou, P. Georgiou & D. Ballas (2010). 

Active tectonics in the Hellenic volcanic arc: The 

Kolumbo submarine volcanic zone. Bull. Geol. Soc. 

Greece, XLIII, 2,1056-1063. 

Sigurdsson, H., S. Carey, M. Alexandri, G. Vougioukalakis, 

K. L. Croff, C. Roman, D. Sakellariou, C. Anagnostou, G. 

Rousakis, C. Ioakim, A. Gogou, D. Ballas, T. Misaridis, 

and P. Nomikou 2006. Marine investigations of Greece’s 

Santorini volcanic field. EOS, 87(34):337, 342. 

Reck, H. (1936). Santorini. –Der Werdergang eines 

Inselvulcans und sein Ausbruch 1925-1928. Dietrich 

Reimer, Berlin, 3 vols. MANCA. 

166


EARTHQUAKE 

ARCHAEOLOGY 



THE DYNAMIC ANALYSIS OF MULTI-DRUM ANCIENT STRUCTURES UNDER 

EARTHQUAKE EXCITATIONS 

Loizos Papaloizou (1), Petros Komodromos (2) 

(1) University of Cyprus, Department of Civil and Environmental Engineering, School of Engineering, University of Cyprus 

75 Kallipoleos Street, PO Box 20537, 1678 Nicosia, Cyprus. Email: loizop@ucy.ac.cy 

(2). University of Cyprus, Department of Civil and Environmental Engineering, School of Engineering, University of Cyprus 

75 Kallipoleos Street, PO Box 20537, 1678 Nicosia, Cyprus. Email: komodromos@ucy.ac.cy 

Abstract (The dynamic analysis of multi-drum ancient structures under earthquake excitations): The seismic behaviour of 

ancient monumental structures with monolithic or multi-drum classical columns and colonnade systems in two rows is investigated. 

In particular, the Discrete Element Method (DEM) is utilized in the study of ancient columns under strong ground excitations, by 

simulating the individual rock blocks as distinct bodies. A specialized software application is developed, using a modern objectoriented 

programming language, in order to enable the effective simulation of multi-drum columns and colonnades. A number of 

parametric studies is performed in order to investigate the effect of excitation characteristics on the behaviour of multi-drum 

columns under harmonic and earthquake excitations. The simulations reveal that the columns and colonnades have the capacity 

to successfully withstand earthquakes excitations that were selected from regions where these monuments are often built. 

Key words: Ancient Columns, Earthquake, DEM. 

INTRODUCTION 

Strong earthquakes are common causes of 

destruction of ancient monuments, such as classical 

columns and colonnades. Ancient classical columns 

of great archaeological significance can be 

abundantly found in high seismicity areas in the 

Eastern Mediterranean. Multi-drum columns are 

constructed of stone blocks that are placed on top of 

each other, with or without connecting material 

between the individual blocks. The seismic behaviour 

of these structures exhibits complicated rocking and 

sliding phenomena between the individual blocks of 

the structure that very rarely appear in modern 

structures. 

Entablature 

Column 

Cornice 

Frieze 

Architrave 

Capital 

Shaft 

Fig. 1: Architecture of a typical classical monument. 

In ancient Greece the temples formed the most 

important class of buildings erected during that era 

and can be classified into three "Orders of 

167 

Architecture", the Doric, Ionic and Corinthian order. 

An "order" in Greek architecture consists of the 

column, including the base and the capital, and the 

entablature (Fig. 1). The entablature is divided into 

the architrave (lower part), the frieze (middle part) 

and the cornice (upper part). The differences among 

these three orders refer on the dimensions, 

proportions, mouldings and decorations of the 

various parts. 

Today, the remains of most of these temples are 

often limited to series of columns with an entablature 

or only an architrave, and in some cases only 

standalone columns. The investigation of the seismic 

behaviour of such monuments is scientifically 

interesting, as it involves complicated rocking and 

sliding responses of the individual rock blocks. The 

understanding of the seismic behaviour of these 

structures contributes to the rational assessment of 

efforts for their structural rehabilitation and may also 

reveal some information about past earthquakes that 

had struck the respective region. 

Ancient monuments, compared to modern structures, 

have been exposed to large numbers of strong 

seismic events throughout the many centuries of their 

life spans. Those that survived have successfully 

withstood a natural seismic testing that lasted for 

several centuries. Thus, it is important to understand 

the mechanisms that enabled them to avoid 

structural collapse and destruction during strong 

earthquakes. Since analytical study of such multiblock 

structures under strong earthquake excitations 

is practically difficult, if not impossible for large 

numbers of blocks, while laboratory tests are very 

difficult and costly, numerical methods can be used 

to simulate their seismic response.


EARTHQUAKE 

ARCHAEOLOGY 



very extensive review of the literature on the usage 

of numerical methods for the analysis of monuments 

until 1993 was published by Beskos (1993). The 

dynamic behaviour of infinitely rigid bodies during 

horizontal excitations was studied by Housner 

(1963), while, later on, other researchers (Psycharis 

et al., 2003, Pompei et al., 1998, Makris & Zhang, 

1998, Manos et al., 2001, Komodromos et al., 2008) 

investigated further, both analytically and 

experimentally, the required conditions to overturn 

rigid bodies. Such structures can be simulated 

utilizing the Discrete Element Methods (DEM), which 

have been specifically developed for systems with 

distinct bodies that can move freely in space and 

interact with each other with contact forces through 

an automatic and efficient recognition of contacts. 

Research efforts to use the DEM in the simulation of 

ancient structures have already shown promising 

results, motivating further utilization of this method. 

Recent research work based on commercial DEM 

software applications (Psycharis et al., 2003, 

Papantonopoulos, 2002), demonstrated that the DEM 

can be reliably used for the analysis of such 

structures, although they reported a sensitivity of the 

response to small perturbations of the characteristics 

of the structure or the excitation. However, similar 

sensitivity has also been observed in experiments 

with classical columns (Mouzakis et al., 2002). 

Hence, it is important to perform large numbers of 

simulations with varying earthquake characteristics 

and design parameters to properly assess and 

interpret the simulation results. 

Latest research studies in the fields of 

paleoseismology and archaeoseismology (Hinzen et 

al., 2010, Caputo et al., 2011) investigate the 

damage in ancient monument structures and propose 

various quantitative models to test the seismogenic 

hypothesis of observed damage. Papaloizou and 

Komodromos (2011) used the Discrete Element 

Method (DEM) as well as a modern object-oriented 

design and programming approach, in order to 

examine the simulation of multi-drum columns and 

colonnades under harmonic and earthquake 

excitations. 

A custom-made DEM software has been specifically 

designed (Papaloizou and Komodromos, 2011) and 

implemented to enable efficient performance of large 

numbers of numerical simulations with varying 

parameters, modelling these structures with 

independent distinct bodies, as they are constructed 

in practice. Such simulations allow us to assess the 

influence of different earthquake characteristics as 

well as the various mechanical and geometrical 

parameters of these structures on their seismic 

responses. 

NUMERICAL ANALYSIS 

A large number of parametric studies is performed in 

order to investigate the effect of excitation 

characteristics as well as the influence of geometrical 

and mechanical characteristics of multi-drum 

columns on their behaviour and response under 

168 

harmonic and earthquake excitations. Colonnade 

systems with two rows of columns, one over the 

other, are examined with earthquake ground motions. 

T=36.3 sec 

T=37.0 sec 

Fig. 2: Mexico City Earthquake (scaled 1.5 Times). 

T=4.4 sec 

T=12.2 sec 

Fig. 3: Athens Earthquake (scaled 9 Times). 

Parametric studies have been performed by varying 

the excitation frequency and acceleration, as well as 

the friction coefficient and the geometric


EARTHQUAKE 

ARCHAEOLOGY 



characteristics of the simulated columns and 

colonnades in order to assess the influence of these 

parameters in the seismic response of the structure. 

Fig. 2 and Fig. 3 show snapshots (in different time 

steps) of the response of such structural systems for 

the Mexico City and the Athens Earthquakes, 

respectively (Table 1). The results show that the 

frequency content of the ground motion affects 

significantly the response. The displacements of the 

upper level columns in respect to the displacements 

of the lower level column are affected by the 

frequency content of the excitation. 

Specifically, for the Mexico City earthquake, which 

has low predominant frequencies, the upper level 

columns overturn. For the Athens Earthquake, which 

has higher predominant frequencies, and for a much 

larger maximum ground acceleration the upper level 

columns are not affected by the seismic excitation. 

Date 

and 

Time 

9/19/1995 

(13:19CT) 

9/7/1999 

(11:56:50) 

Earthquake 

Component 

Mexico City 

(270) 

Athens 

(N46) 

PGA 

(m/sec2) 

Predominant 

Frequencies 

(Hz) 

0.98 0.45-0.53 

3.01 4.1-8.3 

Table 1. List of earthquake records that have been used in 

the analyses. 

Moreover, for low frequency harmonic excitations, 

the exhibited response is dominated by rocking, while 

sliding prevails in cases of excitations with very high 

frequencies. In between the two extremes, the 

response contains both rocking and sliding 

phenomena. 

Furthermore, the results indicate that the required 

acceleration to initiate rocking or sliding decreases as 

the excitation frequency increases. The acceleration 

that is needed to overturn the column also increases 

as the frequency increases. 

By examining the stability of multi-drum columns and 

colonnades for earthquakes that were selected from 

regions, where these monuments are often built, 

such as the Eastern Mediterranean regions, the 

simulations reveal that the columns have the capacity 

to successfully withstand strong earthquakes. The 

required acceleration to overturn a column decreases 

as the predominant frequency of the earthquake 

decreases. 

The investigation of the dynamic response of such 

monumental structures, combined with the research 

fields of paleoseismology and archaeoseismology, 

may reveal certain information from past strong 

earthquakes that have struck the respective regions. 

The investigation of the response of multi-drum 

structures under different ground motions can help in 

defining the frequency content of old destructive 

earthquakes. 

Acknowledgements: The authors would like to thank the 

Cyprus Research Promotion Foundation for funding the 

active research project entitled “Investigation for the 

protection of ancient multi-drum columns and colonnades 

from strong earthquakes (// 

0609()/23)”. 

References 

Beskos D. (1993), “Use of finite and boundary elements in 

the analysis of monuments and special structures”, 

Association of Civil Engineers of Greece, 217. 

Caputo, R., Hinzen, K.-G., Liberatore, D., Schreiber, S., 

Helly, B., Tziafalias, A., (2011), “Quantitative 

archaeoseismological investigation of the Great Theatre 

of Larissa, Greece “Bulletin of Earthquake Engineering. 9 

(2), pp. 347-366. 

Hinzen, K.-G., Fleischer, C., Reamer, S.K., Schreiber, S., 

Schütte, S., Yerli, B., (2010), “Quantitative methods in 

archaeoseismology” Quaternary International, DOI: 

10.1016/j.quaint.2010.11.006. 

Housner G.W (1963), ‘The behavior of inverted pendulum 

structures during earthquakes”. Bulletin of seismological 

Society of America, 53, 403-417. 

Komodromos P., Papaloizou L., Polycarpou P., (2008) 

“Simulation of the response of ancient columns under 

harmonic and earthquake excitations”, Engineering 

Structures 30(8), 2154-2164. 

Makris N. and Zhang J., (2001) “Rocking response of 

anchored blocks under pulse-type motions”, Engineering 

Mechanics 127, 484-493. 

Manos G.C., Demosthenous M., Kourtides V., 

Hatzigeorgiou A., (2001) “Dynamic and earthquake 

behavior of models of ancient columns and colonnades 

with or without energy absorbtions systems”, 

Proceedings of Second Greek National Conference on 

Earthquake Engineering and Seismology, 1, 257-276. 

Mouzakis H., Psycharis I., Papastamatiou D., Carydis P., 

Papantonopoulos C., and Zambas C., (2002). 

“Experimental investigation of the earthquake response 

of a model of a marble classical column”, Earthquake 

Engineering and Structural Dynamics 31, 1681-1698. 

Papaloizou, L., Komodromos, P., (2011) “Investigating the 

seismic response of ancient multi-drum colonnades with 

two rows of columns using an object-oriented designed 

software”, Advances in Engineering Software, DOI: 

10.1016/j.advengsoft.2011.05.030. 

Papantonopoulos C., Psycharis I., Papastamatiou D., 

Lemos J. and Mouzakis H., (2002). “Numerical prediction 

of the earthquake response of classical columns using 

the distinct element method”, Earthquake Engineering 

and Structural Dynamics 31, 1699-1717. 

Pompei A., Scalia A. and Sumbatyan M.A. (1998), 

“Dynamics of Rigid Block due to Horizontal Ground 

Motion”, Engineering Mechanics 124, 713-717. 

Psycharis I., Lemos J., Papastamatiou D., Zambas C. and 

Papantonopoulos C. (2003), “Numerical study of the 

seismic behaviour of a part of the Parthenon Pronaos”, 

Earthquake Engineering & Structural Dynamics 32, 2063- 

2084. 

169


EARTHQUAKE 

ARCHAEOLOGY 



NEOTECTONIC AND ACTIVE DIVERGING RATES OF EXTENSION IN THE NORTHERN 

AND SOUTHERN HELLENIDES ACROSS THE CENTRAL HELLENIC SHEAR ZONE 

Papanikolau, Dimitrios (1, Royden, Leigh (2), Vassilakis, Emmanuel (1) 

(1) School of Geology and Geoenvironment, Department of Dynamics, Tectonics and Applied Geology, National and Kapodistrian 

University of Athens, 15784, Greece. Email: dpapan@geol.uoa.gr 

(2) Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, 54-826, Cambridge, MA 

02139, USA 

Abstract (Neotectonic and active diverging rates of extension in the Northern and Southern Hellenides across the Central 

Hellenic Shear Zone): Present day location and geometry of the Hellenic arc and trench system is only a small portion of the 

previously developed Hellenic arc that created the Hellenides orogenic system. The timing of differentiation is constrained in Late 

Miocene, when the arc was divided in a northern and a southern segment. This is based on: a) the dating of the last compressive 

structures observed all along the Hellenides during Oligocene to Middle-Late Miocene, b) on the time of initiation of the Kephalonia 

transform fault, c) on the time of opening of the North Aegean Basin and d) on the time of opening of new arc parallel basins in the 

south and new transverse basins in the central shear zone, separating the rapidly moving southwestward Hellenic subduction 

system from the slowly converging system of the Northern Hellenides. The driving mechanism of the arc differentiation is the 

heterogeneity produced by the different subducting slabs in the north (continental) and in the south (oceanic) and the resulted 

shear zone because of the retreating plate boundary producing a roll back mechanism in the present arc and trench system. The 

extension produced in the upper plate has resulted in the subsidence of the Aegean Sea and the creation of several neotectonic 

basins in southern continental Greece in contrast to the absence of new basins in the northern segment since Late Miocene. 

Crustal thickness, structural profiles, earthquake mechanisms and GPS rates are compared in the two segments, showing 

significant extension in the Southern Hellenides. 

Key words: Hellenides, neotectonic extension, slab roll-back 

INTRODUCTION 

The active portion of the Hellenic orogen is an 

arcuate belt reaching from the central Adriatic Sea 

southwards and eastwards to western Anatolia 

(McKenzie, 1978; Le Pichon and Angelier, 1979; 

Picha, 2002; Reilinger et. al., 2006). Within the 

northwest-trending portion of the Hellenic belt, two 

segments can be distinguished by subduction rate 

(slow in the northwest, fast in the southeast) and 

water depth of the foreland lithosphere (shallow in 

the northwest, deep in the southeast; e.g. Dewey and 

Sengor, 1979; Finetti, 1982; McKenzie, 1978; Le 

Pichon and Angelier, 1979, Fig. 1). These segments, 

dextrally offset from one another near the island of 

Kephalonia, have been referred as the Northern and 

Southern Hellenides (Papanikolaou and Royden, 

2007; Royden and Papanikolaou, 2011). 

THE SEPARATION OF THE NORTHERN AND 

SOUTHERN HELLENIDES BY THE CENTRAL 

HELLENIC SHEAR ZONE 

Several years of GPS measurements indicate a 

convergence rate of ~5-10 mm/yr across the 

Northern Hellenic subduction boundary, as measured 

between stations on the subducting plate (Apulia) 

and on the over-riding plate in northern Greece 

(Hollenstein et al., 2008; Bennett et al., 2008; 

Vassilakis et al., 2011) and seismic activity and focal 

solutions for local earthquakes attest to continuing 

convergence in this region (e.g. Louvari et al., 1999; 

Papazachos et al., 2000). Based on evidence from 

170 

seismology, morphology and industry seismic data, 

the active thrust front of the Northern Hellenides lies 

just west of the Ionian islands of Corfu and Paxos 

(Monopolis & Bruneton, 1982; Vassilakis et al., 2011) 

(Fig. 1). 

Even though gravity data indicate that the basement 

is flexed downward beneath the thrust front and the 

resulting depression filled with sedimentary foredeep 

deposits (Moretti and Royden, 1988), no trench is 

present in the bathymetry along the Northern 

Hellenides. For ease of reference we will refer to this 

zone of convergence as the Northern Hellenic trench, 

despite the fact that the trench has been entirely filled 

with sediments. The lithosphere entering the 

Northern Hellenic trench is continental or transitional 

in character, with a crustal thickness of ~25-30 km 

(Morelli et al, 1975; Marone et al, 2003, Cassinis et. 

al., 2003). Modern water depths near the thrust belt 

are generally ~1 km or less and overlie a shallow 

water sedimentary sequence of Triassic through 

Pliocene age (Jacobshagen et al., 1978). 

On the contrary the same GPS data indicate a 

convergence rate of ~35 mm/yr across the Southern 

Hellenides, as measured between Africa and points 

in the over-riding (Aegean) domain (McClusky et al., 

2000; Reilinger et al., 2006). Behind the Southern 

Hellenides, a Benioff zone reaches to ~150 km depth 

(e.g. Papazachos et al., 2000) and an active volcanic 

arc is present ~200 km behind the trench (Fytikas et 

al., 1984). The zone along which basement is 

subducted beneath the Hellenides lies ~50 km west


EARTHQUAKE 

ARCHAEOLOGY 



of the southwestern coast of the Peloponnesus, 

passing beneath the deepest portions of the Hellenic 

trench (Fig. 1). Here, earthquake hypocenters and 

gravity data indicate that the depth to basement is 

~12-15 km depth (Royden 1993; Hirn et al., 1997; 

Clément et al., 2000; Sachpazi et al., 2000). Thrust 

faults and folds also occur within the accretionary 

prism outboard of the trench, over a width of several 

hundred kilometers to the Mediterranean ridge (Kopf 

et al., 2003), but thrusting here involves only 

sedimentary cover detached above the basement. 

The crust beneath the Ionian Sea is almost certainly 

oceanic, probably Triassic or Jurassic in age, and 

consists of approximately 8 km of crystalline crust 

overlain by 6-10 km of sedimentary cover (Makris, 

1985; Kopf et al., 2003). The water depth throughout 

much of the Ionian Sea region is 3-4 km, with the 

deepest depths of up to 5 km occurring along the 

Southern Hellenic Trench. 

the Northern Hellenides to ~3-4 km in front of the 

Southern Hellenides, particularly near the west coast 

of mainland Greece (Fig. 1). Here the Hellenic 

subduction boundary (trench) appears to be dextrally 

offset by ~100-150 km across the Kephalonia 

transform zone, although the precise offset is difficult 

to determine due to north-south variations in water 

depth and sediment thickness and due to the fact 

that some of the offset of the northern and southern 

segments is taken up by clockwise rotation of crustal 

units adjacent to the south side of the transform 

(Vassilakis et al., 2011). Northeastward, the 

Kephalonia transform zone extends into mainland 

Greece to merge with the broadly defined zone of 

dextral and extensional deformation in the Central 

Hellenic Shear Zone (Papanikolaou and Royden, 

2007; see also discussions by Goldsworthy et al., 

2002; Armijo et al., 1996; Roberts and Jackson, 

1991). 

Global P-wave tomography indicates a northeastdipping 

zone of high P-wave speeds beneath the 

Southern and Northern Hellenides (Spakman et al., 

1993; van der Hilst et al., 1997; Karason and van der 

Hilst, 2001; Wortel and Spakman, 2000; Suckale et 

al., 2009). Behind the Southern Hellenides, the 

subducted lithosphere appears to reach to the base 

of the upper mantle, perhaps deeper. A northeastdipping 

zone of high P-wave speed also exists north 

of the Kephalonia transform zone, but here the 

velocity contrast with the surrounding mantle is not 

as large as beneath the Southern Hellenides. 

COMPARISON OF TECTONIC STRUCTURES AND 

DEFORMATION RATES BETWEEN THE 

NORTHERN AND SOUTHERN HELLENIDES 

Fig. 1: Selected GPS velocities from McClusky et al. [2000] 

in a reference frame that minimizes velocities in the Aegean 

region. Some data were omitted in areas where high data 

density obscured the velocity pattern, mainly near the North 

Anatolian Fault zone. Shading indicates the zones of active 

oblique extension that bound the northwestern and 

northeastern margins of the largely undeforming Aegean 

domain. 

The northern and southern segments of the Hellenic 

subduction boundary are separated by the 

Kephalonia transform zone (e.g. Dewey and Sengor, 

1979; Finetti, 1982; Kahle and Muller, 1998; Kahle et 

al., 1995; Hollenstein et al., 2008), across which GPS 

data indicate ~25 mm/yr of dextral motion (Fig. 1). 

The Kephalonia transform separates the slowly 

subducting, continental foreland of the northern 

Hellenides from the rapidly moving upper plate of the 

southern Hellenides. Focal solutions from 

earthquakes located along the Kephalonia transform 

show right-slip on steep southwest-striking fault 

planes and also thrust faulting along northeaststriking 

fault planes. 

The Kephalonia transform coincides closely with a 

change in foreland water depth from ~1 km in front of 

171 

The difference of the tectonic structure and of the 

deformation rates in the Northern and Southern 

Hellenides can be studied in two time periods: the 

recent neotectonic period of the last 5 million years 

during the Pliocene-Pleistocene and the active 

deformation of the last 14000 years during the 

Holocene incorporating also the actual observations 

and measurements. The main points showing the 

diverging rates of extension in the two segments of 

the Hellenides are the following. 

1) The width of the Hellenic peninsula from the Ionian 

sea at the west to the Aegean Sea at the east, is 

longer approximately by 60 km across the southern 

profile which corresponds to the offset of the trench 

across the Kephalonia transform since the Aegean 

coast is rather rectilinear from the Olympus foothills 

to the Southern Evia and Andros-Tinos-Mykonos 

Cycladic islands. 

2) The crustal thickness in the Southern Hellenides is 

much thinner (less than 25 km) than in the Northern 

Hellenides (up to 45 km) (Makris, 1977; 1985). 

3) The tectonic structure of the Southern Hellenides 

is characterised by the occurrence of the External 

Tectonometamorphic Belt in Peloponnesus in the 

form of tectonic windows, which resulted from 

extensional detachments during Miocene


EARTHQUAKE 

ARCHAEOLOGY 



(Papanikolaou and Royden, 2007; Papanikolaou et 

al., 2009). Thus, the structure in the Northern 

Hellenides is rather simple with a regular nappe 

emplacement from the area west of the Olympus 

tectonic window up to the Ionian Sea in contrast to 

the more complex structure in the Southern 

Hellenides, where there are new metamorphic 

geotectonic units (such as the Mani and Arna units) 

appearing below the non metamorphosed nappes 

(such as the Tripolis and Pindos units) (Papanikolaou 

and Vassilakis, 2010). 

4) A number of arc parallel extensional neotectonic 

basins oriented NW-SE occur along the transverse 

profile of the Southern Hellenides (e.g. in Southern 

Peloponnesus) as described by Papanikolaou et al. 

1988 (Fig. 2). In contrast, no neotectonic basins are 

developed in the transverse profile of the Northern 

Hellenides with the exception of the Ioannina basin 

(the only extensive Plio-Quaternary structure west of 

the Oligo-Miocene Mesohellenic basin). The 

difference of extension in the two segments within 

the upper plate is more than 20%. 

5) The last onshore compressive structures in the 

Northern Hellenides are observed in Corfu and Parga 

areas, involving Late Miocene and early Pliocene 

sedimentary formations (including the Messinian 

evaporites) and in Kephalonia and Zakynthos islands 

in the Southern Hellenides with a dextral offset 

similar to the Kephalonia transform but somewhat 

smaller. 

6) The transition from compression to extension as 

determined by earthquake mechanisms occurs in 

central Epirus in the Northern Hellenides and in the 

shallow marine zone between the Ionian islands and 

mainland Greece in the Southern Hellenides. 

7) The rate of upper plate extension across the belt is 

~3-5mm/yr in the Northern Hellenides and more than 

30mm/yr in the Southern Hellenides. 

EPILOGUE 

All arguments displayed above point to the same 

conclusion of differentiation between the northern 

and southern Hellenides, regarding the amount of the 

extension both during the neotectonic period and the 

present day deformation. In many cases this 

differentiation can be quantified and the results are 

quite impressive showing that a major complex 

structure such as the Central Hellenic Shear Zone is 

acting as an oblique transform structure of the 

Hellenides by dividing the Hellenic peninsula and 

changing the present morphology. 

Acknowledgements: This research was supported by the 

Continental Dynamics Program at NSF, grant EAR- 

0409373. 

Fig. 2: In this vertically exaggerated (x10) SW-NE section across the Southern Hellenides the post-nappestacking extensional 

structures are indicated. 

172


EARTHQUAKE 

ARCHAEOLOGY 



References 

Armijo, R., et al. (1996), Quaternary evolution of the Corinth 

Rift and its implications for the late Cenozoic evolution of 

the Aegean, Geophys. J. Int, 126, 11-53. 

Bennett, R. A., et al. (2008), Eocene to present subduction 

of southern Adria mantle lithosphere beneath the 

Dinarides, Geology, 36(1), 3-6. 

Cassinis, R., et al. (2003), The deep crustal structure of Italy 

and surrounding areas from seismic refraction data: a 

new synthesis, Bollettino di Geofisica Teorica ed 

Applicata, 122, 365-375. 

Clément, C., et al. (2000), Seismic structure and the active 

Hellenic subduction in the Ionian islands, 

Tectonophysics, 329(1-4), 141-156. 

Dewey, J. F., and C. engör (1979), Aegean and 

surrounding regions: Complex multiplate and continuum 

tectonics in a convergent zone, Geological Society of 

America Bulletin, 90(1), 84-92. 

Finetti, I. (1982), Structure, stratigraphy and evolution of 

central Mediterranean (Pelagian Sea, Ionian Sea), 

Bollettino di Geofisica Teorica ed Applicata, 24(96), 247- 

312. 

Fytikas, M., et al. (1984), Tertiary to Quaternary evolution of 

volcanism in the Aegean region, in The geological 

evolution of the eastern Mediterranean, edited by J. 

Dixon and A. Robertson, pp. 687-701, Geological Society 

of London, London. 

Goldsworthy, M., et al. (2002), The continuity of active fault 

systems in Greece, Geophysical Journal International, 

148(3), 596-618. 

Hirn, A., et al. (1997), Increase in melt fraction along a 

south-north traverse below the Tibetan Plateau: evidence 

from seismology, Tectonophysics, 273(1-2), 17-30. 

Hollenstein, C., et al. (2008), Crustal motion and 

deformation in Greece from a decade of GPS 

measurements, 1993-2003, Tectonophysics, 449(1-4), 

17-40. 

Jacobshagen, V. (1986), Geologie von Griechenland, 363 

pp. 

Kahle, H. G., and S. Mueller (1998), Structure and 

dynamics of the Eurasian-African/Arabian plate boundary 

system: Objectives, tasks and resources of the 

WEGENER group, Journal of Geodynamics, 25(3-4), 

303-325. 

Kahle, H. G., et al. (1995), The strain field in northwestern 

Greece and the Ionian Islands: results inferred from GPS 

measurements, Tectonophysics, 249, 41-52. 

Kopf, A., et al. (2003), The Mediterranean Ridge: A mass 

balance across the fastest growing accretionary complex 

on Earth, J. Geophys. Res., 108(B8), 2372. 

Le Pichon, X., and J. Angelier (1979), The Hellenic arc and 

trench system: a key to the neotectonic evolution of the 

eastern Mediterranean area, Tectonophysics, 60, 1-42. 

Louvari, E., et al. (1999), The Cephalonia Transform Fault 

and its extension to western Lefkada Island (Greece), 


Makris, J. (1976), A dynamic model of the Hellenic Arc 

deduced from Geophysical data, Tectonophysics, 36, 

339-346. 

Makris, J. (1985), Geophysics and geodynamic implications 

for the evolution of the Hellenides, in Geological 

Evolution of the Mediterranean Basin, edited by D. J. 

Stanley and F. Wezel, pp. 231–248, Springer, Berlin 

Marone, F., et al. (2003), Joint inversion of local, regional 

and teleseismic data for crustal thickness in the Eurasia- 

Africa plate boundary region, Geophysical Journal 

International, 154(2), 499-514. 

McClusky, S., et al. (2000), Global Positioning System 

constraints on plate kinematics and dynamic in the 

eastern Mediterranean and Caucasus, Journal of 

Geophysical Research, 105(B3), 5695-5719. 

173 

McKenzie, D. (1978), Active tectonics of the Alpine- 

Himalayan belt: the Aegean Sea and surrounding 

regions, Geophysical Journal of the Royal Astronomical 

Society, 55(1), 217-254. 

Monopolis, D., and A. Bruneton (1982), Ionian sea (Western 

Greece): Its structural outline deduced from drilling and 

geophysical data, Tectonophysics, 83(3-4), 227-242. 

Morelli, C., et al. (1975), Bathymetry, gravity and magnetism 

in the Strait of Sicily and in the Ionian Sea, Bollettino di 

Geofisica Teorica ed Applicata, 17, 39-58. 

Moretti, I., and L. Royden (1988), Deflection, Gravity- 

Anomalies and Tectonics of Doubly Subducted 

Continental Lithosphere - Adriatic and Ionian Seas, 

Tectonics, 7(4), 875-893. 

Papanikolaou, D., (2010), Major palaeogeographic, tectonic 

and geodynamic changes from the last stage of the 

Hellenides to the actual Hellenic arc and trench system, 

Bulletin Geological Society of Greece, XLIII(1), 72-85. 

Papanikolaou, D., and E. Vassilakis (2010), Thrust faults 

and extensional detachment faults in Cretan tectonostratigraphy: 

Implications for Middle Miocene extension, 

Tectonophysics, 488(1-4), 233-247. 

Papanikolaou, D., and L. Royden (2007), Disruption of the 

Hellenic arc: Late Miocene extensional detachment faults 

and steep Pliocene-Quaternary normal faults - Or what 

happened at Corinth?, Tectonics, 26 (TC5003). 

Papanikolaou, D., et al. (1988), A comparative study of 

neotectonic basins across the Hellenic Arc: The 

Messiniakos, Argolikos, Saronikos and Southern Evoikos 

Gulfs, Basin Research, 1, 167-176. 

Papanikolaou, D., et al. (2009), The Itea-Amfissa 

detachment: a pre-Corinth rift Miocene extensional 

structure in central Greece, Geological Society, London, 

Special Publications, 311(1), 293-310. 

Papazachos, B. C., et al. (2000), The geometry of the 

Wadati-Benioff zone and lithospheric kinematics in the 

Hellenic arc, Tectonophysics, 319, 275-300. 

Picha, F. J. (2002), Late Orogenic Strike-Slip Faulting and 

Escape Tectonics in Frontal Dinarides-Hellenides, 

Croatia, Yugoslavia, Albania, and Greece, AAPG Bulletin, 

86(9), 1659-1671. 

Reilinger, R., et al. (2006), GPS constraints on continental 

deformation in the Africa-Arabia-Eurasia continental 

collision zone and implications for the dynamics of plate 

interactions, J. Geophys. Res., 111. 

Roberts, S., and J. Jackson (1991), Active normal faulting in 

central Greece: an overview, Geological Society, London, 

Special Publications, 56(1), 125-142. 

Royden, L. H. (1993), Evolution of retreating subduction 

boundaries formed during continental collision, Tectonics, 

12(3), 629-638. 

Royden, L., and D. Papanikolaou (2011), Slab 

segmentation and late Cenozoic disruption of the Hellenic 

arc, Geochem. Geophys. Geosyst., 12(3), Q03010. 

Sachpazi, M., et al. (2000), Western Hellenic subduction 

and Cephalonia Transform: local earthquakes and plate 

transport and strain, Tectonophysics, 319(4), 301-319. 

Spakman, W., et al. (1993), Travel-time tomography of the 

European-Mediterranean mantle down to 1400 km, 

Physics of The Earth and Planetary Interiors, 79(1-2), 3- 

74. 

Suckale, J., et al. (2009), High-resolution seismic imaging of 

the western Hellenic subduction zone using teleseismic 

scattered waves, Geophysical Journal International, 

178(2), 775-791. 

van der Hilst, R. D., et al. (1997), Evidence for deep mantle 

circulation from global tomography, Nature, 386(6625), 

578-584. 

Vassilakis, Emm., et al. (2011), Kinematic Links from the 

Hellenic Trench to the Western Gulf of Corinth. Earth & 

Planetary Science Letters, 303, 108-120. 

Wortel, M. J. R., and W. Spakman (2000), Subduction and 

Slab Detachment in the Mediterranean-Carpathian 

Region, Science, 290(5498), 1910-1917.


EARTHQUAKE 

ARCHAEOLOGY 



CLUSTERING AND ANTICLUSTERING IN THE SOUTHERN APENNINES AS EVIDENCED 

FROM GEOLOGICAL FAULT SLIP-RATE SEISMIC HAZARD MAPS AND THE 

HISTORICAL RECORD 

Papanikolau Ioannis D. (1,2), Gerald Roberts (2) 

(1) Mineralogy-Geology Laboratory, Department of Earth and Atmoshperic Sciences, Agricultural University of Athens, Iera Odos 

75, 118-55, Athens, Greece 

(2) Department of Earth and Planetary Sciences, Birkbeck and University College London, Gower Street, WC1E 7HX. UK. 

i.papanikolaou@ucl.ac.uk, gerald.roberts@ucl.ac.uk 

Abstract (Clustering and anticlustering in the Southern Apennines as evidenced from geological fault slip-rate seismic 

hazard maps and the historical record): Geological fault slip rate seismic hazard maps were constructed in the area of the 

southern Apennines. By comparing short-term throw-rates (e.g. historical record) with longer-term throw-rates (e.g. offset 

postglacial geomorphic features), one might be able to differentiate areas that are currently in a cluster of earthquakes from areas 

that are currently in an anticluster, providing important insights into seismic hazards. In the southern Apennines, there seems to be 

an inconsistency between the historical record and the longer-term rates extracted from the geological fault slip-rate hazard maps, 

implying that the historical record is not representative of the longer-term throw-rates. This occurs because the northern part of the 

study area (Irpinia, northern Basilicata) appears to be ahead of its longer-term seismicity rate, thus perhaps in a temporal 

clustering period, whereas the southern part (Pollino, southern Basilicata) appears to be below its longer-term rate, implying that it 

is in a temporal anticlustering period. The latter agrees with the so-called Pollino seismic gap. 

Key words: Southern Italy, Pollino, active faults 

INTRODUCTION 

Gaps on seismicity maps are areas that constantly 

attract the interest of the scientific community. One 

fundamental question that puzzles scientists 

worldwide is whether a seismic gap on a seismicity 

map that is based on instrumental/historical data, 

could represent: a) an area of genuinely low longterm 

seismicity, leading to a correct low hazard 

interpretation; or b) a poor or incomplete historical 

record, suggesting that the region has experienced 

large earthquakes in the past, but they have not been 

recorded because the catalogues were either too 

short or incomplete; or c) a quiescent period in an 

area characterised by temporal clustering followed by 

a long recurrence interval (e.g. Marco et al. 1996). 

Scenario a suggests a low hazard whereas, 

scenarios b and c imply a high hazard potential. In 

particular, in scenario b the time elapsed since the 

last event, exceeds the duration of historic 

catalogues, implying that the region could probably 

have entered the last stage of the seismic cycle. In 

scenario c, if a perceived seismic gap coincides with 

a quiescent interval, then the region sooner or later 

will enter a period of earthquake spatial and temporal 

clustering, implying a very high potential hazard. 

However, without knowledge of the long-term 

seismicity record scenarios b and c are difficult to 

differentiate from scenario a. Geological data have 

the potential to extend the history of slip on a fault 

back many thousands of years, a time span that 

generally encompasses a large number of 

earthquake cycles (e.g. Yeats and Prentice, 1996), 

and thus elucidates the long-term pattern of fault-slip. 

As a result, multi-cycle seismic hazard maps based 

174 

on geological fault slip-rate data provide a way to 

visualize this by depicting the long-term deformation 

pattern of slip. 

Fig.1.Seismic hazard map of Italy derived from the historical 

earthquake record (GNDT-SSN, 2001). This map shows 

intensities that have a 90% probability of not being 

exceeded in a certain time period (e.g. 50 years) assuming 

a 475-year return period. White box shows the area of the 

seismic hazard map constructed during this study, whereas 

the dash white circle shows the Pollino seismic gap. 

In this paper we present such hazard maps for the 

region of the southern Apennines. Then we weigh 

our map against the historical record by comparing


EARTHQUAKE 

ARCHAEOLOGY 



the number of recorded earthquakes in the last 5-7 

centuries with that implied by slip-rates on active 

normal faults averaged over 18 kyrs, a time period 

which contains numerous seismic cycles. This 

comparison demonstrates that the long history of 

earthquakes in the Italian Apennines, which is 

certainly complete for 5-7 centuries (Boschi et al. 

1995) may indeed contain evidence for earthquake 

clustering, implying that the historical record is 

unrepresentative of the long-term deformation 

pattern. 

data from published trench-site palaeoseismic data 

were also utilised (e.g. Pantosti et al. 1993, Michetti 

et al. 1997). In these maps each fault ruptures in 

"floating" earthquakes (e.g. WGCEP 2002), which 

are distributed around a mean magnitude of Ms=6.5. 

THE AREA OF THE SOUTHERN APPENNINES 

AND THE POLLINO SEISMIC GAP 

A so-called “seismic gap” appears on a historical 

seismicity map in the Pollino region in the southern 

Italy (Fig.1). In the Pollino area where several active 

faults exist, no large magnitude historical 

earthquakes are known, but palaeoseismological 

studies clearly demonstrate the occurrence of 

surface faulting events (Michetti et al., 1997; Cinti et 

al., 1997; Michetti et al. 2000). However, seismic 

hazard and probabilistic maps based on the historical 

record (Slejko et al., 1998; Faenza and Pierdominici, 

2007) regard this area as an area of genuinely low 

hazard, disregarding the geological evidence. We 

suggest that knowledge of the long-term rates of 

surface faulting derived from offset geology is 

essential before such an area can be defined as low 

or high hazard. 

SEISMIC HAZARD MAP FROM GEOLOGICAL 

FAULT SLIP-RATE DATA 

Methodology 

Fault throw-rates are firstly converted into earthquake 

frequencies, assuming that each fault ruptures in 

"floating" earthquakes, which are distributed around a 

mean magnitude of fixed size. Then, this information 

is turned into a hazard map after using: i) empirical 

relationships between coseismic slip values, rupture 

lengths and earthquake magnitudes (Wells and 

Coppersmith 1994), ii) empirical relationships 

between earthquake magnitudes and intensity 

distributions that define the shapes and sizes of IX 

and VIII isoseismals (Grandori et al. 1991, D’ Amico 

et al. 1999) and iii) a simple attenuation/amplification 

functions for seismic shaking on bedrock compared 

to flysch and basin-filling sediments. The final 

product is a high spatial resolution seismic hazard 

map showing how many times each location has 

been shaken at a certain intensity value (e.g. 

intensity IX) over a fixed time period (e.g. since the 

last glaciation), which can be easily transformed into 

a map of recurrence intervals (see Papanikolaou 

2003 and Roberts et al., 2004). 

Data input 

Fault geometry and fault throw-rates have been 

extracted from Papanikolaou and Roberts (2007) and 

references therein (Fig.2). Throw rates are mainly 

extracted from post-glacial fault scarps, using as a 

reference the last major glacial retreat phase that 

initiated 18,000 years ago and left a clear imprint in 

the topography (Palmentola et al. 1990); other input 

175 

Fig. 2: (a) Location map of the study area, showing active 

faults that have displaced Holocene deposits, (b) Map of 

major active faults of the Southern Apennines in UTM 

coordinate frame (Papanikolaou and Roberts 2007). The 

bedrock map was modified and simplified from the CNR 

(1990) map. VF - Volturara Fault; IrF - Irpinia Fault; AntIrF - 

Antithetic Irpinia Fault; SGrF - San Gregorio Fault; AlF - 

Alburni Fault; VDF - Vallo di Diano Fault; VAF - Val' D Agri 

Fault; MaF - Maratea Fault; MAF - Monte Alpi Fault; MeF - 

Mercure Fault; PF - Pollino Fault, CaF – Castrovillari fault 

(from Cinti et al., 2002), CiF – Civita fault (from Michetti et 

al., 1997). 

Results 

Figure 3 shows the frequency of shaking for intensity 

VIII assuming homogeneous bedrock geology, a 

circular pattern of energy release and a 25 km radius 

for the VIII intensity. The highest hazard is observed 

in the centre of the map which receives enough 

energy to shake at intensity VIII or higher about 80 

times and decreases smoothly towards the tips of the 

fault array (about 20-30 times).


EARTHQUAKE 

ARCHAEOLOGY 



normalised long-term shaking values and the 

historical record both for intensities IX and VIII. 

Instead, there is a geographic variation, where 

earthquake shaking from historical seismicity for 

towns in the north area is ahead of the rate expected 

given modeling of the 18 kyrs record of fault slip, with 

the opposite relationship in the south close to the socalled 

"Pollino Gap". This suggests earthquakes 

have been clustered in time in the north during 

historical times and anticlustered in the south 

(Scenario c of the Introduction). 

Fig. 3: Seismic hazard map showing the frequency of 

shaking for intensity VIII assuming homogeneous bedrock 

geology, a circular pattern of energy release and a 25 km 

radius for the VIII isoseismal. 

Figure 4 shows how many times a locality receives 

enough energy to shake at intensities IX in 18 kyrs, 

assuming a 12.5 km radius of isoseismal IX and a 

simple attenuation/amplification function, where 

bedrock Mesozoic-Cenozoic limestones shake at a 

single intensity value less than the 

Quaternary/flysch/foredeep deposits at the same 

epicentral distance. In particular, the area of highest 

shaking frequency is located in the hangingwall 

centre of the Val’ D’ Agri fault, which will receive 

enough energy to shake at intensity IX or higher up 

to 36 times in 18 kyrs. On the other hand, the 

hangingwall centres of the Volturara and the Pollino 

distal faults will shake only about 10-15 times at 

intensities IX because these faults have lower sliprates. 

Therefore, the recurrence interval for intensity 

IX or higher is estimated to be as short as 500 yrs for 

localities in the hangingwall centre of the Val' D' Agri 

fault and as long as 1200-1800 yrs in the hangingwall 

centres of the Pollino and the Volturara faults. 

COMPARISON WITH THE HISTORICAL RECORD 

To investigate whether rates of earthquake 

occurrence measured using the historical record (e.g. 

442 years) are representative of longer time periods 

(18 kyrs) where the effects of clustering may be 

averaged out, the recurrence intervals extracted from 

these maps are compared with the historical record. 

Overall, 24 towns were chosen that are uniformly 

distributed in the entire area and most of them have 

numerous records of earthquake shaking dating back 

to the sixteenth and seventeenth century (Table 1). 

There is no correlation (R 2


EARTHQUAKE 

ARCHAEOLOGY 



not ruptured during this time period and other faults 

that could be in a temporal clustering period. 

Town 

(see also 

Fig. 3) 

Predictions 

from Fig.3 and 

Fig.4: 

Number of 

times in 18 

kyrs 

Table 1 

Normalised values 

Number of times 

In 442 yrs In 

654yrs 

Historical 

Record 

(Number of 

times each 

locality was 

shaken 

at stated 

intensity) 

IX VIII IX VIII IX VIII IX VIII 

Buccino 28 49 1 1 1 2 1 3 

Balvano 12 44 0 1 0 2 3 1 

Padula 31 68 1 2 1 2 1 0 

Atena 

20 70 1 2 1 3 2 2 

Lucana 

Tito 0 25 0 1 0 1 4 0 

Salerno 0 0 0 0 0 0 0 0 

Potenza 0 0 0 0 0 0 0 3 

Sicignano d’ 

Alburni 

25 45 1 1 1 2 1 1 

Eboli 0 10 0 0 0 0 0 0 

Cassano allo 

Ionio 

3 15 0 0 0 1 0 0 

Maratea 18 48 0 1 1 2 0 1 

Laino 

Castello 

18 37 0 1 1 1 0 0 

Rotonda 13 37 0 1 0 1 0 0 

Castrovillari 12 22 0 1 0 1 0 0 

Sant' 

0 0 0 0 0 0 1 0 

Arcangelo 

Montemurro 8 42 0 1 0 2 1 0 

Sanza 8 65 0 2 0 2 0 1 

Lagonegro 18 70 0 2 1 3 1 0 

Lauria 20 55 1 1 1 2 0 2 

Lioni 23 32 1 1 1 1 3 1 

Teora 23 35 1 1 1 1 4 0 

Viggiano 21 46 1 1 1 2 1 0 

Montella 0 15 0 0 0 1 0 3 

Polla 15 65 0 2 1 2 2 2 

It is evident that the northwestern part of the study 

area (Irpinia and northern Basilicata) is ahead of its 

longer-term shaking record, indicating that this area 

may be in a temporal earthquake cluster. Towns like 

the Balvano, Atena Lucana, Tito, Lioni, Teora, Polla 

have experienced more intensity IX (in particular) and 

VIII events during the last five to seven centuries 

compared to their longer-term shaking record derived 

from the maps (Table 1, Fig. 4). On the other hand, 

the southeastern part of the study area (Pollino, 

southern Basilicata) may be in a temporary 

anticlustering process. Towns like the Lauria, 

Lagonegro, Sanza, Castrovillari, Rotonda, Laino 

Castello and Maratea have experienced less events 

(particularly for intensity VIII) compared to the longerterm 

shaking record (Table 1, Fig.3, Fig. 4). The 

latter can potentially explain why paleoseismic data 

extracted from the Pollino region (Michetti et al., 

1997; Cinti et al., 1997) are in conflict with the 

historical record, with paleo surface ruptures found in 

a region with no major recorded historical events. 

References 

Boschi, E., et al. (1995). Catalogo dei forti terremoti in Italia 

dal 461 a.C. al 1980, Istituto Nazionale di Geofisica - 

177 

SGA Storia Geofisica Ambiente, printed by Grafica 

Ragno, Ozzano Emilia, BO, Italy, 973 pp. 

Cinti, F., L. Cucci, D. Pantosti, G. D’Addezio and M. 

Meghraoui (1997). A major seismogenic fault in a “silent 

area”: the Castrovillari fault (southern Italy), Geophysical 

Journal International, 130, 595-605. 

CNR (1990). Structural model of Italy, 1:500000, edited by 

Bigi,G., Cosentino, D., Parotto, M., Sartori, R., and 

Scandone, P., Consiglio Nazionale delle Richerche. 

S.E.L.C.A. Florence. 

D’Amico, V., Albarello, D. and Mantovani, E. (1999). A 

distribution-Free Analysis of Magnitude-Intensity 

Relationships: an Application to the Mediterranean 

Region. Phys. Chem. Earth 24, 517-521. 

GNDT-SSN Albarello, D.et al. (2001). New seismic hazard 

maps of the Italian territory. Gruppo Nazionale per la 

Difesa dai Terremoti- Servizio Seismico Nazionale 

Website:http://www.dstn.it/ssn/PROG/2000/carte_pericol 

osita/mcs_e.gif. 

Grandori, G., Drei, A., Perotti, F. and Tagliani, A. (1991). 

Macroseismic intensity versus epicentral distance: the 

case of Central Italy. Tectonophysics 193, 165-171. 

Faenza, L. and Pierdominici, S. (2007). Statistical 

occurrence analysis and spatio-temporal distribution of 

earthquakes in the Apennines (Italy). Tectonophysics 

439, 13-31. 

Marco, S., Stein, M., Agnon, A., and Ron, H. (1996). Longterm 

earthquake clustering: a 50000 year paleoseimic 

record in the Dead Sea Graben. Journal of Geophysical 

Research 101, 6179-6191. 

Michetti, A.M., L. Ferreli, L. Serva and E. Vittori (1997). 

Geological evidence for strong historical earthquakes in 

an “aseismic” region: the Pollino case (Southern Italy), 

Journal of Geodynamics 24, 1-4, 67-86. 

Michetti. A.M., Ferreli, L., Esposito, E., Porfido, S., Blumetti, 

A-M, Vittori, E., Serva, L., and Roberts, G.P. (2000). 

Ground effects during the 9 September 1998, Mw=5.6, 

Lauria earthquake and the seismic potential of the 

“Aseismic” Pollino region in Southern Italy. Seismological 

Research Letters 71, 31-46. 

Palmentola, G., Acquafredda, and Fiore, S. (1990). A new 

correlation of the Glacial Moraines in the Southern 

Apennines, Italy. Geomorphology 3, 1-8. 

Pantosti, D., Schwartz, D.P., and Valensise, G. (1993). 

Paleoseismology along the 1980 surface rupture of the 

Irpinia fault: implications for the Earthquake recurrence in 

the Southern Apennines, Italy. Journal of Geophysical 

Research 98, 6561-6577. 

Papanikolaou, I.D., (2003). Generation of high resolution 

seismic hazard maps through integration of earthquake 

geology, fault mechanics theory and GIS techniques in 

extensional tectonic settings. Unpublished Ph.D thesis, 

University of London, 437pp. 

Papanikolaou, I.D., and Roberts, G.P. (2007). Geometry, 

kinematics and deformation rates along the active normal 

fault system in the southern Apennines: Implications for 

fault growth. Journal of Structural Geology 29, 166-188. 

Roberts, G.P., Cowie, P., Papanikolaou, I., and Michetti, 

A.M., (2004). Fault scaling relationships, deformation 

rates and seismic hazards: An example from the Lazio- 

Abruzzo Apennines, central Italy. Journal of Structural 

Geology 26, 377-398. 

Slejko, D., Peruzza, L., and Rebez, A. (1998). Seismic 

hazard maps of Italy. Annali di Geofisica 41, 183-214. 

Yeats, R.S., and Prentice, C.S. (1996). Introduction to 

special section: Paleoseismology. Journal of Geophysical 

Research 101, 5847-5853. 

Working Group on California Earthquake Probabilities, 

(2002). Earthquake probabilities in the San Francisco bay 

region: 2002-2031. USGS Open-File Report 03-214. 

Wells, D.L., Coppersmith, K., (1994). New Empirical 



Bull. Seismological Society of America 84, 974-1002.


EARTHQUAKE 

ARCHAEOLOGY 



THE SPARTA FAULT, SOUTHERN GREECE: TECTONIC GEOMORPHOLOGY, SEISMIC 

HAZARD MAPPING AND CONDITIONAL PROBABILITIES 

Papanikolau Ioannis D. (1,2), Gerald Roberts (2), Georgios Deligiannakis (1,3), Athina Sakellariou (1), Emmanuel Vassilakis (3) 



(2) AON Benfield UCL Hazard Research Centre, Department of Earth Sciences Birkbeck College and University College 

London, WC 1E 6BT,London UK, Email: (i.papanikolaou@ucl.ac.uk) 

(3) Laboratory of Natural Hazards, Faculty of Geology and Geoenvironment, National and Kapodistrian University of Athens, 

Panepistimioupolis, 15784, Athens, Greece 

Abstract (The Sparta fault, southern Greece: Tectonic geomorphology, seismic hazard mapping and conditional 

probabilities): The Sparta Fault system is a major structure approximately 64km long that bounds the eastern flanks of the 

Taygetos Mountain front (2.400m) and shapes the present-day Sparta basin. This fault is examined and described in terms of its 

geometry, segmentation, drainage pattern and postglacial finite throw, emphasizing also how these parameters vary along strike 

the fault. Based on fault throw-rates and the bedrock geology, geological data can offer both a qualitative and quantitative 

approach of the expected hazard distribution. This is achieved by the construction of a seismic hazard map based on fault throwrates 

that shows the number of times a locality receives enough energy to shake at a certain intensity value, extracting a locality 

specific long-term earthquake recurrence record. The Sparta fault was activated in 464 B.C., devastating the city of Sparta. Since 

no other major earthquake has been generated by this system since 464 B.C., a future event could be imminent. As a result, not 

only time-independent but also time-dependent probabilities, which follow the concept of the seismic cycle, have been calculated 

for the city of Sparta. 

Key words: Lakonia, slip-rates, time-dependent probabilities, Taygetos, 

INTRODUCTION 

In 464 B.C. a large earthquake devastated the city of 

Sparta (~20000 fatalities), causing great social unrest 

(Papazachos and Papazachou, 1997). This event is 

regarded as the oldest well-defined event in the 

Hellenic historical record (Papazachos and 

Papazachou, 1997). However, the area is 

characterized by low seismicity over the last 25 

centuries (Papanastassiou, 1999) and no other major 

event has occurred in the town of Sparta since 464 

B.C, suggesting that a future event could be 

imminent. This is also supported by cosmogenic 

dating techniques applied on the Sparta bedrock 

scarp showing that this fault ruptured repeatedly (six 

times over the last 13kyrs), with time intervals 

ranging from 500-4500yrs (Benedetti et al., 2002). 

This fault is studied in detail based on its postglacial 

scarp, the analysis of the drainage network and the 

major catchments that are influenced by footwall 

uplift. 

most impressive imprint in the topography showing 

sings of recent intense activity. The southern 

segment can be divided into two patches that in the 

past were probably two individual structures that are 

now hard-linked. This fault exhibits an impressive 

postglacial scarp that can be traced for many km 

(Fig.3). In particular, from the village of Anogia up to 

the area of Mystras, it is continuous and has an 8- 

12m high scarp (Fig.4). 

THE SPARTA FAULT SYSTEM 

The Sparta fault system (Fig. 1), bounds the eastern 

part of the Taygetos Mt (2.407m) and shapes the 

western boundary of the Sparta basin (Fig.2). It 

trends NNW-SSE and has a length of 64km. Its 

southern tip is located close to the Gerakari 

catchment approximately 2-3km SW from the 

Potamia village, whereas its northern tip towards the 

Alfios river, a couple of km westwards from the 

Kamaritsa village in the Megalopolis basin. Two 

major faults are traced within this structure (Fig.1). 

The northern segment is about 14 km long and 

characterized by lower slip-rates. On the other hand, 

the southern segment is 50km long and leaves the 

Fig. 1: Simplified geological map of the study area. It shows 

the segmentation pattern as well as the localities of the 

studied catchments profiles. 

178


EARTHQUAKE 

ARCHAEOLOGY 



Fig.2: Distant view of the Sparta fault. It uplifts the Taygetos 

Mt on its footwall and shapes the basin of Sparta towards its 

hanging wall. 

through time. Whittaker et al. (2008) showed that 

rivers with drainage areas greater than 10km 2 and 

crossing faults that have undergone an increase in 

throw rate within the last 1Myrs have significant longprofile 

convexities. They also established that this 

relationship holds for throw rate variation along strike 

the same fault segment, as well as between faults. 

Moreover, Boulton & Whittaker (2009) suggested that 

rivers crossing active faults are undergoing a 

transient response to ongoing tectonic uplift and this 

interpretation is supported by typical signals of 

transience such as gorge formation and hill slope 

rejuvenation within the convex reach. 

Fig.3: View of the post-glacial scarp of the Sparta fault in 

the Kalyvia-Sochas locality. 

Fig.4: Topographic profile perpendicular to the postglacial 

scarp near the village of Anogia. It exhibits an 8.2m high 

scarp. 

TECTONIC GEOMORPHOLOGICAL ANALYSIS; 

CATCHMENTS AND TECTONIC UPLIFT 

The Evrotas river flows through the Sparta Basin, 

parallel to the Sparta Fault, trending NNW-SSE, 

while the secondary branches of this fluvial system 

consist of transient rivers which flow perpendicular to 

the main structure, indicating a strong tectonic 

influence to the drainage pattern. The combination of 

fault parallel and fault perpendicular flow is 

characteristic of active normal faulting settings. 

Kirby & Whipple (2003) demonstrated that 

tectonically unperturbed “equilibrium” fluvial long 

profiles are typically smooth and concave-up. 

However, upland rivers are also sensitive to alongstream 

variations in differential uplift (potentially 

leading to changes in the profile concavity or 

steepness index) and also to changes in uplift rate 

179 

Nine fluvial long profiles of the transient rivers 

crossing different segments of the Sparta fault were 

studied in order to examine the longitudinal convexity 

and its variation along strike. Such profiles were also 

compared to the longitudinal profiles of rivers that are 

not influenced by any fault. Furthermore, in order to 

examine the transience of the streams across the 

Sparta fault, cross sections perpendicular to the river 

flow in the headwaters and within the downstream 

convex reach were analyzed. 

Elevation(m) 

Elevation(m) 

Elevation(m) 

Elevation(m) 

Elevation (m) 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

0 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

0 

1200 

1000 

800 

600 

400 

200 

0 

F3AgiosKonstantinos 

0 2 4 6 8 10 12 

DownstreamDistance(km) 

F4Kastorio 

0 2 4 6 8 10 12 


F7KalyviaSochas 

0 2 4 6 8 10 12 


1200 

1000 

F9Potamia 

800 

600 

400 

200 

0 

1800 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

0 2 4 6 8 10 12 14 16 18 20 

0 


F3Ag.Konstantinos 

F4Kastorio 

F7KalyviaSochas 

F9Potamia 

0 5 10 15 20 25 

Downstream Distance(km) 

Fig.5: a) Long profiles of catchments crossing perpendicular 

the Sparta Fault. Locality names are shown geographically 

in Fig.1. b) Comparison of long profiles in the same graph of 

the above catchments showing significant differences in 

longitudinal convexities along strike the faults. 

a 

b


EARTHQUAKE 

ARCHAEOLOGY 



While river long profile convexities can lead to the 

observation of the ongoing tectonic uplift, differential 

erosion of geological formations may also cause the 

same pattern. Thus, in order to exclude the latter 

phenomena and emphasize the tectonic uplift, the 

rivers to be examined must fulfill some restrictions. 

Whittaker, et al. (2008) suggest that the selected 

rivers should discharge a drainage basin larger than 

10km 2 above fault and the upstream length should be 

at least 5km. These restrictions are applicable in the 

south and central segments of Sparta fault. However, 

in the northern segment of Sparta fault the streams 

are not long enough to fulfill these criteria, due to the 

proximity of the watershed of Evrotas Basin to the 

SW segment of the fault. 

Catchments crossing the North, Central and South 

parts of the Sparta Fault were grouped and studied 

separately. Qualitative analysis of long profiles 

showed a significant difference in longitudinal 

convexity between the Central and both the South 

and North parts of the fault, leading to the conclusion 

of varying uplift rate along strike. A convex reach of 

205m height in Potamia catchment long profile 

(southeast part of the Sparta Fault) can be observed 

although it seems to have propagated upstream in 

relation to the fault. This could happen as the 

channel successively adjusts to the imposed uplift 

field (Whipple & Tucker, 2002). Moreover, Kalyvia 

Sochas fan deposits were extensively examined by 

Pope et al. (2003). Kalyvia Sochas catchment long 

profile revealed a convex reach of 246m height, 

which is in contact with the Sparta fault, in contrast to 

Potamia catchment’s convex reach. Agios 

Konstantinos catchment seems to have a concaveup 

channel profile, possibly indicating a constant slip 

rate (Whittaker, et al. 2008). On the other hand, 

towards the central part of the Sparta fault, were no 

fan deposits and talus cones appear and the finite 

throw is smaller, the Kastorio catchment’s convex 

reach height outreaches 590m, as measured from 

the fault (Fig. 5a,5b). 

In addition, the normalized steepness index, k sn , 

using a reference concavity of 0.45, was calculated 

for six catchments crossing all Sparta fault parts, as 

well as for the two catchments crossing the antithetic 

structure and two catchments that cross no fault 

(localities shown in Figure 1). The k sn rates for the 

catchments closer to the tips of the Sparta fault (F3- 

Agios Konstantinos and F9-Potamia) were 81 and 

82.7 respectively, while in the central part the 

steepness rates are higher and vary from 98.5 to 114 

(98.5


EARTHQUAKE 

ARCHAEOLOGY 



town of Sparta. A considerably higher timedependent 

probability of 2,14% over the next 30 

years and 3,55% over the next 50 years has been 

calculated. The time dependent probability follows 

the seismic cycle concept (WGCEP 2002) and 

exhibits higher values because the elapsed time 

since the last event (2475yrs) has exceeded the 

mean recurrence interval (1792±458yrs). However, 

due to the irregularity of earthquake time intervals of 

the Sparta fault (Benedetti et al. 2002) and the 

introduction of a high sigma value (=0.64) this 

difference is noteworthy, but not substantial. 

Fig.6: Seismic hazard map for the Sparta Basin, showing 

how many times a locality receives enough energy to shake 

at intensities IX in 15±3kyrs, after considering the bedrock 

geology and assuming a circular pattern of energy release, 

with a 18 km radius of isoseismal IX. 

Fig. 7: Diagram showing the probability density for the town 

of Sparta immediately after the 464 B.C. event and the 

conditional probability density considering that no event 

occurred since 2011. A sigma value of 0.64 is used. 

Acknowledgements: The Institution of State Scholarships 

of Greece is thanked for support. 

References 

Benedetti, L, et al. (2002). Post-glacial slip history of the 

Sparta fault (Greece) determined by 36 Cl cosmogenic 

dating: Evidence for non-periodic earthquakes. Geophys. 

Res. Letters 29, 10.1029/2001GL014510. 

Boulton, S.J., Whittaker, A.C., (2009). Quantifying the slip 

rates, spatial distribution and evolution of active normal 

faults from geomorphic analysis: Field examples from an 

oblique-extensional graben, southern Turkey. 


Kirby, E., Whipple K.X., (2003). Distribution of active rock 

uplift along the eastern margin of the Tibetan Plateau: 

Inferences from bedrock channel longitudinal profiles. 

Journal of Geophysical Research, 108(B4/2217), pp.1-24 

Papaioannou, Ch., (1984). Attenuation of seismic intensities 

and seismic hazard in the area of Greece. Ph.D. Thesis, 

University of Thessaloniki, 200pp. 

Papanastassiou, D. (1999). Seismic hazard assessment in 

the area of Mystras-Sparta, south Peloponnesus, 

Greece, based on local seismotectonic, seismic, geologic 

information and on different models of rupture 

propagation. Natural Hazards 18, 237-251. 

Papanikolaou, I.D., (2003). Generation of high resolution 

seismic hazard maps through integration of earthquake 

geology, fault mechanics theory and GIS techniques in 

extensional tectonic settings. Unpublished Ph.D thesis, 

University of London, 437pp. 

Papazachos, B. C., Papaioannou, (1997). The 

macroseismic field of the Balkan area. Journal of 

Seismology 1, 181-201. 

Papazachos B. C. and Papazachou C., (1997). The 

earthquakes of Greece. Thessaloniki: Ziti Publications. 

Pope, R.J., Wilkinson, K.N., and Millington, A.C. (2003). 

Human and climatic impact on Late Quaternary 

deposition in the Sparta basin piedmont: Evidence from 

alluvial fan systems. Geoarcheology: An International 

Journal 18, 685-724. 

Psonis, K., Latsoudas, C., (1983). 1:50000 Geological Map 

“Xirokampion”, IGME. 

Psonis, K., (1990). 1:50000 Geological Map “Sparti”, IGME. 

Roberts, G.P., Cowie, P., Papanikolaou, I., and Michetti, 

A.M., (2004). Fault scaling relationships, deformation 

rates and seismic hazards: An example from the Lazio- 

Abruzzo Apennines, central Italy. Journal of Structural 

Geology 26, 377-398. 

Theodoulidis, N.P. (1991). Contribution to the study of 

strong motion in Greece. Ph.D. Thesis, University of 

Thessaloniki, 500pp. 

Whipple, K.X., Tucker, G.E., (2002). Implications of 

sediment-flux-dependent river incision models for 

landscape evolution. Journal of Geophysical Research, 

107(B2), pp.1-20. 

Whittaker, A.C., Attal, M., Cowie, P.A., Tucker, G.E. & 

Roberts, G., (2008). Decoding temporal and spatial 

patterns of fault uplift using transient river long profiles. 


Working Group on California Earthquake Probabilities, 

(2002). Earthquake probabilities in the San Francisco bay 

region: 2002-2031. USGS Open-File Report 03-214. 

Wells, D.L., Coppersmith, K. (1994). New Empirical 



Bull. Seismological Society of America 84, 974-1002. 

Zindros, G., Exindavelonis, P. (2002). 1:50000 Geological 

Map “Kollinae”, IGME. 

181


EARTHQUAKE 

ARCHAEOLOGY 



ACTIVE FAULTING TOWARDS THE EASTERN TIP OF THE CORINTH CANAL: STUDIED 

THROUGH SURFACE OBSERVATIONS, BOREHOLE DATA AND 

PALEOENVIRONMENTAL INTERPRETATIONS 

Papanikolau Ioannis D. (1, 4, Maria Triantaphyllou (2), Aggelos Pallikarakis (1,3), Georgios Migiros (1) 



(2) Hist. Geology-Paleontology Department, Faculty of Geology and Geoenvironment, National and Kapodistrian University of 

Athens, Panepistimiopolis 15784, Athens, Greece 

(3) Laboratory of Natural Hazards, Faculty of Geology and Geoenvironment, National and Kapodistrian University of Athens, 

Panepistimioupolis, 15784, Athens, Greece 

(4) AON Benfield UCL Hazard Research Centre, Department of Earth Sciences, University College London, WC 1E 6BT, London 

UK, Email: i.papanikolaou@ucl.ac.uk 

Abstract (Active faulting towards the Eastern tip of the Corinth Canal: Studied through surface observations, borehole 

data and paleoenvironmental interpretations): The most important active fault that crosses the eastern tip of the Corinth Canal 

is studied in detail, involving surface observations, borehole data and peleoenvironmental interpretations. This fault intersects 

and/or is parallel and at short distances with major infrastructure facilities such as the Athens-Corinth highway, the railway and the 

Corinth Canal. A positive remark referring to the seismic hazard assessment is that this fault has a limited length (~5km) and thus 

it can not produce extensive and large displacement (>20cm) from primary surfaces ruptures. Moreover, borehole data and 

correlation between the footwall and hangingwall horizons, show no significant offset over the last 200ka, confirming that no 

significant displacements have been accumulated and/or has a very low slip-rate. Paleoenvironmental interpretations based on 

borehole data show a very complex sedimentation history during the Upper Pleistocene that involves subaerial exposure, 

backshore, paralic, lagoonal, shallow marine environments and possibly even some lake sediments. 

Key words: Isthmus, Saronic Gulf, seismic hazards, Isthmia - Kalamaki 

INTRODUCTION 

The Gulf of Corinth is one of the fastest extending 

regions worldwide with up to 20mm/yr that diminishes 

to 8 and 4 mm/yr towards its eastern end (Billiris et 

al. 1991; Briole et al. 2000). Our study area lies 

eastwards from the major active faults that shape the 

present-day eastern part of the Corinth Gulf, towards 

the Corinth Isthmus. In particular, it lies towards the 

eastern tip of the Corinth Canal (Fig.1). The area of 

the Isthmus divides the Corinth and Saronic Gulfs. 

The Corinth Canal is a major infrastructure project 

6.3km long, 8m deep and with up to 60m high slopes 

that often experience landslides (Marinos and 

Tsiambaos 2008, Papantoniou et al. 2008). 

The Canal was constructed 120 years ago and offers 

a unique opportunity to visualise the faults, forming 

an impressive mega-trench. More than 40 faults can 

be recognised, most of them normal and oblique 

normal (Freyberg 1973). Some of them are overlain 

by undeformed late Pleistocene strata and are clearly 

not a hazard threat. The majority of them do not 

displace the topography and are of limited length, 

therefore are considered as secondary structures 

with low slip-rates. However, towards the eastern tip 

of the Canal we traced a significant ENE-WSW 

trending fault that downthrows towards the SSE and 

appears to displace the modern ground surface. The 

hazard potential of this fault is studied in detail based 

on surface observations and the analysis of borehole 

data and discussed in this paper. 

182 

Fig.1: Digital elevation map showing the major and 

secondary faults in the study area (modified from Bornovas 

et al. 1972, Gaitanakis et al. 1985, Papanikolaou et al. 

1989, Papanikolaou et al. 1996, Roberts et al. 2009). 

FAULT MAPPING AND SURFACE 

OBSERVATIONS 

This fault lies in the hangingwall of the Kechries and 

Loutraki faults and in the footwall of the South 

Alkyonides fault that was activated during the 1981 

earthquake sequence. Our 1:5000 mapping shows 

that the study fault is extended for at least 1km 

southwards and 4km northwards of the Canal (Fig.1, 

Fig.2). This fault controls the topography of the area, 

producing 150m of topographic variation towards its


EARTHQUAKE 

ARCHAEOLOGY 



characteristics we suggest that the fault activity has 

progressively migrated eastwards towards our 

studied fault, which forms the predominant present 

day structure that influences the topography. 

Moreover, we identified some secondary active fault 

planes of the studied fault in the immediate 

hangingwall that offset the entire sedimentary column 

(Fig.2, Fig.4). These secondary fault planes 

outcropping in the immediate hangingwall, extend the 

width of the fault zone up to 50m. 

Fig.2 Simplified geological map showing the studied fault. A 

detailed map showing the main and secondary fault planes 

and the boreholes near the Canal crossing. 

Fig.4: View of the secondary fault plane of the principal fault 

on the southern part of the Canal. It also created a landslide 

in the past. 

Fig. 3: View of the main fault plane about 50m northwards 

the Canal (right). Close up view of the free face and the 

striations developed on the fault plane (left). This is a high 

angle normal fault dipping 65 o towards the SSE. 

centre, between the footwall and the hangingwall, 

and bounds recent alluvial sediments. Moreover, it 

offsets Upper Pleistocene sediments, exhibits a clear 

fault plane and a free face with striations (Fig.3), 

indicating recent activation. It strikes at 075 o -255 o 

(ENE-WSW trending) and is a high angle normal fault 

dipping at 65 o towards the SSE. Striations measured 

on the fault plane are plunging at 55 o towards the SE 

(145 o ), confirming that it is an almost pure normal 

fault. Strata exposed on the immediate footwall are 

Upper Pleistocene sediments, dipping about 10-20 o 

to the NNW, indicating also that this tilt is due to the 

recent fault activity. 

The fault plane has also been traced about 650m 

southwards the Canal where it trends at 070 o -250 o 

and dips 70 o towards the SSE (Fig.5). The crossing 

of the Isthmus and its intersection with the fault, is 

characterized by the presence of semicohesive 

Upper Pleistocene sediments of variable grain sizes 

(from cm thick pebbles to clay) with an abundance of 

ostracods that indicate a recent brackish marine 

shallow water environment (e.g. Krstic & Dermitzakis, 

1981). In addition, to the NW of the main fault plane 

and over a 350m distance we identified several fault 

planes parallel to the main fault trace. Based on their 

183 

Fig. 5: View of the fault about 650m southwards the Canal. 

Arrows point to the fault plane displayed in the photo 

towards the right. It trends 070 o -250 o and dips 70 o towards 

the SSE. 

Fig. 6: Location map showing the levelling route, the 

benchmarks and their post-1981 displacements (redrawn 

from Mariolakos and Stiros 1987). Benchmark 77 the only 

one that subsided by 4cm, lies at the immediate hangingwall 

of the fault and benchmark 79 that was uplifted by 2cm on its 

immediate footwall, implying that our studied fault may have 

been passively ruptured during the 1981 earthquake.


EARTHQUAKE 

ARCHAEOLOGY 



It is interesting to note that the area of the Canal 

uplifted by approximately 2cm from the 1981 

earthquake based on a levelling campaign 

(Mariolakos and Stiros 1987) except for one 

benchmark that subsided by 4cm and lies at the 

immediate hangingwall of our studied fault (Fig. 6). 

The latter implies that our studied fault may have 

been passively ruptured during the 1981 earthquake. 

PALEOENVIRONMENTAL INTERPRETATIONS 

Borehole data 

Seven boreholes have been recovered on the 

easterm side of the Corinth Canal; three of them at 

the hanging wall up to 70 m deep and four at the 

footwall up to 57 m deep (Figure 2). Two of the 

longest boreholes were selected for detailed 

micropaleontological-paleoenvironmental analyses, 

while the rest were maintained for geotechnical 

analyses. Approximately 100 samples were collected 

from these boreholes (45 from borehole Bh-3 and 57 

from borehole Bh-7). Each sample (20 gr dry weight) 

was treated with H 2 O 2 to remove the organic matter, 

and subsequently was washed through a 125m 

sieve and dried at 60 o C. A subset of each sample 

was obtained using an Otto microsplitter to obtain 

aliquots of at least 200 benthic foraminifers. The 

microfauna were identified under a Leica APO S8 

stereoscope. A scanning electron microscope 

analysis (SEM Jeol JSM 6360, Dept. of Hist. 

Geology-Paleontology) was used for taxonomical 

purposes. 

Lithostratigraphic units 

In each borehole lithological alterations of sand, clay, 

clayey sand, conglomerate, marl, fractions of 

limestone, even topsoil have been recognized. It is 

important that among these layers no significant 

correlation was found except for a conglomerate 

horizon that lies on top of a coral colony. The latter 

strongly indicates that there are major lateral 

alterations and stratigraphic variations. Between the 

two boreholes we identified a layer that can be easily 

correlated in both of them. This is a conglomerate 

layer 5.5 m thick in Bh-3 and 3 m thick in Bh-7. 

Micropaleontological analysis 

Samples from Bh-3 and Bh-7 were analyzed in order 

to identify foraminiferal assemblages (Figure 7). The 

studied benthic foraminiferal assemblages are 

relatively abundant and moderately preserved. The 

identified foraminiferal species were grouped in 

euryhaline forms mainly represented by Ammonia 

spp. large and small sized, Elphidium spp., 

Haynesina spp., Aubignyna perlucida and marine 

foraminiferal indicators, mostly including miliolids, 

and in a lesser degree full-marine species grouping 

Asterigerinata, Neoconorbina, Rosalina spp., 

Planorbulina mediterranensis. A series of different 

depositional environments were recognized through 

micropaleontological analysis of the benthic 

foraminiferal fauna (e.g. Triantaphyllou et al., 2003; 

Goiran et al., 2011) in both boreholes. In Bh-3 and in 

Bh-7 we found approximately the same pattern of 

alternations, but with slight differences. 

184 

In Bh- 3 the microfauna indicates a shallow marine 

environment within the upper 10 m thick layer, 

partially influenced by fresh water, that turns to an 

open lagoon. Beneath this layer, the next 6 m were 

described as a paralic or even a backshore 

environment, with no or few broken foraminiferal 

specimens. The underlying layers were characterized 

as shallow marine to open lagoon environment, 

indicating that sea level had changed once more. In 

these sediments apart from foraminiferal fauna, a 

colony of Cladocora corals was found. This layer is 

approximately 6 m thick. Beneath these sediments 

the environment changes to paralic-backshore for the 

next 9 m and beneath them, again a 6 m thick layer 

is described as shallow marine. Down to the base of 

Bh-3 at 70 m depth, we encounter two more 

alternations of the depositional environment from 

shallow marine to coastal. 

At the top layers of Bh-7 the microfauna indicates a 

shallow marine to paralic environment. These 

sediments were 4 m thick while beneath them we 

encountered a 3 m thick layer with no microfauna at 

all; indicating the presence of a paralic or backshore 

environment. Below this layer we determine a 6 m 

thick layer described as shallow marine/paralic. The 

micropaleontological analysis provides clear 

evidence for the presence of a closed lagoon in the 

underlying 1 m thick layer. In the following 10 m we 

encountered at least 4 alternations between the 

depositional environment, from shallow marine to 

closed lagoon and back to shallow marine-paralic 

conditions. At approx. 23 m depth the environment 

changed again to closed lagoon, that followed once 

more by shallow marine with fresh water input, that 

turned to a closed lagoon for a short interval. Below 

this point no microfauna was found indicating a 

paralic/backshore environment. It is significant that at 

18m depth we found a colony of Cladocora corals, 

occurring approximately at the same height as in Bh- 

3. The beginning of fault’s deformation zone was 

observed at ca. 33m depth and ended at 

approximately 45m depth. In the following 12m, 

dense alternations of cataclastites and fragments of 

limestones occur. From that point until the end of Bh- 

7 at 56 m depth, no foraminifera specimens are 

traced, indicating the lack of full marine conditions. 

The presence of ostracods (Cyprideis spp.) at this 

level implies a brackish –oligohaline 

paleoenvironment. 

Presence of corals and other shallow marine 

foraminifera shows that marine sediments were 

deposited at glacioeustatic sea level highstands. 

Thus, we try to correlate these with the global 

glacioeustatic sea level curve and the known uplift 

rate from neighbouring corals (0.3mm/yr uplift rate 

from 200ka corals in the centre of the Canal about 2 

km westwards from our locality based on Collier et 

al. 1992). Within our boreholes we have a succession 

of lowstand and highstand deposits. Lowstand 

deposits such as subaerial deposits and top soil, are 

marking unconforminites with marine highstand 

deposits. Relative dates for these lowstand and 

highstand deposits can be derived from the sea level


EARTHQUAKE 

ARCHAEOLOGY 



curve. Therefore, it is clear from our records that the 

marine to subaerial transitions form when the location 

emerges above sea level either due to uplift and/or 

due to falling global sea level. As a result, marine 

sediments are expected at 125ka for several 

thousand years, then possibly during a short period 

at 175ka and then at a prolong period at 200ka 

(Fig.8). 

Fig.7: View of the stratigraphic columns and the 

depositional environment of the two boreholes (Bh3 located 

on the immediate footwall and Bh7 located on the 

immediate hangingwall of the studied fault). 

Fig.8: Sea level curve from Sidall et al. 2003 and the 

expected depositional environment based on the 0.3mm/yr 

uplift rate from the neighbouring dated corals of Collier et al. 

(1992). 

DISCUSSION- CONCLUSIONS 

We have studied a normal fault that trends ENE- 

WSW and downthrows towards the SSE. The current 

study confirms that the studied fault is active and 

forms a secondary structure that accommodates 

displacement between major E-W trending faults. 

This fault crosses an area, where major infrastructure 

facilities are based, such as the Corinth Canal, the 

highway and the railway. A positive remark referring 

to the seismic hazard assessment and design is that 

this fault has a limited length (~5km) and thus it can 

not produce extensive primary surfaces ruptures. 

Primary surface ruptures from this fault are not 

expected to exceed 20cm of displacement. In 

addition, correlation of the same horizons based on 

borehole data from the footwall and hanging wall of 

the fault, show an offset of 4±2m of the corals that 

based on the expected depositional environment are 

probably 200kyrs old. Therefore, no significant 

displacements have been accumulated over the last 

200kyrs, implying that it is a very low slip-rate fault. 

Finally, borehole data exhibit a complex 

paleoenvironment with major lateral and temporal 

alterations and stratigraphic variations as a result of 

the interplay between the sea level fluctuations and 

the tectonic activity that resembles to the nearby 

Perachora peninsula (Roberts et al. 2009). 

References 

Billiris, H., et al. (1991), Geodetic determination of tectonic 

deformation in central Greece from 1900 to 1988. Nature, 

350, 124– 129, doi:10.1038/350124a0. 

Bornovas, J., Lalechos, N. and Filipakis, N. (1972). 

1:50.000 scale geological map, Sheet “Korinthos”. 

Institute of Geology and Mineral Exploration 

Briole, P., A. Rigo, H. Lyon-Caen, J. C. Ruegg, K. 

Papazissi, C. Mitsakaki, A. Balodimou, G. Veis, D. 

Hatzfeld, and A. Deschamps, Active deformation of the 

Corinth rift, Greece: Results from repeated Global 

Positioning System surveys between 1990 and 1995, J. 

Geophys. Res., Solid Earth, 21, 25,605– 25,625, 2000. 

Collier, R.E.L., Leeder, R.M., Rowe, P. and Atkinson, T. 

(1992). Rates of tectonic uplift in the Corinth and Megara 

basins, Central Greece, Tectonics, 1159-1167. 

Gaitanakis, P., Mettos, A., and Fytikas, M. (1985). 1:50.000 

scale geological map, Sheet “Sofikon”. Institute of 

Geology and Mineral Exploration. 

Goiran, J.P., Pavlopoulos, K., Fouache, E., Triantaphyllou, 

M.V., Etienne, R., 2011. Piraeus, the ancient island of 

Athens: Evidence from Holocene sediments and 

historical archives. Geology, doi:10.1130/G31818.1. 

Freyberg, V. (1973). Geologie des Isthmus von Korinth, 

Erlangen Geologische Ablhandlungen, Heft 95, Junge 

und Sohn, Universitats Buchdruckerei Erlangen, 183pp. 

Krstic, N., Dermitzakis, M.D., (1981). Pleistocene Fauna 

from a section in the channel of Corinth (Greece). Ann. 

Geol. du Pays Hellen. 30 (2): 473-499. 

Mariolakos, I. and Stiros, S.C. (1987). Quaternary 

deformation of the Isthmus and Gulf of Corinth (Greece). 

Geology 15, 225-228. 

Marinos, P. and Tsiambaos, G. (2008). The Geotechnics of 

Corinth Canal: A review. Bulletin of the Geological 

Society of Greece XXXXI/I, 7-15. 

Papanikolaou, D., Chronis, G., Lykousis, V., Pavlakis, P., 

Roussakis, G., and Syskakis, D. (1989). 1:100.000 scale, 

Offshore Neotectonic Map the Saronic Gulf. Earthquake 

Planning and Protection Organization, National Centre for 

Marine Research, University of Athens. 

Papanikolaou, D., Logos, E., Lozios, S. and Sideris, Ch. 

(1996). 1:100.000 Neotectonic Map of Korinthos Sheet, 

Earthquake Planning and Protection Organization, 

Athens. 

Papantoniou, L., Rozos, D., Migiros, G. (2008). Engineering 

geological conditions and slope failures along the Corinth 

Canal. Bull. Geol. Soc. Greece, Vol. XXXVI/I, 17-24. 

Roberts, G. P., S. L. Houghton, C. Underwood, I. 

Papanikolaou, P. A. Cowie, P. van Calsteren, T. Wigley, 

F. J. Cooper, and J. M. McArthur (2009). Localization of 

Quaternary slip rates in an active rift in 10 5 

years: An 

example from central Greece constrained by 234U-230Th 

coral dates from uplifted paleoshorelines, Journal of 

Geophys. Res. 114, B10406, doi:10.1029/2008JB0058 

Triantaphyllou, M.V., Pavlopoulos, K., Tsourou, Th. & 

Dermitzakis M.D., 2003. Brackish marsh benthic 

microfauna and paleoenvironmental changes during the 

last 6.000 years on the coastal plain of Marathon (SE 

Greece). Rivista Italiana Paleontologia et Stratigafia, 109 

(3), 539-547. 

Siddall, M., E. J. Rohling, A. Almogi-Labin, C. Hemleben, D. 

Meischner, Schmelzer, and D. A. Smeed (2003), Sealevel 

fluctuations during the last glacial cycle. Nature 423, 

853 – 858. 

185


EARTHQUAKE 

ARCHAEOLOGY 



ROMAN AQUEDUCTS AS INDICATORS OF HISTORICALLY ACTIVE FAULTS 

IN THE MEDITERRANEAN BASIN 

Passchier, Cees W. (1), Gilbert Wiplinger (2), Gül Sürmelihindi (1), Paul Kessener (3), Talip Güngör (4) 

(1) Department of Earth Sciences, Gutenberg University , 55099 Mainz, Germany. Email: cpasschi@uni-mainz.de, 

surmelih@uni-mainz.de 

(2) Österreichisches Archäologisches Institut, Vienna, Austria. Email: gilbert.wiplinger@gmail.com 

(3) van Slichtenhorststraat 13, NL 6524 JH Nijmegen, Netherlands. Email: lenl@euronet.nl 

(4) Department of Geological Engineering, Dokuz Eylül University, Izmir, Turkey. Email: talip.gungor@deu.edu.tr 

Abstract (Roman aqueducts as indicators of historically active faults in the Mediterranean Basin): Roman aqueducts 

are a potential source of information on infrequently active faults in the Mediterranean Basin. We have localised more than 

1300 such aqueducts from the literature. Carbonate deposits in aqueduct channels can help to date and characterise fault 

activity in detail. As an example, the Deirmendere aqueduct near Ephesos, Turkey is described, which was apparently 

displaced and tilted by up to 2 metres a short time after it came into use. The aqueduct was subsequently repaired by 

constructing a second channel at the correct elevation running next to the original one over a length of several km. 

Key words: roman aqueduct, sinter, carbonate deposits, fault, fault scarp 

INTRODUCTION 

Major fault zones such as the San Andreas and 

North Anatolian fault systems have been extensively 

studied, and the local risk of earthquakes and their 

frequency is understood to some extend. Smaller 

faults, however, which may produce strong 

earthquakes at very long time intervals, may go 

undetected and may pose a similar or greater risk, as 

local building regulations are not adapted. Examples 

are the Basel earthquake of 1356 (Lambert et al. 

2005) and the Lisbon earthquake of 1755 (Mendez- 

Victor 2008). For a better understanding of 

earthquake risks and neotectonic processes, it is 

useful to determine which faults show this kind of 

long-term activity. 

Several attempts have been made in the past to 

localize infrequently active faults mainly by the study 

of historical records and paleoseismic events. Here, 

Roman aqueducts 1 are presented as a new tool for 

earthquake studies to enhance our knowledge of 

infrequently active faults in the Mediterranean Basin. 

ANCIENT AQUEDUCTS 

Ancient aqueducts are gravity driven water transport 

systems consisting of roofed masonry channels or 

piped conduits running from perannial springs to 

cities, villages or farms (Fig. 1,2). Meandering along 

the contour lines of the terrain, they gradually 

descend from source to destination (Hodge 1992, 

Wikander 2000, Grewe 1986). Aqueducts that served 

large cities frequently acquired considerable lengths, 

the 250 km Constantinople aqueduct at the top of the 

list (Grewe 1986, Hodge 1992, Çeçen 1996). 

1 We use the term “aqueduct” to represent the entire water 

transporting system of channel, piped conduit, intermediate 

basins, tunnels and substructures, not just the bridges. 

186 

Fig. 1. Presently known roman aqueducts in central 

Italy, showing the dense network of long channels, 

several of which cross major faults. ROMAQ Database. 

Aqueducts can be valuable for the study of 

historically active faults for several reasons: 

(1) they are long, anastomosing objects that 

cover a large area, and are much more likely to 

be cut by an active fault than any other structure; 

(2) they comprise bridges, inverted siphons and 

tunnels which are vulnerable to earthquake 

damage; (3) they may serve as an earthquake 

recording device for over 2000 years, i.e. from the 

construction date to present day; (4) they 

generally slope down with a minor and rather 

constant gradient allowing a detailed 

reconstruction of the uplift pattern produced by 

movement on faults; (5) since most aqueducts 

are linked with cities where archaeological work is 

in progress or has been done, the building and 

operation of the aqueduct can usually be dated


EARTHQUAKE 

ARCHAEOLOGY 



with an accuracy of decades or less; (6) many 

aqueducts contain carbonate deposits (Fig. 2) which 

record seasonal and abrupt changes in water 

chemistry, temperature and depth. Even if an 

aqueduct is not directly cut by an active fault, a 

seismic event can leave traces in the carbonate 

deposits that can be dated. 

Ancient aqueducts are found on all continents, 

including the Aztec and Inca Empires of South 

America (Wright et al. 2006), but the largest 

concentration is in the Mediterranean Basin. Most are 

of Roman Imperial age (50-300 AD) when hundreds 

of aqueducts were built within a short period of time, 

often close together (Fig. 1)(Bedon 1997; Hodge 

1992; Wikander 2000). More than 1300 Greek and 

Roman aqueducts are known from the literature. 

A number of aqueducts are known to have been 

damaged by earthquakes. An aqueduct near 

Venafro (Italy) was apparently damaged by three 

major earthquakes on what seemed to be a minor 

fault (Galli and Naso 2009; Galli et al. 2010); the 

major 1349 earthquake that damaged many cities 

in central Italy, known from medieval records, can 

now be traced to this fault (Galli et al. 2010). The 

aqueduct of Nimes (France) was damaged by an 

earthquake in the third or fourth century in a zone 

that was thought to be free of recent tectonic 

activity (Levret et al. 2008; Volant et al. 2009). 

Marra et al. (2004) reported on a small aqueduct 

near Rome that was deformed by an earthquake. 

Damage to the aqueducts of Petra (Jordan) could 

be attributed to an earthquake of AD 363 

(Bellwald et al. 2003). The aqueduct at Aphamea 

(Syria) was destroyed by earthquakes in 526 and 

528 AD and subsequently repaired (Balty 1987). 

At Al Hativ (Syria) the Dead Sea fault displaced 

an aqueduct by up to 13 metres in several 

seismic events (Meghraoui et al. 2003; Sbeinati et 

al. 2010). 

Fig. 2. Cross section of the roman aqueduct channel 

of Cologne, showing typical elements. The buried 

masonry channel is covered on the inside by a red 

layer of waterproof cement (“opus signinum”). 

Carbonate on the walls tapers upwards because of 

variable and gradually rising water level. In the 

Cologne case, the deposits indicate that the aqueduct 

ran for approximately 170 years (Fig. 3). 

187 

Fig. 3. Sinter from the Cologne aqueduct. Note typical 

laminated deposition. Height of sample 12 cm. 

THE ROLE OF CARBONATE DEPOSITS 

Many aqueducts were damaged by earthquakes 

during or after their active life, but it is usually 

hard to date these events, or to understand how 

many times the channels were damaged and 

repaired (cf. Meghraoui et al. 2003; Sbeinati et al. 

2010). In such cases, carbonate deposits in 

roman aqueducts may give clues. 

Many aqueducts carried water from karstic 

aquafers containing carbonates, which deposited 

on the walls of the channel. (Fig. 3)(Grewe 1986, 

Guendon & Vaudoir 2000, Garczynski et al. 2005, 

Guendon & Leveau 2005, Dubar 2006). This 

travertine like material, also known as sinter,


EARTHQUAKE 

ARCHAEOLOGY 



could, over time, acquire considerable thickness 

corresponding to the period that the aqueduct was 

running, from decades to as long as 800 years 

(Sbeinati et al. 2010). Earthquakes have interfered 

with these deposits in several ways. 

1 – Fault motion when an aqueduct is running. 

In the most dramatic case, the channel is cut by the 

fault and water flow is interrupted. Even if the 

channel was not destroyed, fracturing of the walls 

and sinter deposits could occur, with subsequent 

growth of new sinter over the fractures (Carbon et al. 

2005). In this way, the seismic event can be precisely 

localised in the sinter stratigraphy. Careful 

examination of the damaged site can reveal much 

about earthquake displacement and dating 

(Meghraoui et al. 2003; Volant et al. 2009; Galli et al. 

2010; Carbon et al. 2005). Volant et al. (2009) found 

deformation twins in calcite in sinter of the Nimes 

aqueduct, which they attribute to earthquake 

damage. 

Seismic events may also leave marks in sinter away 

from the damaged site, or where the active fault did 

not cross the aqueduct. Local damage to the channel 

or to bridges can lead to water loss and thinner 

deposits downstream after the earthquake (Levret et 

al. 2008; Carbon et al. 2005). In case a section of 

aqueduct was ruptured and subsequently repaired, 

sinter deposits in this section will differ from those 

upstream from the damaged site. Fragments of 

building material and speleothems fallen into the 

channel, and clastic sediments may be covered by 

sinter. Commonly, a new layer of waterproof cement 

is put on top of the deposits when repairing the 

channel. If the rupture site is lost due to later 

destruction or erosion, deposits downstream from the 

site may still differ from those upstream, and a 

relative dating of the earthquake can thus be 

established. A further item to be explored is the 

possibility that an earthquake may temporarily or 

permanently cause changes in the chemical 

composition of the water due to alterations in the 

cave aquifer system that fed the aqueduct. 

2 – Fault motion after the aqueduct has stopped 

running. 

If faults cut an aqueduct after it ceased to 

operate, dating of the event is possible by 

classical trenching and dating techniques used for 

active faults (Sbeinati et al. 2010). In this case, 

the aqueduct serves to give a minimum age for 

fault slip, and careful geodetic survey of the 

deformed channel can give information about 

vertical ground motion on both sides of the fault. 

THE DEIRMENDERE AQUEDUCT OF 

EPHESOS 

In recent years Gilbert Wiplinger of the 

Österreichisches Archaeologisches Institut (ÖAI) 

and an international team of researchers 

investigated the Deirmendere aqueduct of 

Ephesos, Turkey. This aqueduct, one of seven 

that served the ancient city, took its waters from a 

spring along a nappe contact to the SE (Fig. 4; 

Wiplinger 2008). The 37,5 km aqueduct was built 

as a masonry vaulted channel 80 cm wide and 

over 240 cm high and was equipped with tunnels 

and bridges. For a stretch of over 10 km, two 

channels appeared to run parallel to each other 

(Figs. 4,5). It was found that the channels were 

not built simultaneously but one some years after 

the other, the younger channel being partly built 

on top of the older one, so that both channels 

cannot have carried water at the same time. 

Halfway this 10 km stretch the aqueduct crosses 

a valley by means of the two story Bahçecikboaz 

bridge, where the level of both channels coincide. 

However, at the upstream end the younger 

channel’s floor is positioned 2 meters above the 

older one, while at the downstream end, close to 

where an extended tunnel section starts, it is 2 

meters below it. (Wiplinger 2008, Figs. 

4,5). 

Fig. 4. Double channel of the Deirmendere aqueduct 

looking north. The lower channel is the older one and 

was partly demolished to build the upper channel on the 

right. Location shown in Figure 5. 

188 

Fig. 5. a - Schematic geological map of the area south 

of Ephesos. The double channel section lies in the 

central marble unit; b - schematic height profile of the 

Deirmendere aqueduct showing the section with two 

channels. Geology after Çetnkaplan (2002).


EARTHQUAKE 

ARCHAEOLOGY 



Initially it was thought that survey errors by the 

Roman engineers would explain this feature 

(Wiplinger 2008). However, recent work has shown 

that the earlier channel has a layer of sinter several 

mm wide, which indicates that it must have carried 

water for several years. This makes survey errors 

unlikely. The younger channel has carbonate 

deposits up to 30 cm wide accounting for many 

decades of running water up to the final stage of the 

aqueduct. Geological reconaissance early 2011 has 

localised a major fault scarp near the upstream 

starting point of the twin section. Whether the 

northern, downstream end of the twin section also 

coincides with a fault is currently under investigation. 

We presently think of the following scenario. 

The Deirmendere aqueduct was cut by at least one 

major fault after it has been in operation for only a 

few years, probably in the second half of the second 

century AD. The fault produced a scarp of 

approximately 2 metres high, interrupting the water 

flow and altering the slope of the aqueduct over a 

stretch of 10 km (Fig. 5). To restore the aqueduct, a 

new channel parallel to the disrupted one had to be 

built at the proper level. The aqueduct channel and 

the geology surrounding the Deirmendere aqueduct 

will be further investigated in 2011. 

Acknowledgements: The authors thank the Geocycles 

Cluster and the University of Mainz for financial support. 

The Österreichisches Archaeologisches Institut provided 

crucial logistic and general support. 

References 

Balty, J. C., (1987). Problèmes de l'eau a Apamée de Syrie. 

In: Les Travaux de la Maison de l'Orient. L'Homme et 

l'eau en Mediterranée et au Proche Orient. (Louis, P, 

Métral, F., Métral , J. eds). 4, 9-23. 

Bedon, R., (1997). Les aqueducs de la gaule romaine et 

des regions voisines. Pulim, Limoges. 

Bellwald,U., M. Al-Huneidi, A. Salihi, D. Keller, R. Naser, & 

D. Al-Eisawi, (2003). The Petra Siq, Nabataean hydrology 

uncovered. Petra National Trust. 

Carbon, D., G. Fabre, P. Volant, J.-L. Fiches, A. Levret & P. 

Combes, (2005). L’aqueduc de Nîmes dans la haute 

Vistrenque: analyse interdisciplinaire d’un tronçon 

souterrain. Gallia 62, 69-86. 

Çeçen, K., (1996). Roma su yollarnn en uzunu (The 

Longest Roman Water Supply Line). 

Çetnkaplan, M., (2002).Tertiary high pressure / low 

temperature metamorphism in the Mesozoic cover series 

of the Menderes massif and correlation with the cycladic 

crystalline complex. Unpublished PhD. Thesis, Izmir. 

Dubar, M., (2006). Recherche et interprétation climatique 

des microcycles du concrétionnement travertineux de 

l'aqueduc romain de fréjus (var, France) Quaternaire 17, 

79-85 

Galli, P.A.C., Naso J.A., (2009), Unmasking the 1349 

earthquake source (southern Italy); paleoseismological 

and archaeoseismological indications from the Aquae 

Iuliae Fault, Journal of Structural Geology 31,128–149. 

Galli, P.A.C., Giocoli, A., Naso, J.A., Piscitelli, S., Rizzo, E., 

Capini, S. & Scaroina, L., (2010). Faulting of the 

Roman aqueduct of Venafrum (southern Italy): 

Methods of investigation, results, and seismotectonic 

implications. In: Ancient Earthquakes (Sintubin, M., 

Stewart, I.S., Niemi, T.M., Altunel, E. eds). Geological 

Society of America Special Paper 471, 233-242. 

Garczynski, P, J. Foucras & M. Dubar, (2005). 

Láqueduc dÁntipolis dit de la Boullide (Alpes 

Maritimes). Gallia 62, 13-34 

Grewe, K., (1986). Atlas der römischen 

Wasserleitungen nach Köln. Rheinland Verlag, Koeln. 

Guendon, J-L., Vaudoir, J., (2000). Concrétions et 

fonctionement de láqueduc: étude morphostratigraphique. 

In: Láqueduc de Nîmes et le Pont du 

Gard. (Fabré, G., Fiches, J-L., Paillet, J-L. eds). 

CNRS Editions. 263-271 

Guendon, J-L., Leveau, Ph., (2005). Dépots carbonatés 

et fonctionement des aqueducs romains. Gallia, 62, 

87-96 

Hodge, A. T., (1992). Roman aqueducts and water 

supply, Duckworth, London. 

Lambert J., T. Winter, T.J. Dewez & P. Sabourault, 

(2005). New hypotheses on the maximum damage 

area of the 1356 Basel earthquake (Switzerland), 

Quaternary Science Reviews, 24, 383–401. 

Levret, A, Ph. Volant, D. Carbon, D Combescure, T. 

Verdel, A. Piant, O. Scotti & Ph. Laurent, (2008). L’ 

aqueduc romain de Nimes (France): nouveaux 

résultats en archéosimicité. In: Archéosismicité et 

Vulnérabilité: patrimoine bâti et société. (Levret, A. 

ed),165-188 

Marra, F., P. Montone, M. Pirro & E. Boschi, (2004). 

Evidence of tectonics on a Roman aqueduct system 

(II-III century AD) near Rome, Italy. Journal of 

Structural Geology 26, 679-690. 

Meghraoui, M., F. Gomez, R. Sbeinati, J. Van der 

Woerd, M. Mouty, A. Nasser Darkal, Y. Radwan, I. 

Layyous, H. Al Najjar, R. Darawcheh, F. Hijazi, R. Al- 

Ghazzi & M. Barazangi, (2003). Evidence for 830 

years of seismic quiescence from palaeoseismology, 

archaeoseismology and historical seismicity along the 

Dead Sea fault in Syria. Earth and Planetary Science 

Letters 210, 35-52. 

Mendez-Victor, L. A., C. S. Oliveira & J. Azevedo, 

(2008). The 1755 Lisbon Earthquake: revisited. 

Springer Verlag. 

ROMAQ. Unpublished database. Passchier, C. W., Van 

Opstal, D., Schram, W. (eds.) Internet publication 

pending. 

Sbeinati, M., M. Meghraoui, G. Suleyman, F. Gomez, 

P. Grootes, M. Nadeau, H. Al Najjar & R. Al Ghazzi, 

(2010). Timing of earthquake ruptures at the Al Harif 

Roman aqueduct (Dead Sea fault, Syria) from 

archaeoseismology and paleoseismology. Geological 

Society of America Special Papers 471, 243-267. 

Volant, Ph., A. Levret, D. Carbon, O. Scotti, D. 

Combescure, T. Verdel, A. Piant, Ph. Laurent, (2009). 

An archaeo-seismological study of the Nîmes Roman 

aqueduct, France: indirect evidence for an M6 seismic 

event? Natural Hazards 49, 53-77. 

Wikander, O., (2000). Handbook of ancient water 

technology. Brill, Leiden. 

Wiplinger, G., (2008). The Deirmendere Aqueduct to 

Ephesus. Anodos 8, 393-400. 

Wright, K. R.,G. F. McEwan & R. M. Wright, (2006). 

Tipon: water engineering masterpiece of the Inca 

empire. ASCE Press. 

189


EARTHQUAKE 

ARCHAEOLOGY 



TESTING ARCHAEOSEISMOLOGICAL TECHNIQUES WITH INSTRUMENTAL SEISMIC 

DATA CAUSED BY THE Mw 5.1 LORCA EARTHQUAKE (5-11-2011, SE OF SPAIN) 

Pérez-López (1)*, R., J.L. Giner-Robles(2), M.A. Rodríguez-Pascua (1), F. Martín-González(3), J. García Mayordomo (1) 

J.A. Álvarez-Gómez(4)(5), M.J. Rodríguez-Peces(4), J.M. Insua-Árévalo (4), J. J. Martínez-Díaz (4) and P.G. Silva(6) 

(1) IGME – Instituto Geológico y Minero de España, Área de Investigación en Riesgos Geológicos. Madrid, SPAIN. Email: 

r.perez@igme.es , m.rodríguez@igme.es, julian.garcia@igme.es 

(2) Departamento de Geología. UAM, Universidad Autónoma de Madrid. SPAIN. Jorge.giner@uam.es 

(3) Universidad Rey Juan Carlos I, Madrid. SPAIN 

(4) Departamento de Geodinámica. Facultad de Ciencias Geológicas. Universidad Complutense de Madrid. SPAIN. Email: 

jmartinez@geo.ucm.es 

(5) Instituto de Hidráulica Ambiental "IH-Cantabria", E.T.S.I. Caminos, Canales y Puertos, Universidad de Cantabria. Spain. 

(6) Departamento de Geología. Universidad de Salamanca. Email: pgsilva@usal.es 

* Corresponding author 

Abstract (Testing archaeoseismological techniques with instrumental seismic data caused by the Mw 5.1 Lorca 

Earthquake, 5-11-2011, SE of Spain): the earthquake of the 11 th of May of 2011 was located near of the village of Lorca (SE of 

Spain) and with a magnitude (Mw) of 5.1. Despite of the small sized- earthquake, it generated a relative high seismic intensity (VII 

EMS), nine people were killed and almost 14,000 were left from their homes. The total estimated cost of this earthquake is 75 

millions of Euros. A systematic analysis of damaged ancient buildings was performed with the aim to check archaeoseismological 

new techniques. Namely, we have described different Earthquake Archaeological Effects (EAE) affecting both monumental 

buildings as modern edifices. The interest is the possibility to compare EAEs with focal parameters of the earthquake and the 

geometry of the seismic source, the Alhama de Murcia Fault (FAM).The preliminary results displays an almost N-S seismic wave 

anysotropy obtained from different EAE recorded at eight different monumental buildings. 

Key words: archaeoseismology, instrumental earthquake, EAE, Lorca, SE of Spain 

THE LORCA EARTHQUAKE OF THE 5-11-2011 

The 11 th of May, 2011, a small and shallow 

earthquake struck the village of Lorca (92,000 

inhabitants approximately). This earthquake hit the 

city at 18:50 h (local time), such as you can see in 

the broken clock of the tower of the San Francisco 

church (also so-called the church of “El Paso Azúl”). 

The recorded magnitude by the Spanish Instituto 

Geográfico Nacional, IGN, was 5.1 Mw (www.ign.es). 

Nevertheless, the maximum value of the peak ground 

acceleration (PGA) was 0.41g, recorded at the 

basement of the ancient jailhouse of the city. Also, 5 

mm of GPS movement northward was recorded in 

the basement of the Fireman Station. Outliers values 

for this small magnitude. 

Unfortunately, this small earthquake killed nine 

people, two from the collapse of one modern building 

located at La Viña neighbourhood, southward of the 

city, and the others due to the falling of the hanging 

decoration from the outer façade. Besides, almost 

14,000 people were moved from their homes and 

300 of them still live in camp sites. The city was 

dramatically collapsed and during the following 

twelve hours the fatality ruled the city. 

A few hours after the main shock, a group of 

earthquake geologists grew up from diverse Spanish 

institutions, and they went to the disaster zone with 

the aim of finding answers to these questions: (1) 

what was the seismic source of this earthquake? Or 

better, which is the causative fault of the earthquake? 

(2) Is there evidence of surface rupture for such a 

minor event? 

190 

Fig. 1: Up. Seismic serie of the Lorca earthquake recorded 

by the IGN. The size of the point is related to the 

earthquake magnitude. The colour indicates the day of the 

occurrence. The focal mechanism solutions show two 

strike-slips with reverse component. The red lines indicate 

the main active faults. The FAM (Alhama de Murcia major 

Fault) is a NE-SW reverse fault with oblique component, 

with a total length of ca 90 km (Martínez-Díaz, 2002). 

Down: frequency grey bars indicate the number of 

earthquakes recorded in 3 h. 

Empirical models predict small rupture area (i.e. ca 

12 km 2 , Wells & Coppersmith, 1994), (3) Can we 

compile an Earthquake Environmental Effect


EARTHQUAKE 

ARCHAEOLOGY 



catalogue from this shaking? Rock fall, ground cracks 

affecting soils and rocks, changes of temperature in 

hot springs and thermal aquifers, liquefactions, and 

so, and finally, (4) Can we do an archaeoseismictype 

analysis from the monumental building damage 

within the village? In this case we tried to make a 

compilation of the maximum number of the 

Earthquake Archaeological Effects (see Rodríguez- 

Pascua et al., 2011 for a comprehensive description 

of EAEs). The interest of this last analysis is the 

correlation between EAEs and focal and geometrical 

parameters of the earthquake. 

Despite the low magnitude of the earthquake, we 

have enough geological and seismological data to 

find a preliminary answer to these questions: high 

seismic intensity (VII EMS, www.ign.es), high 

building damage (almost 200 damaged buildings are 

catalogued), we have an active fault with different 

outstanding palaeoseismic studies (the Alhama de 

Murcia Fault, FAM)(Silva et al., 1997, Martínez-Díaz 

1998)(see location in Fig.1). 

Consequently, the aim of this work is to check EAEs 

usefulness by comparing the damage anisotropy 

caused by the main shock of the earthquake (Mw 

5.1), and performed along the old town of the historic 

village of Lorca. The technique used here is widely 

described in Giner-Robles et al. (2009). 

SEISMOTECTONICS OF THE EARTHQUAKE 

A precursor event of Mw 4.5 (www.ign.es) occurred 3 

hours before of the main shock of 5.1. The epicentre 

location of both earthquakes coincides with the 

mapped active fault of Alhama de Murcia (FAM)( 

Bousquet & Montenat, 1974))(Fig.1). The location of 

the epicentres on the FAM suggests that a small 

segment of 3 x 3 km could have broken. Moreover, 

the focal mechanism solutions indicate either a NE- 

SW fault plane (in coincidence with the FAM 

trending) or NW-SE fault plane. The geometry of both 

mechanisms shows an oblique fault with reverse 

component. This geometry is in agreement with the 

geology and palaeoseismological studies on the FAM 

(Silva et al., 1997; Martínez Díaz et al., 2001). The 

estimated depth of the earthquake lies between 3 

and 5 km (www.ign.es). 

In the following six days, more than 100 aftershocks 

with a maximum magnitude of 3.9 were recorded 

(www.ign.es). The temporal occurrence of 

aftershocks shows cluster behaviour in small groups 

instead of a slight decay of the number of aftershocks 

from the main shock (Fig. 1). 

GEOLOGICAL EFFECTS (EEE) OF THE 

EARTHQUAKE 

A shallow earthquake (2 km deep) and high seismic 

intensity (VII EMS) suggested the possibility to have 

a surface rupture associated to the earthquake. With 

this aim, we performed a field trip across the fault 

trace of FAM to catalogue Earthquake Environmental 

Effects (EEE) according to the ESI 07 scale (Michetti 

et al., 2007). 

191 

Fig. 2. Rock fall triggered by the Lorca earthquake and 

located at Las Estancias Sierra, southwestwardly of the 

epicentre location. 

The main effect was rock falling located in steep 

carbonatic cliffs in the nearby (


EARTHQUAKE 

ARCHAEOLOGY 



casualties, particularly on the habitations located 

right under the cliff. A quick visual inspection 

suggested the perfect state of the measures, 

although this observation should be confirmed after a 

systematic and detailed inspection. The approximate 

volume of rock falling blocks is lesser than 10 3 m 3 . 

Fig. 3. Partial collapse of the easternmost point of Lorca’s 

castle cliff. A detached rock-block produced damage in a 

near house. 

ANALISIS OF EAE: DAMAGE DESCRIPTION 

The earthquake of Lorca (SE Spain) was responsible 

for a large amount of damage and seismic intensity 

on a wide range of buildings in the city of Lorca, 

including the historical buildings. Aerial view shows a 

concentration of damage in the highest towers, 

mainly affecting the arches, buttresses canopies, 

bollards etc. Rotations also appear in decorative 

elements such as bollards and obelisks, like the 

obelisk in the San Francisco church. 

The most relevant key of this archaeoseismic study is 

the correlation between the instrumental information 

of this earthquake with damage in historical and 

modern buildings. This information also can be 

correlated with seismic and geological parameters, 

such as the magnitude and focal mechanism of 

earthquake, the seismogenic fault and site effects 

related to the geology of Lorca. 

Lorca's earthquake has not produced widespread 

buildings collapses (only two buildings collapsed), 

during the inspection we have recognized, classified 

and described more than one hundred effects of the 

earthquake on buildings and structures, similar to 

those described in the work of Giner-Robles et al., 

2009, Rodríguez-Pascua et al., 2011. These authors 

defined the Earthquake Archaeological Effects, 

(commonly known as EAE). 

The EAE describes and quantifies the coseismic 

deformation in archaeological sites and historic 

buildings. After Giner-Robles et al., 2009 and 

Rodríguez-Pascua et al., 2011, they classified 

according to EAE: (1) permanent deformation of the 

surface (2) temporary deformation by the seismic 

shaking during the earthquake. 

192 

Fig. 4. Damage in the tower and main buildings of the 

Clarisas´ Convent. (a) Destroyed tower, (b), collapsed 

roof of the main edifice, (c) and (d) X-fractures in the East 

and West sides of the annexed building. 

Figure 4 shows an example of the type of analysis 

that we have performed in the village. We have 

located the damaged building and we have described 

all of EAEs recognized. In this case, the Clarisas’s 

Convent displayed two types of EAEs: collapsed roof 

and walls, and penetrative conjugated faults affecting 

oriented sides of annexed buildings. Other type of 

study involves the orientation of fallen key stones in 

arches. Fig. 5 is a clear example of arches showing 

evidence of damage anisotropy obeying the 

mechanism proposed by Giner et al., 2009. 

All this preliminary information has been represented 

in a rose diagram (Fig. 6), showing the main 

orientation in a NW-SE trend, N150º-160ºE. We 

interpret these direction NW-SE with azimuth from 

the SE. 

PRELIMINARY RESULTS 

- Alhama-Murcia Fault (FAM) is the structure with 

greater evidence of Quaternary activity in the area: 

paleoseismic activity (Mw > 6.0) over the last 1000 

years, associated with thermal springs and a wellrecognized 

surface trace. Destructive historical 

seismicity located along the trace during the XVII, 

XVIII and XIX centuries were reported in chronicles. 

FAM has a trace that it is parallel to one of the nodal 

planes of focal mechanisms obtained for the 

earthquake. The oblique (reverse - sinistral 

movement) movement of the fault is consistent with 

the focal mechanism solution.


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 6. Rose diagram of the seismic shaking, using the 

EAEs. Preliminary results (After Martínez-Díaz et al., 

2011). 

References 

Fig. 5. Example of analysis of deformation structures and 

the results of the orientation assessment of the strain (San 

Juan’s Church) (After Martínez-Díaz et al., 2011). 

- The high seismic intensity experienced by the town 

of Lorca (intensity VII EMS-98 scale, data IGN) 

associated with a magnitude 5.1 Mw, may be due to 

the earthquake spread from the Sierra de la Tercia 

(epicentral area) to the SW. The lack of geological 

effects to the east of the epicenter (La tercia Range 

and Lower Guadalentin Valley) support the possible 

existence of rupture directivity southwest. 

-The seismic wave propagation supports the 

directionality of the FAM rupture spread from the 

epicentral area, crossing the city of Lorca. This 

reason associated with the shallowness of the 

earthquake, would explain the high seismic intensity 

and peak accelerations of 0.36 g (IGN data) recorded 

in the accelerometer of the old prison of Lorca 

(located in the downtown). However, the high value 

of PGA could be also related to site effects. 

- The archaeoseismic data (more than a hundred 

values) suggest an origin of the deformation 

associated with a nearby seismic field, implying that 

the main earthquake rupture occurred beneath the 

historic city of Lorca because the faulting subsurface 

rupturing runs below the Lorca village. 

- The ESI 07 macroseismic classification (Michetti et 

al. 2007) for this earthquake is between VI and VII, 

according to the features observed. Hence, the 

interest of this small-sized earthquake is that it could 

be used as a lower limit for the instrumental 

calibration of the scale ESI 07. 

Acknowledgements: Thanks are given to the Spanish 

Instituto Geografico Nacional (IGN) for providing 

instrumental data of the earthquake. Also a special mention 

to the people of Lorca which helped us even with their 

homes collapsed and all people which showed friendly and 

kindness against the earthquake disaster. 

Bousquet, J.C. y Montenat, C. (1974). Presence décrochements 

NE-SW plio-quaternaires dans les Cordillères 

Bétiques Orientales (Espagne). Extension et signification 

général. C. R. Acad. Sci. Paris 278: 2617-2620. 

Giner-Robles, J.L. M.A. Rodríguez-Pascua, R. Pérez- 

López, P.G. Silva, T. Bardají, C. Grützner and K. 

Reicherter. (2009). Structural Analysis of Earthquake 

Archaeological Effects (EAE): Baelo Claudia Examples 

(Cádiz, South Spain), vol. 2,130p. Instituto Geológico y 

Minero de España, Madrid. Dep. Leg. M-27206-2009. 

Martínez Díaz, J.J., E. Masana , J.L. Hernández-Enrile & P. 

Santanach. (2001). Evidence for coseismic events of 

recurrent prehistoric deformation along the Alhama de 

Murcia Fault, Southeastern Spain. Geol. Acta. 36 (3-4): 

315-327. 

Martínez Díaz, J.J. (1998). Neotectónica y Tectónica Activa 

del Oeste de Murcia y sur de Almería (Cordillera Bética). 

Tesis Doctoral. Universidad Complutense Madrid. 470p. 

Martínez-Díaz, J. J. (2002). Stress field variety related to 

fault interaction in a reverse oblique-slip fault: the Alhama 

de Murcia Fault, Betic Cordillera, Spain. Tectonophysics, 

356: 291-305. 

Martínez Díaz J. J.; et al. (2011). Geological Preliminary 

Field Report of The Lorca Earthquake (5.1 Mw, 11th May 

2011). Instituto Geológico y Minero de España (IGME), 

Spanish Group of Active Tectonics, Paleosismicity and 

Risks, Universidad Complutense de Madrid (UCM), 

Universidad Autónoma de Madrid (UAM) and Universidad 

Rey Juan Carlos de Madrid (URJC). 45p. 

Michetti A.M., Esposito E., Guerrieri L., Porfido S., Serva L., 

Tatevossian R., Vittori E., Audemard F., Azuma T., 

Clague J., Comerci V., Gurpinar A., Mc Calpin J., 

Mohammadioun B., Morner N.A., Ota Y. & Roghozin E. 

(2007). Intensity Scale ESI 2007. In. Guerrieri L. & Vittori 

E. (Eds.): Memorie Descrittive Carta Geologica. d’Italia., 

vol. 74, Servizio Geologico d’Italia – Dipartimento Difesa 

del Suolo, APAT, Roma, 53 p. 

Rodríguez-Pascua M.A., R. Pérez-López, J.L. Giner- 

Robles, P.G. Silva, V.H. Garduño-Monroy and K. 

Reicherter. (2011). A Comprehensive Classification of 



Roman and Mesoamerican cultures. Quaternary 

International, doi:10.1016/j.quaint.2011.04.044. 

Silva, P.G., Goy, J.L., Zazo, C., Lario, J., Bardají, T. (1997). 

Paleoseismic indications along "aseismic" fault segments 

in the Guadalentín depresion (SE Spain). J. 

Geodynamics 24(1-4), 105-115. 

Wells, D.L. & Coppersmith, K.J. (1994). New empirical 


width, rupture area, and surface displacement. Bull. 

Seismol. Soc. Amer. 84, 974-1002. 

193


EARTHQUAKE 

ARCHAEOLOGY 



FRONTIERS OF EARTHQUAKE ARCHAEOLOGY: THE OLYMPIA AND SAMICUM CASES 

(PELOPONNESE, GREECE) 

Reicherter, Klaus (1) 

(1) RWTH Aachen University, Inst. of Neotectonics and Natural Hazards, Aachen, 52056, Germany, email: 

k.reicherter@nug.rwth-aachen.de 

Abstract (Frontiers of earthquake archaeology: the Olympia and Samicum cases (Peloponnese, Greece)): The ancient 

Olympia on the western Peloponnese/Greece is worldwide known as the place where for 1170 years the classical Olympic Games 

took place. Furthermore, to domino-like toppled column drums of the Zeus temple are famous and widely accepted as earthquakerelated 

structural damage the 522/551 AD events. The Peloponnese peninsula is characterized by frequent seismicity within the 

Euroasian-African convergence zone (Hellenic arc and backarc system) by mainly E-W trending normal faults. 

Herein, some examples of the heterogeneous structural deformation are given observed in the ancient Olympia including an 

interpretation. Also, we show deformation patterns at Samicum, a Hellenistic fortified village (acropolis) with a cyclopean masonry 

situated on top of the coastal Lapithas mountain ridge, close to Olympia. Both examples may be interpreted as archaeoseismic 

evidence, however, an on-fault palaeoseismological approach and assessment is missing in this part of Greece. 

Key words: earthquake geology and archaeology, Olympia, Samicum, structural deformation 

INTRODUCTION 

This paper deals with the combination of well-known 

historic sites and their earthquake-related damage. 

By a number of ancient historians (see references) 

we know about the destruction of ancient Olympia by 

several earthquakes (early 4 th cent. AD, 522 AD, 551 

AD) and by warfare and vandalism. On the other 

hand, the close-by fortified village Samicum was 

described as abandoned by Pausanias in the 2 nd 

cent. AD (Fig. 1). This seems rather strange taking 

into account the strategic position of the village on 

top of the mountain Lapithas. We apply the 

classification of Earthquake Archaeological Effects of 

Rodríguez-Pascua et al. (2011; EAE) to describe and 

classify the observed damage. The investigation of 

past earthquakes is commonly split into several 

disciplines, which cover Instrumental Seismology, 

Historical Seismology, Archaeoseismology and 

Earthquake Geology or Palaeoseismology depending 

on information available (e.g. Caputo and Helly, 

2008). However, the ambiguity of data and results of 

distinct disciplines (Niemi, 2008), e.g. of written 

sources and recent observations, results in 

limitations of the seismic hazard assessment of an 

area (Nur, 2008). 

The Elis region is characterized by a huge number of 

neotectonic faults, which have been intensely studied 

(Fig. 2; Lekkas et al., 1993; Lekkas et al., 2000, 

Papanikolaou et al., 2007; Fountoulis and 

Mariolakos, 2008). The seismicity in the study area 

on the western Peloponnese peninsula is high. The 

last earthquake with M 5.5 struck the Pyrgos area on 

the 23 rd of March in 1993 with maximum intensities of 

VIII (e.g. Lekkas et al., 2000) associated with EEE 

(Earthquake Environmental Effects on the INQUA 

ESI-scale, Papanikolaou et al., 2009) like 

liquefaction, landslides, ground fractures and 

changes in the aquifer level. 

STUDY AREAS 

Fig. 1: Study area of ancient sites in Greece, inset 

shows location of Fig. 2. 

194


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 2: Study area of ancient sites on the Peloponnese peninsula, Greece, with major faults (compiled from Lekkas et al., 1993; 

Lekkas et al., 2000, Papanikolaou et al., 2007; Fountoulis and Mariolakos, 2008). 

Ancient Olympia 

The ancient Olympia is a famous World Heritage site 

located in the Peloponnesian province of Elis in the 

west of the peninsula (Figs.1 and 2). Known as the 

place of the classical Olympic Games, the sanctuary 

(Altis) consists of several buildings, which spread 

between the Kladeos creek and Mount Kronos. The 

Altis resembles a variety of buildings of different use 

was forgotten by people until 1829 AD. One of the 

most famous examples of EAE are the domino-style 

fallen columns of the Temple of Zeus (5 th cent. BC), 

which made it on the cover of a book entitled 

“Archaeoseismology” (see Stiros, 1996). The former 

Temple of Zeus, today known as Temple of Hera (it 

was rededicated in the 5 th cent. BC), is reported to 

have been destroyed by an earthquake in the early 

Date Event 

590-580 BC Construction of the Temple of Hera 

472-456 BC Construction of the Temple of Zeus 

c. 175 BC Earthquake? and repair of the Zeus 

statue, roof and columns (Pausanias, 

IV,31,6 and Dinsmoor, 1941) 

sculptures of the Temple of Zeus 

(Dinsmoor, 1941), which are replaced 

36 BC Earthquake? Second-hand roof tile – 

repair work (Dinsmoor, 1941) 

3 th cent. AD Series of earthquakes? 

early 4 th 

Earthquake (destroys Temple of Hera) 

cent. AD 

394 AD Pillage of Olympia by the Goths 

426 AD Theodosius orders destruction of the 

Temple of Zeus 

522 AD Earthquake (destroys Temple of Zeus) 

in combination with floods of the 

Alfeios and Kladeos, and landslide 

from the Kronos hill? (date unclear 

could be 551 AD) 

551 AD Earthquake (destroys Temple of Zeus) 

in combination with floods of the 

Alfeios and Kladeos, and landslide 

from the Kronos hill? 

Tab. 1: Timetable of con-/ and destruction observed in 

Olympia (compiled with different sources) 

and age during more than 3000 years (since c. 2500 

BP until 6 th cent. AD). Finally, the area was flooded 

and covered by sediments of the Kladeos creek and 

Alfeios river (Fouache and Pavlopoulos, 2010) and 

195 

Fig. 3: Fallen columns of the Temple of Zeus, Olympia 

(Classical period) 

4 th cent. AD, and was never rebuilt. Table 1 sums the 

construction and destruction events of the Zeus and 

Hera temples, several dates of possible destruction 

are ambiguous and not stated by exact dating, 

furthermore the causative fault(s) are not known. 

Samicum 

The fortified village of Samicum (or Macistus, ancient 

Kato Samia, however there is discussion, Pausanias 

and Polybius mention only Samicum, and Xenophon 

only Macistus) is situated in the southern part of the 

Elis region (Ilia, Triphylia) on top of the Lapithas 

mountain ridge, south of mouth of the Alfeios river


EARTHQUAKE 

ARCHAEOLOGY 



(Perseus Digital Library, 2011). Most probably 

Samicum was founded in the last 5 th cent. BC 

(classical Hellenistic period) and was occupied until 

the 2 nd cent. AD (Roman period). Samicum was 

occupied by the Aetolian Polysperchon against the 

Arcadians in 244 BC, and was later taken by Philip, 

in 219 BC. According to archaeologists the Samicum 

site has to be considered as one of the most 

important cities in Elis because of the strategic 

position on the mountain top, controlling the Kaiafas 

pass, lagoons and springs. 

The village is surrounded by a so-called pseudopolygonal 

wall of c. 1500 m length in a trapezoid 

shape with several towers, thus forming an acropolis. 

The cyclopean masonry of the wall surrounding 

Samicum is characterized by polygonal limestones, 

which fit each other with precision and minimal 

clearance between the stones and no use of mortar, 

little gaps are filled with shaped small stones. 

via the vaulted archway of c. 200 BC, dropped key 

stones can be observed (Fig. 4). 

Finally, the Phillipeion, a 338 BC construction, shows 

nice corner breakouts (Fig. 5), described as “dipping 

corners” by Rodríguez-Pascua et al. (2010) in the 

EAE. 

OBSERVATIONS 

Ancient Olympia 

Walking around in the Altis of Ancient Olympia 

reveals a good opportunity to also study earthquakerelated 

structural damage. On the other hand, a lot of 

reconstruction work and replacement limits full 

pleasure, which is further diminished by “slow” 

deformation due to landslides in the Kronos hill area. 

The most prominent features are the domino-like 

fallen columns of the Temple of Zeus (Fig. 3). 

Already Stiros (1996) stated that one of the columns 

is known to have been fallen during a storm (possibly 

after its reconstruction in the 19 th cent. AD). Another 

peculiar observation is that the colonnade columns 

fell towards the N- and the S-side of the temple´s 

long axis. The spaces between individual column 

drums are filled with sediment yielding ceramics. 

Fig. 5: Phillipeion, Olympia, c. 338 BC (Hellenistic 

period) 

However it is remarkable that the Roman ruins do not 

show such peculiar features of deformation. After an 

earthquake in the early Byzantine period (6 th cent. 

AD), the Altis was covered by a “mass of yellow earth 

spread over the entire area after a cloudburst 

perhaps in the 7 th cent. AD” (Dinsmoor, 1973). 

Samicum 

After reaching the ruins of Samicum on the Lapithas 

mountain enjoy the nice view around including the 

Volax and Kaifas lagoons. Little of the central part is 

excavated, the most impressive feature of Samicum 

is the cyclopean masonry of the city wall with several 

fallen towers and gates. Towards the south a steep 

cliff frames the fortified village. The wall and its 

blocks show various indicators of seismic damage. 

Among them are moved and rotated blocks, corner 

break-outs, collapsed watch towers (Figs. 6 and 7). 

Up to now these damages are not described and 

related to earthquakes, also there is no historic 

account for earthquake damage in Samicum. But 

Pausanias refers to his journey to Samicum, where 

Fig. 4: Vaulted entrance to the stadium Olympia (Crypt), 

approx. 200 BC (Hellenistic period) 

Dinsmoor (1941) observed to different styles of 

column drum beds, on which the connection to each 

other is observable: one with Lewis Holes, one 

without (but with the Empolion, the central hole). Of 

175 drums analyzed only 30 drums yielded Lewis 

Holes (Dinsmoor, 1941). This and repair work at 

columns as well as repair clamps at the western 

corners, lead Dinsmoor (1941) to postulate a 175 BC 

earthquake damage. Further on, entering the stadium 

196 

Fig. 6: Samicum, wall damage (Hellenistic period)


EARTHQUAKE 

ARCHAEOLOGY 



he found the village abandoned and destroyed in the 

2 nd cent. AD. 

Fig. 7: Samicum, wall damage and tower collapse 

(Hellenistic period) 

CONCLUSIONS 

We presented EAE of the ancient Greek sites of 

Olympia and Samicum, both have in common that 

the structural damage up to now has not been 

mapped, characterized and catalogued in detail. 

Also, the structural damage has not been classified 

and differentiated in the cause of damage. The 

Samicum case is most probably a “one-event” 

earthquake deformation. Whereas Ancient Olympia 

has obviously suffered from several earthquakes and 

other destructions (pillage, landslides and flooding). 

Both cases have also in common that the causative 

fault(s) for the earthquakes have not been detected 

yet, palaeoseismological studies are missing. Recent 

seismicity, such as the 1993 Pyrgos event (EEE 

description in Papanikolaou et al., 2009) shows that 

even moderate earthquakes (M 6.5) can cause 

severe structural damage and secondary effects as 

landslides and liquefaction. So, as a conclusion both 

cases reveal clearly the frontiers of 

archaeoseismology, and are at the moment that what 

Sintubin et al. (2008) pointed out: just a good story. 

References 

Caputo, R., Helly, B. (2008). The use of distinct disciplines 

to investigate past earthquakes. Tectonophysics, 

453(14), 719. 

Dinsmoor, W.B. (1941). An archaeological earthquake at 

Olympia. Am. J. Archaeol., 45/3, 399-427. 

Dinsmoor, W.B. (1973). The architecture of Ancient Greece. 

An account of its historic development. Biblio and Tannen, 

424 pp. 

Fouache, E., Pavlopoulos, K. (2010). The interplay between 

environment and people from Neolithic to Classical times 

in Greece and Albania. In: Martini, I.P. (ed.). Landscapes 

and Societies: Selected Cases, Chapter 10, 155-166. 

Fountoulis, I., Mariolakos, I. (2008). Neotectonic folds in the 

central-western Peloponnese, Greece. Z. dt. Ges. 

Geowiss., 159/3, 485-494. 

Lekkas, E., Papanikolaou, D., Fountoulis, L. (1993). 

Neotectonic map of Greece, Pyrgos-Tropaia sheets, scale 

1:100.000, Research Project of eth Univ. of Athens, Dept. 

of Geology, Division of Dynamic, Tectonic and Applied 

Geology, Athens. Available online at: 

http://elekkas.gr/en/research/mapping.html 

Lekkas, E., Fountoulis, I., Papanikolaou, D. (2000). Intensity 

distribution and neotectonic macrostructure Pyrgos 

Earthquake data (26 March 1993, Greece). Nat. Haz., 21, 

19-33. 

Niemi, T. (2008). Historical Earthquake Catalogues and 

Archaeological Data: Avoiding Circular Reasoning. 

Seismological Research Letters, 79(2), 289. 

Nur, A. (2008). Apocalypse. Earthquakes, Archaeology, and 

the Wrath of God. Princeton University Press, Princeton. 

Papanikolaou, D., Fountoulis, I., Metaxas, C. (2007): Active 

faults, deformation rates and Quaternary paleogeography 

at Kyparissiakos Gulf (SW Greece) deduced from 

onshore and offshore data. Quat. Int., 171-172, 14–30. 

Papanikolaou, I.D., Papanikolaou, D.I., Lekkas, E. (2009). 

Advances and limitations of the Environmental Seismic 

Intensity scale (ESI 2007) regarding near-field and farfiled 

effects from recent earthquakes in Greece: 

implication for the seismic hazard assessment. In: 

Reicherter, K., Michetti, A.M., Silva, P.G. (eds.). 

Paleoseismology: Historical and prehistorical records of 

earthquake ground effects for seismic hazard 

assessment. J. Geol. Soc. London Spec. Publ., 316: 11- 

30. 

Pausanias (c. 115-180 AD). Pausanias Description of 

Greece with an English Translation by W.H.S. Jones, 

Litt.D., and H.A. Ormerod, M.A., in 4 Volumes. 

Cambridge, MA, Harvard University Press; London, 

William Heinemann Ltd. 1918. Online at 

http://www.perseus.tufts.edu. 

Perseus Digital Library, lasted visit 5/2011: 

http://www.perseus.tufts.edu 

Polybius (c. 200 – 120 BC). Histories. Evelyn S. 

Shuckburgh (translator). London, New York. Macmillan. 

1889. Reprint Bloomington 1962. Online at 


Rodríguez-Pascua, M.A., Pérez-López, R., Giner-Robles, 

J.L., Silva, P.G., Garduño-Monroy, V.H., Reicherter, K. 

(2011, in press). A comprehensive classification of 



Roman and Mesoamerican cultures. Quat. Int. 

Sintubin, M., Stewart, I. S., Niemi, T. and Altunel, E. (2008). 

Earthquake Archaeology Just a Good Story? 

Seismological Research Letters, 79(6), 767768. 

Stiros, S.C. (1996). Identification of Earthquakes from 

Archaeological Data: Methodology, Criteria and 

Limitations. In: Archaeoseismology (edited by Stiros, S. C. 

and Jones, R. E.). Fitch Laboratory Occasional Paper 7. 

Institute of Geology and Mineral Exploration and the 

British School at Athens, Athens, 129152. 

Strabo (c. 63 BC - 29 AD), Geography. Online at 


Xenophon (c. 430-354 BC), Hellenica, 3.2. Online at 


197


EARTHQUAKE 

ARCHAEOLOGY 



REGIONAL STRAIN-RATES ON ACTIVE NORMAL FAULTS AND VARIABILITY IN THE 

SEISMIC CYCLE: AN EXAMPLE FROM THE ITALIAN APENNINES 

Roberts, Gerald (1, Joanna Faure Walker (1), Patience Cowie (2), Richard Phillips (3), Ken McCaffrey (4), Ioannis Papanikolaou 

(5), Max Wilkinson (4), Alessandro Michetti (6), Peter Sammonds (1). 

(1) Department of Earth and Planetary Sciences, Birkbeck/UCL, University of London, Gower Street, WC1E 7HX. UK. 

gerald.roberts@ucl.ac.uk 

(2) Department of Earth Sciences, University of Bergen, Postboks 7800, NO-5020 BERGEN 

(3) Institute of Geophysics and Tectonics, School of Earth and Environment, University of Leeds, LS2 9LT, UK 

(4) Department of Earth Sciences, Durham University, Science Labs, Durham DH1 3LE. UK. 

(5) Mineralogy-Geology Lab of the Agricultural University of Athens, Greece. 

(6) Università degli Studi dell'Insubria - Sede di Como, Facoltà di Scienze MM. FF. NN., Dipartimento di Scienze Chimiche e 

Ambientali, Via Valleggio, 11 - 22100 Como – Italy. 

Abstract(Regional strain-rates on active normal faults and variability in the seismic cycle: an example from the Italian 

Apennines): The rate at which a fault slips fundamentally determines the seismic hazard because average earthquake recurrence 

intervals tend to decrease as slip rates increase. Slip-rates on different faults within a growing fault system are not distributed 

randomly in space and time, because slip-rates must accommodate the regional strain-rate. Here we show how slip-rates vary 

temporally and spatially along the Italian Apennines, constrained by offsets across fault scarps dated with 36 Cl in situ cosmogenic 

dating. We map regional variations in strain-rate, investigating natural variability in the seismic cycle. 

Key words: Fault scarps, regional strain-rates, cosmogenic dating, seismic-cycle. 

INTRODUCTION 

Palaeoseismologists and earthquake geologists 

should try to move forward from simply characterising 

slip on single faults. Instead, they should try to 

characterise slip across systems of faults as it is 

clear that faults interact through stress transfer, and 

this controls regional patterns of slip-rate (the 

geography of seismic hazard), and the recurrence 

intervals for earthquakes on specific faults. The 

regional approach is needed to complement the 

regional strain-rate databases provided through GPS 

geodesy and instrumental seismicity. We should 

attempt to constrain how the geography of seismic 

hazard implied over 10 2 -10 3 years (palaeoseismology 

and earthquake geology) compares with that over 10- 

100 years (geodesy and instrumental seismicity). We 

expect differences between data from these different 

timescales to reveal temporal and spatial variability in 

the seismic cycle (e.g. Faure Walker et al. 2010). 

Fault scarps and 36 Cl cosmogenic dating 

The Italian Apennines are characterised by fault 

scarps that offset and deform deposits and landforms 

formed during the last glacial maximum. The offsets 

have accumulated since 15 ±3 ka shown by studies 

of tephrachronology and 14 C dating (Giraudi and 

Frezzotti 1995, 1997) and this is confirmed by 36 Cl in 

situ cosmogenic dating (Palumbo et al. Schlagenhauf 

2009, Schlagenhauf et al. 2010), and our own 

ongoing cosmogenic dating. The scarps are 

widespread and allow a regional study of strain rates 

and natural variability in the seismic cycle. 

198 

Figure 1. A regional strain-rate map for the Italian 

Apennines constrained by offsets of 15 ±3 ka 

features across fault scarps. 

Regional Strain rates 

A key unknown is how far a fault can stray from its 

long-term slip rate, both at a timescale equivalent to 

the interseismic period (10 2 -10 3 years), and over 

timescales equivalent to several seismic cycles (10 3 - 

10 4 years). The lack of such knowledge impedes our 

ability to perform probabilistic seismic hazard


EARTHQUAKE 

ARCHAEOLOGY 



assessments and understand the underlying physics 

that controls repeated earthquake slip. In order to 

study the existence of possible deficits or surpluses 

of geodetic and earthquake strain in the Italian 

Apennines compared to 15 ±3 kyr multi-seismic-cycle 

strain-rates, horizontal strain-rates are calculated 

using slip-vectors from striated faults and offsets of 

Late Pleistocene-Holocene landforms and sediments, 

using an adaptation of the Kostrov equations (Faure 

Walker 2010). 

Strain-rates calculated over 15 ±3 kyr within 5km x 

5km grid squares vary from zero up to 2.34 ±0.54 x 

10 -7 yr -1 , 3.69 ±1.33 x 10 -8 yr -1 , and 1.20 ±0.41 x 10 -7 

yr -1 in the central Apennines Lazio-Abruzzo region, 

the Molise-North Campania region, and the southern 

Apennines South Campania-Basilicata region, 

respectively. The data resolve variations in strain 

orientations and magnitudes along the strike of 

individual faults. Strain-rates over a time period of 15 

±3 kyrs from 5km x 5km grid squares integrated over 

an area of 1.28 x 10 4 km 2 (80 km x 160 km), show 

the horizontal strain-rate of the central Apennines is 

1.18 (+0.12/-0.04) x 10 -8 yr -1 parallel to the regional 

principal strain direction (043-223 o ±1 o ). In Molise 

and North Campania, the horizontal principal strainrate 

calculated over an area of 5 x 10 3 km 2 (50 km x 

100 km) is 2.11 (+1:14/-0:16) x 10 -9 yr -1 along the 

principal horizontal strain direction (039-219 o ±3 o ). 

Within the southern Apennines region within an area 

of 8 x 10 3 km 2 (50 km x 160 km), the average 

horizontal principal strain-rate is 3.70 ±0:26 x 10 -9 yr -1 

along the horizontal principal strain direction (044- 

224 o ±2 o ). 

Strain-rates calculated within 5 x 5 km and 20 x 20 

km grid squares, and at a regional scale, are highest 

in the central Apennines, medial in the southern 

Apennines and lowest Molise and North Campania. 

At the regional length-scale, the strain-rates are 

comparable in direction and magnitude to strain-rates 

calculated using GPS. Smaller areas (~2000-7000 

km 2 ), corresponding to polygons defined by geodesy 

campaigns (126 years) and seismic moment 

summations (700 years) show higher 10 2 yr strainrates 

than 10 4 yr strain-rates in some areas, with the 

opposite situation in other areas where seismic 

moment release rates in large (> Ms 6.0) magnitude 

historical earthquakes have been reported to be as 

low as zero. High strain-rates over 15 ±3 kyr in 

places occur where instrumental seismicity rates 

have been low. This demonstrates that strain-rates 

vary spatially on the length-scale of 10-100 km and 

on a timescale between 10-100 yrs and 10 4 yrs in the 

Italian Apennines. 

The multi seismic cycle strain-rates are used to 

calculate earthquake recurrence intervals for a given 

earthquake slip magnitude, at the scale of individual 

seismic sources; these value are compared to 

palaeoseismic data. The results are used to discuss 

spatial and temporal earthquake clustering and the 

natural variability of the seismic cycle. 

Acknowledgements: We thank the Natural Environment 

Research Council for funding. NERC Urgency Grant 

NE/H003266/1. A LiDAR and field study of surface rupture 

and post-seismic slip for the 6 th 

April 2009 L’Aquila 

Earthquake (M6.3). Dr. K. McCaffrey, Dr. G. P. Roberts, 

Prof. P. Cowie. (April 2009-May 2010). NERC Standard 

Grant NE/E01545X/1. Testing Theoretical models for 

Earthquake Clustering using 36 Cl Cosmogenic Exposure 

Dating of Active Normal Faults in Central Italy. Prof. P. 

Cowie, Dr. G. P. Roberts, Dr. K. McCaffrey (October 2007- 

2010). 

References 

Faure Walker, J., Mechanics of continental extension from 

Quaternary strain field in the Italian Apennines. 

Unpublished PhD Thesis, University College London. 

2010. 

Faure Walker, J., Roberts, G. P., Sammonds, P., Cowie, P. 

A., Comparison of earthquakes strains over 10 2 and 10 4 

year timescales: insights into variability in the seismic 

cycle in the central Apennines, Italy. Journal of 

Geophysical Research VOL.115, B10418, 

doi:10.1029/2009JB006462, 2010 

Giraudi, C., Frezzotti, M., 1995. Paleoseismicity in the Gran 

Sasso Massif (Abruzzo, Central Italy). Quaternary 

International, 25, 81-93. 

Giraudi, C., Frezzotti, M., 1997. Late Pleistocene glacial 

events in the central Apennines, Italy. Quaternary 

Research, 48, 280-290. 

Palumbo, L., L. Benedetti, D. Bourles, A. Cinque, R. Finkel, 

2004. Slip history of the Magnola fault (Apennines, 

Central Italy) from 

36 Cl surface exposure dating: 

evidence for strong earthquakes over the Holocene, 

Earth and Planetary Science Letters, 225, 163-176. 

Schlagenhauf, A., Identification des forts seismes passes 

sur les failles normales actives de la region Lazio- 

Abruzzo (Italie centrale) par “datations cosmogeniques” 

(36Cl) de leurs escarpments. These, Docteur de 

l’Universite Joseph Fourier, Grenoble, France. 2009. 

Schlagenhauf, A., et al. Using Chlorine-36 cosmonuclides to 

recover earthquake histories on limestone normal fault 

scarps : a reappraisal of methodology and 

interpretations. Geophys. J. Int., doi:10.1111/j.1365- 

246X.2010.04622.x 2010. 

199


EARTHQUAKE 

ARCHAEOLOGY 



THE VARIABILITY OF ALONG-STRIKE CO-SEISMIC SLIP: A NEW EXAMPLE FROM THE 

IMPERIAL FAULT OF SOUTHERN CALIFORNIA 

Rockwell, Thomas K., (1,2, Yann Klinger (2) 

(1) Geological Sciences, San Diego State University, San Diego, CA 92182 USA, trockwell@geology.sdsu.edu 

(2) Institute du Physique du Globe, Paris 

Abstract (The variability of along–strike co-seismic slip: A new example from the Imperial Fault of Southern California): 

The 1940 M7 earthquake on the Imperial fault in southern California and northern Baja California produced 70 km of surface 

rupture along a N37W striking right lateral fault. This rupture crossed cultivated fields with long alignments of crop and tree rows, 

along with an orthogonal pattern of canals and roads, and 15 km of the rupture was photographed from the air in high-resolution 

stereo immediately after the earthquake. We determined the precise scale of these vintage aerial photographs and used them to 

assess the short spatial variability of slip along strike. We also analyzed 3 km of very high-resolution aerial photography taken the 

day after the 1979 rupture of the northern half of the Imperial fault. We find that lateral slip varies substantially along strike by 

more than 30% over distances of tens to hundreds of meters. These results are similar to the variability determined after the 1999 

Izmit and Duzce ruptures in Turkey, and the 2010 El Mayor-Cucapah rupture in northern Baja California. 

Key words: surface rupture, Imperial Fault, rupture variability, rupture distribution 

INTRODUCTION 

The short spatial variability of co-seismic slip along 

strike for large strike-slip fault ruptures has generally 

been attributed to the inability of geologists to 

measure off-fault deformation. However, after the 

1999 Izmit and Duzce earthquakes in Turkey, 

Rockwell et al. (2002) demonstrated from offset 

groves of trees that both off-fault near-field 

deformation and lateral variations of co-seismic slip 

along fault strike were significant variables in making 

field measurements of displacement after large 

earthquakes. New optical correlation techniques 

applied to the 2010 El Mayor-Cucapah surface 

rupture in northern Baja California (Hudnut et al., 

2010) confirm that there are both short and long 

wavelength along-strike variations in lateral slip as 

well as substantial off-fault deformation such that 

field measurements of lateral slip consistently 

underestimated the amount of strike-slip, even where 

basement rock is close to the surface and alluvial 

cover is thin. 

In this work, we present new measurements for the 

1940 and 1979 surface ruptures along the Imperial 

fault in southern California (Figure 1), made from 

high-resolution aerial photography. We measured 

over 600 new displacements along 15 km of the 1940 

rupture, and over 300 new displacements along 3 km 

of the 1979 rupture, which are the extents of highresolution 

aerial photography for these ruptures. 

Measurements are mostly on long, linear cultural 

features and crop rows, and are spaced between one 

to tens of meters apart, allowing for resolution of 

variations in lateral slip. We present these new 

measurements and conclude that slip variations are 

real aspects of co-seismic slip along large strike-slip 

faults. We also show that the slip distribution reported 

for the 1940 earthquake underestimates the 

maximum displacement, as field measurements after 

the earthquake were made on widely spaced cultural 

200 

features, such as roads and canals, and the zone of 

maximum displacement was not visited. 

These observations point to the need for field 

geologists to over sample displacement data after an 

earthquake, rather then make random, widely spaced 

measurements that may not represent the actual coseismic 

slip distribution. 

DETERMINATION OF SCALE OF THE AERIAL 

PHOTOGRAPHY 

We determined the scale for the aerial photographs 

by measuring the distance between field boundaries 

(~400 m) and other man-made structures and then 

comparing them to the identical features in Google 

Earth. We made measurements only in the central 

(50%) portion of the stereo pairs, as the affects of 

parallax are minimal in this part of the imagery. We 

then constructed a scale that is accurate to within 

1%. We then used the scaling tool in Adobe 

Illustrator to construct a smaller scale to measure 

offset cultural features. 

The largest uncertainty in both the 1940 and 1979 

data sets is the determination of the alignment of a 

crop row, tree row, road, canal, or other feature. We 

used a straight-line segment to place along a feature 

on one side of the fault, and copied the segment to 

maintain a parallel line for the same feature across 

the fault. The line width is about 10 cm, which is 

about the smallest increment of displacement that 

could be measured for the 1940 imagery. For the 

1979 photography, we could commonly measure to 

about 5 cm. A significant source of uncertainty is the 

actual placement of the line along a feature, which is 

dependent on how straight the feature was, and how 

far it could be extended beyond the near-field fault 

zone. In most cases with crop rows, the plow lines 

were found to be remarkably straight outward from 

the fault for tens to hundreds of meters, and straight 

line segments were easy to place along the middle or


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 1: Location map of the Imperial fault in southern California and northern Baja California. Note that the 1940 earthquake 

ruptured the entire length of the Imperial fault, whereas the 1979 earthquake only ruptured the northern half. Also note the 

paleoseismic sites along the fault, and the other historical surface ruptures. 

edge of a crop row. For the 1979 imagery, we 

estimate that small crop rows had placement 

uncertainties of only 10-15 cm, whereas for the 1940 

imagery, the uncertainties are larger and perhaps as 

much as 0.5 m. In affect, this provided hundreds of 

alignment arrays to be measured along the strike of 

the fault to assess lateral slip variability, with 

reasonably small uncertainty estimates. 

OBSERVATIONS 

edges of a road adjacent to a field was offset about 7 

m (Figure 3). Slip decreased 

in both directions to between 5 and 6.5 m, similar to 

the variability in lateral displacement that we 

observed along the entire photographed portion of 

the rupture. 

Comparison to J.P. Buwalda’s field measurements 

from his field notes 

As a first test of our methods, we located the sites of 

Buwalda’s field measurements that were made 

immediately after the 1940 rupture, of which there 

were only six along the 15 km of rupture captured in 

the aerial photography. Figure 2 shows our slip 

estimates versus Buwalda’s field measurements; 

they are in close agreement except for the All 

American Canal (site B1a) where Buwalda reports 

only a single strand, whereas two strands are clearly 

evident in the imagery and cause displacement of the 

canal, resulting in our larger estimate. 

Maximum Displacement 

Maximum displacement for the 1940 earthquake 

exceeded the reported 6 m of lateral displacement by 

about a meter. In the same general vicinity as 

Sharp’s 6 m measurement, we determined that both 

201 

Fig. 2: Comparison of J.P. Buwalda’s field 

measurements with our estimates from analysis of 

aerial photography. 

Slip Distribution 

Figure 4 shows the slip distribution for the 1940 

earthquake from our new measurements, along with


EARTHQUAKE 

ARCHAEOLOGY 



J.P. Buwalda’s 1940 field measurements in red 

(corrected to a bearing of 323 o ). We also include 

measurements by R.V. Sharp on preserved features 

north and south of the international border (black 

Fig. 3: The 1940 rupture along the Imperial fault north of 

the international border. The All American Canal is just 

meters north of the border. About 2 km NW from the 

border, a road/field boundary is offset about 7 m, the 

maximum displacement that we found along the 1940 

rupture. R.V. Sharp measure about 6 m near the border, 

whereas J.P. Buwalda measured only about 5 m across 

one of two strands. The detail shows the offset, with 

scale included, at our measurement points 38 and 39. 

dots). Finally, we surveyed a line of telephone poles 

that were established prior to the 1940 earthquake at 

Tamaulipas (formerly Cucapa) and resolved 2.7 m of 

right-lateral strike-slip (blue star). Our new 

observations increase the number of slip 

observations for the 1940 surface rupture by over an 

order of magnitude, although the majority are on a 

section of the fault that represents only 20-25% of the 

full length of the rupture. 

An important observation on slip distribution is that 

by Rockwell et al. (2002) for the 1999 Izmit and 

Duzce earthquakes. This degree of variability has 

been observed for many previous earthquakes, 

including the 1999 Hector Mine earthquake (Treiman 

et al., 2002), the 1992 Landers earthquake (Sieh et 

al., 1993), the 1987 Superstition Hills earthquake 

(Sharp et al., 1989), and many others. However, in 

most previous cases, slip measurements were made 

on small geomorphic features such as channel 

margins, channel thalwegs, alluvial bars, canyon 

walls, etc., none of which were linear for any distance 

from the fault. Consequently, near-field off-fault 

deformation could not be assessed and it was 

generally assumed that some of the variability was 

the result of non-quantified offset. In contrast, 

measurement of long crop rows, tree lines, roads, 

fences and other long cultural features along the 

1940 surface rupture demonstrates that significant 

lateral variations in displacement are real, and that 

they occur over short spatial dimensions. For 

instance, we determined displacement for adjacent 

rows of trees and crops for entire fields. In some 

cases, as in figure 5, the variability occurs at about 

our estimated resolution of displacement uncertainty, 

a half meter. In this case, we measured offset of 

individual tree lines to vary by about 1 m over a 

lateral distance of a few hundred meters, similar to 

the variability on offset tree lines in the Izmit 

earthquake, but one could argue that within the 

stated uncertainty, these measurements agree. 

However, other examples, such as the offset crop 

rows in figure 6, we measured the offsets to +10 cm, 

and estimate the uncertainty to about +20 cm. In this 

case, lateral slip varied from zero to over a meter 

along a several hundred meter section of rupture. 

The overall degree of variability along strike is 

Fig. 4: Slip distribution for the 1940 Imperial fault rupture. Our new observations are between the international border (zero 

point) and about km 14 north of the border. J.P. Buwalda’s field measurements are plotted in red, R.V. Sharp’s estimates of 

displacement are plotted as black dots, and our measurement of an offset telephone line is the blue star. 

there are significant lateral variations in displacement 

over short spatial dimensions, similar to that reported 

202 

evident in the slip distribution curve in figure 4. Areas


EARTHQUAKE 

ARCHAEOLOGY 



of larger offset tend to have greater variations in 

displacement, even though resolution of displaced 

features and estimates of uncertainty remain about 

the same. From these observations, we conclude 

Fig. 5: Offset grove of trees between 2.1 and 2.4 km NW 

of the international border. A dot was placed on the 

center of each tree, and the dots were regressed to 

resolve lateral displacement. Uncertainty is estimated at 

about 0.5 m. 

much of this is rarely accounted for in the 

measurement of non-linear geomorphic features but 

it is clearly evident for the Imperial ruptures. 

Based on the collection of nearly 1000 new 

displacement measurements from the 1940 and 1979 

surface ruptures along the Imperial fault, several 

generalizations can be made. First, the sparse field 

data from the 1940 earthquake, nearly all of which 

was collected at convenient road crossings, missed 

the maximum displacement of about 7m, as well as 

the maximum displacement along strike, although 

most of the field observations fall near the average 

displacement for a section of fault. It is evident from 

this analysis that dense collection of displacement 

data is required to quantify the maximum 

displacement along a rupture, and that sparse data 

likely miss such displacement peaks and other 

details of the rupture. 

Second, lateral variations along strike are substantial 

and show similar variability to that documented with 

survey data after several recent earthquakes. Offfault 

warping is evident in the bending of crop rows, 

and some of this bending may be missed in near-field 

field mapping of a rupture after an earthquake. Using 

long agricultural features as closely spaced 

alignment arrays allows for both the assessment of 

total slip at a point along the fault, as well as the 

lateral variability of displacement along the fault. The 

variability that we document here is similar in 

magnitude and spatial scales to that documented 

after the Izmit and Duzce earthquakes, as well as 

more recently with optical imaging techniques after 

the 2010 El Mayor-Cucapah earthquake in northern 

Baja California (Hudnut et al., 2010). 

References 

Fig. 6: Offset crop rows between 11.7 and 12 km NW of 

the international border. Crops rows could be measured 

to 10-15 cm resolution, but offsets vary by over a meter. 

that lateral variability of displacement typically varies 

by as much as 30% along a section of rupture. 

DISCUSSION AND CONCLUSIONS 

Until the advent of pre- and post-earthquake 

comparisons of LiDAR and optical imagery data, 

most measurements of lateral displacement along 

strike-slip faults after large earthquakes were 

conducted on nearfield, non-linear features such as 

rills, stream channels, channel margins, bars, and 

other common geomorphic features (Clark et al., 

1972; Sharp, 1982; Sieh et al., 1993; Treiman et al., 

2002; Barka et al., 2002). Surveying of long cultural 

features such as tree rows, telephone pole arrays, 

and fence lines demonstrate that significant off-fault 

warping can account for a substantial amount of the 

near-field strike slip (Rockwell et al., 2002), and 

Barka, A., Akyuz, H. S. et al. 2002. The surface rupture and 

slip distribution of the 17 August 1999 Izmit earthquake 

(M7.4), North Anatolian fault. Bulletin of the 

Seismological Society of America (Special Issue on the 

1999 Izmit and Duzce, Turkey, Earthquakes, N. Toksoz 

(ed.)), 92(1), 43–60. 

Clark, M.M., 1972, Surface rupture along the Coyote Creek 

fault, in The Borrego Mountain Earthquake of April 9, 

1968, U.S. Geol. Sur. Prof. Pap. 787, p. 55-87. 

Hudnut, K.W., J.M. Fletcher, T.K. Rockwell, J.J. Gonzalez- 

Garcia, O. Teran, and S.O. Akciz, 2010, Earthquake 

rupture complexity evidence from field observations. EOS 

Transactions, Fall AGU, T51E-02. 

Rockwell, T. K., S. Lindvall, T. Dawson, R. Langridge, W. 

Lettis, Y. Klinger, 2002, Lateral offsets on surveyed 

cultural features resulting from the 1999 Izmit and Duzce 

earthquakes, Turkey. Bulletin of the Seismological 

Society of America, v. 92, no. 1, pp. 79-94. 

Sharp, R.V., 1982, Comparison of 1979 surface faulting with 

earlier displacements in the Imperial Valley, in The 

Imperial Valley, California earthquake of October 15, 

1979: U.S. Geological Survey Professional Paper 1254, 

p. 213-221. 

Sieh, K., L. Jones, et al., 1993, Near-field investigations of 

the Landers earthquake sequence, April to July 1992, 

Science v. 260, pp. 171–176. 

Treiman, J.A., Kendrick, K.J., Bryant, W.A., Rockwell, T.K., 

and McGill, S.F., 2002, Primary surface rupture 

associated with the Mw 7.1 16 October 1999 Hector Mine 

earthquake, San Bernardino County, California. Bulletin 

of the Seismological Society of America, v. 92, no. 4, pp. 

1171-1191. 

203


EARTHQUAKE 

ARCHAEOLOGY 



EARTHQUAKE ARCHAEOLOGICAL EFFECTS GENERATED BY THE LISBON 

EARTHQUAKE (FIRST OF NOVEMBER 1755) IN THE CORIA´S CATHEDRAL (CÁCERES, 

WESTERN SPAIN) 

Rodríguez-Pascua M.A. (1), P.G. Silva (2), Perucha Atienza, M.A. (1), J.L. Giner-Robles (3), R. Pérez-López (1) 

(1) Instituto Geológico y Minero de España. Street Ríos Rosas, 23. 28003-Madrid. SPAIN. E-mail: ma.rodriguez@igme.es, 

ma.perucha@igme.es ; r.perez@igme.es 

(2) Dpto. Geología, Escuela Politécnica Superior de Ávila, Universidad de Salamanca. Avda. Hornos Caleros, 50. 05003-Ávila. 

pgsilva@usal.es 

(3) Dpto. Geología. Facultad de Ciencias. Universidad Autónoma de Madrid. Cantoblanco. Tres Cantos. Madrid. SPAIN. E-mail: 

jlginer@gmail.es 

Abstract (Earthquake Archaeological Effects generated by the Lisbon Earthquake (first of November 1755) in the Coria´s 

Cathedral (Cáceres, western Spain): The Lisbon Earthquake was the most destructive earthquake in the Western European 

history. This earthquake affected the entire Iberian Peninsula and the city of Lisbon completely collapsed. The intensity of this 

earthquake was X (EMS-1998) and damaged the historical buildings of Spain. These effects are preserved in historical buildings, 

like the Coria’s Cathedral. The damage observed in this cathedral is described by using the new classification of Earthquake 

Archaeological Effects (EAE), with the aim to study both strain structures and the seismic wave pattern. 

Key words: Lisbon Earthquake (1755), Coria´s Cathedral, Earthquake Archaeological Effects (EAE). 

INTRODUCTION 

The Lisbon earthquake (November 1 st , 1755) is the 

largest earthquake that struck Western Europe in 

historic times. This earthquake affected the 

population in a physical sense and also changed the 

knowledge of the scientific origin of earthquakes. 

This date was the starting point of foundation of 

modern seismology. The Lisbon earthquake affected 

the entire Iberian Peninsula and North Africa. Its 

shaking was felt in Central European countries like 

Germany (Martínez Solares, 2001). Nowadays the 

epicentre of the earthquake is still the subject of 

scientific debate (Gutscher, 2005), although the 

approximate location is widely assumed at the 

southwest of San Vicente Cape, regardless of the 

exact fault that produced it. The maximum intensity 

of this earthquake was X (EMS-1998) (Martínez 

Solares and Mezcua, 2002), being located in 

southern Portugal (Algarve Coast). Some of those 

seismic intensity effects from the earthquake are still 

visible in the historic heritage all around the Iberian 

Peninsula, such as churches and cathedrals. This is 

the case study of the Coria´s Cathedral (Cáceres, 

central part of Spain). The building suffered structural 

damage, even collapse the cupola of the tower. 

GEOGRAPHICAL SETTING 

The town of Coria is located to the NW of the city of 

Caceres in the central Western part of Spain and 

near of the border with Portugal. The isoseismal map 

of the Lisbon Earthquake (Martínez-Solares, 2001) 

suggests that Coria and surroundings were affected 

by a seism intensity VI (Fig. 1). The main damage in 

the city was produced in the cathedral, in which the 

collapse of the cupola of the tower killed thirteen 

people. 

Fig. 1: Location of the town of Coria on the isoseismal 

map of the 1755 Lisbon earthquake (Intensity scale 

EMS-1998) (after Martínez Solares, 2001). 

METHODOLOGY 

Some of the effects of the earthquake in the 

cathedral of Coria were documented by the Dean of 

the Cathedral (Martínez-Vázquez, 1999) and are 

available for public consult in the archive of the 

Cathedral. Hence, we have elaborated a list of 

structural damage using the classification of 

Earthquake Archaeological Effects (EAE) 

(Rodríguez-Pascua et al., 2011). However, we have 

to bear in mind that the Cathedral is also affected by 

geotechnical problems that could mask the effects of 

the earthquake of Lisbon. Those have to be 

discriminated by using existing documentation, both 

the earthquake and previous geotechnical studies 

(Martínez-Vázquez, 1999). 

204


EARTHQUAKE 

ARCHAEOLOGY 



Rotated and displaced drums in columns and 

pinnacles: 

12- Clockwise rotations of the pinnacles 

(hexagonal pyramidal base) of the north 

front of the Cathedral (Fig. 6). 

13- Collapse of the tower’s cupola. 

Displaced masonry blocks: 

14- Sinistral displacement of masonry blocks in 

the columns of the organ (Fig. 7). 

15- Displacement of masonry blocks in the 

tower’s balcony. 

Fig. 2: Plan of the Coria´s Cathedral. The red 

numbers are the location of the different EAEs (see 

the text for further explanation). 

DATA 

According to the classification of damage proposed 

by the EAE scale (Rodríguez-Pascua et al., 2011), 

the effects of the Lisbon earthquake catalogued in 

the Cathedral of Coria are the following (see Fig 2 for 

spatial location): 

Penetrative fractures in masonry blocks: 

1- Cracks in the East front. Existing prior to the 

earthquake and enhanced by this. 

2- Cracks in the lintel of the north entrance and 

in the north front of the Cathedral. 

3- Cracks in the west main front separated by 

the central column which differentiates the 

two entrance arches of the door (Fig. 3). 

4- Cracks in the ceilings and vaults of the 

central nave of the building. 

Dropped key stones in arches or lintels in windows 

and doors: 

5- Arches of the bell tower with horizontal 

displacement of key stones in the segments 

to the arcs of the N front of the tower. 

6- Arch of the north entrance with dropped 

lintel. 

Collapsed walls or balustrades: 

7- Collapsed balustrade and pinnacles of the 

“Relics Balcony” (Fig. 4). 

8- Fall down of the balustrade’s pinnacles in 

the south terrace of the Cathedral (Fig. 5A). 

9- Fall down of the balustrade’s pinnacles in 

the north and south front of the Cathedral 

(Fig. 5B). 

10- Collapsed balustrades of the balconies in 

the south front subsequently repaired with 

bricks. 

11- Fall down of the tower’s pinnacles. 

Fig. 3: Cracks in the W main front of the Coria´s 

Cathedral. A) View before restoration; B) view after 

restoration in 2009. 


The intense seismic damage of the Coria Cathedral 

could be attributed to its topographic location: on the 

cliff edge overlooking the Alagón River’s flood plain. 

The average orientation of cracks (NE-SW) to about 

45° to the central axis of the building, suggest a 

sinistral shear in the building. All of this data 

(checked by epoch documents) could be applied in 

other buildings affected by the Lisbon Earthquake. 

205


EARTHQUAKE 

ARCHAEOLOGY 



As a principal geological effect of the Lisbon 

Earthquake within the area, the river modified his 

course and changed the channel by moving to the 

south. Consequently, the medieval “Stone Bridge” 

was out of use and today you can ask yourself why 

medieval people had built a heavy stone bridge in the 

middle of the cultivation land? 

Fig. 5: A) Fall down of the balustrade´s pinnacles in 

the south terrace of the Cathedral B) decapitated 

pinnacles in the north front of the Cathedral. 

Fig. 4: Collapsed balustrade and pinnacles of the 

“Relics Balcony”. 

Acknowledgements: Thanks are given to the project 

ACI2009-1037 (MICINN). 

References 

Gutscher, M.A. 2005. What caused the Great Lisbon 

Earthquake? Science, 305: 1247-1248. 

Martínez-Solares, J.M. 2001. Los efectos en España del 

terremoto de Lisboa. Instituto Geográfico Nacional. 

Ministerio de Fomento. 756 pp. 

Martínez-Solares, J.M. & Mezcua, J. 2002. Catálogo 

sísmico de la Península Ibérica (880 a.C.-1900). Instituto 

Geográfico Nacional. Ministerio de Fomento. 756 pp. 

Martínez-Vázquez, F. 1999. El terremoto de Lisboa y la 

Catedral de Coria (Vicisitudes del Cabildo) 1755-1759. 

Colección Temas Cauriaciences. Vol. V. Ed. 

Ayuntamiento de Coria. 187 pp. 

Rodríguez-Pascua, M.A.; Pérez-López, R.; Giner-Robles, 

J.L.; Silva, P.G.; Garduño-Monroy, V.H. & Reicherter, K. 

2011. A comprehensive classification of Earthquake 

Archaeological Effects (EAE) in archaeoseismology: 

Application to ancient remains of Roman and 

Mesoamerican cultures. Quaternary International, DOI: 

10.1016/j.quaint.2011.04.044. 

Fig. 6: Clockwise rotations of the pinnacles 

(hexagonal pyramidal base) of the north front of the 

Cathedral. 

Fig. 7: Sinistral displacement of masonry blocks in the 

columns of the organ. 

206


EARTHQUAKE 

ARCHAEOLOGY 



NEOTECTONIC OF THE LONGITUDINAL FAULT SYSTEM IN SOUTHERN 

COSTA RICA 

Rojas, Wilfredo (1), Nury Simfors-Morales (2), Luis Sanez (3), Åke Sivertun (4) 

(1, 3) Escuela Centroamericana de Geología, University of Costa Rica. Email: wrojas@geologia.ucr.ac.cr 

(2, 4) Department of Computer Science, University of Linköping, Sweden. Email: nury.simfors2@comhem.se 

Abstract (Neotectonic of the longitudinal fault system in Southern Costa Rica): Southern Costa Rica is one of the most 

seismically active areas in Costa Rica because the subduction between Cocos, Caribbean and Nazca Plates. Several active fault 

zones go also across the study area. One of this fault zones is the WNW-ESE Longitudinal fault system (LFS) which is the longest 

fault in southern Costa Rica. The goal of this study is to find a suitable place to use it as a sanitary landfill site. The results indicate 

that there are several geological structures in the study area that indicate deformation due active faulting but more detail studies 

are necessary in order to understand better the seismogenic potential of this fault. Future work in southern Costa Rica will take 

into account these issues. 

Keywords: Trenching, Neotectonics, Longitudinal, Subduction 

INTRODUCTION 

Most of the seismic activity in southern Costa Rica is 

due to the subduction between Cocos, the Caribbean 

and the Nazca Plates. However, the area is also 

characterized by several active fault systems which 

cut across southern Costa Rica, causing also seismic 

activity and serious damage during strong 

earthquakes. The Longitudinal Fault System (LFS) is 

the longest fault system in southern Costa Rica with 

a length of about 167.2 km (Mann & Corrigan, 1990), 

(Kolarsky et. al., 1995), (Cowan et al., 2001) and 

(Montero, 1998) (Fig. 1). It is an ESE-striking obliquereverse 

fault, parallel to the volcanic arc in Costa 

Rica and western Panama (Montero, 1994). The 

recurrence of earthquake events of this fault is not 

well known, while the fault slip rate may be roughly 

estimated at 15 mm/year in Costa Rica and 10 

mm/year in Panamá (Cowan et al., 2001). 

The objective of this preliminary work is to find a 

suitable place to use it as a sanitary landfill site. Two 

trenches were open on the outskirts of Rio Claro, 

southern Costa Rica (Fig. 1). The results indicate that 

there are several geological structures in one of the 

excavated trenches that indicate deformation due to 

active faulting. Two other fault scarps on the way to 

Ciudad Nelly cited in this work were also visited. 

They could be potential trenching targets in case 

more studies will be done in the future. Taking into 

account the importance of this fault in southern Costa 

Rica and western Panama is recommended to do 

more studies and apply dating techniques in order to 

better understand the seismogenic potential of the 

fault and the damage it could pose to the 

communities. The University of Costa Rica (UCR) 

has a seismic monitoring project in southern Costa 

Rica which is studying the spatial and temporal 

seismicity and earthquake rupture processes in the 

region. The LFZ and other important faults in the 

region will be studied in more detail in a close future. 

Fig. 1: Digital Elevation Model of Costa Rica showing the 

location of southern Costa Rica. In the map two triangles 

show the location of Rio Claro and Ciudad Nelly. The 

trenches were excavated in Rio Claro (see: Fig.3). The fault 

scarp follows toward the NE and ca. 1 kilometer from the 

trench side the fault scarp can be seen clearly along the 

smooth plain of southern Costa Rica ( Fig. 4 and 5). On the 

outskirts of Ciudad Nelly was also found two interesting 

places where the LFZ could be studied in the near future 

(Fig. 6). (Map courtesy of University of Costa Rica). 

207


EARTHQUAKE 

ARCHAEOLOGY 



HISTORIC SEISMICITY IN SOUTHERN COSTA 

RICA 

The spatial distribution of the main historical 

subduction related earthquakes (1803, 1854, 1904, 

1941 and 1983) which have affected southern Costa 

Rica over the past 200 years have been generated 

by the subduction of the Cocos plate along its 

interface with Panama Microplate. Although most of 

the historic earthquakes are subduction related, few 

of them are due to faulting. Subduction earthquakes 

rupture at shallow focal depths between 28 and 34 

km and magnitudes around 7.0 to 7.5 (Rojas et al., 

1993) and (Morales-Simfors et al., submitted). 

According to the historic earthquakes there is a 

medium-term probability of a strong earthquake Mw 

7.3 in the cited area in the near future. The 

recurrence interval of strong events (Mw ≥ 7.0) is in 

the order of 45 years. In case a big earthquake 

occurs in southern Costa Rica, damage is expected 

with a Mercalli Intensity of VIII (Rojas 2008) and this 

may cause significant casualties in the communities. 

Fig. 2: Simplified geological map of southern Costa Rica. It 

shows sedimentary rocks in brown (Mesozoic-Paleozoic), 

Volcanic Rocks in green (Miocene-Pliocene), Volcanic 

Quaternary rocks in dark green. The small map in the left 

corner shows the major fault zones in the study area (Maps 

courtesy of UCR). 

THE LONGITUDINAL FAULT SYSTEM (LFZ) 

The WNW-ESE trending Longitudinal fault system 

extends from western Panama to central Costa Rica 

in the west (Mann & Corrigan, 1990) (Fig.2). The slip 

rate of the Longitudinal fault zone may be


EARTHQUAKE 

ARCHAEOLOGY 



One of the trenches presents a clear evidence of 

geological structures due to active faulting. 

Fig. 5: A. View of the fault scarp at NE of the trench site at 

Rio Claro (Coordinates: 569.405 291.319, Piedras Blancas 

Sheet, 1.50000). B. The scarp is smaller than the scarp 

close to the trench in figure 2. This fault scarp is ca. 2 

meters high and it extends several kilometers along the 

plain in Rio Claro. C. The fault trace goes almost 

perpendicular across the road. It is possible to see a narrow 

change in the dip of the road. 

The Fig. 3B and C shows a close view of the grayblue 

deformed clay layer in the trench which is 

overturned and faulted by two thrust faults and other 

secondary normal faults but in order to do a better 

interpretation of the tectono-stratigraphic sequence of 

this site, a detail study of the fault is necessary. The 

LFZ and other faults in the region appears to be in 

obvious proximity with important infrastructure such 

as hospitals, schools, roads, highways etc. in 

southern Costa Rica and western Panama, In order 

to understand the active tectonics in the area, more 

detail studies are necessary to evaluate the 

seismogenic potential of the fault and answer other 

questions about the age of the fault scarps, 

earthquake recurrence and seismic hazard in 

southern Costa Rica. 

Acknowledgements: The authors thank Peter Bryer for 

helping us with the English manuscript. We are also grateful 

to the Municipality of Golfito, southern Costa Rica and the 

Osa-Golfo Dulce Research Program (PIOSA), University of 

Costa Rica 

References 

Fig. 6: View of fault scarp at Abrojo, Costa Rica 

(Coordinates: 582.785 and 286146, Canoas topographic 

sheet, 1 50000). The scarp goes almost parallel to the 

highway from Costa Rica to Panama. 

The second place was 1 kilometer away from the 

trench site, where the fault scarp can be seen clearly 

and follows towards the NE several kilometers along 

the smooth plains in southern Costa Rica (Fig. 4). 

The displacement of this scarp is ca. 2 meters (Fig. 

5). The third point was outside Ciudad Nelly 30 

kilometers to the SE of the trench side where it is 

possible to see several geological structures from the 

main highway that communicate Costa Rica with 

Panama. In this site the scarp is ca 2 meters high 

(Fig. 6). 

RESULTS AND RECOMMENDATIONS 

Two trenches were excavated in the study area in 

order to find a suitable site for a sanitary landfill site. 

Cowan H, M., N. Machette., K.M. Haller & R.L. Dart (2001). 

Map and Database of Quaternary Faults and Folds in 

Panama and Its Offshore Regions. Open-File Report 98- 

779. 41 pp. 

Fisher, D. M., T.W, P. Gardner., P. Sak; J. Marshall & M. 

Protti (2001). Uplift patterns in the Forearc of the Middle 

America Trench, Costa Rica: implications for mass 

balance and fore-arc kinematics. Eos, Transactions of the 

American Geophysical Union, 82, F1151. 

Kolarsky, R.A. & P. Mann (1995). Structure and 

neotectonics of an oblique-subduction margin, 

southwestern Panama. In: Geologic and tectonic 

development of the Caribbean Plate boundary in 

Southern Central America: (Mann, P. ed.). Geological 

Society of America Special Paper 295, p. 131-157. 

Mann, P. & J.D, Corrigan (1990). Model for late Neogene 

deformation in Panama: Geology, 8, 558-562. 

Montero, W. (1994). Neotectonics and related stress 

distribution in a subduction-collisional zone, Costa Rica. 

In: Geology of an evolving island arc (Seyfried, H. and 

Hellmann, W, eds.). Profil (University of Stuttgart, 

Germany, 7, 125-141. 

Montero, W., P. Denyer., R. Barquero., G.E. Alvarado., H, 

Cowan., M.N, Machette., K.M, Haller & R.L.Dart (1998). 

Map and database of Quaternary faults and folds in 

Costa Rica and its offshore regions: U.S. Geological 

Survey Open-File Report 98-481, 63 p., 1 plate 

(1:750,000 scale). 

Mora, S (1979). Estudio geológico de una parte de la región 

sureste del Valle del General, Provincia Puntarenas, 

Costa Rica: San José, University of Costa Rica, unpubl. 

Senior Undergraduate thesis, 185 p. 

Morales-Simfors, N; et. al. A contribution to the study of the 

Osa-Golfo Dulce Seismicity, Costa Rica: March 11-14 

earthquakes. Int. J. Earth Sci. (Submitted). 

Rojas, W., Bungum, H & C, Lindholm (1993). Historical and 

recent earthquakes in Central America. Revista 

Geológica de América Central, 16, 5-21. 

Rojas, W. (2008). Informe anual del proyecto de 

investigación de monitoreo sísmico de Zona Sur de 

Costa Rica VI-113-A7-170, RSN, San José, 13 pp. 

209


EARTHQUAKE 

ARCHAEOLOGY 



GEODETIC STUDIES IN THE ZAFARRAYA FAULT (BETIC CORDILLERAS) 

Ruano, Patricia (1,2), Antonio J. Gil (3), Jesús Galindo-Zaldívar (1,2), Gracia Rodríguez-Caderot (4), María Clara de Lacy (3), 

Antonio M. Ruiz (3), María Jesús Borque (3), Juan A. Armenteros (3), Antonio Herrera (3), Antonio Jabaloy (1), Angel C. López- 

Garrido (2), Antonio Pedrera (5), Carlos Sanz de Galdeano (2) 

(1) Dept. Geodinámica, Universidad de Granada. Campus Fuentenueva 18071-Granada, Spain. Email: pruano@ugr.es 

(2). Instituto Andaluz de Ciencias de la Tierra, Facultad de Ciencias, 18071 Granada, Spain 

(3). Dept. Ing. Cartográfica, Geodésica y Fotogrametría. E.P.S., Universidad de Jaén, Campus Las Lagunillas, 23071 Jaén, 

Spain. 

(4). Dept. Astronomía y Geodesia, Universidad Complutense de Madrid, Ciudad Universitaria s/n, 28040 Madrid, Spain. 

(5). Instituto Geológico y Minero de España. C/ Alcázar del Genil, 4. 18006 Granada, Spain. 

Abstract (Geodetic studies in the Zafarraya fault (Betic cordilleras): The Zafarraya Fault Zone constituted the seismic source 

of the 1884 Andalucian earthquake with an estimated magnitude of 6.5 – 7. It is a normal fault zone located in the northern limb of 

the Sierra Tejeda antiform and intersects the Internal and External Zone boundary of the Betic Cordillera. Two non- permanent 

GPS networks made up of 16 sites have been installed and surveyed in September 2004 and February and July 2010 in order to 

constrain the present-day motion of these structures. Result in 6 years elapsed time may help to discuss about coseismic and 

aseismic motions of the fault and elastic deformation before faulting. 

Key words: GPS network, Zafarraya Fault, Sierra Tejeda antiform, present-day deformation 

THE ZAFARRAYA FAULT AND SIERRA TEJEDA 

ANTIFORM 

The Betic Cordillera is located in the Western 

Mediterranean built up by the distributed 

deformations related to the Eurasian-African plate 

boundary and characterized by a present day 

moderate seismicity. 

The Zafarraya Fault Zone is located in the Southwest 

of Granada Basin, which is one of the most important 

Neogene intramountain basins in the Cordillera that 

cover the Internal and External zones boundary. It is 

placed at the northern limb of the Sierra Tejeda 

antiform, and could be interpreted as a collapse 

structure developed along the external arch of the 

uplifted fold. 

The main normal fault trends roughly E-W and 

includes other minor faults like Llanos de la Dona 

Fault. The whole fault system constitutes a set of 

fractures extending 23 km in length and 4 km in width 

that runs obliquely to the boundary between the 

External and Internal zones. The main fault segment 

has associated a rectilinear mountain front formed in 

Jurassic limestones of the External Zone and Triassic 

marbles of the Internal Zone. This mountain front is 

divided into several segments of E-W to NW-SE 

orientations. The kinematics of the fault is mainly 

normal with a minor dextral component as can be 

deduced by the observed slicken-lines. This fault 

controls the main features of the Zafarraya polje, an 

endorheic area developed in its hanging wall, and 

filled by Tortonian to Quaternary sediments (López- 

Chicano et al., 2002). The throw of this fault zone is 

1500 m, considering the displacement of the 

External-Internal zone boundary. Taking into account 

210 

the regional geological setting, most of the slip 

probably occurred since the Tortonian. The Zafarraya 

Fault was the causative fault of the largest historical 

earthquake registered in Spain (the Andalusian 

Earthquake, December 25, 1884) with a maximum 

intensity of X (MSK scale) from which a magnitude of 

6.5 - 7 has been calculated (Muñoz and Udías, 1981) 

with an estimated total rupture of 16 km. 

The slip-rate of the fault calculated from geological 

markers (10 Ma, Tortonian) is 0.125 mm/yr (Sanz de 

Galdeano et al., 2003) and 0.17 mm/yr (Reicherter et 

al., 2003). Whereas, from paleoseismological studies 

the slip-rate estimated is 0.35 mm/yr (Reicherter et 

al., 2003), 0.3 and 0.45 mm/yr with a recurrence 

period of 2-3 kyr for major, surface rupturing 

earthquakes (Reicherter et al., 2010). 

THE GPS NETWORKS 

Two non-permanent GPS networks were installed in 

Zafarraya Fault Zone and Sierra Tejeda antiform in 

2004. The 16 GPS sites (Fig. 1) are located in a local 

and a regional network that extends up to the coast 

line in order to study the local motion along the 

Zafarraya fault and the regional development of the 

Sierra Tejeda antiform. These networks were made 

up of sixteen reinforced concrete pillars anchored to 

rock with an embedded forced centring system to 

assure that the antennas are placed exactly at the 

same position in different reoccupations. The local 

network comprises the sites 811, 812 and 816, 

located on the hanging wall of the fault. Most of them 

were built up on Jurassic limestones of the External 

Zones. The sites 810, 813, 814 and 815 are located 

in the footwall, and were build-up mainly on 

limestones and marbles. Although these sites are


EARTHQUAKE 

ARCHAEOLOGY 



located across the contact between External and 

Internal Zones, this contact is inactive at Present, 

and the southern part of the network may constitute a 

reference for the activity of the most recent Zafarraya 

Fault. The regional network that extends from the site 

800, located on Jurassic limestones of the External 

Zones, crosses southwards the Zafarraya Fault and 

the contact between External and Internal Zones and 

reaches the uppermost part of Sierra Tejeda in the 

site 850, located on Triassic Alpujarride marbles. The 

southern part of the network is located along the 

southern limb of Sierra Tejeda, which is deformed by 

NW-SE oriented normal faults, and reaches the coast 

line (sites 890 and 880, builds up respectively on 

Alpujarride metapelites and marbles). The regional 

network also covers the WNW periclinal end of the 

Sierra Tejeda antiform (Fig. 1). 

All sites meet the following requirements: no 

obstruction above 15 degrees; no high power lines 

nearby; easily accessible. 

The first survey was done in September 2004 

(Borque et al. 2005) and a. second survey in 

February 2010, for local network and in July 2010 for 

regional one. 

The GPS constellation was tracked throughout a 

three-day campaign with 24-hour sessions per day in 

the local network and a six-day campaign in two 

settles with 2 shared sites in the regional one. 

For data acquisition we used 6 dual frequency carrier 

phase GPS receivers Leica System 1200, consisting 

of GX1230 receivers and AX1202 antennas. 

The GPS data processing was performed by using 

Bernese 5.0 software in the following way: single 

sessions were computed in multibaseline mode. The 

first step (preprocessing) related to receivers clocks 

calibration, performed by code pseudoranges, and 

detection and repair of cycle slips and removal of 

outliers, was carried out simultaneously for L1 and L2 

data. The final solution for each session was 

obtained using the iono-free observable with precise 

ephemeris and absolute antenna phase centre 

variation files. The fixed solution of the coordinates 

was estimated using the QIF method to fix integer 

ambiguities. Troposphere parameters every two 

hours were estimated. 

Fig. 1: Geological map and GPS network locations of Sierra Tejeda and Zafarraya fault. 

211


EARTHQUAKE 

ARCHAEOLOGY 




These geodetic surveys provide new insights on the 

Zafarraya fault (ZF) and Sierra Tejeda antiform 

(STA). Slow-motion faults need long periods of 

measurements to have suitable information. 

The obtained results for both networks suggest a 

slow to moderated motion of the structures during the 

last 6 years. In Figure 2 are shown the present-day 

crustal deformation rates in terms of the annual 

WNW trend. There are small differences due to the 

local activity of tectonic structures. To facilitate the 

interpretation of the presented results apart from a 

Eurasia fixed reference frame the results are 

presented (blue arrows, Fig. 2) in a frame were we fix 

the station 810, shared by both networks and located 

southward ZafarRaya fault. This way, it is easy to 

appreciate the motion of ZF and STA. South of the 

ZF most of the statiofs show very small northward 

motion that sugest a very slow activity of STA that 

constitutes the footwall of the fault. However, in the 

Fig. 2: GPS Velocity vectors in mm/yr with standard error ellipses. Black arrows: Eurasia fixed reference frame; Blue 

arrows: Site 810 fixed reference frame. The seismicity from September 2004 to July 2010 from IGN data base is shown. 

velocity vectors with Eurasia fixed reference frame. 

These deformation rates are in order of several mm 

per year. Although velocity vectors must be quoted 

with its standard ellipses even largest of some annual 

velocity vector, all the vectors are consistent with a 

hanging wall the behavior od the sites is variable. 

While site 812 and the western end underline its 

dextral character, sites 811 and 816 may show a 

combination of motion with a cover NW-SE normal 

fault producing active NW-SE extension. The sites 

212


EARTHQUAKE 

ARCHAEOLOGY 



820 and 830 located in its eastern area show its 

normal motion. Toward the North, the vector of site 

800 could be related to the dextral motion of the 

Llanos de la Dona Faults. 

From these results a NW-SE to N-S extension up to 

2 mm/yr could be estimated for the Zafarraya fault 

zone. These rates are higher than those estimated by 

paleoseismological studies. Moreover, the 

paleoseismic data suggests that recent activity is 

higher that neotectonic activity. However, it is not 

well-constrained the importance of creep motions in 

the fault and the accommodation of elastic 

deformation in the area before seismic activity. 

The historical and instrumental (Fig. 2) data support 

that Zafarraya fault has a seismic activity with 

discontinuous slip-rate. The occurrence of at least 

two seismic series in the area (Sep 2005 and Aug- 

Sep 2007), during elapsed time between surveys, 

could be the responsible of these highest rates that 

confirm the coseismic behavior of the ZF. 

The present and future results will be very significant 

for a better understanding of the active tectonic 

structure interaction providing short-term mode of 

deformation and slip rates of one the most active 

sectors of the Betic Cordilleras. 

Acknowledgements: Projects CSD2006-00041, CGL- 

2008-03474-E/BTE, CGL2010-21048, P09-RNM-5388 

AYA2010-15501, and research groups of Junta de 

Andalucía RNM282, RNM370 and RNM148 are 

acknowledged. 

References 

Borque, M.J., Galindo-Zaldívar, J., Gil, A.J., Jabaloy, A., 

Lacy, M.C., López, A.C., Rodríguez-Caderot, G., Ruiz, 

A.M., Sanz de Galdeano, C. (2005): Establishment of a 

non-permanent GPS network to monitor the deformation 

in Zafarraya Fault and Sierra Tejeda Antiform (Spain). 

Física de la Tierra, 17, 23-31. 

López Chicano, M.; Calvache, M. L.; Martín-Rosales, W. y 

Gisbert, J. (2002). Conditioning factors in flooding of 

karstic poljes—the case of the Zafarraya polje (South 

Spain). Catena, 49, 331-352. 

Muñoz, D., Udias, A. (1981) - El Terremoto de Andalucía 

del 25 de Diciembre de 1884. Instituto Geográfico 

Nacional, Madrid. 

Reicherter, K., Jabaloy, A., Galindo-Zaldívar, J., Ruano, P., 

Becker-Heidmann, P., Morales, P., Reiss, S. & González- 

Lodeiro, F. (2003) Repeated palaeoseismic activity of the 

Ventas de Zafarraya fault (S Spain) and its relation with 

the 1884 Andalusian earthquake. Int. J. Earth Sci., 92, 

912-922. 

Reicherter, K., Jabaloy, A., Galindo-Zaldívar, J., Becker- 

Heidmann, P. & Sanz de Galdeano, C. (2010). Trenching 

results of the Ventas de Zafarraya Fault (Southern 

Spain). In: Contribución de la Geología al Análisis de la 

Peligrosidad Sísmica (J.M. Insua y F. Martín-González, 

eds.). Sigüenza (Guadalajara, España). 125. 

Sanz de Galdeano, C., Pelaéz Montilla, J.A. & López 

Casado, C. (2003) Seismic Potential of the Main Active 

Faults in the Granada Basin (Southern Spain). Pure and 

Applied Geophysics, 160, 1537-1556. 

213


EARTHQUAKE 

ARCHAEOLOGY 



NEOTECTONIC ACTIVITY OF THE GRANADA BASIN – 

NEW EVIDENCE FROM THE PADUL-NIGÜELAS FAULT ZONE 

Rudersdorf, Andreas, Jochen Hürtgen, Christoph Grützner, Klaus Reicherter 

Lehr- und Forschungsgebiet Neotektonik und Georisiken, RWTH Aachen, Lochnerstraße 4-20, 52056 Aachen (Germany). 

E-Mail: andreas.rudersdorf@rwth-aachen.de , jochen.huertgen@rwth-aachen.de 

Abstract (Neotectonic activity of the Granada Basin – New evidence from the Padul – Nigelas fault zones): The Padul- 

Nigüelas Fault Zone (PNFZ) is situated at the south-western mountain front of the Sierra Nevada (Spain) in an extensive regime 

and belongs to the internal zone of the Betic Cordilleras. The aim of this study is a collection of new evidence for neotectonic 

activity of the fault zone with classical geological field work and modern geophysical methods, such as ground penetrating radar 

(GPR). Among an apparently existing bed rock fault scarp with triangular facets, other evidences, such as deeply incised valleys 

and faults in the colluvial wedges, are present in the PNFZ. The preliminary results of our recent field work have shown that the 

synsedimentary faults within the colluvial sediments seem to propagate basinwards and the bed rock fault is only exhumed due to 

erosion for the studied segment (west of Marchena). We will use further GPR data and geomorphologic indices to gather further 

evidences of neotectonic activity of the PNFZ. 

Key words: Active faulting, ground penetrating radar (GPR), neotectonics, Granada Basin 

INTRODUCTION 

The Granada Basin is a seismically active 

intramontane basin in the Betic Cordillera of 

Andalucía, Southern Spain. Earthquakes have been 

documented both in the historic and the instrumental 

record (Reicherter (2001) among others). Situated in 

an extensive regime, normal faults delimit basins of 

Neogene and Quaternary ages. In this study, modern 

geophysical methods like ground penetrating radar 

(GPR) have been combined with classical geologic 

field work to gain new insight into neotectonic activity 

of the Padul-Nigüelas Fault Zone. 

sedimentary succession from Tortonian ages (11.6 - 

7.2 Ma BP) on and consist of calcarenites, 

evaporites, terrigenous clastics as well as carbonate 

intercalations. The youngest sequences are 

composed of Quaternary colluvial and alluvial 

deposits. The footwall, the Alpujárride complex, 

comprises several units, such as metapelites, 

metapsammites and quartzites at the base and 

carbonate rocks as uppermost formation. 

(Azañón et al., 2002) 

GEOLOGY 

The convergence of the Eurasian and African plates 

resulted in coupled extension and compression 

during the Oligocene and Miocene whereas recent 

stress fields designate a NW-SE oblique 

convergence with a velocity of 4 mm/a (McClusky et 

al., 2003). As a result the Betic Cordillera orogeny 

formed a highly folded mountain belt simultaneously 

with associated normal faults. Uplift and exhumation 

resulted in erosion of stacked lithological units 

(Galindo-Zaldívar et al., 2003). The Betic Cordilleras 

are subdivided into an External Zone and an Internal 

Zone, which is subdivided into the Malaguide- 

Complex, the Alpujárride-Complex and the Nevado- 

Filábride-Complex because of different metamorphic 

facies. 

The Padul-Nigüelas Fault Zone (PNFZ), situated in 

the internal zone of the Betic Cordilleras, is part of a 

NW-SE trending system of normal faults adjacent to 

sedimentary basins. They usually show a 

Fig. 1: Geological Map of the southern Iberian region (after 

Reicherter & Peters, 2005). 

The PNFZ (Fig. 2) is well documented in the 

westernmost and easternmost parts whereas faults 

are circumspectly indicated in the central part north 

of Dúrcal. Due to the hard-rock carbonate lithology of 

the Alpujarrian basement, active faults in this 

segment are commonly preserved as bedrock faultscarps. 

214


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 2: Geological Map of the Padul-Nigüelas Fault Zone, simplified and composed from Sanz de Galdeano et al. (1975) and 

González Donoso et al. (1978). 

METHODS 

Apart from classical field methods, such as 

geological mapping, we used geophysical survey 

methods like ground penetrating radar and laser 

distance measurements to get a detailed image of 

both subsurface and outcropping structures in order 

to give evidence for historic and recent seismic 

Fig. 3: Illustration of a laser distance measurement. 

Exhumed scarp near the natural monument “Nigüelas 

Fault”. 

215 

activities in the Padul-Nigüelas Fault Zone. 

Laser distance measurements 

Laser distance measurements (LDM) have been 

done with a compact rangefinder (LTI TruPulse 360) 

with an integrated tilt sensor and compass to 

measure slope, horizontal distance, vertical distance, 

inclination and azimuth of points in outcrops. LDM 

was used (1) to describe undulating scarp 

geometries and their dimensions and (2) to get exact 

relationships between several faults in outcrops of 

colluvial wedges and alluvial fans. The method of 

LDM is illustrated in Fig. 3. The rangefinder is 

mounted on a tripod and the position of the tripod is 

stationary during a measurement. With a narrow grid, 

it is possible to get detailed information about 

exhumed scarp planes or faults in colluvial wedges. 

We have recorded these measurements at several 

sites in the PNFZ. 

Ground penetrating radar 

Ground penetrating radar is a non-invasive 

geophysical method which uses electromagnetic 

waves for shallow subsurface surveys. Due to 

changing magnetic and electric properties in the 

underground, reflections of transmitted waves can be 

registered and the two way travel time (TWT, in ns) 

of the waves gives information about the depth of a 

reflector. We used a 270 MHz antenna to record the 

raw data, which have to be interpreted by various 

filtering and correction tools. According to subsurface 

conditions (permittivity, conductivity, presence of 

water-saturated or clay-rich sediments), the 

penetration depth varies between 5 and 10 m with 

high-resolution data. In our study area, we especially


EARTHQUAKE 

ARCHAEOLOGY 



investigated the colluvial wedges and alluvial fans of 

the Padul-Nigüelas Fault Zone with the aim to find 

basinwards trending faults in the subsurface. 

PRELIMINARY RESULTS 

The 3-week field work for this study has just been 

completed in the end of April 2011 and the evaluation 

and processing of recorded data is still in progress. 

Therefore, we present some preliminary results, 

especially from the central part of the Padul-Nigüelas 

Fault Zone on the Marchena fan. 

During our field work, we have found several obvious 

evidences which suggest geologically recent fault 

activity in the PNFZ, such as triangular facets, deeply 

incised channels, faultward dipping layers and faults 

in both colluvial wedges and alluvial fans at 

numerous sites within the PNFZ. These features 

were also observed and documented by other 

authors (e.g. Alfaro et al. (2001), Galindo-Zaldívar 

(2003)). 

The morphology of the PNFZ is characterized by the 

exhumed fault scarp, developed on the Alpujarrian 

bedrock. We could confirm the observations from 

Alfaro et al. (2001), who has already mentioned that 

the scarp is exhumed due to erosion and not due to 

fault activity. The exhumed fault scarps of the PNFZ 

are specified by lateral as well as downslope 

undulations of the main fault plane surface in a 

widespread range. The dip angle varies in a range 

from 20° to 65° and the dip direction from 190° to 

260°. These undulations were documented and 

combined with laser distance measurements on 

several sites along the exhumed fault scarp of the 

entire fault zone. 

Due to the apparently inactive main fault, our 

investigations in the field brought the colluvial 

wedges and alluvial fans in front of the PNFZ into 

focus. We gathered multiple profiles of faults in the 

colluvium and alluvium by classical field methods and 

by laser distance measurements. We have found 

active as well as buried faults in the sediments (see 

also Hamdouni et al. (2008)). Most of the outcrops 

are created naturally by strong river incision into the 

sediments and are situated adjacent to the bedrock 

fault scarp. 

For this state of our study, we have concentrated the 

field-work on the widespread alluvial fan deposits 

located between the two segments of the Padul- 

Nigüelas Fault Zone developed on the Alpujarrian 

bedrock north of Marchena, which is axially dissected 

Fig. 4: Outcrop of an incised channel in the alluvial and 

colluvial fan in the north of Marchena. A: Photograph of the 

colluvial sediments and existing faults. B: Sketch of the 

same outcrop with a few traceable layers. 

by the river Dúrcal. We collected several GPR 

profiles and documented several outcrop profiles in 

the proximal colluvial wedges developed at fault-fan 

contacts. One example is located in the northern part 

of the Marchena fan (Fig. 2). Fig. 4 shows a 

photograph of the outcrop (A) with interpreted faults 

in the sediments and a sketch (B) with the simplified 

geometry and certain marker horizons. On the basis 

of these horizons, we can preliminary determine the 

offset of the larger faults between a few cm and 

1.8 m. Noticeable features in the faults are aligned 

clasts with their longitudinal axis along slip direction 

and carbonate-coated clasts, which indicate water 

circulation on the faults. Some of the faults are 

traceable up to the surface and build small scarps in 

the topology. 

Fig. 5: GPR profile on the upper surface of the outcrop (Fig. 4). The profile is not topographically corrected and a velocity of 

0.1 m/ns is assumed for the time-depth conversion. Grey box: excerpt of the outcrop. Black lines: interpreted faults. 

216


EARTHQUAKE 

ARCHAEOLOGY 



In addition to the documentation of the faults in the 

colluvial wedge, we achieved a GPR profile on the 

upper surface of the outcrop (Fig. 5). The processed 

data is not topographically corrected, but first 

interpretations allow inferring a few faults in the 

profile. An exact correlation between the outcrop and 

the radar profile was not done so far. This profile as 

well as profiles in other studies (Reicherter et al. 

(2001)) demonstrates that GPR surveys are a useful 

method to locate faults and faultward dipping layers 

within colluvial wedges along presumably active 

range-front faults, such as the case of the Padul- 

Nigüelas Fault Zone. 

The above mentioned features also occur at sites 

NW of El Puntal, Padul and Nigüelas which points to 

similar fault activity in this stage of our research. 

DISCUSSION & OUTLOOK 

The Padul-Nigüelas Fault Zone has been examined 

extensively for decades and is considered as active 

since instrumental and historical seismicity indicate 

several earthquakes in the Granada Basin. In our 

field study we found the fault scarps exposed 

because of erosion instead of fault activity. The 

colluvial wedges are not undisturbed, they show both 

buried and active faults, therefore we consider 

basinwards trending fault activities for this sector, as 

well as for the entire PNFZ. 

Since our study is not finished, GPR data and 

geomorphologic indices will be used to give further 

evidence for either active or inactive faults in the 

Padul-Nigüelas Fault Zone. 

References 

Alfaro, P., J. Galindo-Zaldívar, A. Jabaloy, A.C. López- 

Garrido & C. Sanz de Galdeano, (2001). Evidence for the 

activity and paleoseismicity of the Padul fault (Betic 

Cordillera, southern Spain). Acta Geologica Hispanica 36 

(3-4), 283-295. 

Azañón, J.M., J. Galindo-Zaldívar, V. García-Dueñas and A. 

Jabaloy (2002). Alpine tectonics II. Betic Cordillera and 

Balearic Islands. In: W. Gibbons and T. Moreno (eds.), 

The Geology of Spain 16, 401-416. 

Galindo-Zaldívar, J., A.J. Gil, M.J. Borque, F. González- 

Lodeiro, A. Jabaloy, C. Marín-Lechado, P. Ruano, C. 

Sanz de Galdeano (2003). Active faulting in the internal 

zones of the central Betic Cordilleras (SE, Spain). Jorunal 

of Geodynamics 36, 239-250. 

González Donoso, J.M., V. Garcá-Dueñas, J.A. Gallegos, J. 

Avidad Castañeda (1978). Mapa Geologico de España - 

1041 Dúrcal. Instituto Geológico y Minero de España. 

Hamdouni, R., Irigaray, C., Fernández, T., Chacón, J., 

Keller, E.A. (2008). Assessment of relative active 

tectonics, southwest border of the Sierra Nevada 

(southern Spain). Geomorphology 96, 150-173. 

McClusky, S., R. Reilinger, S. Mahmoud, D. Ben Sari, A. 

Tealeb (2003). GPS constraints on Africa (Nubia) and 

Arabia plate motions. Geophysical Journal International 

155 (1), 126-138. 

Reicherter, K.R. (2001). Paleoseismologic advances in the 

Granada basin (Betic Cordilleras, southern Spain). Acta 

Geologica Hispanica 36 (3-4), 267-281. 

Reicherter, K., G. Peters (2005). Neotectonic evolution of 

the Central Betic Cordilleras (Southern Spain). 

Tectonophysics 405,191-212. 

Sanz de Galdeano, C., J.M. González Donoso, J.A. 

Gallegos (1975). Mapa Geologico de España - 1026 

Padul. Instituto Geológico y Minero de España. 

217


EARTHQUAKE 

ARCHAEOLOGY 



HOLOCENE SEAFLOOR FAULTING IN THE GULF OF CORINTH: THE POTENTIAL FOR 

UNDERWATER PALEOSEISMOLOGY 

Sakellariou, Dimitris (1), Lykousis Vasilis (1), Rousakis Grigoris (1) 

(1). Institute of Oceanography, Hellenic Centre for Marine Research, 47km Athens-Sounio Ave., 19013 Anavyssos, Greece. 

Email: sakell@ath.hcmr.gr, vlikou@ath.hcmr.gr, rousakis@ath.hcmr.gr 

Abstract (Holocene seafloor faulting in the Gulf of Corinth: the potential for underwater paleoseismology): The techniques 

used for the marine geological – geophysical investigation of the seafloor of the Gulf of Corinth were suitable for the mapping of 

the offshore faults and for the detection of recent, Holocene faulting activity. The available seismic data provide clear evidence that 

several faults have moved repeatedly in Holocene times and have produced cumulative offsets of up to several meters during the 

last 14-13 kyrs. The next step in the investigation of the offshore faulting in the Gulf of Corinth will be to use higher resolution 

methods to perform on-fault seismic profiling and to recognize individual earthquake ruptures along the faults on the seafloor of the 

Gulf of Corinth. 

Key words: seafloor faults, Holocene earthquakes, recent movements 

INTRODUCTION 

The Gulf of Corinth is an active continental rift 

developed within the stretching Aegean region, 

perpendicular to the alpine Pindos mountain chain. It 

is a 100 km-long rift characterized by high extension 

rates in N-S direction (currently up to 20 mm/yr, 

Clarke et al. 1998), localized mostly within the 

narrow, 15-20 km wide, marine basin. With a very 

high seismicity and more than ten earthquakes of 

magnitude M>6 in the last 50 years, the Gulf of 

Corinth is an ideal site to study active tectonics and 

recent fault movements. 

The offshore fault pattern shown in Fig. 1 has been 

recognized on numerous seismic reflection profiles 

Fig. 1: Landsat image with main active offshore faults in the Gulf of Corinth rift. Swath bathymetry after Alexandri et al. (2003) 

and Nomikou et al. (this volume) 

218


EARTHQUAKE 

ARCHAEOLOGY 



(airgun, boomer, 3.5 kHz) acquired during several 

cruises of R/V Aegaeo between 1995-2005 and R/V 

Alkyon recently (Sakellariou et al., 2001; 2004; 2007; 

Lykousis et al., 2007). More marine geophysical 

studies have been conducted by other Greek or 

international teams and have contributed to the 

understanding of the structure of the Gulf of Corinth 

(Stefatos et al., 2002; Moreti et al., 2004; Zelt et al., 

2004; McNeill et al., 2005; Bell et al., 2008) 

continuous sequence of Holocene mud and sand 

turbidites. Correlation of the seismic data with the 

sedimentary sequence recovered in the cores 

enabled absolute dating of basin-wide reflectors, 

which are frequently offset by faults. Careful and 

detailed analysis of the shallow seismic stratigraphy 

led Lykousis et al. (2007) to estimate Holocene slip 

rates of the basin bounding faults. 

Although the scope of the above campaigns was 

mostly to study the neotectonic structure of the rift 

with relatively low to medium resolution techniques, 

many seafloor faults were identified to have moved in 

Holocene or even in historic times. The aim of this 

paper is to shed light these faults and explore the 

potential for high resolution underwater paleoseismological 

studies in the Gulf of Corinth. 

LECHAION GULF 

Sakellariou et al. (2004b) showed that the Lechaion 

Gulf is currently a half graben developed on the 

hangingwall of the south-facing on/offshore Loutraki 

fault. Antithetic, north facing faults run E-W on the 

shallow southern shelf of the Gulf. Recent Boomer 

profiles and older Airgun 10 in 3 and 5 in 3 seismic 

profiles show that these faults cut through the 10- 

15m thick Holocene deposits. 

The Boomer profile of Fig. 2 shows one of the northfacing 

athithetic faults of the Lechaion Gulf 

crosscutting the submerged Last Glacial Maximum 

(LGM) landscape and the Holocene drape. The offset 

of 6-7 m must have been accumulated after the 

inundation of the LGM Corinth lake about 14 ka ago. 

Fig. 2: Boomer profile, 175 Joules, 1-2,5 kHz, Lechaion 

Gulf. 

CENTRAL GULF OF CORINTH 

A dense grid of Airgun 10 in 3 single channel 

reflection profiles revealed the shallow structure and 

the seismic stratigraphy of the Central Gulf of Corinth 

deep basin (Lykousis et al., 2007). Gravity coring 

from R/V Aegaeo and long piston coring from R/V 

Marion Dufresne (Moreti et al. 2004) validated the 

interpretation of the seismic data and recovered a 

219 

Fig. 3. Airgun 10in 3 

single channel reflection profile 

through the deep basin of the Central Gulf of Corinth, 

south of Itea bay. Left=South, right=North. Top: Raw 

profile. Down: Interpreted profile and line drawing. The 

yellow line marks the 14 ka old interface between LGM 

lacustrine and Holocene marine sedimentation. Slip 

rates of the individual faults have been estimated from 

the vertical offset (Lykousis et al., 2007). 

Vertical slip rates of 0.6-0.7 m/ka have been 

calculated for the intra-basin faults, while rates of 2.4- 

3.7 m/ka were estimated for the basin bounding 

faults. Still, the resolution of the technique is not 

sufficient high to allow recognition of individual 

earthquakes and rupturing events. 

The deep basin of the Gulf of Corinth is an ideal site 

for on-fault and off-fault underwater paleoseismological 

studies. The geophysical record of the 

shallow subseafloor sedimentary sequence with the 

continuous succession of mud and sand turbidites, 

potential well defined seismic reflectors, may resolve 

characteristic structures associated with individual 

fault ruptures. High resolution seismics with deeptowed 

vehicles in combination with carefully selected 

coring sites on the hangingwall and footwall of the 

faults will provide recognition and dating of dislocated 

layers and thus define individual earthquakes. 

WESTERN GULF OF CORINTH 

After the 1995 Aegion earthquake the western part of 

the Gulf of Corinth has been the site of intensive onand 

offshore surveys by Greek and international 

teams. Numerous field surveys and marine 

geological-geophysical campaigns have been 

conducted and yielded very wealth data sets on


EARTHQUAKE 

ARCHAEOLOGY 



active tectonics and faulting. Onshore 

paleioseismological studies on the Heliki Fault and 

the broad region of Aegion provide new knowledge 

and understanding of the fault behaviour onshore. 

Systematic offshore seismic reflection profiling 

enabled detailed mapping of the seafloor faults of the 

western Gulf of Corinth. Nevertheless, as in the rest 

part of the Gulf, more precise data on the activity of 

specific offshore faults has been only occasionally 

gained. Two examples are given here below. 

Medium resolution Airgun 5 in 3 single channel 

seismic reflection profiling has been conducted on 

the shelf and upper slope off Aegion – Diakofto 

region. A couple of the shot profiles crossed the 

eastward offshore prolongation of the well studied 

onshore Aegion fault (Fig. 4). 

the western Gulf of Corinth and show significant fault 

movements in Holocene in the basin. 

More precisely, significant fault movements have 

been detected on the south facing South Eratini fault, 

between Psaromyta Cape and Aegion (Fig. 5). The 

Airgun profile of Fig. 5 shows the multiplefaulting of 

the basin between Psaromyta Cape (North) and 

Aegion (South). The 3.5 kHz profiles through the 

South Eratini Fault indicate that the stratified 

Holocene drape of the basin has been vertically 

dislocated by about 5 m. 

Fig. 4: Medium resolution Airgun 5 in 3 

single channel 

seismic reflection profiling crosscutting the Aigion fault. 

Upper image: Map of Aigion area with the trace of the 

Aigion fault, the location of the profile and the 

sedimentological description and radiometric ages of an 

offshore drilling (Schwartz & Tziavos, 1979). Middle: raw 

profile. Down: Interpretation of the seismic profile. The 

north-facing Aigion fault and an antithetic, south-facing 

fault are marked in red. The yellow reflector marks the 

interface between clayey silt above and cobbles-pebbles 

below, dated to about 1o ka. Note that the Aegion fault 

offset vertically the yellow reflector by about 8 m and the 

seafloor by about 4 m. 

Careful interpretation of the seismic profiles from the 

Aegion shelf along with the sedimentological data 

from the offshore drillings (Schwartz & Tziavos, 

1979) indicate that the Aegion fault has produced a 

cumulative vertical offset of at least 4m during the 

last 10 kyrs. 

Fig. 5: Airgun 10 in 3 seismic profile (top) and 3.5 kHz 

profile (bottom) through the South Eratini Fault. Note 

that the 3.5 kHz profile shows vertical offset of the 

Holocene drape by about 5.5-4.5 m. 

CONCLUSIONS 

The techniques used for the marine geological – 

geophysical investigation of the seafloor of the Gulf 

of Corinth were suitable for the mapping of the 

offshore faults and for the detection of recent, 

Holocene faulting activity. The available seismic data 

provide clear evidence that several faults have 

moved repeatedly in Holocene times and have 

produced cumulative offsets of up to several meters 

during the last 14-13 ka. The next step in the 

investigation of the offshore faulting in the Gulf of 

Corinth will be to use higher resolution methods to 

perform on-fault seismic profiling and to recognize 

individual earthquake ruptures along the faults on the 

seafloor of the Gulf of Corinth. 

Further on, Airgun 10 in 3 seismic profiles and 3.5 kHz 

profiles have been acquired from the deep basin of 

220


EARTHQUAKE 

ARCHAEOLOGY 



References 

Alexandri M., Nomikou P., Ballas D., Lykousis V. & 

Sakellariou D. (2003): Swath bathymetry map of Gulf of 

Corinth. Geoph. Res. Abstracts, Vol. 5, 14268, EGS 2003 

Bell, R., McNeill, L.C., Bull, J.M., and Henstock, T.J., 2008. 

Evolution of the western Gulf of Corinth continental rift, 

Greece. Geological Society of America Bulletin, 120, 156- 

178. 

Clarke, P.J., Davies, R.R., England, P.C., Parson, B., 

Billiris, H., Paradissis, D., Veis, G., Cross, P.A., Denys, 

P.H., Ashkenazi, V., Bingley, R., Kahle, H.-G., Muller, M.- 

V., and Briole, P., 1998, Crustal strain in central Greece 

from repeated GPS measurements in the interval 1989- 

1997: Geophysical Journal International, 135, 195-214. 

Lykousis, V., Sakellariou, D., Moretti, I., and Kaberi, H., 

2007, Late Quaternary basin evolution of the Gulf of 

Corinth: Sequence stratigraphy, sedimentation, fault slip 

and subsidence rates: Tectonophysics, 440, 29-51 

Leeder, M.R., Portman, C., Andrews, J.E., Collier, R.E.L., 

Finch, E., Gawthorpe, R.L., McNeill, L.C., Perez-Arlucea, 

M., and Rowe, P., 2005, Normal faulting and crustal 

deformation: Alkyonides Gulf and Perachora peninsula, 

eastern Gulf of Corinth rift basin, Greece: Journal 

Geological Society of London, 162, 549-561. 

McNeill, L., Cotterill, C., Stefatos, A., Henstock, T., Bull, J., 

Collier, R., Papatheoderou, G., Ferentinos G., and Hicks, 

S. 2005, Active faulting within the offshore western Gulf 

of Corinth, Greece: implications for models of continental 

rift deformation: Geology, 33, 241-244. 

Moretti, I., Lykousis, V., Sakellariou, D., Reynaud, J.-Y., 

Benziane, B., and Prinzhoffer, A., 2004, Sedimentation 

and subsidence rate in the Gulf of Corinth: what we learn 

from Marion Dufresne’s long-piston coring: Comptes 

Rendus Geoscience, 336, 291-299. 

Nomikou, P., Alexandri, M., Lykousis, V., Sakellariou, D. & 

Ballas, D. (2011). Swath bathymetry and morphological 

slope analysis of the Corinth Gulf. 2 nd INQUA-IGCP-567 

International Workshop on Active Tectonics, Earthquake 

Geology, Archaeology and Engineering, Corinth, Greece 

(2011), this volume 

Sakellariou, D., Lykousis, V. & Papanikolaou, D. (2001) 

Active faulting in the Gulf of Corinth, Greece. In: 36th 

CIESM Congress Proceedings, 36 pp. 43. 

Sakellariou, D.,Kaberi, H. & Lykousis,V. (2004) Infuence of 

active tectonics on the recent sedimentation of the Gulf of 

Corinth basin. In: 10th International Congress of Greek 

Geological Society, Abstracts 15-17 April, p. 232-233, 

Thessaloniki. 

Sakellariou D., Fountoulis I. & Lykousis V. (2004b). 

Lechaion Gulf: the last descendant of the Proto-Gulf-of- 

Corinth basin. 5 th 

Int. Symp. On East. Mediterranean 

Geology, Proceedings 2/881-884, Thessaloniki 

Sakellariou, D., Lykousis, V., Alexandri, S., Kaberi, H., 

Rousakis, G., Nomikou, P., Georgiou, P., and Ballas, D., 

2007, Faulting, seismic-stratigraphic architecture and late 

Quaternary evolution of the Gulf of Alkyonides basin – 

East Gulf of Corinth, Central Greece: Basin Research, 

19, 273-295. 

Schwarz M.L. & Tziavos C. (1979): Geology in the search 

for Ancient Helice. J. Field Archaeology, 6, p. 243-252 

Stefatos, A., Papatheoderou, G., Ferentinos, G., Leeder, 

M., and Collier, R., 2002, Active offshore faults in the Gulf 

of Corinth, Greece: Their seismotectonic significance: 

Basin Research, 14, 487-502. 

Zelt, B.C., Taylor, B., Weiss, J.R., Goodliffe, A.M., 

Sachpazi, M., and Hirn, A., 2004. Streamer tomography 

velocity models for the Gulf of Corinth and Gulf of Itea, 

Greece. Geophys. J. Int., 159, 333-346. 

221


EARTHQUAKE 

ARCHAEOLOGY 



DAMAGE ASSESSMENT IN ARCHAEOSEISMOLOGY: 

METHODS AND APPLICATION TO THE ARCHAEOLOGICAL ZONE COLOGNE, 

GERMANY 

Schreiber,Stephan (1, Klaus-G. Hinzen (1) 

(1) Earthquake Geology and Archaeoseismology Group, Cologne University, Vinzenz-Pallotti-Str. 26, 51429 Bergisch 

Gladbach. GERMANY. Email: stephan.schreiber@uni-koeln.de 

Abstract (Damage Assessment in Archaeoseismology: Methods and Application to the Archaeological Zone Cologne, 

Germany): A work flow to investigate archaeoseismological problems including the use of modern surveying equipment and 

quantitative methods was developed. We exemplify a comprehensive damage assessment in the historic city center of Cologne, 

Germany. During the construction of an underground museum (Archeological Zone Cologne), parts of the Roman and medieval 

city are being excavated. The remains exhibit structural damages spread over an area of 150 x 200 m. The structures were 

mapped using a phase-based 3D laser scanner. The acquired data were analyzed and the results lead to a quantitative damage 

database for the investigation area. 

Key words: Damage Analysis, Laser Scanning, Quantitative Methods, Roman 

INTRODUCTION 

One of the first steps in modern archaeoseismology 

is the precise mapping of damages with potential 

seismogenic origin. In addition to the archaeoseismologic 

working scheme proposed by Galadini et 

al. (2006) and specified by Hinzen et al. (2009) and 

Hinzen (2011), Schreiber & Hinzen (2011) presented 

a site-specific modification for the situation in 

Cologne (Fig.1). The investigation of the archaeological 

site is structured into four major categories: 

(1) investigation of the constructions, (2) surface 

topography, (3) subsurface conditions, and (4) geological 

setting of the region. 

Fig.1: Working scheme for the investigation of an 

archaeological site with four major thematic groups (after 

Schreiber & Hinzen, 2011. 

METHODS 

This contribution focuses on the first category and 

shows how a comprehensive quantitative dataset is 

collected, analyzed, and prepared for further modelling 

and estimation of the damage cause. 

Data Acquisition & Processing 

222 

In this study the constructions were mapped using a 

phase-based 3D laser scanner (FARO Photon80). 

The scanner emits a permanent bundled infrared 

beam towards the measurement target via a rotating 

mirror. The target reflects the beam and the reflected 

signal is detected by the measuring device. The 

scanner records the phase shift between transmitted 

and received signal and calculates the distance to 

the target. Combined with instrumental parameters 

including the position of the rotating mirror during the 

measurement and the recording position of the 

scanner, the Cartesian coordinates are calculated for 

each discrete reflection point. The resolution of 

0.00076° in horizontal and 0.009° in the vertical 

direction, which correlates to 0.13 mm and 1.57 mm 

at a distance of 10 m, allows a distance resolution in 

the range of 1-2 millimeters (Schreiber et al., 2011b). 

Due to the fast data acquisition rate of 120.000 pt/s 

measurement times are short and ongoing 

excavations are not significantly disturbed. The data 

were processed with the software JRC 3D 

Reconstructor 2 (Sequieira et al., 1999, Sgrenzaroli & 

Wolfart, 2002) which is a capable tool to handle large 

3D point clouds. 

After the application of different automatic filters to 

remove erroneous points (e.g. points at edges or 

points acquired in the open sky) and the manual 

cleaning of the point clouds to remove vegetation or 

modern structures, the scans were merged into 

models of substructures of the investigation area. 

Due to smaller file sizes (5-20% of the overall data 

volume) these “submodels” are easier to handle in 

the analysis phase. 

Damage Analysis 

The substructure models were used to perform a 

detailed analysis of the scanned structures and their


EARTHQUAKE 

ARCHAEOLOGY 



damages. The structural damages were (1) identified, 

(2) located within the global reference system, (3) 

classified (rotation, displacement, tilting, cracking), 

and (4) quantified with damage-dependent methods. 

All results were merged into a damage database. 

The damage database is linked to a construction 

inventory, where information about the buildings from 

(1) previous and recent archaeological excavations, 

(2) city annals and historical reports and (3) historical 

maps were collected. In addition to these archaeological 

and historical sources, information about the 

construction type, the building material and the 

function of the building was collected. 

WORKING AREA 

The working area is located in the historic city center 

of Cologne, on the top of the slope to a former side 

arm of the Rhine River (Fig. 2). The excavations are 

part of the Archaeological Zone Cologne (AZC), a 

large museum area, which will exhibit remains from 

2000 years of history after its completion. During 

several building phases the constructions expanded 

over the edge of the slope, towards the Rhine River, 

so the excavated buildings are founded partially in 

artificial fillings. The topography, the complexity of 

the construction ground, the position next to a large 

river and the seismic potential of the region suggest 

several possible causes for the observed damages. 

The following paragraphs present three examples of 

the data acquisition and damage analysis methods 

applied to different structures in a rather complex 

excavation environment. Since 2008 over 200 laser 

scans were collected in the AZC resulting in a point 

cloud of 2.4 billion individual points. This data were 

combined to ten models of substructures within the 

AZC. The analysis of the data leads to a damage 

database with currently 2000 detected damages. 

Example: Praetorium 

The praetorium, the palace of the Roman governor 

was the main administrative building of the province 

germania inferior. The first building phase of the 

praetorium started with the beginning of the first 

century AD followed by three more phases (end of 

the 1st century, end of the 2nd century and 4th 

century). The structure was completely demolished in 

the 9th century (Gechter & Schütte, 2000). The 

foundations of the northern half of this former 90 m 

long and 25 m high building were excavated in the 

1950s and 1970s. We mapped the remains of the 

construction in 97 scans from different height levels 

and combined them to four models of the main parts 

of the building. 

The models were used to identify, locate and classify 

different damages. The main damage type in the 

praetorium is the tilting of walls; 822 tilt 

measurements were made on virtual cross-sections 

of the point cloud, which were exported to CAD 

software. Figure 3 shows the distribution of 

directional tilting found in the praetorium. 

Fig.2:(Left) Aerial photo of the historic city center of Cologne 

with the main structures of the excavations. The white line 

indicates the planned museum area. (Right) 3D surface model 

of the natural ground. The position of the Roman city wall is 

indicated and the black rectangle marks the excavation area 

(Schreiber et al., 2011a). Circles give the location of borings 

used to construct the model. The former sidearm of the Rhine 

River is marked with the dotted white line. 

223 

Fig.3: Floor plan of the praetorium with directional tilting 

vectors. Color indicates the orientation sector (blue: west, 

green: east, red: north, yellow: south). The vector length 

gives the amount of tilting. 

The eastern part of the praetorium is tilted eastward. 

The distribution of tilting correlates well to a change 

in the subsurface where the constructions expand 

over the edge of the former slope towards the Rhine 

River and are partly founded in artificial fillings. The 

maximum observed tilt is 15.1° towards east. Average 

tilt of all eastward tilted walls is 3.1°± 2.4°. 

Maximum tilt towards west is 12.9° with an average 

of 1.5°± 1.8°, and the maximum tilts towards north 

and south are 5.6° and 11.4°, respectively. In


EARTHQUAKE 

ARCHAEOLOGY 



addition to tilting, 104 cracks with opening widths of 

up to 25 cm were quantified from the point cloud; 39 

horizontal displacements in the walls with opening 

width up to 9 cm were detected. It is the first time the 

documentation focuses on the damage of the 

buildings. Before only some of the damages were 

documented as a side product in archaeological 

drawings; however, not quantified. 

Example: Roman Well 

A 12.5 m deep Roman well is located west of the 

Cologne city hall (Fig.2). The well was mapped in two 

campaigns with 20 scans. The scans were acquired 

using a special retaining device. The scanner was 

installed upside-down and lowered down the well 

with an electrical winch; the position was fixed using 

air-pressured stamps (Fleischer et al. 2010). 

block using the position of the plane normal. Figure 4 

shows the tilting vectors for each block on top of a 

unrolled cylindrical scan of the inside of the well. The 

damage analysis showed that the block layers are 

tilted up to 11° within the damage zone (layers 

between –8 m and –9 m in Fig. 4). The average tilting 

of the discrete blocks is 4.2°± 3.3°. The maximum tilt 

is 15.8°. 

Example: Roman Sewer 

The 180 m long sewer located north of the excavation 

area is part of the complex Roman dewatering 

system of the city. The eastern and western sections 

of the sewer were built with different construction 

techniques. The latter is made of large tuff blocks. 

This 120 m long section was scanned with 25 scans. 

Damages include displaced blocks in the roof (Fig. 5) 

and shell-like spallings on the sidewalls. 

Fig.4: Flat projection of the wall of the Roman well. Arrows 

give the direction of tilting and rotation of the normal of 

virtual planes, fitted to the front side of each block. The 

length of the arrows gives the amount of tilting. The white 

points are the centre points of the fitted planes connected 

by the black lines indicating the trend of the layer centers 

(after Fleischer et al., 2010). The bars at the right give the 

locations of the repaired section, the deformed section and 

the main damage area. 

The structure shows two sections (Fig. 4): The upper 

5.5 m were reconstructed probably during medieval 

times, while the lower 7 m remained deformed. For 

the analysis of the orientation, virtual planes were 

fitted in a least square sense to the front sides of 338 

blocks in 33 layers. This technique allowed the 

quantification of the tilting and the rotation of each 

224 

Fig.5: (Upper) Displacement of blocks in the vault of the 

western part of the Roman sewer. (Lower) Cracking of the 

vault in the eastern part of the sewer. In the bottom, the top 

of the filling of the sewer is still in place. The inset shows a 

cross-section of the point cloud when the sediment filling of 

the sewer was still in place. The green curve shows the 

original shape of the vault, the red line follows the rotated 

and displaced vault in the damaged area. 

The eastern section, which ends in an outlet through 

the eastern Roman city wall of Cologne, was 

scanned in an emergency campaign before the 

excavations were stopped due to critical static conditions 

of the sewer. Therefore only four scans were


EARTHQUAKE 

ARCHAEOLOGY 



made so far. In this part the vault of the sewer is 

made of opus caementicium. Here massive shell-like 

spallings were found on the bottom parts of the 

sidewalls. The roof shows a large crack with 

displacements of up to 15 cm on a length of minimum 

10 m (Fig. 5). Currently the damages in the western 

part are investigated in detail using virtual crosssections 

in 1 m intervals. For the eastern part 

additional scans are necessary to perform a 

complete damage analysis. 

CONCLUSIONS 

The application of 3D laserscanning in the AZC 

shows that this technique is an appropriate tool 

especially when complex structures are analyzed. 

The main challenge during the mapping of the 

Cologne praetorium was the widespread and complex 

damage pattern, which required a large amount 

of scans. In the Roman well the vertical shaft, 

required a complete different approach. The scanner 

was mounted upside down in order to avoid shadowing 

effects and the time consuming construction of 

different plateaus in the shaft. The scanning of the 

sewer required a careful selection of scanning 

positions to avoid shadowing effects due to steep 

angles in the only 1.6 m wide structure. It could be 

shown that the analysis of the damages needs a sitespecific 

approach in order to quantify the different 

types of damages. In comparison to traditional 

mapping techniques laser scanning provides a large 

amount of high-resolution data in a short period of 

time. In case of fragile conditions of the excavated 

objects, repeated scanning can virtually conservate 

the original conditions before these are alterated due 

to ongoing restoration and/or excavation. The 

accuracy of the measurements allows a detailed 

analysis even when the original finding is no longer 

accessible. 

The results of the damage analysis are the base for 

further studies on the damage cause. The type and 

the amount of the damages can be used to narrow 

down possible damage scenarios. However, as 

shown in Figure 1, the mapping and the analysis of 

the damages are one of the first steps in the 

workflow. Further steps, e.g. the surface and 

subsurface analysis (Schreiber et. al 2011a) and the 

modelling of scenarios (e.g. Hinzen et al., 2010, 

2011b) are necessary for an accurate investigation of 

an archaeoseismological problem. 

Acknowledgements: 

We are grateful to the staff of the Archaeological Zone 

Cologne. We thank the staff of the Seismological Station 

Bensberg, especially Claus Fleischer for his dedication to 

the fieldwork 

This work is beeing financed by Deutsche Forschungsgemeinschaft 

DFG (HI 660 / 2-1). 

References 

Fleischer, C., K.-G. Hinzen and S. Schreiber, (2010). 

Laserscanning eines römischen Brunnens in der 

Archäologischen Zone Köln. AVN Allgemeine 

Vermessungs Nachrichten 5/2010, 176-181. 

Galadini, F., K.-G. Hinzen and S. Stiros, (2006). 

Archaeoseismology: Methodological Issues and 

Procedure. Journal of Seismology 10, no. 4, 395-414. 

Gechter, M. and S. Schütte, (1999). Ursprung und 

Voraussetzung des Mittelalterlichen Rathauses und 

seiner Umgebung. Stadtspuren 22, 69-195. 

Hinzen, K.-G., C. Fleischer, S.K. Reamer, S. Schreiber and 

S. Schütte, (2009). Quantitative Methods in 

Archaeoseismology. In: Archaeoseismology and 

Palaeoseismology in the Alpine-Himalayan Collisional 

Zone (Pérez-López, R., Grützner, C., Lario, J., 

Reicherter, K., Silva, P.G. eds). Baelo Claudia, Spain, 50- 

51. 

Hinzen, K.-G., S. Schreiber and B.Yerli, (2010). The Lycian 

Sarcophagus of Arttumpara, Pinara (Turkey) -Testing 

Seismogenic and Anthropogenic Damage Scenarios. 


3148-3164. 

Hinzen, K.-G., (2011). Earthquake, Archaeoseismology. In: 

Encyclopedia of Solid Earth Geophysics (Gupta, H. K., 

ed.) Springer, Heidelberg, DOI 10.1007/978-90-481- 

8702-7. 

Hinzen, K.-G., H. Kehmeier, S. Schreiber and S.K. Reamer, 

(2011). Test Of Earthquakes And Rockfall As Possible 

Cause Of Damage To A Roman Mausoleum In Pinara, 

Sw Turkey. In: Proceedings of 2nd INQUA-IGCP-567 

International Workshop on Active Tectonics, Earthquake 

Geology, Archaeology and Engineering (Papanikolaou, 

I.D., Reicherter, K., Vött, A., Silva, P.G. eds.) Corinth, 

Greece, submitted. 

Schreiber, S., I. Wiosna, M. Wegner and K.-G. Hinzen, 

(2011a). Archeoseismological Study in the Historic City 

Center of Cologne, Germany. 71. Jahrestagung der 

Deutschen Geophysikalischen Gesellschaft, 21.- 

24.2.2011, Cologne, Germany. 

Schreiber, S., K.-G. Hinzen and C. Fleischer (2011b). 

Excavation Parallel Laser Scanning of a Medieval 

Cesspit in the Archaeological Zone Cologne, Germany. 

Journal of Computing and Cultural Heritage, submitted. 

Schreiber, S. and K.-G. Hinzen, (2011). Archeoseismological 

Investigations in the Historic Center of 

Cologne, Germany. Seismological Research Letters 82, 

No. 2, 334. 

Sequieira, V., K. Ng, E. Wolfart, J.G.M. Goncalves and D. 

Hogg, (1999). Automated Reconstruction of 3D Models 

from Real Environments. ISPRS J. Photogramm. 54, 1- 

22. 

Sgrenzaroli, M. and E. Wolfart, (2002). Accurate Texture- 

Mapped 3D models for Documentation, Surveying and 

Presentation Purposes. In: CIPA WG 6 International 

Workshop on Scanning For Cultural Heritage Recording, 

Greece, Corfu, 1-2 September 2002. 

225


EARTHQUAKE 

ARCHAEOLOGY 



EARTHQUAKE TRIGGERING, CLUSTERING, AND THE SYNCHRONIZATION OF FAULTS 

Christopher H. Scholz 

Lamont-Doherty Earth Observatory, Columbia University. Palisades, NY. USA 

Abstract (Earthquake Triggering, Clustering, and the Synchronization of Faults): Large earthquakes are sometimes 

observed to trigger other large earthquakes on nearby faults. The magnitudes of the calculated Coulomb stress transfers 

presumed to cause the triggering are 10 -2 – 10 -3 the earthquake stress drop and the triggering delay times are similarly small with 

respect to the natural recurrence time of the earthquakes. This requires that both faults be simultaneously very close to the ends of 

their seismic cycles. Paleoseismological data show that for the same regions prior earthquakes have occurred in clusters in space 

and time separated by long quiescent periods. Both observations suggest that synchronization is occurring between faults. Theory 

and observations indicate that synchronization can occur between nearby faults with positive stress coupling and intrinsic slip 

velocities within an entrainment threshold. In the south Iceland seismic zone, the central Nevada seismic belt and the eastern 

California shear zone several synchronous clusters that apparently act independently, can be recognized. This behaviour is the 3D 

equivalent of the phase locking those results in the seismic cycle being dominated by large characteristic earthquakes, and for 

synchronization of fault segments along a single fault. Rupture patterns of repeated individual earthquakes or earthquake clusters 

are not identical in either the 2D or 3D cases. The state of this system, which exhibits strong indications of synchrony without exact 

repetition, may be called fuzzy synchrony. 

226


EARTHQUAKE 

ARCHAEOLOGY 



RELIEF PRODUCTION, UPLIFT AND ACTIVE TECTONICS IN THE GIBRALTAR ARC 

(SOUTH SPAIN) FROM THE LATE TORTONIAN TO THE PRESENT 

Silva, Pablo G. (1, Alex Ribó (1), Moises Martín Betancor (2), Pedro Huerta (1), M. Ángeles Perucha (3), Cari Zazo (4), Jose L. 

Goy (1), Cristino J. Dabrio (5), Teresa Bardají (6) 


SPAIN. pgsilva@usal.es, aribo@usal.es, phuerta@usal.es, joselgoy@usal.es 

(2) Dpto. Cartografía y Expresión Gráfica en Ingeniería, Universidad de Las Palmas de Gran Canaria, Campus Tafira Baja 

35017-Las Palmas de Gran Canaria, SPAIN. mmartin@dcegi.ulpgc.es 

(3) Instituto Geológico y Minero de España. Ríos Rosas, 23. 28003-Madrid. SPAIN. ma.perucha @igme.es 

(4) Dpto. Geología, Museo Nacional de Ciencias Naturales (CSIC), 28006-Madrid SPAIN 

(5) Dpto. Estratigrafía. Facultad CC. Geológicas. Universidad Complutense de Madrid. 28040-Madrid. SPAIN. 

(6) Dpto. Geología, Universidad de Alcalá de Henares, 28871- Alcalá de Henares, Madrid SPAIN 

Abstract (Relief production, uplift and active tectonics in the Gibraltar Arc (South Spain) from the Late Tortonian to the 

present): This work analyzes the topographic evolution of the Gibraltar Arc (Betic Cordillera) from the Tortonian to the present. 

This first preliminary approach is based on the morphometric analysis of drainage basins and the identification of stratigraphic 

markers for ancient positions of the sea-level. Computed data indicate that mean thickness of eroded materials by fluvial erosion is 

593 m, generating an overall isostatic rebound of about 495 m. These data fit with computed mean elevations for individual basins, 

suggesting a quasi-isostatic equilibrium in the zone. However uplift is differentially distributed indicating that the main tectonic 

structures outcropping in the zone (ancient Betic thrusts and main tectonic contacts) are accommodating differential uplift, and 

therefore have been actives throughout Late Neogene to Quaternary times. 

Key words: Differential Uplift, Geophysical relief, Erosional unloading, Gibraltar Arc, Betic Cordillera, Spain. 

INTRODUCTION 

Isostatic uplift in response to erosional unloading of 

single faulted range fronts, mountain ranges and 

entire cordilleras has been proved to be a key 

process for the long-term landscape evolution in 

tectonically active areas (e.g. Molnar and England, 

1990; Bishop, 2007; Fernández-Ibáñez et al., 2010). 

Denudational unloading promotes relevant 

lithospheric upward flow, and subsequent surface 

uplift, controlling the topographic evolution of active 

orogens and backfeeding the rhythms of mass rocktransfer 

between ranges, local sedimentary basins 

and oceans/seas (e.g. Watts, 2001). Some authors 

envisage this process to explain regional surface 

uplift in relation to active fluvial erosion in mountain 

chains (Gilchrist et al., 1994; Small and Anderson, 

1998; Brocklerhurst & Whipple, 2002), oceanic 

islands (Menéndez et al., 2008) or even single 

faulted-fronts (Fernández-Ibáñez et al., 2010). On the 

other hand, aside of the backfeeded contribution of 

tectonic and isostatic uplift, cyclic or unique events of 

base-level drop can affect to the related rates of 

erosion and uplift. This study is a preliminary 

analysis focused on the topographic development of 

the Gibraltar Arc (Betic Cordillera, South Spain) since 

Late Neogene times, and it is part of a large project 

within the framework of the EUROCORES- 

TOPOMED Program on the topographic evolution of 

the Betic-Rif Cordillera in the Western Mediterranean 

Basin. 

GEOLOGICAL BACKGROUND 

The development of the Betic–Rif orogen on top of 

the Africa–Iberia plate boundary across the transition 

227 

zone between the Atlantic Ocean and the 

Mediterranean Sea resulted in a complex structural 

framework mostly developed during the Alpine 

orogeny. The role of fluvial unloading within the 

northern segment of the Gibraltar Arc is analyzed by 

means of the morphometric analysis of the main 

“direct” river basins draining towards the Gulf of 

Cádiz (west) and the Alboran Basin (East) from the 

fluvial outlets of the Guadalete river (Cádiz)) up to 

Motril (Granada). The study comprises the analysis 

of the computed bulk volume of rock-mass removed 

by erosion since the Tortonian until the present, as 

well as the analysis of this same parameter for 

discrete time-windows: Tortonian, Messinian, 

Pliocene and Quaternary. This last sequenced 

analysis will allow establishing time-dependent 

differential uplift, but also discrete temporal pulses of 

uplift for the whole studied zone, especially before 

and after the “Messinian Salinity Crisis”. This event 

occurred between 5.96 and 5.33 Myr ago triggering 

the disconnection of the Mediterranean Sea from the 

Atlantic Ocean in response to the final stages of 

build-up of the Gibraltar Arc, as well as a rapid 

episode of desiccation in the Mediterranean basin 

with a related sea-level drop of about 1000 metres in 

the western Mediterranean (e.g. Blanc, 2006). In fact, 

recent studies (i.e. Iribarren et al., 2009) indicate that 

the continentalization (i.e. emersion or uplift) of this 

sector of the Betic Cordillera took place around 5.3 

Myr ago, probably linked to this event. In fact 

computed data of sedimentary budget (5,490 

km 3 /Myr) and sedimentation rates (0.18 mm/yr) 

within the Alboran Basin during the Late Neogene 

were higher than those recorded for previous and


EARTHQUAKE 

ARCHAEOLOGY 



subsequent time-periods. The Messinian sea-level 

drop generated an unusual topographic scenario 

along the growing emerged tectonic wedge of the 

Gibraltar Arc, with an oceanic (Atlantic) foreland 

basin to the West and a large desiccated basin to the 

East. This situation gave place to an abrupt 

asymmetry in the base-level at both sides of the 

growing tectonic wedge and therefore, triggered 

asymmetric fluvial erosion, probably backfeeding 

differential uplift in mountain building. The 

subsequent Zanclean Flooding (García Castellanos 

et al., 2010) inundated again the entire 

Mediterranean basin in an instantaneous time-span 

of about 2 years dissipating a gravitational potential 

energy of about 1.63x10 22 J and balancing again 

base-levels at both sides of the Gibraltar Arc. Sea 

level reached a maximum of ca. + 70 m during the 

Pliocene previous to the eventual growing of the 

Antarctica ice sheet. 

ANALYSIS OF RELIEF AND UPLIFT 

A digital elevation model of 40x40m pixel resolution 

has been used in order to extract main ridgelines and 

drainage basins of the Gibraltar Arc from Cádiz 

(West) to Motril (East), differentiating 50 individual 

drainage basins larger than 18 km 2 over a bulk area 

of 14552 km 2 , which excludes the western slope 

basins draining towards the Guadalquivir Basin. 11 

drainage basins dissect the western Atlantic slope of 

the Gibraltar Arc with an area of 5849 km 2 , and 39 

ones are located in the Mediterranean slope over an 

area of 9063 km 2 . For a preliminary estimation of the 

volume of removed materials by fluvial erosion a preincision 

surface was computed for each individual 

basin from the present elevation of their ridgelines 

perimeters and representative erosional surfaces. 

The results indicate eroded budgets of 8632.9 km 3 

for the bulk of the studied section of the Gibraltar Arc, 

of which 836.7 km 3 corresponds to the Atlantic slope 

and 7765.2 km 3 for Mediterranean one. Dividing the 

obtained removed volumes by the present drained 

areas results in the theoretical equivalent thickness 

of removed materials or “geophysical relief” (Small 

and Andersson, 1998), which for the bulk studied 

area is of 593 m. 

In a second step hypsometric curves were computed 

for each individual drainage basin in order to obtain 

the distribution and values of mean and maximum 

present elevations for the studied basins, resulting in 

a mean value for the whole studied area of 411.4 m 

above the sea-level, but with a clear asymmetry 

between the Atlantic (108 m) and Mediterranean 

(492.3 m) slope of the Gibraltar Arc, with computed 

percentual weights over the total analyzed area 

always above of 70% and therefore statistically 

representative. Comparing the values of mean 

elevations and computed geophysical reliefs offer 

well fit linear regressions, with correlation coefficients 

close to 0.9 for the whole studied zone and their 

zonations. This fact implies a good correlation 

between the present elevation and the theoretical 

thickness of removed materials by fluvial erosion, 

suggesting near-isostatic equilibrium conditions for 

the zone. From this approach, in a third step the zone 

was divided in individual crustal/lithospheric sectors 

corresponding to the analyzed basins in order to 

calculate the consequent uplift in response to 

erosional unloading following the Airy-Heiskannen 

isostatic model. Different approaches have been 

applied in this study in order to obtain the isostatic 

response (e.g. Gilchrist et al., 1994; Fernández- 

Ibáñez et al., 2010), but the best fitting ratios 

between computed uplift and present mean elevation 

Fig. 1: Digital Elevation Model of the Gibraltar Arc (Western Betic Cordillera) displaying the main elements of the present relief, 

drainage basins and stratigraphic markers (littoral sediments) analyzed in this study. Mapped Betic thrusts and normal faults 

(conventional symbols) are those controlling differential uplift from the Messinian. A: Atlantic Slope; M: Mediterranean Slope, 

subdivided in Frontal Flysch (Mf), Eastern (Me) and Western (Mw) analyzed sectors. Gub: Guadalete Neogene basin; Rb: 

Ronda Neogene basin; S-Az: Setenil-Antequera zone; Grb: Granada Neogene basin; Gfb: Guadalfeo Neogene basin. Numbers 

corresponds to the individual analyzed drainage basins. 

228


EARTHQUAKE 

ARCHAEOLOGY 



correspond to the application of the Gilchrist et al. 

(1994) isostatic constant, which offer linear 

regression lines with correlation indexes up to 88%. 

This fact suggests that isostatic compensation is 

more related to crust/mantle models rather to 

lithosphere/ asthenosphere models, but also that the 

studied zone should be in near-isostatic equilibrium 

at present. The application of crust/mantle isostatic 

constant yields a value of +495 m for the mean uplift 

of the whole investigated area of, but differentially 

distributed from the Atlantic flysch sector (+74.8 m), 

the Mediterranean flysch sector (+170.4 m), the 

Mediterranean mantle-core complex of the Ronda 

Range sector (+417.2 m) and the Mediterranean 

sector west to the Málaga basin along the Mijares 

Range (+592.8 m). However, taking into account the 

present mean maximum elevation of the analyzed 

area (1195.8 m), maximum denudational isostatic 

rebound represent the 41-49% of the recorded uplift 

since the Tortonian, the remaining 59-51 % should 

be achieved by the African-Eurasian plates 

convergence. 

Implementation of digital geological data obtained 

from the IGME data base (GEODE) in the digital 

elevation model allows the extraction of 

representative linear and punctual stratigraphic 

markers for different Late Neogene sea-level 

positions. The geological data implemented in the 

DEM only correspond to littoral, shallow marine and 

proximal delta units and the intersection of their 

upper contacts with the present relief sketch or depict 

approximate paleo-shorelines with a confidence of ± 

40m forced by the DEM resolution. These relative 

paleoshorelines has been obtained for different timewindows 

corresponding to the Late Tortonian, 

Messinian, Lower Pliocene, Upper Pliocene and 

Lower Pleistocene. These appear identified by 

means of different colours in the map of figure 1. 

The analysis of these stratigraphic markers is still in 

progress, but the preliminary analysis of their 

spatial/elevation distribution clearly indicates that the 

main stage of relief production in the studied zone 

was attained after the relevant Messinian sea-level 

drop (5.96 Myr). Most of the littoral Messinian 

depostis identified in the GEODE data base in the 

Mediterranean slope of the studied zone may mainly 

correspond to younger Pliocene sediments. 

Therefore “true” Messinian deposits in the studied 

area are only present northwards and westwards of 

the main ridgeline of the Gibraltar Arc (Fig. 1) along 

the ancient Atlantic front of the growing tectonic 

wedge. In this zone Messinian, but also Tortonian 

deposits, occur at ridgeline locations on the 

headwaters of the present drainage basins, which 

obviously indicate that main fluvial dissection stages 

are post-Messinian. 

The occurrence of sediments of these ages in the 

present headwaters of the Mediterranean slope is 

consequence of Messinnian and post-Messinian 

aggressive headward erosion of the largest fluvial 

basins trespassing the ancient main ridgeline of the 

Gibraltar Arc (Fig. 2) such as those corresponding to 

the Guadalfeo, Guadairo, Guadarranque and 

229 

Guadalhorce rivers (Fig.1). This ridgeline was 

already emerged during the Tortonian limiting the 

Noegene basins of Ronda and Granada by the south. 

The main emerged ridgeline was only interrupted 

(inundated) during the pre-Messinian times in the 

Málaga Basin along the present Guadalhorce 

corridior, the unique location in which there are 

outcrops of littoral/shallow marine Tortonian 

sediments eastwards the ancient ridgeline of the 

Gibraltar Arc within the studied zone (Fig. 1). The 

Pliocene and Pleistocene littoral materials in the 

Mediterranean simply partially filled the main fluvial 

basins generated during the Messinian. These 

deposits never appear in ridgeline positions along the 

entire Mediterranean slope. 

The elevation distribution of post-Messinian sealevels 

markers strongly support that the 

aforementioned differential uplift is being conducted 

along the more relevant NNE-SSW or N-S thrust 

contacts developed during the final stages of the 

construction of the Gibraltar Arc broadly delineated in 

Fig. 2: Conceptual Model for the interaction of fluvial 

unloading and gravitational unloading triggering the 

subsequent differential uplift along the Gibraltar Arc during 

the Messinian sea-level drop (rates from Iribarren et al., 

2009). 

Figure 1. These tectonic structures can be 

considered as active ones in which a vertical partition 

of tectonic and isostatic movements seems to occur 

in order to distribute regional differential uplift. In the 

central sector of the Gibraltar Strait these tectonic 

structures controlled the generation of N-S subsiding 

troughs (controlled by normal faulting) during the 

Pliocene, clearly control the distribution of differential 

uplift of the Last Interglacial marine terraces and are 

presently related with instrumental seismicity (Zazo 

et al., 1999; Silva et al., 2006). 

CONCLUSIONS 

The analysis of relief of the Gibraltar Arc presented in 

this work is a preliminary approach to decode the 

evolution of the onshore topography of this sector of 

the Betic Cordillera from the Tortonian to the present. 

Initial computed data indicate that the equivalent 

thickness of erosional unloading (geophysical relief) 

is of 593 m for the whole studied area. Isostatic uplift 

response to fluvial unloading is of 495 m, but 

differentially distributed, with maximum values in the 

eastern studied sector south of Sierra Nevada 

(Motril) and minimum values in the western Atlantic 

sector (Cádiz). Subsequent analyses will be also


EARTHQUAKE 

ARCHAEOLOGY 



focused in the relationships of seismotectonic activity 

of the more outstanding North-South tectonic 

structures accommodating differential uplift in the 

central-axial sector of the Gibraltar Arc. 

The sequential analysis of eroded volumes during 

discrete periods will allow to establish theoretical 

values of uplift for the different studied drainage 

basins and sectors. The preliminary results in this 

study indicate an accumulated mean uplift of +495 m 

since the end of the Tortonian (c.a. 7.2 Myr) which 

suggest broad mean uplift rates of 0.068 mm/yr for 

this time-period, clearly decreasing towards the 

Gibraltar Strait area where uplift rates are of ca. 0.01- 

0.02 mm/yr. The obtained values are within the range 

of the estimated uplift rates by other authors, and 

other methods, for the whole Betic Cordillera (i.e. 

Iribarren et al., 2009) which is of 0.05 mm/yr, but 

decreasing to 0.02 mm/yr to the Gibraltar Strait (Fig. 

2). 

A final preliminary remark is that isostasy triggered 

by mass-rock transfer between the growing tectonic 

wedge of the Gibraltar Arc and two the adjacent 

marine basins can explain about the 40-50% of the 

recorded uplift since the end of the Tortonian. On the 

other hand, the volume of post-Tortonian sediments 

accumulated in the Alboran basin indicates mean 

rates of sedimentation of about 0.18-0.24 mm/yr (Fig. 

2). The conversion of these sedimentation rates to 

erosion rates will allow checking the uplift response 

by means of different isostatic crust/mantle and 

lithosphere/asthenosphere conceptual models. 

These models will also take into account the isostatic 

rebound related to sea-water unloading in the 

Mediterranean slope triggered by the relevant sealevel 

drop occurred during the Messinian (Fig.2). 

Acknowledgements: This work has been supported by the 

EUROCORES-TOPOMED PROGRAM through the Spanish 

research project CGL2008-03474-E/BTE (CSIC) produced 

by the TopoMed-Spain Onshore Research Group of the 

University of Salamanca. Also financial support has been 

obtained from the research projects CGL08-03998BTE, 

CGL08-04000BTE 

References 

Blanc, P. L. (2006). Improved modelling of the Messinian 

Salinity Crisis and conceptual implications. Palaeogeogr. 

Palaeoclimatol. Palaeoecol. 238, 349–372. 

Bishop, P. (2007). Long-term landscape evolution: Linking 

tectonics and surface processes. Earth Surface 

Processes and Landforms, 32, 329–365. 

Brocklehurst, S.H., Whipple, K.W. (2002). Glacial erosion 

and relief production in the Eastern Sierra Nevada, 

California. Geomorphology 42 , 1-24. 

Fernández-Ibáñez, F., Pérez-Peña, J.V., Azor, A., Soto,J. 

Azañón, J.M. (2010). Normal faulting driven by 

denudational isostatic rebound. Geology, 38, 643-646. 

Garcia-Castellanos, D., Estrada, F., Jiménez-Munt, I., 

Gorini,C., Fernández, M., Vergés, J. & De Vicente, R. 

(2009). Catastrophic flood of the Mediterranean after the 

Messinian salinity crisis. Nature, 462, 778-782. 

Gilchrist, A.R., Kooi, H. & Beaumont, C. (1994). Post- 

Gondwana geomorphic evolution of southwestern Africa: 

Implications for the controls on landscape development 

from observations and numerical experiments: Journal of 

Geophysical Research, 99, 12,211–12,228. 

Iribarren, L., Vergés, J. & Fernandez, M. (2001). Sediment 

supply from the Betic–Rif orogen to basins through 

Neogene. Tectonophysics, 475, 68–84 

Menéndez, I., Silva, P.G., Martín Betancor, M., Pérez 

Torrado, F.J., Guillou, H. & Scaillet; S. (2008). Fluvial 

dissection, isostatic uplift, and geomorphological 

evolution of volcanic islands (Gran Canaria, Canary 

Islands, Spain). Geomorphology, 102, 189 – 203. 

Molnar, P. & England, P. (1990). Late Cenozoic uplift of 

Mountain ranges and global climate change: Chicken or 

egg?: Nature, 346,,29– 34. 

Silva, P.G., Goy, J.L., Zazo, C., Bardají, T., Lario, J., 

Somoza, L., Luque, L. & González Fernández, F.M. 

(2006). Neotectonic fault mapping at the Gibraltar Strait 

Tunnel area, Bolonia Bay (South Spain). Eng. Geology, 

84, 31–47 

Small, E. & Anderson, R. (1998). Pleistocene relief 

production in Laramide mountain ranges, western United 

States: Geology, 26, 123–126. 

Watts, A.B. (2001). Isostasy and flexure of the lithosphere. 

Cambridge University Press, Cambridge, UK, pp. 458. 

Zazo, C., Silva, P.G., Goy, J.L., Hillaire-Marcel, C. Lario, J. 

Bardají, T. & González, A. (1999). Coastal uplift in 

continental collision plate boundaries: Data from the Last 

interglacial marine terraces of the Gibraltar Strait area 

(South Spain). Tectonophysics, 301, 95-119. 

230


EARTHQUAKE 

ARCHAEOLOGY 



REASSESSING ANCIENT EARTHQUAKES ON MINOAN CRETE 

GETTING RID OF CATASTROPHISM 

Sintubin, Manuel (1, Simon Jusseret (2,1), Jan Driessen (2) 

(1) Geodynamics & Geofluids Research Group, Department of Earth & Environmental Sciences, Katholieke Universiteit Leuven, 

Celestijnenlaan 200E, B-3001 Leuven. BELGIUM. Email: manuel.sintubin@ees.kuleuven.be 

(2) Aegean Interdisciplinary Studies Research Group (AegIS-CEMA-INCAL), Université Catholique de Louvain, Place B. Pascal 

1, B-1348 Louvain-la-Neuve. BELGIUM. Email: jan.driessen@uclouvain.be; simon.jusseret@uclouvain.be 

Abstract(Assessing ancient earthquake on Minoan Crete. Getting rid of catastrophism): Early in the 20 th century Arthur 

Evans invoked catastrophic earthquakes to explain the destruction encountered in the Palace of Knossos. Ever since, these 

earthquake catastrophes have been used as reference events to which structural damage or destruction layers on Minoan sites 

are indiscriminately attributed in archaeological and archaeoseismological publications. However, structural damage to 

archaeological remains cannot be unequivocally attributed to earthquakes. A detailed analysis of Late Minoan contexts moreover 

reveals that multiple moderate earthquakes occurred during this 200-year period (c. 1600 – 1400 BC). All evidence suggests that 

earthquakes did not play a crucial role in Minoan history and did certainly not cause the decline of Minoan civilization. This 

reassessment of the Minoan case clearly demonstrates that earthquakes in itself are incapable of causing the collapse of a 

civilization. 

Key words: Minoan Crete, Late Bronze Age, catastrophism, earthquake archaeology 

Early in the 20 th century Arthur Evans invoked 

catastrophic earthquakes to explain the destructions 

encountered during the excavations of the Minoan 

Palace at Knossos, Crete (Fig. 1). Such seismic 

catastrophes were considered to have ended the 

Protopalatial period (c. 1700 BC), to have been 

responsible for the “great destruction” during the 

Neopalatial period (c. 1600 BC), and to have 

ultimately caused the collapse of Minoan civilization 

around 1450 BC (Evans, 1928). 

Ever since, these earthquake catastrophes have 

been taken for granted and used as reference events 

to which structural damage to buildings and other 

cultural remains or earthquake-related destruction 

layers are indiscriminately attributed in 

archaeological (e.g., Sakellarakis & Sapouna- 

Sakellaraki, 1991) and archaeoseismological 

publications (e.g., Monaco & Tortorici, 2004; 

Vallianou, 1996). But is there any reliable evidence to 

support the existence of such catastrophic 

earthquakes and can we parameterize them? 

Fig. 1: The “house of the fallen blocks” at the palatial 

site of Knossos, a particular damage interpreted by 

Evans (1928) to have been caused by a catastrophic 

earthquake. 

231 

First, it is extremely difficult to attribute unequivocally 

structural damage to Minoan archaeological remains 

to earthquakes. In most cases it cannot be excluded 

that other physical and/or anthropogenic agents have 

generated the damage observed (cf., Driessen, 

1995). A macroseismological parameterization of 

these ancient earthquakes based on the detailed 

archaeological record remains a very challenging 

prospect. 

Secondly, a detailed analysis of Late Minoan (c. 1600 

– 1400 BC) archaeological contexts (cf., Driessen & 

Macdonald, 1997) shows that earthquake-related 

damage, repairs, adjustments (e.g., Driessen, 1987) 

and/or abandonment are all rather isolated and local 

phenomena within and not necessarily 

contemporaneous between the different sites. This 

evidence reveals that most probably multiple 

moderate earthquakes occurred during this 200-year 

time period, rather comparable to today’s seismicity 

of the island. 

There is seemingly only consistent archaeological 

evidence for widespread, earthquake-related damage 

on Crete, as well as on Thera (Santorini), Kos, and


EARTHQUAKE 

ARCHAEOLOGY 



Rhodes at c. 1600 BC (transition of ceramic stage 

MM IIIB to LM IA), known as the “great destruction” 

(Evans, 1928). This potentially “catastrophic” event 

is, however, followed by a sudden increase of the 

number of secondary sites, in particular in eastern 

Crete (cf., Driessen & Macdonald, 1997), new palatial 

architecture with new architectural (anti-seismic?) 

elements, such as ‘pier-and-door partitioning’ (cf., 

Driessen, 1987), and the greatest construction 

program of any prehistoric era in the Aegean. The 

heyday of Minoan civilisation followed this major 

seismic event. 

Between c. 1520 and c. 1480 BC (ceramic stage LM 

IA) this prolific building program came to an abrupt 

end. Monumental buildings were left unfinished. This 

heralded the demise of Minoan society. In a period of 

one to two generations – from c. 1480 to c. 1425 BC 

(ceramic stage LM IB) – a wave of fire destruction 

raged over the island. Settlements were abandoned, 

population density declined. Both a “crisis 

architecture” (cf., Driessen, 1995) and a “crisis cult” 

took hold of Minoan society. Evidence for 

earthquake-related damage in that period indicates 

multiple moderate earthquakes affecting local 

communities rather than island-wide destructive 

events. Most of the destruction was indeed caused 

by man. All evidence indicates that the socio-political 

and economic landscape of Minoan society 

completely disintegrated and collapsed in that period, 

leaving behind a “failed state”. The power vacuum 

was later – during ceramic stage LM II (c. 1425 to 

c.1400 BC) – filled by the Mycenaens. 

Although Late Minoan society can clearly be 

characterized as a society in crisis (cf. Driessen & 

Macdonald, 1997), no hard evidence exists to link 

this societal decline with (catastrophic) earthquakes. 

Even the existence of such seismic catastrophes 

during Minoan history – except for the c. 1600 BC 

event – should be questioned. Minoans lived with 

earthquakes, very much as modern Cretans do. 

Earthquakes did not play a crucial role in Minoan 

history and did definitively not cause the decline of 

Minoan civilization. At most, they added some extra 

stress to a society already in crisis. 

This reassessment of the Minoan case illustrates that 

earthquakes, irrespective of their magnitude and 

recurrence, provoke different societal responses, 

largely depending on the political, social, economic 

and military context. Earthquakes in itself are 

incapable of causing the collapse of a community, let 

alone a civilization. It’s therefore time to get rid of the 

catastrophism that has burdened earthquake 

archaeology for too long (e.g., Nur, 2008; Nur & 

Cline, 2000). 

Acknowledgements: This contribution frames in the 

UNESCO-IUGS funded International Geoscience 

Programme IGCP 567 Earthquake Archaeology. S. 

Jusseret is currently a Intercommunity Scientific 

Postdoctoral Collaborator of the Francqui Foundation 

(Belgium) at the K.U.Leuven. 

References 

Driessen, J. (1987). Earthquake-Resistant Construction and 

the Wrath of the "Earth-Shaker". Journal of the Society of 

Architectural Historians 46, 171-178. 

Driessen, J. (1995). "Crisis Architecture". Some 

observations on Architectural Adaptations as Immediate 

Responses to Changing Socio-Cultural Conditions. Topoi 

5, 63-88. 

Driessen, J., Macdonald, C. F. (1997). The Troubled Island. 

Minoan Crete before and after the Santorini Eruption, 

Liège & Austin. 

Evans, A. (1928). The Palace of Minos, part II., London. 

Monaco, C., Tortorici, L. (2004). Faulting and effects of 

earthquakes on Minoan archaeological sites in Crete 

(Greece). Tectonophysics 382, 103-116. 

Nur, A. (2008). Apocalypse. Earthquakes, Archaeology, and 

the Wrath of God. Princeton University Press, Princeton. 

Nur, A, Cline, E. H. (2000). Poseidon's Horses: Plate 

Tectonics and Earthquake Storms in the Late Bronze Age 

Aegean and Eastern Mediterranean. Journal of 

Archaeological Science 27, 43-63. 

Sakellarakis, J. A., Sapouna-Sakellaraki, E. (1991). 

Archanes. Ekdotike Athenon S.A., Athens. 

Vallianou, D. (1996). New Evidence of Earthquake 

Destructions in Late Minoan Crete. In: 

Archaeoseismology (edited by Stiros, S. C. & Jones, R. 

E.). Fitch Laboratory Occasional Paper 7. Institute of 

Geology & Mineral Exploration & The British Scholl at 

Athens, Athens, 153-167. 

232


EARTHQUAKE 

ARCHAEOLOGY 



PALEOTSUNAMIS EVIDENCE FROM A COMBINED INLAND AND OFFSHORE STUDY IN 

THE AUGUSTA BAY AREA (EASTERN SICILY, ITALY) 

Smedile, Alessandra (1, Paolo Marco De Martini (1), Daniela Pantosti (1) 

(1) Istituto Nazionale di Geofisica e Vulcanologia. Via di Vigna Murata, 605. 00143- Roma. ITALY. 

Email: alessandra.smedile@ingv.it, paolomarco.demartini@ingv.it, daniela.pantosti@ingv.it 

Abstract (Paleotsunamis evidence from a combined inland and offshore study in the Augusta bay area (eastern Sicily, 

Italy): We present the geological evidence for a 4000 year-long record of repeated tsunami events along the coast of the Augusta 

Bay derived from the combination of inland and offshore cores data. The research was carried out through a multi-theme approach 

which benefited from an long historical record that we used to guide detailed geomorphologic, geologic, and geophysical surveys, 

that were performed both inland and offshore. These surveys served to locate and characterize the best sites were to perform the 

coring collection that where studied through and several laboratory analyses. The Augusta Bay represents a unique case study 

because besides showing a long record of paleotsunami events, it contains the first inland and offshore evidence for the Crete AD 

365, and Santorini Late Minoan tsunamis in the central-western Mediterranean area. 

Key words: Eastern Sicily, tsunami deposits, AD 365 Crete and Santorini events. 

Eastern Sicily (Italy) was repeatedly affected by 

tsunami waves related to large local historical 

earthquakes (e.g. 1908, 1693, 1169) (CPTI Working 

group, 2004) as well as to far-field sources (e.g. AD 

365 Crete earthquake) (Jerome, 380). Along the 

eastern Sicily coasts, we selected the Augusta Bay, a 

natural gulf about 15 km wide and with a 25 km-long 

shoreline (Fig.1), as the key area of this study. In 

fact, it is one of the locations where both information 

available from historical written reports on tsunami 

effects (hit localities, inundated areas and run-up 

distribution) (Gerardi et al., 2008) and local 

geomorphology suggest it is very favorable for the 

research of the geological signature of past tsunamis. 

Well-targeted sediment samples have been collected 

both inland and offshore through coring at different 

depths. Small ponds, marshes and lagoons 

characterize the coastal area, while a relatively wide 

continental shelf with a thick late-Holocene record 

has been investigated offshore through the 

acquisition of a tight grid of CHIRP-sonar profiles. 

The integrated interpretation of the geophysical and 

geological data has been carried out in order to 

recognize, date and correlate key-layers in the 

sediment column that may be directly or indirectly 

related to tsunami events. 

A total number of 22 cores were collected inland at 

two different sites with a maximum distance of 530 m 

from the present coastline (De Martini et al., 2010). 

The dominant fine to very fine stratigraphy is 

intercalated by at least 6 high-energy depositional 

layers, repeatedly found in several cores. These 

relatively thin (about 10 cm) single massive and 

structureless beds with abrupt erosional lower 

contact are made of coarse to fine sand and present 

a bioclastic component (sometimes predominant) 

made of microfauna (benthic and planktonic 

233 

foraminifera, from both shallow and open marine 

environment) and shell fragments both suggestive of 

a marine origin. Chronological constraints on the age 

of these deposits are based on 8 AMS radiocarbon 

datings and on the attribution of a tephra layer to the 

122 BC Etna eruption. For the marine shell samples 

(details in De Martini et al., 2010 and Smedile et al., 

2011), measured C14 ages were dendrochronologically 

corrected using a marine calibration 

curve that incorporate a time-dependent global ocean 

reservoir correction of about 400 yrs (Reimer et al., 

2009). Moreover, the marine palaeo-reservoir effect 

was subjected to the local effect (R offset) that, in 

the Mediterranean Sea, appears to be constant for 

the past 6 or 7 ka (Reimer and McCormac, 2002). 

The appropriate R offset can be selected from the 

Chrono Marine reservoir Database (Reimer and 

Reimer, 2001). 

Figure 1: The Augusta Bay area in eastern Sicily, Italy. 

The Augusta Hospital and Priolo Reserve sites are 

marked with white empty rectangles, while a white box 

locates the offshore coring site (MS-06). Two 

panoramic pictures of the in-land investigated sites are 

also shown.


EARTHQUAKE 

ARCHAEOLOGY 



On the basis of the combination of all the data 

collected, the inland sequence spans the last 4100 

yrs. In terms of tsunami timing, we could list them as 

follow (PR= Priolo Reserve site; AU= Augusta 

Hospital site): younger than 1420–1690 AD (PR-01), 

650–770 AD (AU-00), 160–320 AD (PR-02), 600–400 

BC (AU-01), 800–600 BC (PR-03), 975–800 BC (AU- 

02) and 2100–1635 BC (PR-04). 

Three of the tsunami deposits found at the Priolo 

Reserve site may be associated with historical 

tsunamis: PR-01 with the 1693 local event, PR-02 

with the 365 AD Crete event and PR-04 with the ca. 

3600 BP Santorini event. 

The offshore record was derived from a 6.7 m-long 

piston-core sampled 2 km offshore the Augusta 

harbor at 72 m depth (Fig.1). The core study includes 

X-ray imaging, isotopic dating, tephrachronology, 

grain-size and foraminiferal analyses (Smedile et al., 

2011). The homogeneous sequence of dark grey 

mud is interrupted at -2.9 m b.s.f. (below sea floor) by 

the same Etna tephra deposit found inland. Through 

the analysis of tephrostratigraphy, radiocarbon dating 

and radioactive tracers, the entire core sequence has 

been dated back to the last 4500 yrs. Furthermore, 

we estimated an age range for the 12 high energy 

intervals as follows: E1 (AD 1820-1920), E2 (AD 

1430-1810), E3 (AD 930-1170), E4 (AD 590-800), E5 

(AD 430-660), Ex (AD 90-370), E6 (BC 350-130), E7 

(BC 580-320), E8 (BC 660-400), E9 (BC 800-560), 

E10 (BC 1130-810), E11 (BC 1720-1200). Moreover, 

the quantitative micropaleontological (on the benthic 

foraminifera assemblage) and sedimentological 

analyses highlighted 12 anomalous layers marked by 

high concentration of displaced epiphytic foraminifera 

(species growing in vegetated substrates like the 

Posidonia oceanica) and subtle grain size changes. 

These anomalous layers are likely to have been 

caused by high-energy events, with tsunamis (backwash 

wave) as best candidates. This hypothesis is 

also supported by the fact that the ages of 5 of these 

peculiar layers coincide with that of historical 

tsunamis [1908 (E1), 1693 (E2), 1169 (E3), AD 365 

Crete (Ex) and ca. 3600 BP Santorini (E11)]. 

Moreover, to better detail and replicate the MS06 

results at least for the recentmost sequence, new 

cores were collected in the northern part the Augusta 

Bay. These cores were sampled in order to define a 

W-E oriented transect along the shelf from 58 to 110 

m water depth. 

affected by intermittent erosional and sedimentation 

events as well as by antrophic activities . 

Acknowledgements: This work was funded by the Italian 

Dipartimento della Protezione Civile in the frame of the 

2004–2006 and 2007–2009 agreements with Istituto 

Nazionale di Geofisica e Geofisica e Vulcanologia — INGV 

with contribution of the EU Transfer project. We have to 

warmly thank M.S. Barbano, F. Gerardi, C. Pirrotta from 

Catania University, L. Gasperini, L. Bellucci, A. Polonia from 

ISMAR-CNR Bologna and P. Del Carlo and E. Boschi from 

INGV for their enthusiastic and productive collaboration. 

References 

CPTI Working group, 2004. Catalogo Parametrico dei 

Terremoti Italiani, version 2004 (CPTI04). INGV, 

Bologna. http://emidius.mi.ingv.it/CPTI/ 

De Martini, P.M., Barbano, M.S., Smedile, A., Gerardi, F., 

Pantosti, D., Del Carlo, P., Pirrotta, C, (2010). A 4000 yrs 

long record of tsunami deposits along the coast of the 

Augusta Bay (eastern Sicily, Italy): paleoseismological 

implications. Marine Geology, 276, 42-57, doi: 

10.1016/j.margeo.2010.07.005 

Gerardi, F., Barbano, M.S., De Martini, P.M., Pantosti, D., 

2008. Discrimination of tsunami sources (earthquake vs. 

landslide) on the basis of historical data in eastern Sicily 

and southern Calabria. Bulletin of the Seismological 

Society of America, 98 (6), 2795–2805. 

Smedile A., P.M. De Martini, D. Pantosti, L. Bellucci, P. Del 

Carlo, L. Gasperini, C. Pirrotta, A. Polonia, E. Boschi 

(2011). Possible tsunamis signatures from an integrated 

study in the Augusta Bay offshore (Eastern Sicily–Italy), 

Marine Geology, 281, 1-13, doi: 

10.1016/j.margeo.2011.01.002. 

Reimer, P.J., McCormac, F.G., 2002. Marine radiocarbon 

reservoir corrections for the Mediterranean and Aegean 

Seas. Radiocarbon 44, 159–166. 

Reimer, P.J., Reimer, R.W., 2001. A marine reservoir 

correction database and on-line interface. Radiocarbon 

43, 461–463 suppl. mat.URL: http://www.calib.org. 

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, 

J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., 

Burr, G.S., Edwards, R.L., Friedrich,M., Grootes, 

P.M.,Guilderson, T.P.,Hajdas, I., Heaton, T.J., Hogg, 

A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac, 

F.G., Manning, S.W., Reimer, R.W., Richards, D.A., 

Southon, J.R., Talamo, S., Turney, C.S.M., van der 

Plicht, J., Weyhenmeyer, C.E., 2009. IntCal09 and 

Marine09 radiocarbon age calibration curves, 0–50, 000 

years cal BP. Radiocarbon 51 (4), 1111–1150. 

The Augusta Bay represents a unique case study 

because it allows a comparison between geological 

(both inland and offshore) and historical records. For 

the 365 AD Crete tsunami and the Late Minoan 

Santorini event, our findings represent the first 

inland-offshore evidence in the central-western 

Mediterranean area. On the basis of these results we 

can propose, for the past 4 ka in the Augusta Bay, an 

inland and offshore geologic average tsunami 

recurrence interval of about 550-650 and 320 years, 

respectively. This difference is conceivably due to the 

better preservation of the stratigraphic record in the 

offshore with respect to coastal areas, commonly 

234


EARTHQUAKE 

ARCHAEOLOGY 



BULGARIAN NATIONAL DIGITAL SEISMOLOGICAL NETWORK 

Solakov D. (1), L.Dimitrova (1), S.Nokolova (1), St. Stoyanov (1), S. Simeonova (1), L. Zimakov (2), L.Khaikin (3) 

(1) Dpt. Seismology, National Institute of Geophysics, Geodesy and Geography. Acad. G. Bonchev str.,bl. 3, 1113 – Sofia, 

BULGARIA, Emails: dimos@geophys.bas.bg, lidim@geophys.bas.bg, sbnik@geophys.bas.bg, stoyan@geophys.bas.bg, 

stelas@geophys.bas 

(2) Refraction Technology, Inc., USA. l.zimakov@reftek.com 

(3) Russia, leonidh@ramblert.ru 

Abstract (Bulgarian national digital seismological network): The Bulgarian National Digital Seismological Network (BNDSN) 

consists of: National Data Centre (NDC); 13 stations equipped with RefTek Seismic Recorders DAS130 and 1 station equipped 

with Quanterra 680; broadband seismometers and accelerometers. The real-time data acquisition and processing are performed 

by a system for signal detection; evaluation of the signal parameters; phase identification and association; source estimation. 

Seismic interactive analysis system and Early warning system are running in the NDC also. Modern digital equipment installed at 

Bulgarian seismic stations, careful selection of the software packages running in the data centre proved to be suitable choice for 

the purposes of BNDSN – to ensure reliable automatic localization of the seismic events and rapid notification of the governmental 

authorities in case of felt earthquakes on the territory of Bulgaria, to provide a modern basis for seismological studies in Bulgaria. 

Key words: Bulgarian National Digital Seismological Network, digital equipment, data processing 

INTRODUCTION 

In 2005, the Geophysical Institute (with governmental 

support) performed overall modernization of the 

Bulgarian National Seismological Network (BNDSN) 

(Fig.1). Modern digital equipment and broadband 

seismometers were installed in all stations. 

Data acquisition 

Real-time data transfer from seismic stations to NDC 

is realized via Virtual Private Network (VPN) of the 

Bulgarian Telecommunication Company (BTC) 

(Fig.2) with the following characteristics: 64kbps 

baud rate from each individual site to the digital 

network of BTC; high security – closed Internet 

access; broadband 2Mbps optical line established 

between the NDC and the Centre of Communication 

Company in Sofia. This solution proved to be very 

stable and has ensured a reliable communication 

system through the six years of exploitation. 

Fig. 1: Bulgarian National Seismological Network 

At present the upgraded network consists of a 

National Data Centre (NDC), 13 stations equipped 

with RefTek High Resolution Broadband Seismic 

Recorders – model DAS 130-01/3, 1 station 

equipped with Quanterra 680 (installed in 1996 in 

station VTS by project PLATO1/MEDNET as a 

station from VBB Mediterranean Network) and 

sensors: very broadband - STS2, STS1, KS2000, 

RefTek151/120; broadband - CMG 3ESPC, 

CMG40T; short-period - S13 and accelerometers 

RefTek 131/03. 

Fig.2. Real-time data transfer from seismic stations to NDC 

via Virtual Private Network (VPN) and three layer local 

network 

235


EARTHQUAKE 

ARCHAEOLOGY 



The network availability exceeded 99.99%. The 

communication interruptions did not cause any data 

loss at the NDC. The data is backed up in the field 

station recorder 4Mb RAM memory and is 

retransmitted to the NDC immediately after the 

communication link is re-established. The recorders 

are equipped with 2 compact flash storages able to 

save more than 1 month long data. The data from the 

flash disks can be downloaded remotely using FTP. 

The data acquisition and processing hardware 

redundancy at the National Data Center is achieved 

by two clustered SUN Fire V240 servers with 

swappable software module and two Blade 1500 

Workstations. In the case of server failure the 

swappable module re-directs the real-time data flow. 

To secure the acquisition, processing and data 

storage processes a three layer local network is 

designed at the NDC. First layer incorporates 

acquisition and real-time processing equipment, the 

second layer consists of interactive processing and 

archiving modules and external users belong to the 

third layer (Fig.2). Real-time data acquisition is 

performed using REFTEK’s full duplex errorcorrection 

protocol RTPD. Data from the Quanterra 

recorder and foreign stations are fed into RTPD in 

real-time via SeisComP/SeedLink protocol. Using 

SeisComP/SeedLink software the NDC transfers 

real-time data to INGV, Roma and NEIC, USA. The 

BNDSN is a part of European Virtual Broadband 

Seismograph Network and the NDC transmits realtime 

data to the ORFEUS Data Centers. Regional 

real-time data exchange with Romania, Macedonia, 

Serbia and Greece is established at the National 

Data Center also. The data flow from neighbor 

countries is incorporated in the real-time data stream 

in the NDC and is used together with BNDSN data 

for localization of the local, regional and distance 

seismic events. 

Data processing 

Data processing is performed by the Seismic 

Network Data Processor (SNDP) (Haikin and 

Kushnir, 2005; Haikin et al., 2009) software package 

running on the both Servers. SNDP includes 

subsystems: 

Real-time subsystem (RTS_SNDP) – for signal 

detection; evaluation of the signal parameters; 

phase identification and association; source 

estimation; 

Seismic analysis subsystem (SAS_SNDP) – for 

interactive data processing; 

Early warning subsystem (EWS_SNDP) - based 

on the first arrived P-phases. 

The signal detection process is performed by 

traditional STA/LTA detection algorithm. Input data 

streams are band-pass filtered with band limits 

described in a parametric file. The filter parameters of 

the detectors are defined on the base of previously 

evaluated ambient noise at the seismic stations. 

The annual noise distribution at the BNDSN stations 

is presented on Fig. 3 (the 14 figures around the 

map). 

Fig. 3. Evaluation of ambient seismic noise at BNDSN stations 

236


EARTHQUAKE 

ARCHAEOLOGY 



For the analysis purposes, the range of recorded 

periods from 0.1s to 100s is divided into two subperiods: 

1. The periods longer than 1s, where the 

microseisms are spread; 

2. The periods shorter than 1s, where the most 

common noise source is the man-made one. 

The results show that: 

– For the periods greater then 1s the mode 

curve of Power Spectral Density (PSD) closely tracks 

the New Low Noise Model (NLNM) (Fig.3) and even 

low magnitude events should be reliably detected. 

There is no need to filter the signal in this frequency 

range and no need to add parameters in the 

parametric detector file. 

– For the periods lower then 1s the 

distribution of PSD has unique model for every 

station depending on specific noise sources 

(Dimitrova, 2009). The most significant noise sources 

are often related to human activities at or near the 

Earth´s surface. In this frequency range the filter 

parameters of detectors are selected on the base of 

the corresponding station noise model (Fig.3). 

The output of the Real-time subsystem data 

processing is a daily bulletin with the hypocentral 

determinations. The event localizations are 

visualized on a map and an e-mail is sent to the list 

of subscribers. In case of felt (M>=2.5) or damaging 

earthquake on the territory of Bulgaria the NDC 

issues information to the government authority, mass 

media and broad public. The NDC in close 

cooperation with the Civil defense carries out 

macroseismic investigation in the epicentral region. 

The interactive processing of the seismic event 

parameters and magnitude determinations are 

performed by manual graphic analysis of the 

seismogram (Fig. 4). An advantage of SAS_SNDP is 

easy access to the automatic arrivals and waveforms 

collected in a disk loop with capacity more than 365 

days. Additional advantage of the subsystem is the 

ability to operate with the waveforms for rapid 

manual relocation of the events in time interval close 

to real time. 

Fig. 4. Interactive data processing by means of Seismic analysis subsystem – SAS_SNDP - tuning of the seismic phase 

parameters 

The Early warning subsystem provides messages on 

the based of first arrived P-phases within 5 to 8 

seconds. An alarm module is switched on and the 

epicenter of the strong event is visualized on a map. 

Some extra modules for network command/control, 

state-of-health network monitoring and data archiving 

are running as well in the National Data Centre. 

Three types of archives are produced in the NDC - 

two continuous - miniSEED format and RefTek 

237 

PASSCALl format; and one event oriented in CSS3.0 

scheme format. 

CONCLUSIONS 

Modern seismological equipment installed at 

Bulgarian seismic stations, carefully selected and 

developed software packages for data processing 

proved to be suitable choice for the purposes of 

BNDSN and NDC. Currently, the NDC and BNDSN 

allow reliable automatic localization of magnitude


EARTHQUAKE 

ARCHAEOLOGY 



events ML>=1.5 within the network, and ML>=2.5 at 

regional distances. 

Presently BNDSN is a world-class broadband digital 

network providing reliable, high quality real-time 

seismic monitoring and rapid earthquake information 

to both scientific communities and authorities in 

Bulgaria and international community for seismic 

hazard mitigation and research. 

Acknowledgements: The modernization of the BNDSN 

was carried out within the projects: Decision SB - 

3/04.05.2005 of Permanent Commission for Prevention of 

the Population from Natural Disasters, Technological 

Accidents and Catastrophes (PCPNDTAC), Bulgaria, 

Contract N IKI -11/01.09.2005 with Bulgarian National 

Science Fund, Ministry of Education and Science: 

“Environmental Monitoring Implement for Risk Assessment 

of natural and man-made hazard (EMIRA)”. 

References: 

Dimitrova L., (2009). Noise level on selected digital stations 

of the National Operative Teleseismic System for Seismic 

Information (NOTSSI). Comptes rendus de l’Academie 

bulgare des Sciences. Vol 62, No4, pp 515-520. 

Haikin, L. and A. Kushnir, (2005). Seismic Network Data 

Processor (SNDP) - Comprehensive Software for UNIX 

Networks. SYNAPSE Science Center, . 122. 

Haikin L., L. Begun and M. Morskov, (2009). Haikin-SNDM- 

Applications. Seismic Network Data 

Monitoring.Comprehensive Software for UNIX Networks. 

Introductory Software manual. Version 2.0. 

238


EARTHQUAKE 

ARCHAEOLOGY 



PALEOSEISMIC STUDY OF THE SUDETIC MARGINAL FAULT AT THE LOCALITY BÍLÁ 

VODA (BOHEMIAN MASSIF) 

Štpaníková Petra (1, Nývlt Daniel (2), Hók Jozef (3), Dohnal Jií (4) 

(1). Dpt. Engineering, Institute of Rock Structure and Mechanics, Academy of Sciences of the Czech Republic, V Holešovikách 

41, 182 09 Prague 8.CZECH REPUBLIC. E-mail: stepancikova@irsm.cas.cz 

(2). Czech Geological Survey, Brno Branch, Leitnerova 22, 658 69 Brno, CZECH REPUBLIC, E-mail: daniel.nyvlt@geology.cz 

(3). Dpt. Geology and Paleontology, Faculty of Natural Sciences, Comenius University in Bratislava, Mlynská dolina, 842 15 

Bratislava 4, SLOVAKIA, E-mail: hok@fns.uniba.sk 

(4). Faculty of Sciences, Charles University in Prague, Inst. Hydrogeology, Engineering Geology and Applied Geophysics, 

Prague Albertov 6, Prague 2, 128 43, CZECH REPUBLIC, E-mail: dohnalj@natur.cuni.cz 

Abstract (Paleoseismic study of the Sudetic Marginal Fault at the locality Bílá Voda (Bohemian massif): The study area is 

situated in the north-eastern part of the Bohemian Massif and comprises SE part of the Sudetic Marginal Fault zone (SMF) with 

pronounced mountain front, which borders the Sudeten Mts. Paleoseismological trenches at the locality of Bílá Voda municipality 

performed across the SMF were carried out in order to study potential prehistoric morphogenic earthquakes responsible for the 

present-day morphology. Geomorphologic analysis and electrical resistivity tomography enabled to position the trenches at 

suitable site. The trenches revealed prevailing strike-slip character of the late Quaternary faulting. Faulted deposits were dated by 

OSL and radiocarbon dating methods and the results are discussed here. At least four or five events creating a surface rupture 

during late Quaternary (Holocene) were inferred. Due to strike-slip character, the horizontal displacement and the related slip-rate 

on the SMF is still under investigation. From the previous study, minimum magnitude M 6.3 is expected for the SMF. 

Key words: active tectonics, trenching, paleoseismology, Sudetic Marginal Fault 

The study area is situated in the north-eastern limit of 

the Bohemian Massif in the Czech Republic (central 

Europe). The studied Sudetic Marginal Fault (SMF) is 

a part of the Elbe Fault System, which comprises 

WNW-striking zone from SE North Sea to the front of 

Outer Carpathian nappes and disrupts the Variscan 

central Europe at the length of several hundreds of 

kilometers. The SMF is about 200km long and 

divides the Variscan Sudetic Mountains block from 

upland-like Fore-Sudetic block hosting Cenozoic 

cover. For a length of 130, km it controls the 

pronounced mountain front of the Sudetic Mountains 

(Fig. 1). 

Quaternary activity of the fault has been 

demonstrated e.g. by Middle and Upper Pleistocene 

fluvial terraces that are truncated by the SMF and 

expressed by 5 - 20 m high scarp in their longitudinal 

profiles in the Polish portion of the fault (e.g. 

Krzyszkowski et al., 1995). Local historic 

earthquakes recorded within the SMF had epicentral 

intensity estimated to reach only I 0 =4 -7 (MSK), 

which would correspond to macroseismic magnitude 

M M =3-4.9 (Guterch and Lewandowska-Marciniak, 

2002). They include the events from years 1594 and 

1778 (Zlotoryja/Legnica), and 1615 and 1786 near 

Bardo/Dzieroniów. Other historic earthquakes with 

Fig. 1. Digital elevation model of the morphologically distinctive Sudetic Marginal Fault (SMF) that controls the Sudetic Mountains front. 

Red stars - historic earthquakes, I 0 = epicentral intensity; white square – study area. 

239


EARTHQUAKE 

ARCHAEOLOGY 



more distant epicentres or those being felt in the 

zone parallel to the SMF (Fig. 1) include events in 

1496, 1562, 1786, 1823, 1877, and 1895 (Kárník et 

al., 1958; Olczak, 1962; Pagaczewski, 1972; Guterch 

and Lewandowska-Marciniak, 2002). 

As the historic earthquakes were not large enough to 

create the morphology (I 0 =4-7 MSK; cf. Guerrieri and 

Vittori, 2007), the study of potential presence of 

larger prehistoric earthquakes responsible for the 

growth of the mountain front was carried out. This 

paper presents results from investigation of the 

mountain front at the locality Bílá Voda municipality 

(Czech Republic). The results from three parallel 

trenches (C, D, F) performed across the SMF are 

discussed here. 

In order to determine the position of the fault for the 

following trenching, electric resistivity tomography 

with electrode spacing 2 m was performed (Fig. 2). 

The fault appeared to be expressed by the presence 

of a remarkable resistivity gradient, which divided 

high-resistivity of anticipated metamorphic rocks on 

the footwall and low-resistivity of Miocene deposits 

on the hanging-wall (Štpaníková et al., 2011). This 

was confirmed in the trenches C, D, F (Fig 2b). All 

the trenches revealed a subvertical fault zone striking 

135°-150° dividing Paleozoic crystalline rocks 

(phyllites, schists, granitic aplite) and late Quaternary 

colluvial deposits overlaying warped Miocene 

sediments. The SMF has been generally considered 

as a normal fault having experienced inversion during 

Late Cretaceous/Paleogene and extensional 

reactivation since Miocene during Alpine cycle (c.f. 

Badura et al., 2007). Nevertheless, our trenches 

revealed that the movements on the fault, at least the 

youngest ones, had very probably prevailing strikeslip 

character. The strike-slip on the main marginal 

fault with strike and dip of 135°/75°NE is suggested 

by a flower structure within the 4-m wide zone of 

tectonic breccia and fault gouge displayed both on 

the trench walls and on the floor, and by completely 

different lithologies on the both sides without any 

matching points. 

The down-thrown block is composed by Miocene 

(unit A; Fig 3), strongly kaolinised clayey silty sand 

with a small pebble admixture of local lithology. The 

clay originated by chemical weathering of feldspars 

and granite groundmass. These sediments may be 

correlated with Member C described few km to the 

SE by Štpaníková et al. (2010) and interpreted as 

to be deposited in a fluvio-limnic environment during 

the Carpatian and lower Badenian (i.e. early to mid 

Miocene, ~18–15 Ma). These deposits are 

upwarped, which was indicated also by preceding 

ERT profiles (see Fig 2), and covered by colluvial 

deposits (unit B). The colluvial deposits represent 

matrix-supported intermediate to sandy diamicts with 

gravel clast content between 5 and 25% and sandy 

to silty matrix. Largest clasts are up to 20–25 cm and 

are made of angular to subangular gneiss clasts, 

granodiorite and quartz clasts. The colluvial 

sediments have a sharp erosional base and their 

lowermost part is enriched by Mn oxides as a result 

of manganese precipitation from percolating water 

just above the impermeable Miocene clayey sand. 

Fig. 2. Model resistivity of ERT at the locality Bílá Voda, Wenner–Schlumberger electrode array with logged geology in 

Trench C. a) Resistivity in ohm m; vertical exaggeration=1; iteration 3; RMS error=2.3%. The SMF lies within the 

expressive resistivity gradient between Stations 143 and 145. b) Geology in the trench 1 – Late Quaternary colluvium; 2 

– Miocene lacustrine deposits, 3 – fault zone with tectonic breccia and fault gauge, 4 – micashists, 5 – granitic aplite, 6 

– gneisses, 7 – Late Holocene colluvium. 

240


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 3. Log of the trench F. See text for explanation of units A to E. Unit F – fault zone with tectonic breccia, G – metamorphic 

rocks. 

The colluvium close to the fault is obviously faultrelated 

and is interpreted to be derived from the fault 

scarp created after an earthquake (c.f. McCalpin, 

2009); it includes material from the fault zone and 

particularly clasts of the fault breccia. Although the 

colluvium does not display completely clear discrete 

sedimentological units that could be ascribed to 

individual events, there are some indications; in 

trench F (Fig 3), the part of colluvium that includes 

blackish matrix with breccia clasts have wedge shape 

(B1) and seems to be covered by another wedgeshaped 

colluvial unit (B2), which includes the fault 

rocks at lesser amount. This would indicate at least 

two faulting events, which resulted in the origin of the 

colluvial wedges. This fault-related colluvium is then 

covered by 35–40 cm thick beige to grey lens 

composed of beds of slightly clayey silt, sand and 

granule (unit C). They bear a very fine lamination, 

especially in silty layers. The upper part of the lens is 

dominated by bands of Mn and Fe oxides, which 

percolate fine to medium sand with some granule 

clasts. The sedimentary body is fault-parallel oriented 

and it was very probably a small pool emerging in 

front of the fault scarp. After its deposition the lens 

were tilted and displaced by minor faults (Fig 4), 

which displayed various sense of movement on the 

both walls of trench F. In trench C the deposit is only 

tilted. This must have been a result of another 

younger faulting (3 rd event). Unfortunately, no pollen 

Fig. 4. Tilted and displaced unit C - lens composed of beds 

of clayey silt, sand and granule; trench F. 

241 

or macro-remains were found within the silty 

sediment and only residual sediment carbon was 

dated by AMS 14 C dating. 

We have a set of 3 OSL dates from this colluvial unit 

ranging from 25.8 ± 1.6 ka to 9.5 ± 0.9 ka. We accept 

only the youngest one as reliable age, because the 

age corresponds to ~1 m long lens of sorted sand, 

which imply better zeroing before burial of this 

material. The older dates are probably influenced by 

residual signal. This is supported by 2 radiocarbon 

dates (all radiocarbon ages given here are calibrated 

ages using IntCal09 calibration dataset; Reimer et 

al., 2009) falling into the early Holocene – 7,997 ± 23 

a and 6,433 ± 29 a. Both ages originate from residual 

sediment carbon, as no plant macro-remains were 

found in this unit. The exact age may therefore be 

slightly older than the ages given above. 

The above-mentioned fault-derived colluvium is then 

cut by the youngest faults of the flower structure, 

which would indicate the following (4 th ) faulting event 

unless it was simultaneous with the deformation of 

the silty lens (3 rd event), which is ambiguous. Both 

the colluvium (incl. the silty lens) and the fault zone is 

sealed by banded layers due to superficial flow 

processes (unit D). They create individual bands 3 to 

10 cm thick, which continue >5 m downhill. The 

thickness of bands is much smaller, than it was 

described in other trenches studied along the Sudetic 

Marginal Fault at Vlice (Štpaníková et al., 2010), 

where up to 20 cm thick layers were interpreted to be 

deposited by a sheet gelifluction and ascribed to the 

Late Glacial, which was also supported by 

radiocarbon dating. So, the banded layers here in 

Bílá Voda were considered to be of similar origin and 

age, i.e. from Late Glacial, which would have 

important implication for inferring the age of the 

faulting (c.f. Štpaníková et al., 2009, 2011). 

However, new data on both OSL and radiocarbon 

age of the banded layers here give similar ages of 

2.6 ± 1.2 ka and 2,610 ± 108 a respectively. The 

thickness and dating of the layers here may show on 

the frost creeping of this material, which does not 

imply flow processes connected with seasonal 

thawing of the active permafrost layer, which was not 

present here during the Holocene. Frequent and


EARTHQUAKE 

ARCHAEOLOGY 



usually rapid diurnal frost creep processes may affect 

the superficial layer typically 5–10 cm thick even at 

the warm margin of the solifluction-affected 

environment (Matsuoka, 2001). Therefore, the 

presence of these processes at low altitude as here 

(~360 m) during the Holocene is feasible. 

In trench C on the footwall, these banded layers 

display around 35cm high step just coinciding with 

minor faults with opposite dip. These faults also do 

not show any displaced matching points, so 

horizontal movements with minor reverse component 

might be inferred here. On the hanging wall, these 

banded layers are slightly up-warped, which might 

have been related to a transpression responsible for 

the above–described step in the banded layers. This 

would imply another, potentially 5 th or 4 th event. 

The banded layers are covered by late Holocene 

colluvial deposits usually 0.5 to 1.5 m thick (unit E). 

These are matrix-supported, with the matrix 

composed predominantly of sandy fraction originating 

from the weathering of local Javorník granodiorite, 

which is mixed with silt and mostly pebble to fine 

cobble clasts. Gravel clast typically first cm large are 

mostly subangular to less subrounded and are made 

of Javorník granodiorite, quartz, (ortho)gneiss and 

schist, rather rare are clast of Nordics (originating 

from tills and glaciofluvial deposits left here by 

Pleistocene ice sheets; Nývlt et al., 2011) and pieces 

of bricks pointing to young age. Radiocarbon dates of 

charcoals from these young Holocene colluvial 

deposits range from 1,145 ± 89 a to 757 ± 30 a. On 

the top of these colluvia a cambisol type of recent soil 

is developed with 25–30 cm thick light grey-brown 

A-horizon with ploughed erosional base. 

To summarize, the results from the trenches at the 

Bílá Voda locality show potentially four to five 

movements on the SMF during late Quaternary 

(Holocene). Due to lack of kinematic indicators in the 

trenches perpendicular to prevailing movements, the 

sense of the strike-slip is still under investigation. 

Since the perpendicular trenches across strike-slips 

do not show the amount of horizontal slip, also sliprate 

could not be assessed with reasonable 

confidence. Nevertheless, based on the previous 

trenching (Štpaníková et al., 2010), according to 

empirical relationship ‘magnitude versus maximum 

vertical displacement’ (Wells and Coppersmith, 

1994), the minimum moment magnitude M 6.3 is 

expected on the SMF. 

Acknowledgements: The research was supported by the 

Grant Agency of the Czech Republic No. Czech Science 

Foundation No. 205/08/P521. The work has been 

elaborated within the Institute Research Plan of the Institute 

of Rock Structure and Mechanics, Academy of Sciences of 

the Czech Republic, No. AVOZ30460519. 

of the Bohemian Massif, Central Europe. Acta 

Geodynamica et Geomaterialia 4 (148), 7–29. 

Guerrieri, L., Vittori, E. (Eds). (2007). Intensity scale ESI 

2007. Mem. Descr. Carta Geologica dÍtalia, 74, Servizio 

Geologico dÍtalia – Dipartimento Difesa del Suolo, 

APAT, Rome, Italy, 41pp. 

Guterch, B., Lewandowska-Marciniak, H., (2002). Seismicity 

and seismic hazard in Poland. Folia Quaternaria 73, 85– 

99. 

Kárník, V., Michal, E., Molnár, A., (1958). Erdbebenkatalog 

der Tschechoslowakei bis zum Jahre 1956. Geofysikální 

Sborník 69, 411–598. 

Krzyszkowski, D., Migo, P., Sroka, W., (1995). Neotectonic 

Quaternary history of the Sudetic Marginal fault, SW 

Poland. Folia Quaternaria 66, 73–98. 

Matsuoka, N., (2001). Solifluction rates, processes and 

landforms: a global review. Earth-Science Reviews 55, 

107–134. 

McCalpin, J. (Ed.) (2009). Paleoseismology. Second 

Edition. Academic Press. 613 pp. 

Nývlt, D., Engel, Z., Tyráek, J., (2011). Pleistocene 

Glaciations of Czechia. In: Ehlers, J., Gibbard, P. L., 

Hughes, P. D., (Eds): Quaternary Glaciations – Extent 

and chronology, A closer look. Developments in 

Quaternary Science 15, 37–46, Elsevier. 

Olczak, T., (1962). Seismiczno Polski w okresie 1901– 

1950. Acta Geophysica Polonica 10, 3–11. 

Pagaczewski, J., (1972). Catalogue of Earthquakes in 

Poland in 1000–1970 years. Mat. I Prace Inst. Geofiz., 

vol. 51. 36 pp. 

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, 

J.W., Bertrand, C., Blackwell, P.G., Buck, C.E., Burr, G., 

Cutler, K.B. Damon, P.E., Edwards, R.L., Fairbanks, 

R.G., Friedrich, M., Guilderson, T.P., Hughen, K.A. 

Kromer, B. McCormac, F.G., Manning, S. Bronk Ramsey, 

C., Reimer, R.W., Remmele, S., Southon, J.R., Stuiver, 

M., Talamo, S., Taylor, F.W., van der Plicht, J., & 

Weyhenmeyer, C.E., (2009). Intcal09 and Marine09 

Radiocarbon Age Calibration Curves 0-50,000 Years Cal 

BP. Radiocarbon 51, 1111–1150. 

Štpaníková P., Dohnal J., Pánek T., ój M., Smolková V., 

Šilhán K., (2011). The application of electrical resistivity 

tomography and gravimetric survey as useful tools in an 

active tectonics study of the Sudetic Marginal Fault 

(Bohemian Massif, central Europe). Journal of Applied 

Geophysics 74, 69-80. 

Štpaníková, P., Hók, J., Nývlt, D., (2009). Trenching 

survey on the south-eastern section of the Sudetic 

Marginal Fault (NE Bohemian Massif, intraplate region of 

central Europe). In: Archaeoseismology and 

Palaeoseismology in the Alpine-Himalayan Collisional 

Zone (Pérez-López, R., Grützner, C., Lario, J., 

Reicherter, K., Silva, P.G. eds). Baelo Claudia, Spain, 

149–151. 

Štpaníková, P., Hók, J., Nývlt, D., Dohnal, J., Sýkorová, 

I., Stemberk, J., (2010). Active tectonics research using 

trenching technique on the south-eastern section of the 

Sudetic Marginal Fault (NE Bohemian Massif, central 

Europe). Tectonophysics 485, 269–282. 

Wells, D.L., Coppersmith, K.J., (1994). Empirical 


area, and surface displacement. Bulletin of the 

Seismological Society of America 82, 974–1002. 

References 

Badura, J., Zuchiewicz,W., Štpaníková, P., Przybylski, B., 

Kontny, B., Caco, S., (2007). The Sudetic Marginal 

Fault: a young morphotectonic feature at the NE margin 

242


EARTHQUAKE 

ARCHAEOLOGY 



TOWARD DEVELOPMENT OF A LONG RUPTURE HISTORY OF THE IMPERIAL FAULT IN 

MESQUITE BASIN, IMPERIAL VALLEY, SOUTHERN CALIFORNIA 

Tsang, Rebecca Y. (1, Thomas K. Rockwell (2), Aron J. Meltzner (3), Paula M. Figueiredo (4) 

(1) Dept. of Geol. Sci., San Diego State University, San Diego, California 92182, U.S.A. Email: tsang.y.rebecca@gmail.com 

(2) Dept. of Geol. Sci., San Diego State University, San Diego, California 92182, U.S.A. Email: trockwell@geology.sdsu.edu 

(3) Earth Observatory of Singapore, Nanyang Technological University, Singapore. Email: meltzner@ntu.edu.sg 

(4) IDL / FCUL - Geology Department, Lisbon University, Lisbon, Portugal. Email: pmfigueiredo@fc.ul.pt 

Abstract (Toward Development of a Long Rupture History of the Imperial Fault in Mesquite Basin, Imperial Valley, 

Southern California): We conducted a paleoseismic study on the northern Imperial fault at the Dogwood site in Mesquite Basin, 

southern California, to extend the record of late Holocene surface ruptures. New trench exposures have revealed evidence for up 

to 17 events in the past 1300 years, yielding an average recurrence interval of 80 years, and the large CV suggests the fault 

ruptures in a non-episodic manner. Our data indicate that the connection between lake and earthquake cycles is either very weak 

or non-existent as many of the ruptures occurred during dry periods between lakes. However, there is a strong correlation between 

the earthquake chronologies in the southern San Andreas fault and those on the northern Imperial fault, suggesting surface 

rupture at the southern portion of San Andreas fault may have triggered surface slip on the northern Imperial fault, or vice versa. 

Key words: clustering, Imperial, paleoseismology, recurrence 

INTRODUCTION 

The Imperial fault is a northwest-striking, dextral fault 

located in the Mesquite Basin, an area in which the 

seismicity patterns have been interpreted in terms of 

spreading and transform faulting (Sharp, 1982). The 

North American-Pacific plate boundary rate is 

approximately 45±1 mm/yr for southern California 

(DeMets et al., 1994). The Imperial, along with the 

San Jacinto and Elsinore faults, are considered as 

part of the San Andreas fault system and 

accommodate a significant proportion of the plateboundary 

motion (Hill et al, 1990). The 70-km fault 

crosses the U.S.-Mexico Border and terminates in 

the south at a right stepover to the Cerro Prieto fault, 

and in the north at the Brawley Seismic Zone which 

Fig. 1: Map 

showing the 

major structures 

and the 

Dogwood site in 

the Imperial 

Valley (modified 

from Thomas 

and Rockwell, 

1996). 

243


EARTHQUAKE 

ARCHAEOLOGY 



extends farther north to the Salton Sea (Figure 1). 

The fault was not recognized until the 1940 M L 7.1 

earthquake that was caused by an end-to-end 

rupture propagating mostly to the southeast 

(Buwalda and Richter, 1941; Richter, 1958). Thirtynine 

years later, the fault ruptured again to produce a 

M L 6.6 earthquake (Chavez et al, 1982). Besides 

surface ruptures, the area is also characterized by 

regular aseismic creep and high levels of 

microseismicity (Cohn et al., 1982; Lyons et al.,2002; 

Shearer, 2002). 

In earlier work, we resolved the timing and 

displacement for the six most recent surface 

ruptures, all of which have occurred in the past 500 

years. In this study, we conducted new paleoseismic 

studies on the northern Imperial fault to extend the 

record of late Holocene surface ruptures to better 

understand the behaviour of this plate boundary 

strike-slip fault. Twelve new trench exposures at the 

Dogwood site in Mesquite Basin, near El Centro, 

California have revealed evidence for up to 17 events 

in the past 1300 years. 

The Dogwood Site 

Our study area is located about 10 km north of 

Interstate Highway 8, next to Dogwood Road which 

makes it easily accessible (Figure 2). El Centro has 

an arid climate and is covered by farmland, but the 

lack of vegetation particularly at our site makes 

excavation easier. The Imperial fault at this site is 

expressed surficially as a single strand, which 

Fig. 2: Satellite image of the study area (yellow box) in 

the Imperial Valley, southern California. 

reduces the structural complexity that might mask 

evidence for past surface ruptures (Figure 2). Most 

importantly, the Dogwood site has preserved a 

remarkable stratigraphy of Lake Cahuilla which is an 

ephemeral freshwater body formed when the 

floodwaters of the Colorado River intermittently filled 

the Salton depression during the Holocene (Sharp 

1982a). This allows us to identify event horizons and 

to obtain ages for past surface ruptures by 

radiocarbon dating of detrital charcoal that was 

embedded within the stratigraphy. 

Fig. 3: Composite trench logs of T9B-SE (top) and T4Deep-SE (bottom) showing the remarkable Lake Cahuilla 

stratigraphy and the event horizons identified in these two exposures. 

244


EARTHQUAKE 

ARCHAEOLOGY 



Stratigraphy and Lake Chronology 

The stratigraphy consists of lacustrine, deltaic and 

fluvial sediments that were deposited during the filling 

and drying periods of Lake Cahuilla. Up to a total of 5 

m of stratigraphy below ground surface was exposed 

in our trenches which contains a record of 6 to 7 lake 

episodes. The lacustrine deposits consist of massive 

and bedded clay; the deltaic deposits are 

characterized by interbedded fine silt, silty clay and 

clay; and the fluvial deposits contain planar and 

cross-bedded silt and fine sand. In some exposures, 

the strata above the most recent lake deposition 

(post 1720 AD) comprises of localized flood deposits 

and spoil. A well-sorted sandy unit (Unit 390) of 1 to 

2 m in thickness was preserved in 10 trench 

exposures and was interpreted as a sand blow. Other 

sand blows also appear in five exposures of the 

lower section. The well-bedded stratigraphy at this 

site allows us to resolve the age of past earthquakes 

relative to the stratigraphy rather precisely (Figure 3). 

Calendar years of past earthquakes were determined 

by using the lake chronology that Meltzner and 

Rockwell had developed (unpublished data). They 

submitted over 180 detrital charcoal and peat 

samples from the Lake Cahuilla strata at different 

localities for 14 C analysis. About 50 14 C dates were 

used in the OxCal program to calculate calendar 

years of the different lake episodes. (For details of 

OxCal, please refer to Lienkaemper and Ramsey, 

2009). Identifying and locating an event horizon 

within the lake stratigraphy would then allow us to 

obtain calendar years of past surface ruptures. 

was deposited at the paleosurface rather horizontally, 

forming an angular unconformity with the strata 

below. 

A rating scheme of the paleoseismic indicators for 

each event horizon was devised to better qualify the 

likelihood of a past earthquake. Each indicator 

receives a rating of 1 to 4, with 1 being the weakest 

and 4 being the strongest. For example, an angular 

unconformity is a strong line of evidence hence it has 

a rating of 4. On the contrary, downward growth of 

displacement might be a result of fault dying out 

upward rather than cumulative displacement of 

multiple events on the same fault. Therefore, this 

indicator has the lowest rating of 1 due to the 

ambiguity in interpretation. The frequency (F) of 

these indicators is another factor to determine the 

likelihood of an event. In this method, the value of F 

is calculated by dividing the number of exposures 

where the indicator was observed by the number of 

exposures where the indicator should be observed. 

Finally, R is multiplied by F for the R*F value which 

indicates the likelihood of a paleoearthquake. The 

higher the R*F value, the more likely an event had 

occurred (Figure 5). Event 4 although receives a low 

R*F value, its occurrence was assured by a buried 

offset channel study by Meltzner and Rockwell 

(unpublished data). Among all 17 events, almost all 

events receive an R*F value higher than 1 except for 

Events 4, 8.5, 8.7 and 9. These events could be 

interpreted as unlikely or as events with small 

displacement. 

Evidence for Surface Ruptures 

Common paleoseismic indicators were used to 

identify event horizons for past surface ruptures, 

including upward fault termination, fissure fill, sand 

blow, angular unconformity, colluvial wedges and 

scarp-derived debris, liquefaction pipes and 

downward growth of displacement. For example, in 

Figure 4, several indicators and their geometrical 

relationships define the event horizon for Event 8. A 

fissure fill is bounded by two fault splays terminating 

upward at the same horizon, where a sand blow unit 

Fig. 5: Likelihood of an event based on quality and 

quantity of paleoseismic indicators as seen in the plot of 

average RF values of each event. Higher values 

indicate more likelihood. 

Fig. 4: Paleoseismic indicators used to identify the 

horizon of Event 8 include upward fault termination 

(UT), angular unconformity (AU) and fissure fill (FF). 

245 

CONCLUSION 

We conducted this paleoseismic study on the 

northern Imperial fault at the Dogwood site in 

Mesquite Basin, southern California, to extend the 

record of late Holocene surface ruptures to better 

understand the behavior of this plate boundary strikeslip 

fault. Event horizons in 12 trench exposures were 

identified within the Lake Cahuilla stratigraphy using 

the standard paleoseismic indicators such as angular 

unconformity, fissure fill and upward truncation of 

faults. By using the lake chronology model that 

Meltzner and Rockwell (unpublished data) developed 

for Lake Cahuilla in the Imperial Valley, we


EARTHQUAKE 

ARCHAEOLOGY 



developed an earthquake chronology of the northern 

Imperial fault. Our study has revealed evidence for 

up to 17 events in the past 1300 years, yielding an 

average recurrence interval of about 80 years, albeit 

with a large standard deviation and coefficient of 

variation (Figure 6). We tested the hypothesis of a 

connection between lake and earthquake cycles, 

however, our data suggest that the connection is 

either very weak or non-existent as the majority of 

the ruptures occurred during dry periods between 

lakes. On the other hand, we do find a strong 

correlation between the earthquake chronologies 

determined for the southern San Andreas fault and 

the rupture history on the northern Imperial fault, 

suggesting surface rupture at the southernmost 

portion of San Andreas fault may have triggered 

earthquakes on the northern Imperial fault (Figure 7). 

Fig. 6: Earthquake calendar ages obtained from OxCal. 

The higher the probability density, the more wellconstrained 

the ages are. The height of a peak does not 

give any information about the likelihood of a 

paleoearthquake. 

Acknowledgements: We are grateful for the help we 

received to make this project possible. We thank the 

landowner for permission to excavate on his property and 

Jim Little for the backhoe operation. Field assistants 

including Gayatri Marliyani, Nissa Morton, Barrett Salisbury, 

Mike Buga, Petra Štpaníková, Eulàlia Masana and Katie 

Anderson offered much help in the exposure preparation. 

This project was funded by the National Earthquake 

Hazards Reduction Program (NEHRP). 

References 

Buwalda, J. P., & C. F. Richter (1941). Imperial Valley 

earthquake of May 18, 1940. Geological Society of 

America Bulletin, 52, 1942-1943. 

Chavez, D., J. Gonzales, A. Reyes, M. Medina, C. Duarte, 

J. N. Brune, F. L. Verson, III, R. Simons, L. K. Hutton, P. 

T. German, and C. E. Johnson. (1982). Mainshock 

location and magnitude determination using combined 

U.S. and Mexican location and magnitude determination 

using combined U.S. and Mexican data. The Imperial 

Valley, California, Earthquake, October 15, 1979. United 

States Geological Survey Professional Paper, 1254, 51- 

54. 

Cohn, S. N., C. R. Allen, R. Gilman, & N. R. Goulty. (1982). 

Preearthquake and postearthquake creep on the Imperial 

fault and the Brawley fault zone. United States Geological 

Survey Professional Paper, 161-167. 

Fig. 7: Earthquake chronologies from paleoseismic 

sites on the southern San Andreas fault and the 

northern Imperial fault. The possible earthquakes 

Coa-3 and Coa-6, and the less likely events E8.5, 

E8.7 and E9 are shown in gray text. Correlations of 

events at the three paleoseismic sites are highlighted 

in pink. 

DeMets, C., R. G. Gordon, D. F. Argus, and S. Stein. 

(1994). Effects of recent revisions to the geomagnetic 

reversal time scale on estimates of current plate motions. 

Geophysical Research Letters, 21, 2191-2194. 

Hill, D. P., J. P. Eaton, and L. M. Jones (1990). Seismicity: 

1980-1986, in Wallace, R.E., ed. The San Andreas Fault 

System, U.S. Geol. Surv. Profess. Pap. 1515, 114-151. 

Lienkaemper, J. J., and C. B. Ramsey (2009). OxCal: 

Versatile Tool for Developing Paleoearthquake 

Chronologies. Seismol. Res. Lett., 80, 431-434. 

Lyons, S. N., Y. Bock, and & D. T. Sandwell (2002). Creep 

along the Imperial Fault, southern California, from GPS 

measurements. J. of Geophysical Research, 107 (B10). 

Richter, C. F. (1958). Elementary seismology: San 

Francisco. W. H. Freeman, 768. 

Sharp, R. V. (1982). Tectonic Setting of the Imperial Valley 

Region. United States Geological Survey Professional 

Paper, 5-14. 

Shearer, P. M. (2002). Parallel fault strands at 9-km depth 

resolved on the Imperial Fault, Southern California. 

Geophysical Research Letters, 29 (14). 

Thomas, A. P., and T. K. Rockwell (1996). A 300- to 550- 

year history of sip on the Imperial fault near the U.S.- 

Mexico border: Missing slip at the Imperial fault 

bottleneck. Journal of Geophysical Research, 101, 

5987-5997. 

246


EARTHQUAKE 

ARCHAEOLOGY 



MAPPING PALEO SHORELINES IN LESVOS ISLAND: NEW CONTRIBUTION TO THE 

LATE QUATERNARY RELATIVE SEA LEVEL CHANGES AND TO THE NEOTECTONICS 

OF THE AREA 

Vacchi, Matteo (1,2), Alessio Rovere (1,2), Nickolas Zouros (3), and Marco Firpo (1) 

(1) Department for the Study of the Territory and its Resources, University of Genova, Genova. (Italy) 

(2) SEAMap srl, Seascape Evaluation Assessment and Mapping, Genova (Italy) 

(3) Department of Geography, University of Aegean, Mytylene, Lesvos Island (Greece) 

Abstract (Mapping paleo-shorelines in Lesvos Island: new contribution to the late quaternary relative sea level changes 

and to the neotectonics of the area): This study attempts to delineate the late Quaternary morphotectonic evolution of the 

coastal zone of southern Lesvos Island (NE Aegean Sea) on the basis of detailed field mapping of different sea level indicators. 

The geomorphological analysis of the markers, coupled with radiometric dating, allowed to identify a Late Quaternary regional 

uplift trend controlled by the footwall of a large normal fault (Lesvos Fault) located between Lesvos and Chios Islands. At local 

scale, superimposed to this trend, the area is characterized by many indicators of rapid vertical displacements (uplift and 

subsidence) related to the high seismicity. The most important co-seismic event has been found to have happened between 3365 

and 3924 BP which uplifted of about 0.8 m a large sector of the southern Lesvos coastline. 

Key words: raised shorelines; coastal uplift, Lesvos Island; Relative sea level changes. 

INTRODUCTION 

Due to the geodynamic setting, several sectors of the 

Mediterranean basin exhibit evidence of differential 

vertical movements during the late quaternary 

(Stewart & Morhange, 2009). In particular, several 

papers investigated the coastal morphotectonics of 

the Aegean region, one of the most seismically active 

areas not only at Mediterranean scale (Stiros et al., 

2000; 2009 Palyvos et al., 2008; Cundy et al., 2010). 

In many cases geomorphological approaches 

became very practical solutions to get quantitative 

information on late quaternary uplift and 

paleoseismicity of coastal areas (Palyvos et al. 

2008). For this reason, detailed mapping of paleosea 

level markers has been often used as a tool to 

quantify coastal uplift and relative sea level changes 

in several areas of the Aegean Sea (Pirazzoli et al., 

2004; Stiros et al., 2009). In this study we analyzed 

the coastal geomorphology of the southern sector of 

Lesvos Island, located in the NE Aegean Sea. Here, 

despite several papers already dealt with the 

neotectonics of this area, detailed information on the 

coastal morphotectonics of the Island is lacking. 

Morphological, biological and sedimentary records of 

past sea level were mapped, analyzed and, where 

possible, sampled and dated. 

(Roumelioti et al., 2011).The study area is located in 

the south eastern part of the Island, in the coastal 

area comprised between Tarti and the entrance of 

the Kalloni gulf (Fig.1). From the geomorphological 

point of view, it is mainly characterized by steep cliffs 

alternated with gravel pocket beaches, often with 

beachrocks outcrops. The outcropping lithologies are 

composed by a basement of Alpidic and pre-Alpidic 

rocks, mainly Triassic limestones, marbles and 

schists. In the whole sector the tidal range does not 

exceed 0.2 m (Vousdoukas et al., 2009). 

Marine and coastal processes produce many kinds of 

geomorphologic markers, especially on rocky shores, 

that are related to the contemporary sea-level 

position (Pirazzoli, 2007). Among them tidal notches, 

and the inner edge of benches and shore platform 

are considered precise indicator of paleo sea level 

stands in microtidal environment such as the 

Mediterranean Sea. Endolithic organisms often leave 

morphological signs that can be used as sea-level 

indicator and, at Mediterranean scale, Lithophaga 

boreholes horizontal upper limit represent a good 

biological indicator of former m.s.l. (Laborel and 

Laborel-Deguen, 1994). 

Lesvos islandis located in a geotectonically complex 

area, because directly affected by the North 

Anatolian Fault Zone (NAFZ), its westward 

continuation in the Aegean Sea, known as the North 

Aegean Trough (NAT) and the West Anatolia Graben 

System (WAGS) in Asia Minor with significant 

historical seismicity (Papazachos and Papazachou, 

1997; Papazachos and Kiratzi, 1996). As a result of 

the interaction between those tectonic systems, 

Lesvos Island presents a strong diversity in fault 

setting and is presently characterized by the parallel 

activity of both normal and strike slip faults 

247


EARTHQUAKE 

ARCHAEOLOGY 



Along the platform fragments of marine shells 

(vermetids, serpulids, gastropods) were found still 

preserved in growth position at about .0.5 m a.sl. 

AMS radiocarbon dating were performed on 3 

samples of marine shells (gastropods and serpulids) 

sampled at about 4 km of distance one from the 

other. Calibrated ages dated the samples 

respectively 3365 – 3648 BP (K1 beta); 3460 – 3812 

BP (AF1 beta); 3609 – 3924 BP (AF3 poznan). 

Fig.1. Geographical location and active faults of the study 

area. Focal mechanism represent the > 4M earthquakes 

after the 1928. 

A second level of uplifted wave cut platform was 

mapped along this sector at an elevation ranging 

from 5 to 6 m. This level is recurrent in 4 sites along 

the coastline and is always covered by beach 

deposits rich in marine fauna. Two samples of marine 

shells (Patella ferruginea) were sampled. AMS 

radiometric dating gave calibrated ages of 29160 ± 

150 (V1 beta) and 27410 ± 120 (Ag2 beta). 

A detailed mapping of paleo-shorelines including 

erosional, biological and depositional sea level 

markers was carried out in the island during three 

field campaigns between September 2009 and 

October 2010. Although debates in literature 

(Kellettat, 2006; Desruelles et al., 2009), in this study 

we considered beachrocks as markers of past 

relative sea level mainly because of their strict 

altimetric correlation with other sea level markers 

such as benches, tidal notches or wave cut 

platforms. In fact, the combined analysis of erosive 

and depositional paleo-shoreline indicators is of 

timely importance because, although marine 

erosional forms are more precise indicators of sea 

level in coastal settings, they rarely preserve 

dateable materials, which are more frequent in 

depositional landforms (Pirazzoli, 2005; Pirazzoli, 

2007). 

Markers elevations were measured using a 3 m 

metal bar with centimeter sub-division and in-built 

spirit level in order to achieve better vertical and 

horizontal accuracy. Elevations were measured with 

reference to SL at the time of measurement with a 

maximum error of 10 cm and the markers location 

was measured using a handheld GPS (± 5m). 

Underwater transects were carried out up to – 10 m 

to map and sample the submerged markers (mainly 

notches and beachrocks). 

RESULTS 

The most evident morphological marker was 

represented by a uplifted wave cut platform 

developing for about 6 km eastwards and westward 

to Agios Fokas cape (Fig. 2). 

The platform is continuous (despite the several 

changes of lithologies occurring in the area) and 

presented an inner margin often characterized by an 

abrasion notch with the roof positioned at about + 0.9 

m a.s.l. This level was confirmed by the presence of 

isolated limestones blocks lying on the platform 

showing the presence of lithophaga boreholes up to 

0.8 m above the present sea level. 

Fig. 2. Uplifted wave cut platforms in the sector of Agios 

Fokas and Vrisa. They often showed an inner margin often 

characterized by an abrasion notch with the roof at about + 

0,9 m a.s.l. as indicated by the dashed line in the photo 

above. 

On the vertical limestone cliffs, a modern tidal notch 

was always present (width 50- 70 cm). Recurrent 

uplifted relict of tidal notches were also observed at 

different elevation varying from + 0.7 up to +12 m 

a.s.l. (Fig.3c) locally interrupted where lithological 

conditions are unfavorable for their formation and 

preservation (i.e., schists). 

248


EARTHQUAKE 

ARCHAEOLOGY 



Their lateral continuity, until the village of Melinta 

was however confirmed by examination by boat. 

Morphology and elevation of the notches showed 

discrepancies and altimetric correlations has been 

very difficult. However, a recurrent morphology was 

observed on tidal notches positioned at + 9 to +11 m. 

They always presented a double concavity that 

appeared smoothed in exposed areas and well 

preserved in the sheltered ones (Fig. 3b). In front of 

the village of Melinta a large limestone rock 

presented an uplifted wave cut platform 

characterized by an inner margin at +11.2 m clearly 

indicated by the upper limit of a well preserved 

lithophaga boreholes band (Fig.3a). 

The area of Tarti is characterized by almost vertical 

cliffs carved in Triassic limestones. Evidence of 

Lithophaga boreholes bands were observed on both 

sides of the Tarti cape. The upper limit is located at + 

1.8 m. Any evidence of uplifted erosional markers or 

modern tidal notches was observed despite the 

favorable lithology of the area. On the contrary, a 

submerged tidal notch was mapped all along the 

cape. The maximum concavity was measured at -0,7 

m below the modern sea level (Fig.4). 

DISCUSSION 

Raised sea stands have been identified along the 

whole south eastern coast. Altimetric distribution of 

the different markers pointed out some recurrent 

levels especially in the areas comprised between 

Agios Fokas cape and the village of Melinta. 

The first is positioned at + 0.8 ± 0.2 m. According to 

Pirazzoli, (2007) and Stiros et al., (2009), the good 

preservation of both morphological and biological 

markers suggests a very rapid uplift probably related 

to a co-seismic event. 

The second paleo-shoreline is positioned at + 4.80 ± 

0.2 m and, according with the dating, was referred to 

MIS 3. The highest recurrent sea level stand is 

positioned at 10.9 ± 0.2 m. No radiometric dating was 

possible because the lack of datable material. 

The geographical limit of these recurrent levels has 

been positioned in the village of Plomari. In fact, 

eastward to Plomari, although indicators of raised 

shorelines were still present, however their altimetric 

distribution was significantly different with respect to 

the Agios Fokas and Melinta sectors. The site of Tarti 

is the only one that showed both evidence of uplift 

and subsidence. 

The detailed analysis of morphologies and 

distribution of the markers, coupled with radiometric 

dating, allowed to identify a Late Quaternary regional 

uplift trend controlled by the footwall of a large 

normal fault (Lesvos Fault) located between Lesvos 

and Chios Islands (Mascle & Martin, 1990) (Fig.1). 

Fig.3. Raised shorelines mapped in the area of Kryfty and 

Melinta. a) wave cut platform with inner limit positioned at + 

11.2 and characterized by lithofaga boreholes. b) double 

concavity notch occurring in Kryfty cove. c) raised tidal 

notches carved in the Triassic limestones of Kryfty area. 

The maximum concavity is at + 0.8 m. 

Uplifted beachrocks outcrops were observed along 

the whole SE sector of the island. The beachrocks 

are often organized in multiple slabs reaching up to 

about + 3.1 m a.s.l. The only exception is 

represented by Tarti site where beachrocks 

developed from – 1.1 m up to intertidal zone. 

249 

Fig.4 . Underwater tidal notch occurring all along the Tarti 

cape. 

At local scale, superimposed to this trend, the area is 

characterized by many indicators of rapid vertical 

displacements (uplift and subsidence) related to the 

high seismicity. According to the radiomentric dating 

and geomorphological evidence,an important coseismic 

event has been found to have happened 

between 3365 and 3924 BP. This rapid event uplifted 

of about 0.8 m a large sector of the southern Lesvos 

coastline. 

According to Lambeck (1996) and Siddal et al., 

(2008), the position of the MIS 3 shorelines at about 

+ 5 m on the present position allowed to quantify the


EARTHQUAKE 

ARCHAEOLOGY 



rate of uplift during the late quaternary. However, as 

pointed out by several indicators, the average rate is 

a sum of gradual and co-seismic uplifts and more 

accurate consideration strongly needs further 

radiometric dating. If further studies will confirm our 

data, Lesvos Fault is among one of the less 

documented active fault in this sector of the Aegean 

Sea to have ruptured in recurring earthquake within a 

relatively recent timespan. 

In conclusion, geomorphological markers of past sea 

levels are well known for being one of the main 

means to estimate neotectonic trends and historical 

recurrence and magnitude of co-seismic events. The 

scientific value of this kind of studies is flanked by the 

value assumed for society. The latter is twofold: on 

one side, very active sites like Lesvos Island are a 

potential source of data to tune global sea level 

models and therefore global estimates of future sea 

level rise; on the other side, they have a local value 

in terms of earthquake risk assessment. 

References 

Cundy, A.B., Gaki-Papanastassiou, K., Papanastassiou, D. 

Maroukian, H., Frogley, M.R., Cane, T.(2010) Geological 

and geomorphological evidence of recent coastal uplift 

along a major Hellenic normal fault system (the Kamena 

Vourla fault zone, NW Evoikos Gulf, Greece) Marine 

Geology 271, 156–164 

Desruelles, S., Fouache, É., Ciner, A. Dalongeville, R., 

Pavlopoulos, K., Kosun, E., Coquinot, Y., Potdevin, J.L. 

(2009). Beachrocks and sea level changes since Middle 

Holocene: Comparison between the insular group of 

Mykonos–Delos–Rhenia (Cyclades, Greece) and the 

southern coast of Turkey. Global and Planetary Change 

Volume 66, 1-2, 19-33. 

Kellettat, D. 2006. Beachrock as Sea-Level Indicator? 

Remarks from a Geomorphological Point of View. Journal 

of Coastal Research 22 (6), 1558-1564. 

Mascle, J. ,Martin, L. (1990). Shallow structure and recent 

evolution of the Aegean Sea: a synthesis based on 

continuous reflection profiles. Marine Geology 94, 271– 

299. 

Lambeck, K. (1996) – Sea-level changes and shore-line 

evolution in Aegean Greece since Upper Paleolithic time. 

Antiquity, 70, 588-610. 

Palyvos, N., Lemeille, F., Sorel, D., Pantosti, D. 

Pavlopoulos, K. (2008). Geomorphic and biological 

indicators of paleoseismicity and Holocene uplift rate at a 

coastal normal fault footwall (western Corinth Gulf, 

Greece). Geomorphology 96, 16–38 

Papazachos, C.B. and Kiratzi, A.A., (1996). A detailed study 

of the active crustal deformation in the Aegean and 

surrounding area, Tectonophysics 253(1-2), 129-153. 

Papazachos, B., Papazachou, C., 1997. The earthquakes of 

Greece, Ziti editions. Thessaloniki 

Pirazzoli, P.A. (2005). A review of possible eustatic,isostatic 

and tectonic contributions in eight late-Holocene relative 

sea-level histories from the Mediterranean area. 

Quaternary Science Review 24, 1989–2001 

Pirazzoli, P.A., Stiros S.C., Fontugne M., Arnold M. (2004). 

Holocene and Quaternary uplift in the central part of the 

southern coast of the Corinth Gulf (Greece). Marine 

Geology 212, 35-44 

Pirazzoli, (2007). Sea level studies. Geomorphological 

Indicators. Encyclopedia of Quaternary Science, 2974- 

2983. 

Roumelioti, Z., Kiratzi, A., Benetatos, C., (2011). Time 

Domain Moment Tensors of earthquakes in the broader 

Aegean Sea for the years 2006-2007: the database of the 

Aristotle University of Thessaloniki, Journal of 

Geodynamics 51, 179-189. 

Siddal, M., Rohling, E.J., Thompson, W.G., Waelbroeck, C. 

(2008). Marine isotope stage 3 sea level fluctuation: data 

synthesis and new outlook. Reviews of Geophysics 46, 1- 

29. 

Stewart, I.S. Morhange, C. (2009), Coastal geomorphology 

and sea-level change, in J. C.Woodward (ed.), The 

Physical Geography of the Mediterranean. Oxford 

University Press, Oxford, 385–413 

Stiros, S.C., Laborel, J., Laborel-Deguen, F., Papageorgiou 

S., Evind J., Pirazzoli P.A.. 2000. Seismic coastal uplift in 

a region of subsidence: Holocene raised shorelines of 

Samos Island, Aegean Sea, Greece. Marine Geology 

170, 41-58 

Stiros, S.C. Pirazzoli, P.A. Fontugne, M., (2009). New 

evidence of Holocene coastal uplift in the Strophades 

Islets (W Hellenic Arc, Greece). Marine Geology 267, 

207–211. 

Vousdoukas, M.I., Velegrakis, A.F., Karambas, T.V. (2009). 

Morphology and sedimentology of a microtidal beach with 

beachrocks: Vatera, Lesbos, NE Mediterranean. 

Continental Shelf Research 29, 1937–1947. 

250


EARTHQUAKE 

ARCHAEOLOGY 



BOULDER DEPOSITS IN SOUTHERN LESVOS: AN EVIDENCE OF THE 1949’S CHIOS- 

KARABURUM TSUNAMI? 

Vacchi, Matteo (1), Alessio Rovere(1), Marco Firpo (1) and Nickolas Zouros (2) 

(1) Department for the Study of the Territory and its Resources, University of Genova, Genova. (Italy) 

(2) SEAMap srl, Seascape Evaluation Assessment and Mapping, Genova (Italy) 

(3) Department of Geography, University of Aegean, Mytylene, Lesvos Island (Greece) 

Abstract (Boulder deposits in southern Lesvos: an evidence of the 1949’s Chios-Karaburum tsunami?): Aim of this study 

was to understand the mechanism of deposition of clusters of large boulders, consisting of beachrock slabs, which were found on 

the southern coasts of Lesvos Island (NE Aegean Sea). An integrated study was carried out in order to detect whether transport 

and accumulation can be related to an exceptional storm or a tsunami event. The results provided evidence of a tsunami wave 

impact during the last century. Specifically, the boulder accumulation is likely to have been deposited by the tsunami that was 

triggered by the 6.7M Chios-Karaburum earthquake of 1949. Lesvos island, on the basis of the seismic sources, was already listed 

as tsunami affected area but field evidence was previously lacking. The results represent a new contribution that could be of 

primary importance in assessing the degree of vulnerability of the local coastal communities of southern Lesvos to future tsunami 

events. 

Key words: boulder deposits; catastrophic waves; Lesvos Island; coastal hazard. 

INTRODUCTION 

Identification of boulders displaced and eventually 

transported by tsunami or exceptional storm waves 

plays a crucial role in the assessment of the 

occurrence of past catastrophic events (Goto et al., 

2009). Understanding the nature and impact of these 

events is fundamental in terms of evaluation, 

mitigation and management of current coastal 

hazards. In the Mediterranean, historical tsunamis 

were often reported as consequences of destructive 

earthquakes (Soloviev et al., 2000). In the Ionian and 

Aegean Seas, geomorphological records of tsunamis 

were detected in several areas (Mastronuzzi and 

Sansò., 2000; Scheffers and Kellettat, 2003; 

Scheffers and Scheffers, 2007, Scicchitano et al., 

2007; Scheffers et al., 2008). In particular, the area 

surrounding western Turkey and Greece is among 

the most seismically active and rapidly deforming 

regions in the world (Nyst and Tatcher, 2004), and 

thus has historically been strongly affected by 

tsunami events (Papadopoulos & Chalkis, 1894; 

Soloviev, 1990; Soloviev et al., 2000,Fig. 1). 

The aim of this study is to understand the mechanism 

of deposition of clusters of large boulders (weighing 

up to 17 tons) which were found on the southern 

coasts of Lesvos Island (NE Aegean Sea). In 

particular, studies of boulder morphology, tectonic 

setting, wave climate and historical context were 

carried out in order to detect whether transport and 

accumulation can be related to an exceptional storm 

or a tsunami event. 

The study area (Fig. 1) is located in the southern part 

of the island, near the villages of Plomari and Agios 

Isidoros. From a geomorphological point of view, this 

area is characterized by cliffs alternating with sandy 

251 

to gravel beaches, often with beachrock outcrops. 

The main lithologies are represented by Triassic 

schists and limestones. Due to the southern facing, 

this costal sector is characterized by a relatively low 

wave regime because of the protection from the main 

swells of the Aegean sea generated by northern 

winds (Soukissian et al., 2007). The maximum fetch 

does not exceed 100 nautical miles, and main waves 

(coming from SE) reach maximum off-shore wave 

height of about 1,8 m (Vousdoukas et al., 2009). 

Fig. 1. Geographical location and seismo-tectonic setting of 

the study area. Focal mechanism of the > 4M earthquakes 

after the 1928. The white circle indicates the 1949’s M. 6.7 

event triggering a tsunami wave recorded in Chios and in 

the Karaburun peninsula. Black dots in the small square 

represent the historical tsunamigenic earthquakes 

according to Soloviev, 1990. 

A total of 47 boulders was found in the study area. 

Some of them are organized in clusters, others are 

scattered along the shoreline (Fig. 2). After a first 

general mapping, more detailed measures


EARTHQUAKE 

ARCHAEOLOGY 



concentrated on the 26 boulders showing a major 

axis 1 m. The boulders consisted of beachrock 

slabs (Fig. 2) of which having unit weight of 2.6 tm -3 . 

The boulders had a maximum size of 4.5 × 2.5 × 0.6 

m with a volume of about 7 m 3 and a maximum 

weight of about 17 t. Some peculiar features are 

constant on the majority of the beachrock boulders: 

i) the boulders are scattered up to about 20 m from 

the shoreline and their elevation does not exceed 0.5 

m asl; 

ii) many boulders are upside down, as proved by 

erosive features normally cut upon the upper surface 

of a beachrock and presently facing the ground (Fig. 

2); 

iii) Several boulders, especially in the area of Agios 

Isidoros, are completely buried in the sand (Fig. 2). 

The presence of biogenic encrustations (mainly 

vermetids and serpulids) suggests a mid-sublittoral 

pre-transport position of the boulders (as these 

organisms usually live within this environment, 

Laborel, 1987). 

These settings were validated by the transects 

carried out along the whole seaward extension of the 

beachrock outcrop: the high degree of fracturing 

coupled with the presence of scattered broken pieces 

mainly at the seaward edge of the beachrock outcrop 

(about -3 m) confirmed the hypothesis of an original 

submerged position or a submerged joint bounded 

scenario. In fact, geometrical analysis of the boulders 

and of large holes in the submerged part of the 

beachrock itself revealed a close correspondence 

between the shapes of the boulders and the shapes 

of the holes. On the basis of the pre-transport 

settings individuated, values of storm wave height 

(Hs) and tsunami wave height (Ht) theoretically 

required for the boulder displacement were 

calculated using the equations proposed by Nott, 

(2003), Pignatelli et al., (2009), Benner et al., (2010). 

Fig. 2 a) and b) examples of large boulders mapped in the 

site of Plomari; c) and d) examples of large boulders 

mapped in the site of Agios Isidoros 

DISCUSSION 

Significant discrepancies were observed between the 

different hydrodynamic approaches. For the 

submerged boulders (SMBS), the Nott equation 

calculated storm wave values (Hs) up to 15 m to 

252 

displace the largest boulder. The maximum storm 

wave values (Hs) computed using the Benner et al. 

approach are considerably smaller, slightly 

exceeding 8 m. 

Nott’s joint bounded equation (JBBS) computed 

storm waves values (Hs) exceeding 20 m whereas 

storm wave values derived from Pignatelli et al. 

equations reached maximum values of about 10 m. 

The Aegean Sea is characterized by relatively short 

fetch and relatively small swells, mainly generated 

from northern winds (Soukissian et al., 2007). It is 

very unlikely that waves exceeding 10 m could be 

generated in this sector of the Aegean, especially in 

a south-facing area, characterized by a maximum 

fetch not exceeding 100 nautical miles and by 

maximum off-shore significant wave height (Hs) not 

exceeding 2 m. 

These hydrodynamic results suggest that a tsunami 

could have been responsible for on-shore deposition 

of the boulders, but the discrepancies between the 

different approaches did not provide unambiguous 

evidence. Further analyses were then performed in 

order to test the reliability the tsunami hypothesis. 

Some considerations based both on 

geomorphological indicators and on seismotectonic 

sources, support the hypothesis that the boulder 

accumulation in southern Lesvos was caused by 

tsunami. 

The morphology and lithostructural settings of the 

coastal area play a crucial role in boulder detachment 

and transport by catastrophic waves. Mastronuzzi et 

al., (2006) indicated the presence of layered units 

and bedrock fracturing as important pre-conditions 

for boulder displacement. Coastal sectors 

characterized by beachrock outcrops and affected by 

tsunami wave impact often showed on-shore broken 

slab accumulation (Vött et al., 2007, 2009; Scheffers 

and Scheffers, 2007). In the study area, the slabs 

were probably torn out of the original beachrock unit 

which was already fractured by several seismic 

events affecting the whole coastal sector. Moreover, 

on the majority of boulders, a fragile layer of biogenic 

encrustation was observed. Its preservation is a clear 

indicator of short transport generated by a single 

wave (Mastronuzzi et al., 2006). 

Other evidence can be gathered by subdividing the 

a-axis orientation of the boulders on the basis of the 

relative Hs values. Most of the boulders presenting 

Hs values exceeding 7,5 m (i.e. not compatible with 

calculated extreme storm events) had their elongated 

axis almost perpendicular to the shoreline and 

oriented between 170°N and 200°N. However, other 

boulders are oriented mainly between 130° and 

150°N. This is the direction of the major swells in the 

area (Vousdoukas et al., 2009). This orientation 

pattern reflects a scheme already present in other 

Mediterranean boulder accumulations (Mastronuzzi 

and Sansò, 2004; Scicchitano et al., 2007) and was 

explained with re-orientation by storm waves after a 

tsunami event. According to the previous


EARTHQUAKE 

ARCHAEOLOGY 



considerations, we hypothesize that a single tsunami 

wave displaced all the boulders from the submerged 

position to the shore. Subsequent storm wave 

events, coming from S-SE, reworking the deposited 

boulders were able to re-orient only the smaller 

ones.Further confirmation was provided by the 

literature dealing with Mediterranean regions most 

frequently affected by tsunamis (Papadopoulos and 

Chalkis, 1984; Altinok and Ersoy, 2000; Soloviev et 

al., 2000; Papadopoulos and Foakefs, 2005; Yolsal 

et al., 2007; Papadopoulos, 2009). These papers 

provided spatial distribution of tsunami hazard in the 

Aegean area based on the historical occurrence of 

tsunami events as well as on the triggering seismic 

sources (Papadopoulos and Chalkis, 1984; 

Papazachos and Dimitriou, 1991; Papadopoulos and 

Foakefs, 2005). On the basis of the distribution of 

tsunamigenic earthquakes and of the historical data 

of past events, several authors listed Southern 

Lesvos as an area vulnerable to tsunami events 

(Fig.3) (Papadopoulos and Chalkis, 1984; Soloviev et 

al., 1990 Papadopoulos and Foakefs, 2005; Yolsal et 

al., 2007). This integrated study, ranging from 

geomorphological to seismotectonic data, 

corroborated the tsunami hypothesis and most 

probably ruled out extreme storms as the cause of 

transport and displacement of the boulders. 

Fig. 3 Maps of the tsunami affected area compiled by 

Papadopoulos & Foakefs, 2005; Papadopoulos & Chalkis, 

1984; Yolsal et al., 2007. 

The high seismic activity affecting this sector of the 

Aegean Sea often triggered tsunami waves that 

caused damage to the surrounding coastal areas 

(Papazachos and Papazachou, 1997; Altinok et al., 

2005). Historical data are available on six main 

events which took place 20 th March 1389, 24 th 

November 1772, 13 th November 1856, 19 th January 

1866, 3 rd April 1881 and 23 rd July 1949 (Soloviev et 

al., 2000; Altinok et al., 2005). 

Radiocarbon age determinations were performed on 

two samples of biological encrustation, the first, 

(serpulids) from the Plomari cluster and the second 

(vermetids) from Agios Isidoros cluster. AMS 

radiocarbon age determination indicated that both 

were of recent origin (fraction of modern carbon, 

~1950). Because of this, it was possible to carry out 

an historical investigation with the help of the cultural 

association “to polion” of Plomari. Aged inhabitants of 

the village confirmed the sudden appearance of the 

boulders on the shoreline but the people who were 

interviewed could not recall precise dates. Important 

data were achieved by using an historical photograph 

of Plomari taken in 1896 (Fig. 4). In the same area 

where boulders are presently found the photograph 

shows no traces dislocated blocks. 

This photograph allows us to place a further temporal 

limit on the date of the boulders deposition. In 1896 

there is no evidence of broken beachrock slabs on 

the shoreline. This is consistent with the theory of a 

single pulse wave as the depositional mechanism as 

opposed to a continuous action of the waves on the 

beachrock outcrops. These historical data supported 

the results obtained by radiometric dating, indicating 

a tsunami event that cannot be earlier than the year 

1896. 

253 

Fig. 4 Plomari today and in the 1896. Boulder deposits are 

missing in the 1896’s photograph. 

Analysis of the historical tsunami catalogs isolated 

the Chios-Karaburum earthquake of 1949 as the only 

known event capable of creating the boulder deposit. 

The epicenter of the earthquake (M=6.7 Papazachos 

and Papazachou, 1997) was located offshore of 

Chios Island (Fig.1) at about 40 km from Plomari. 

The tsunami triggered by the earthquake affected the 

coasts of Chios Island, and Karaburum historical 

reports are available in both sectors (Altinok et al., 

2005). 

This paper presents results of geomorphologic traces 

of catastrophic wave impact in a coastal sector


EARTHQUAKE 

ARCHAEOLOGY 



included in the lists of areas affected by tsunami, but 

where field evidence of paleo-tsunamis was 

previously lacking. The enigmatic large boulders 

observable on the southern coast of Lesvos Island 

were investigated without excluding, a priori, any 

possible cause of deposition. The integrated study of 

geomorphological, hydrodynamic, seismotectonic 

and historical data provided evidence of a tsunami 

wave impact during the last century. Specifically, the 

boulder accumulation is likely to have been deposited 

by the tsunami that was triggered by the 6.7M Chios- 

Karaburum earthquake of 1949. 

Plomari (founded in 1849) home to about 6000 

people, is the second largest town of Lesvos Island 

and is the only sizable settlement on the southern 

coast of Lesvos. Its geographical position justifies the 

creation of serious policies related to seismic hazard. 

In fact, Papazachos et al. (2004) indicated the 

possibility of a strong seismic event in the near future 

(up to M 6.5–6.7) in the area of Lesvos and Chios. 

The above results represent a new contribution that 

could be of primary importance in assessing the 

degree of vulnerability of the local coastal 

communities to future tsunami events. 

Acknowledgements: The authors would like to thank Prof. 

G. Mastronuzzi (Bari, IT) for his suggestions and to C.N. 

Bianchi (Genoa, IT) for the identification of biological marine 

encrustation. 

References 

Altinok, Y., Ersoy., S. (2000). Tsunamis Observed on and 

Near the Turkish Coast. Natural Hazards 21, 185–205. 

Altinok, Y., Alpar, B., Ozer, N., Gazioglu, C. 2005. 1881 and 

1949 earthquakes at the Chios-Cesme Strait (Aegean 

Sea) and their relation to tsunamis. Natural Hazard and 

Earth System Science, 5, 717–725. 

Benner, R.; Browne, T.;Brückner, H.; Kelletat, D.; Scheffers, 

A. (2010). Boulder Transport by Waves: Progress in 

Physical Modelling. Zeitschrift für Geomorphologie NF. 

Supp., 54, (3), 127-146. 

Goto, K., Okada, K., Imamura, F., (2009). Characteristics 

and hydrodynamics of boulders transported by storm 

waves at Kudaka Island, Japan. Marine Geology 262, 

14–24 

Laborel J, 1987. Marine biogenic constructions in the 

Mediterranean—A review. Scientific report of Port-Cros 

Natural. Park (France) 13: 97–126. 

Mastronuzzi, G., Pignatelli, C., Sanso, P., (2006). Boulder 

fields: a valuablemorphological indicator of Paleotsunami 

in the Mediterranean Sea. Zeitschrift für Geomorphologie, 

NF 146, 173–194 

Mastronuzzi, G., Sanso, P., (2000). Boulder transport by 

catastrophic waves along the Ionian coast of Apulia 

(Southern Italy). Marine Geology 170, 93–103. 

Mastronuzzi, G., Sansò, P., (2004). Large boulder 

accumulations by extreme waves along the Adriatic 

Coast of southern Apulia (Italy). Quaternary International 

120, 173–184. 

Nyst, M., Thatcher, W. (2004). New constraints on the 

active tectonic deformation of the Aegean. Journal of 

Geophysical Resources 109, B11406, 

doi:10.1029/2003JB002830. 

Nott, J., (2003). Tsunami or storm waves?—Determining 

the origin of a spectacular field of wave emplaced 

boulders using numerical storm surge and wave models 

and hydrodynamic transport equations. Journal of 

Coastal Research 19, 348–356. 

Papadopoulos, G. A. (2009). Tsunamis, in J. C. Woodward 

(ed.), The Physical Geography of the Mediterranean. 

Oxford University Press, Oxford, 493–512. 

Papadopoulos, G. A., Fokaefs, A. (2005). Strong tsunamis 

in the Mediterranean Sea: a re-evaluation. ISET Journal 

of Earthquake Technology, 42, 159–170. 

Papadopoulos, G. A., Imamura, F.: (2001). A proposal for a 

new tsunami intensity scale. ITS 2001 Proc., Session 5, 

N. 5-1, 569–577 

Papadopoulos, G.A., Chalkis, B.J., (1984). Tsunami 

observed in Greece and the surrounding area from 

antiquity up to present times. Marine Geology 56, 309– 

317 

Papazachos B. C., Dimitriou P. P., (1991). Tsunamis In and 

Near Greece and Their Relation to the Earthquake Focal 

Mechanisms. Natural Hazards 4, 161-170. 

Papazachos, B., Papazachou, C., (1997). The earthquakes 

of Greece, Ziti editions. Thessaloniki 

Papazachos, C.B., Karakaisis, G.F., Scordilis, E.M., and 

Papazachos, B.C., (2004). Probabilities of activation of 

seismic faults in critical regions of the Aegean area, 

Geophysical Journal International 159, 679–687. 

Pignatelli, C., Sansò, P., Mastronuzzi, G. (2009). Evaluation 

of tsunami flooding using geomorphologic evidence. 

Marine Geology 260, 6–18. 

Scheffers, A., Kelletat, D., (2003). Sedimentologic and 

geomorphic tsunami imprints worldwide—a review. Earth- 

Science Reviews 63, 83–92. 

Scheffers, A., Scheffers, S.R., (2007). Tsunami deposits on 

the coastline of West Crete (Greece). Earth and 

Planetary Science Letters 259, 613–624. 

Scheffers A., Kelletat, D., Vött A., May, S.M., Scheffers, 

S.R., (2008). Late Holocene tsunami traces on the 

western and southern coastlines of the Peloponnesus 

(Greece). Earth and Planetary Science Letters 269, 271- 

279. 

Scicchitano, G., Monaco, C., Tortorici, L., (2007). Large 

boulder deposits by tsunami waves along the ionian 

coast of south-eastern Sicily (Italy). Marine Geology 238, 

75–91. 

Soloviev, S.L., (1990). Tsunamigenic zones in the 

Mediterranean Sea. Natural Hazards 3, 183–202. 

Soloviev, S.L., Solovieva, O.N., Go, C.N., Kim, C.S., 

Shchetnikov, N.A., (2000). Tsunamis in the 

Mediterranean Sea 2000 B.C.-2000 A.D. Kluwer, 

Dordrecht. 

Soukissian, T.H., Hatzinaki, M., Korres, G.,, Papadopoulos, 

A., Kallos, G., Anadranistakis, E. (2007). Wind wave atlas 

of the Hellenic seas. Hcmr publication, 300 pp. 

Vött, A., Brückner, H., May, M., Lang, F., Herd, R., 

Brockmüller, S., (2007). Strong tsunami impact on the 

Bay of Aghios Nikolaos and its environs (NW Greece) 

during classical-Hellenistic times. Quaternary 

International 181, (1), 105-122 

Vött, A., Brückner, H., Brockmüller, S., Handl, M., May, 

S.M., Gaki-Papanastassiou, K., Herd, R., Lang, F., 

Maroukian, H., Nelle, O,. Papanastassiou, D. (2009). 

Traces of Holocene tsunamis across the Sound of 

Lefkada, NW Greece. Global and Planetary Change 66, 

112-128. 

Vousdoukas, M.I., Velegrakis, A.F., Dimou, K., Zervakis V., 

Conley, D.C. (2009). Wave run-up observations in 

microtidal, sediment-starved pocket beaches of the 

Eastern Mediterranean. Journal of Marine Systems 78, 

537–547. 

Yolsal, S., Taymaz, T., Yalciner, A., C. (2007). 

Understanding tsunamis, potential source regions and 

tsunami-prone mechanisms in the Eastern 

Mediterranean.In: Taymaz, T., Yilmaz, Y. Dilek, Y. (eds) 

The geodynamics of the Aegean and Anatolia. Geological 

Society, London, Special Publications, 291, 201–230. 

254


EARTHQUAKE 

ARCHAEOLOGY 



ACTIVE FAULTING AND EARTHQUAKE-INDUCED SLOPE FAILURES IN 

ARCHEOLOGICAL SITES: CASE STUDY OF DELPHI, GREECE 

Valkaniotis, Sotiris (1, George Papathanassiou, Spyros Pavlides 

(1) Dept of Geology, Aristotle University of Thessaloniki, Greece, email gpapatha@auth.gr 

Abstract (Active faulting and earthquake-induced slope failures in arcacheological sites: case study of Delphi, Greece): 

The archeological area of Delphi is one of the most visited places in Greece. However, as it is described in historical reports, 

severe ground deformations triggered by earthquakes took place. In order to assess the rockfall potential and the relevant hazard 

within this area, the parameters of the 1870 earthquake were taken into account. Afterwards, the GIS-based Newmark’s 

displacement method was applied in order to compute the permanent displacement values in the study area. The outcome 

provided by this study, indicates that the zone located at the archaeological area is characterized as very unstable. Moreover, the 

value of the critical acceleration within this zone is very low and even a small magnitude seismic event could generate rockfalls. 

Thus, the hazard and consequently the risk is extremely high and permanent mitigation measures should be developed. 

Key words: Delphi, earthquake, rockfall, hazard 

INTRODUCTION-GEOLOGICAL SETTING 

The WNW-ESE-trending Gulf of Corinth separates 

the northern coast of the Peloponnese from mainland 

Greece (Figure 1). To the east it is bounded by the 

Isthmus of Corinth through which the Corinth Canal 

passes into the Saronic Gulf. The Gulf of Corinth cuts 

across the NNW-SSE regional trend of the Hellenic 

mountain chain and overlies the Pindus, Parnassos 

and Othris zones of the internal Hellenides. This has 

resulted in a NNW-SSE and NW-SE structural grain 

to the area, which may have had some control on the 

later extensional deformation (McKenzie, 1972; 1978; 

Papazachos & Papadopoulos, 1977; Makris, 1976; 

1978; Le-Pichon & Angelier, 1979; Angelier et al., 

1982; Papadopoulos et al., 1986; Doutsos et al., 

1988; Jackson & McKenzie, 1988). In the northern 

part of the Gulf of Corinth Rift, the mountain chains of 

Elikonas, Parnassos – Giona and part of Pindos are 

situated. 

DELPHI-ARAHOVA-AMFISSA FAULT ZONE 

One of the most important and impressive fault zones 

of the Central Greece area is the Amfissa – Delphi – 

Arahova Fault Zone (Pèchoux, 1977; Sebrier, 1977; 

de Boer & Hale, 2000). Detailed field mapping along 

with a wealth of historical documents and references 

for strong earthquakes back to 373 BC unveil an 

important and complex active fault structure 

(Piccardi, 2000; Valkaniotis, 2009). The fault zone 

extends along the southern rims of Mt Parnassos for 

a total length of more than 25 km and terminates to 

the west in Mt Giona over the plain of Amfissa 

(Figure 2). Normal- to oblique-slip displacement in 

the fault zone (Delphi – Arahova faults) is undergoing 

a intriguing transition to strike-slip displacement in 

the western part (Agia Efthimia fault) documented 

during field mapping (Valkaniotis, 2009). The 

mountainous area between Kifissos Basin and 

Amfissa – Delphi – Arahova Fault zone contains 

smaller fault structures following pre-existing alpine 

fault zones reactivated in the present extensional 

stress regime. 

255 

Fig. 1: Active and possible active faults in Central Greece. 

Fault data from Pavlides et al. (2007, 2008) 

Delphi fault is a normal-oblique normal fault, with 

dipping to the south and strike direction NW-SE to 

WNW-ESE for the western part and W-E for the 

central part (Figure 2). Bedrock lithology is compiled 

by alpine (Jurassic-Creataceous) limestones of 

Parnassos-Giona geotectonic unit. In the hangingwall, 

these limestones are overlain by flysch 

sediments, and a thick unit of Quaternary consistent 

and loose breccias and scree. 

The Delphi fault cuts through the archaeological site 

of the Delphi Oracle. Clearence of recent (historical) 

depostis in the site due to archaeological excavations 

provide a detailed examination of the fault zone 

structure and earthquake fault displacement. The 

fault zone, reaching a thickness of 500-2000 m in 

Delphi-Arahova valley, consists of sub-parallel fault 

planes, with a mean dip of 70-80 0 , converging in a 

depth of ~2 km with the main fault surface of 60 0 dip 

(Valkaniotis, 2009). 

The entire set of monuments that constitute the 

archaeological area of Delphi (Stadium, Sanctuary of 

Apollo, Castalia spring, High School, Temple of 

Athena) are placed in front of the main surface of the 

fault at Delphi, which forms a striking morphological 

vertical scarp. The archaeological site of Delphi is 

situated in an active scree deposition area, with


EARTHQUAKE 

ARCHAEOLOGY 



continuous rock and debris due to the high relief and 

tectonic weakening of the bedrock. Severe damages 

and surface ruptures are reported in Delphi (Piccardi, 

2000; Papadopoulos, 2000; Valkaniotis, 2009) for 

numerous historical earthquakes (373BC, 278BC, 

326AD, 552AD, 1870AD). Rock falls in the vicinity of 

Delphi are reported during numerous other smalllarge 

earthquakes in the broader area. The last 

strong earthquake in the Delphi-Arahova-Amfissa 

fault zone was the 1870 event, with a magnitude of 

6.8 and a possible rupture of the whole fault zone 

(Schmidt, 1879; Ambraseys & Pantelopoulos, 1989; 

Papadopoulos, 2000). 

model of an infinite slope in material having both 

frictional and cohesive strength and is given by: 

c ' tan ' m 

tan ' 

w 

FS (2) 

tsin a tan a tan a 

where ’ is the effective friction angle, c’ is the 

effective cohesion, is the slope angle, is the 

material unit weight, w is the unit weight of water, t is 

the thickness of the mass at right angles to the slope 

and m is the proportion of the slab thickness that is 

saturated. 

A Newmark analysis can be extended to regional 

analysis using GIS software (Miles & Keefer, 2000), 

Arcinfo, by applying equations 1 and 2 to raster data 

layers created for each input variable. 

APPLYING THE NEWMARK APPROACH TO 

DELPHI AREA 

Fig. 2: The Delphi-Arahova-Amfissa Fault Zone. Fault plane 

projections for western and central-east part. After 

Valkaniotis (2009). 

EVALUATING THE EARTHQUAKE-INDUCED 

LANDSLIDE HAZARD 

The potential of a slope failure triggered by an 

earthquake, can be evaluated by three main 

methods; pseudostatic, permanent or statistical 

analysis (Miles & Keefer, 2000). The latter accesses 

the landslide hazard by assuming the past predicts 

the future and the pseudostatic employs a static 

slope stability analysis with the addition of a 

horizontal force modeling the earthquake ground 

motion (Miles & Keefer, 2009). The second analysis, 

proposed by Newmark (1965), provides information 

regarding actual slope stability based on accepted 

characterization of the severity of the earthquake 

shaking (Miles & Keefer, 2009). 

The development of a Newmark analysis requires the 

evaluation of the parameters of the expected 

earthquake shaking and the capability of the 

geological unit to resist this dynamic effect. The latter 

parameter is quantified as the critical acceleration 

(a c ), a threshold ground acceleration necessary to 

overcome basal sliding resistance and initiate 

permanent downslope movement (Jibson, 2007). 

The computation of critical acceleration is based on 

the following equation proposed by Newmark, (1965) 

a ( FS 1) gsina 

(1) 

c 

where FS is the factor of safety, is the angle of the 

sliding surface (slope angle), g is the acceleration of 

gravity. 

According to Jibson et al. (1998), the factor of safety 

is evaluated using a relatively simple limit-equilibrium 

256 

The first step in the Newmark approach is the 

calculation of the factor of safety of the slope using 

equation 2. Therefore, the strength parameters of the 

geological units in the specific area should be 

evaluated. In our study, for the heavily jointed and 

weathered limestone in the study area, mean values 

of 30 o (angle of friction) and 15 kPa (cohesion) were 

employed and 30 o and 10 kPa for the formation of 

flysch, respectively, based on literature review. 

Information from the literature was used to estimate 

the thickness of the failed material parallel to the 

slope. Khazai & Sitar (2000) proposed a correlation 

between slope angle and the thickness of the failed 

mass in which a slope angle between 40-60 o 

corresponds to a thickness of 1m; this was adopted 

in the present study. 

Finally, the value of the topographic slope angle 

(figure 3) was obtained from the DEM of a 30 m grid 

prepared from contour lines on the 1:5000-scale 

topographic maps using ArcInfo Software. In the 

study area the slope angle varies from 0 o to 60 o , 

however in this research we took into consideration 

only the areas where slope angle >20 o . 

Fig. 3: Slope map of the study area. 

Having computed the factor of safety, the next step 

was the estimation of the critical acceleration using 

equation 1. However, it was first necessary to modify 

estimated values of Fs


EARTHQUAKE 

ARCHAEOLOGY 



et al., 1998). In order to achieve this, the cohesion of 

the geological units was increased, mainly in steep 

areas, until a factor of safety equal to one was 

accomplished. In order to avoid unrealistic cohesion 

values, the procedure suggested by Jibson et al. 

(1998) was used and an Fs of 1.001 was assigned to 

the unstable cells (figure 4). 

necessarily precisely predict measured landslide 

displacements in the field. Rather, Newmark's 

displacement is a useful index of how a slope is likely 

to perform during seismic shaking. 

Fig. 6: Distribution of the generated PGA values based on 

the scenario of 1870 earthquake. 

Fig. 4: Map showing the distribution of the fact of safety 

against slope instability. 

The critical acceleration value for each 30 m-grid 

pixel was then computed and a relative map was 

compiled for the whole area of Delphi using the 

spatial analysis of ArcInfo Software. The criticalacceleration 

map (figure 5) can be characterized as a 

seismic landslide susceptibility map as it delineates 

areas prone to slope failure independent of any 

ground-shaking scenario (Jibson et al., 1998). 

The value of the peak ground acceleration (PGA) that 

was employed in our study was estimated using the 

attenuation relationship proposed by Skarlatoudis et 

al. (2003) and by taking into consideration as 

earthquake magnitude the relevant seismic event of 

1870. 

In the present study, the Newmark displacement was 

computed using equation [3] proposed by Jibson 

(2007), valid for a magnitude from 5.3 to 7.6. This 

equation estimates rigid-block Newmark 

displacement as a function of the critical acceleration 

and peak ground acceleration. 

2.335 1.478 

a 

log 2.710 log 1 c ac 

 

D 

n a 

max a 

max 

(3) 

0.424 M 

0.454 

where D n is Newmark displacement in centimeters, 

a c is critical (yield) acceleration in g's, and a max is the 

peak horizontal ground acceleration in g's. 

The distributions of the estimated values of Newmark 

displacement are shown in figure 7. The boundary 

values taken for the applied approach of Newmark's 

displacement were 2, 5, 10, 25, 50 and 100 cm. Sites 

where the estimated Dn is >5 cm can be considered 

as prone to slope failure while failure is unlikely 

where Dn < 5 cm. As it is shown in figure 7, most of 

the study area is considered as a low potential to 

earthquake-induced rockfalls zone since the 

computed newmark displacement ranges between 2 

and 5 cm. However, a “hotspot”, an area of high 

potential to slope failures, is delineated at the eastern 

part of the archeological area of Delphi where the 

estimated displacement is >100 cm. 

Fig. 5: Map showing the distribution of the value of critical 

acceleration 

As an outcome, a PGA contour map of the study 

area was developed based on the computed values 

of ground motion using the Euclidean distance model 

provided by ArcInfo software. As can be seen in 

figure 6, the value of PGA in the study area varies 

from 0.32 to 0.71g. 

The next step in the analysis developed by Newmark 

(1965) is the calculation of the cumulative permanent 

displacement of the slopes for the given level of 

ground shaking. According to Jibson et al. (1998), 

Newmark’s method is based on a fairly simple model 

of rigid-body displacement and thus does not 

257 

Fig. 7: map showing the estimated Newmark displacement 

values in the area of Delphi


EARTHQUAKE 

ARCHAEOLOGY 



CONCLUSIONS 

The goal of this study was the evaluation of the 

potential of rockfall occurrence generated by seismic 

events in the vicinity of the archaeological site of 

Delphi, Greece. In order to achieve this, the scenario 

of 1870 earthquake was taken into account. The 

outcome provided by this research indicates that the 

value of Newmark displacement close to the 

archaeological site is low and the area is considered 

as low potential to earthquake-induced rockfalls. 

However, high values of displacement were 

estimated at the archaeological area where the factor 

of safety against slope failure and the critical 

acceleration are less than 1 and 0.1g, respectively. 

Thus, even a small magnitude seismic event could 

generate rockfalls. 

Acknowledgements: The study of active faults and the 

evaluation of their potential have been financial supported 

by the General Secretariat for research and technology of 

Greece. The authors of this study would like to thank the 

anonymous reviewers for their constructive comments. 

References 

Ambraseys, N. Pantelopoulos, P. (1989). The Fokis 

earthquake of 1 August 1870. Journal European 

Earthquake Engineering, 3, 10-18. 

Angelier, J., Lyberis, N., Le-Pichon, X., Barrier, E., & 

Huchon, P. (1982). The tectonic development of the 

Hellenic Arc and the Sea of Crete: a synthesis. 

Tectonophysics, 86, 159 -196. 

De Boer J.Z., J. R. Hale (2000). The geological origins of 

the oracle at Delphi, Greece, in: McGuire, B. et al., (eds) 

The archaeology of geological catastrophes. Geological 

Society of London, Special Publication, 171, 399–412. 

Doutsos, T., Kontopoulos, N. & Poulimenos, G. (1988): The 

Corinth-Patras rift as the initial stage of continental 

fragmentation behind an active island arc (Greece). Basin 

Research, 1, 177-190. 

Jackson, J.A. McKenzie, D.P. (1988). The relationship 

between plate motions and seismic tensors and the rate 

of active deformation in the Mediterranean and Middle 

East. Geophysical Journal, 93, 45-73. 

Jibson, R.W, (1993). Predicting earthquake-induced 

landslide displacements using Newmark’s sliding block 

analysis. Transp Res Rec 1411:9–17 

Jibson, R.E, (2007). Regression models for estimating 

coseismic landslide displacement. Engineering Geology 

91:209–218 

Jibson, R.W., Keefer, D.K, (1993) Analysis of the seismic 

origin of landslides: examples from the New Madrid 

Seismic Zone. Geological Society America Bulletin 

105(4):521–536 

Jibson, R.E., Harp, E.L., & Michael, J.A, (1998) A method 

for producing digital probabilistic seismic landslide hazard 

maps: an example from the Los Angeles, California area. 

USGS open-file report, pp 98–113 

Keefer, D, (1984) Landslides caused by earthquakes. 

Geological Society America Bulletin 95:406–421 

Keefer, D.K., Wilson, R.C, (1989) Predicting earthquakeinduced 

landslides, with emphasis on arid and semi-arid 

environments. In: Landslides in a semi-arid environment. 

Inland Geological Survey Society, 2:118–149 

Khazai, B., Sitar, N, (2000). Landsliding in native ground: a 

GIS-based approach to regional seismic slope stability 

assessment, 

report. 

http://www.ce.berkeley.edu/*khazai/research 

Le-Pichon, X. Angelier, J. (1979). The Hellenic arc and 

trench system a key to the neotectonic evolution of the 

eastern Mediterranean area. Tectonophysics, 60, 1-42. 

Makris, J. (1976). A dynamic model of the Hellenic arc 

deduced from geophysical data. Tectonophysics, 36, 

339-346. 

Makris, J. (1978). The crust and upper mantle of the 

Aegean region from deep seismic soundings. 


McKenzie, D.P. (1972). Active tectonics of the 

Mediterranean region, Geophysical Journal Research 

Society, 30, 109-182. 

McKenzie, D. (1978). Active tectonics of the Alpine- 

Himalayan belt: the Aegean Sea and surrounding 

regions. Geophysical Journal Research Society, 55, 217- 

254. 

Miles, S.B., Keefer, D.K, (2000). Evaluation of seismic slope 

performance models using a regional case study. 

Environmental Engineering Geoscience 6(1):25–39 

Miles, S.B., Keefer, D.K, (2009). Evaluation of CAMEL– 

Comprehensive Areal Model of Earthquake-induced 

Landslides. Engineering Geology 104: 1–15 

Newmark, N.M, (1965). Effects of earthquake on dams and 

embankments. Geotechnique 15(2):139–160 

Papadopoulos, G.A. (ed.) (2000). Historical earthquakes 

and tsunamis in the Corinth rift, central Greece. National 

Observatory of Athens, Institute of Geodynamics, 

Publication No 12. 128p. 

Papadopoulos, G.A., Kondopoulou, D.P., Leventakis, G.A. 

& Pavlides, S.B. (1986). Seismotectonics of the Aegean 

region. Tectonophysics, 124, 67-84 

Papazachos, B.C. and Papadopoulos, GA. (1977). Deep 

tectonic and associated ore deposits in the Aegean area. 

Proc. 6th Colloq. Geology Aegean Region, 3, 1071-1080. 

Pavlides, S., Chatzipetros, A. & Valkaniotis, S. (2007). 

Seismically Capable Faults In: Aegean Broader Area: 

Criteria For Classification. Geological Society 

Bicentenary Conference: Earth Sciences in the Service of 

Society, 10-12 September 2007, London. Abstract Book, 

124-125. 

Pavlides, S., Chatzipetros, A. & Valkaniotis, S. (2008). 

Active faults of Greece and surroundings. In Proceedings 

of the 33 rd 

International Geological Congress, Oslo, 

Norway. Abstract. 

Pèchoux, P. Y. (1977). Nouvelles remarques sur les 

versants Quaternaires du secteur de Delphes. Revue de 

Gèographie Physique et de Gèologie Dynamique, 

19:83–92 

Piccardi, L. (2000). Active faulting at Delphi, Greece: 

Seismotectonic remarks and a hypothesis for the 

geologic environment of a myth. Geology, 28: 651–654. 

Schmidt, J. (1879). Studien ueber Erdbeben, ed. Georgi, A. 

(Leipzig), 352p. 

Sebrier, M. (1977). Tectonique recente d’une tranversale a 

l’Arc Egeen . These de Docteur, Academie de Versailles, 

Universite de Paris XI, 100p. 

Skarlatoudis, A.A., Papazachos, C.B., Margaris, B.N., 

Theodulidis, N., Papaioannou, Ch., Kalogeras, I., 

Scordilis, E.M., & Karakostas, V, (2003). Empirical peak 

ground-motion predictive relations for shallow 

earthquakes in Greece. Bulletin of Seismological Society 

of America 93(6): 2591–2603 

Valkaniotis, S. (2009). Correlation Between Neotectonic 

Structures and Seismicity in the broader area of Gulf of 

Corinth (Central Greece). Unpublished PhD Thesis, 

Aristotle University of Thessaloniki, 2009, 247pp. 

258


EARTHQUAKE 

ARCHAEOLOGY 



SEDIMENTARY BURIAL OF ANCIENT OLYMPIA (PELOPONNESE, GREECE) 

BY HIGH-ENERGY FLOOD DEPOSITS – THE OLYMPIA TSUNAMI HYPOTHESIS 

Vött, Andreas (1, Fischer, Peter (2), Hadler, Hanna (1), Handl, Mathias (3), Lang, Franziska (4), Ntageretzis, Konstantin (1), 

Willershäuser, Timo (1) 

(1) Institute for Geography, Johannes Gutenberg-Universität, 55099 Mainz, GERMANY. E-Mail: voett@uni-mainz.de 

(2) Institute for Geography, Universität zu Köln, Albertus-Magnus-Platz, 50923 Köln, GERMANY. 

E-mail: peter.fischer@uni-koeln.de 

(3) Philipps-Universität Marburg, Biegenstraße 10, 35037 Marburg, GERMANY 

(4) Department of Classical Archaeology, Technische Universität Darmstadt, El-Lissitzky-Str. 1, 64287 Darmstadt, GERMANY. 

E-mail: flang@klarch.tu-darmstadt.de 

Abstract(Sedimentary burial of ancient Olympia (Peloponnese, Greece) by high-energy flood deposits – The Olympia 

tsunami Hypothesis): Detailed geo-scientific studies were carried out in the Kladeos and lower Alpheios River valleys in order to 

clarify the mystery of the rapid burial of Olympia under 4-6 m of sediments after the 6 th cent. AD and subsequent erosion of the 

Kladeos River by 8-10 m down to the ancient flow level. Sedimentological, geophysical, geochemical and microfaunal analyses 

were conducted along the Olympia terrace by means of 22 vibracores and 70 resistivity tomography transects. Geomorphological 

studies revealed strong discrepancies between the present hydraulic potential of the Kladeos River and the dimension and structure 

of the Olympia terrace. Our results show that the Kladeos River valley and Olympia experienced at least four distinct phases 

of catastrophic high-energy flood events. Sedimentary, geochemical and faunal traces found in the adjacent Basin of Flokas- 

Pelopio clearly document multiple tsunami impact. Identical fingerprints and strong stratigraphical correlations were also detected 

along the Kladeos River beyond the Ridge of Flokas-Platanos. We thus set up and discuss the Olympia Tsunami Hypothesis 

saying that the shallow saddles of the ridge were repeatedly overflowed by tsunami waters and the cult site Olympia was rather 

destroyed by tsunami than by fluvial processes related to the Kladeos River. 

Key words: Olympia, high-energy deposits, tsunami, geoarchaeology 

INTRODUCTION 

Olympia, used as famous cult site for Panhellenic 

games between Archaic times and the 4 th cent. AD, 

is located at the confluence of the Kladeos and Alpheios 

Rivers in the western Peloponnese (Greece). 

The sedimentary burial of ancient Olympia is one of 

the most interesting geoarchaeological mysteries in 

the Mediterranean world. The sedimentological evolution 

since early medieval times shows two different 

steps. After the 6 th cent. AD, the site was covered by 

4-6 m of sediments; subsequently, the nearby 

Kladeos River eroded its bed by 8-10 m approximately 

reaching the level existing during antiquity. 

Previous studies presented different explanations for 

this setting. Büdel (1981), together with Dufaure & 

Fouache 1988, Fouache & Pavlopoulos 2011, are in 

favour of anthropogenically induced soil erosion as 

main factor for enhanced sediment accumulation at 

the mouth of the Kladeos River during phases of 

uncontrolled landuse, especially after the Slavic 

invasion in early medieval times. On the contrary, 

Fountoulis & Mavroulis (2008) hold distinct periods of 

wet climate responsible for accelerated sediment 

accumulation. However, both scenarios do not give 

explanations for the change from accumulation to 

subsequent erosion dynamics within the past 1500 or 

so years. 

259 

Before systematic excavation of the site by the German 

Archaeological Institute started in 1875, the 

archaeological remains of Olympia were integrated 

into a wide terrace, the so called Olympia terrace. 

From a geomorphological point of view, this terrace 

can be traced both downstream the Alpheios River 

towards the present coast of the Gulf of Kyparissia 

and upstream all along the lower and middle Kladeos 

River valley. In the Kladeos area (Fig. 1), the Olympia 

terrace is up to 300-500 m wide. It is present on 

both sides of the Kladeos River with a distance of up 

to 200 m between the opposite terrace faces. Considering 

that the Kladeos River is rather a creek than 

a river with a perennial runoff concentrated within a 

maximum 5-8 m-wide secondary river channel and 

maximum water flow depths of 2-3 m during winter 

and heavy rain events, there is a considerable discrepancy 

between the dimension and the hydraulic 

potential of the Kladeos River on the one hand, and 

the local geomorphology of the Olympia terrace on 

the other hand. 

The objectives of our investigations thus were (i) to 

establish a well-based stratigraphy of the Olympia 

terrace along the Kladeos River by detailed geomorphological 

and sedimentological studies, (ii) to compare 

these results with stratigraphies found along the 

Alpheios River, especially in the adjacent Basin of 

Flokas-Pelopio, and (iii) to find a geomorphological 

model which best explains the hydro-dynamic finger-


EARTHQUAKE 

ARCHAEOLOGY 



print and distribution pattern of the encountered 

sediments. 

METHODS 

We carried out geomorphological mapping of the 

Olympia terrace using topographical and geological 

maps and remote sensing data. Stratigraphical studies 

are based on 22 vibracores, 16 of which were 

drilled in the Kladeos River valley and 6 in the Alpheios 

River valley by means of a handheld Cobra 

mk1 vibracorer (Atlas Copco) and a Nordmeyer drill 

rig (type RS 0/2.3). We used core diameters of 6 and 

5 cm. Maximum coring depth was 17 m below ground 

surface (m b.s.). Earth resistivity tomography (ERT) 

was conducted to study subsurface structures and 

stratigraphies along 70 transects using a multielectrode 

geo-electrical Iris instrument (type Syscal 

Junior Switch 48). Selected sediment samples were 

analysed by their microfossil content. Key cores were 

additionally cored with an inliner system and analysed 

using the X-ray fluorescence technique (XRF). 

We also conducted grain size analyses for selected 

samples. Vibracoring sites and ERT transects were 

measured by means of a differential GPS (type Topcon 

HiPer Pro) with an accuracy of 2 cm or better. A 

local geochronostratigraphical framework was established 

using age estimations of diagnostic ceramic 

fragments, radiocarbon dating and Optically Stimulated 

Luminescence (OSL) approaches. 

RESULTS 

Vibracores ALP 3, 4, 5 and 8, drilled in the environs 

of ancient Olympia, especially on top of the Olympia 

terrace towards the south of the southeastern Roman 

baths and on top of the Olympia terrace to the west 

of the Kladeos, revealed characteristic sequences of 

light brown, silt- and clay-dominated alluvial or colluvial 

deposits with remains of freshwater fauna which 

are repeatedly interrupted by up to four sections out 

of light brown sand and gravel. These coarse-grained 

deposits were found in comparable stratigraphic 

positions in every core and thus indicate synchronous 

flooding of wide areas related to high flow velocities. 

Associated to these deposits, we found 

sedimentary structures such as basal erosional unconformities 

in underlying silt deposits, fining upward 

sequences with mud caps and abundant marine shell 

debris and marine microfauna. XRF analyses revealed 

clear maximum peaks of the Ca/Ti ratio for 

the coarse-grained sections. The calcium content 

documents the input of calcium carbonate from biogenic 

and bedrock sources, the titan content reflects 

terrigenous input by subaerial weathering into the 

sedimentary system. Results were also tested for 

masking and matrix effects which can be excluded as 

major sources of bias. Apart from a distinct and ca. 1 

m-thick palaeosol found on top of a coarse-grained 

section at ca. 4 m b.s. in core ALP 8 (Fig. 2), which 

includes Roman sherds, palaeosols are missing. A 

charcoal fragment from a fining upward sequence out 

of sand and gravel deposited under high-energy 

conditions at site ALP 5 was 14 C AMS radiocarbon 

dated to 585-647 cal AD (2 interval, 3.74 m b.s.). 

Vibracores ALP 12-15 and 19 were drilled in the 

middle Kladeos River valley around the villages of 

Mageiras and Kladeos on top of the Olympia terrace. 

Here, we also found predominating clayey to silty 

deposits accumulated under quiescent to moderate 

flow conditions. These deposits are grey in colour, 

partly include lots of plant material and freshwater 

shell and thus document a permanent water body of 

fluvio-limnic nature. Similar to the situation at Olympia, 

we found up to four intersecting layers of sand 

and gravel, partly grey (base and mid-section), partly 

light brown in colour (top), associated to high-energy 

sediment type structures such as basal unconformities, 

muddy intraclasts, fining upward sequences and 

including abundant faunal remains of marine origin. 

Based on XRF measurements, the Ca/Ti ratio again 

shows clear maximum peaks stratigraphically corresponding 

to the intersecting coarse-grained layers. 

Thus, all along the Kladeos River valley between 

Kladeos, Mageiras and Olympia, the Olympia terrace 

shows a similar inner structure with the individual 

coarse-grained high-energy layers lying in stratigraphically 

consistent positions. The same is true for 

both distal and proximal parts of each specific terrace 

section as documented by ERT transects. 

We thus conclude, in a first step, that the Olympia 

terrace between Kladeos and Olympia documents at 

least four phases of high-energy flood events that 

obviously affected the whole valley bottom to an 

extent far beyond the dimensions of the present 

Kladeos river channel. 

Vibracores 9-11 were drilled at the eastern fringe of 

the Basin of Flokas-Pelopio at a distance of 1 to 2 km 

to the west of the Kladeos River valley across the 

Flokas-Platanos Ridge. Vibracore ALP 10 lies around 

2.2 km distant from the Alpheios River at a right 

angle opposite to its seaward flow direction. Despite 

the vicinity to the Alpheios River, vibracore ALP 9 

does not include any pieces of gravel; it consists of 

homogeneously light brown (top) to grey (base), 

clayey to silty deposits accumulated in a low-energy 

freshwater lake environment. Associated to basal 

unconformities, we found several intersecting layers 

of sand with fining upward sequences and mud caps 

and marine faunal remains which clearly document 

episodic high-energy interruption of the allochthonous 

environment. Vibracore ALP 10, drilled at the 

western hill slope of the Flokas-Platanos Ridge, 

revealed a similar stratigraphic pattern with distinct 

interruptions of limnic (base and mid-section) and 

colluvial (top) deposits by predominantly sandy to 

gravelly high-energy deposits reaching up to 20 m 

above present sea level (m a.s.l.). In each case, at 

least the upper part of the intersecting layer consists 

of brown sediments of mostly terrestrial origin. The 

Ca/Ti ratios for both cores show distinct maximum 

peaks for the intersecting coarse-grained allochthonous 

material (Fig. 3). 

260


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 1: View of the middle Kladeos River valley (foreground) towards the west. The valley is separated from the adjacent 

Basin of Flokas-Pelopio (middleground) by the homonymous ridge. The Alpheios River valley connects the basin and the Gulf 

of Kyparissia (left background) in a direct line. Photo taken by A. Vött, 2011. 

In the more seaward vibracore ALP 7, drilled some 8 

km inland in the midst of the Alpheios River valley 

halfway between ALP 9 and the present coast, we 

found autochthonous marine sand at approx. 2.50 m 

below present sea level (m b.s.l.). Covered by a 10 

m-thick layer of sandy gravel, this unit documents 

that the Gulf of Kyparissia extended far into today’s 

Alpheios valley during the Holocene. 

DISCUSSION 

Sedimentary structures, geochemical fingerprints and 

faunal remains encountered at the western fringe of 

the Basin of Flokas-Platanos indicate episodic highenergy 

marine flooding from the sea side. Stratigraphies 

of cores ALP 9 and 10 clearly show tsunami-type 

marine incursions into a shallow lake and 

runup-backflow sequences at the adjacent hillside, 

respectively. Considering that at least parts of the 

lower Alpheios River valley were flooded by the sea 

during the Holocene and manifold traces of palaeotsunami 

are known from the present coast (Vött 

et al. 2011), these findings are plausible. 

core sections revealed freshwater ostracods in finegrained 

silt-dominated deposits but exclusively marine 

species in high-energy flood sediments. In case, 

these were transported by a mega-Kladeos River 

with a width of several hundreds of meters – dimensions 

which are necessary to explain the consistent 

lateral distribution of this facies across the entire 

Olympia terrace – one would have to expect admixed 

freshwater or even river-borne species. This was not 

the case in the samples which we analysed. However, 

further attempts are needed to fully understand 

the fossil record of the high-energy flood deposits in 

the environs of Olympia. 

Geomorphological studies carried out within our 

project revealed erosion as well as scouring features 

across the lower saddles of the Flokas-Platanos 

Ridge lying at about 60 m a.s.l. which indicate possible 

flow paths across the ridge. 

Considering, however, that both the geochemical and 

the overall stratigraphic patterns of the cores from the 

west of the ridge and from the Kladeos River valley 

itself are principally identical – documenting episodic 

high-energy input of coarse-grained marine sediments 

into prevailing quiescent environments – we 

hypothesize that marine flooding also affected the 

Kladeos River valley and Olympia. Our main arguments 

thus are of sedimentological and geochemical 

nature and based on consistent stratigraphies on 

either side of the Flokas-Platanos Ridge. We call this 

scenario the “Olympia Tsunami Hypothesis”. 

As a major argument against our hypothesis one may 

bring forward the fact that Neogene bedrock units in 

the catchment area of the Kladeos River also include 

conglomerate and sand units, the latter being characterized 

by abundant remains of a Plio-Pleistocene 

marine fauna (IGME 1982). However, especially 

around Olympia and in most parts of the middle 

Kladeos River valley, Plio-Pleistocene marl and not 

sand is the predominant bedrock material provoking 

many landslides (IGME 1982, Christaras et al. 

2002). Moreover, microfaunal analyses of selected 

261 

Fig. 2: Facies distribution pattern of vibracore ALP 8 (38.37 

m a.s.l.) drilled on top of the Olympia terrace to the west of 

the Kladeos River some 250 m to the west of the Kronos 

hill. Several sandy to gravelly high-energy flood type (heft) 

deposits can be seen alternating with fine-grained colluvial 

or (fluvio-)limnic deposits. The mid-core palaeosol dates to 

Roman times. Please note the three fining upward sequences 

encountered in the uppermost heft unit. Photo 

taken by T. Willershäuser, 2010. 

Concerning the elevation which, at a first glance, 

seems to be too high for tsunami flooding, one has to 

consider (i) channelling and accelerating effects of 

the tube-like and ca. 8 km-long lower Alpheios River 

valley during inflow, (ii) backwatering and boosting of 

inundating marine water masses because of blocked 

backflow in the Alpheios River valley itself and at the


EARTHQUAKE 

ARCHAEOLOGY 



breakthrough through Drouva Ridge, where the modern 

Alpheios dam is located, (iii) the arrival of subsequent 

waves of a longer wave train before backflow 

was completely accomplished, as well as (iv) potential 

interim changes in the topography due to landslides 

of the predominating Tertiary marls typical for 

the ridge. From the nearby coast, there is evidence of 

multiple tsunami landfall since the mid-Holocene and 

for tsunami run-up up to 18 m a.s.l. around the ancient 

site Pheia, one of the harbours of Olympia (Vött 

et al. 2011). 

By the Olympia Tsunami Hypothesis, we suggest that 

tsunami waters repeatedly overflowed the lower 

saddles between Flokas and Platanos and then 

partly flowed upstream and partly downstream the 

Kladeos River valley, hereby creating a vast terrace 

structure way above the flow level of the Kladeos at 

that time. Subsequently, tsunami backflow concentrated 

along the Kladeos creek eroding an up to 200 

m-wide gap into the terrace. Tsunami backflow 

through the breakthrough of the Alpheios across the 

Drouva Ridge was hindered and possibly blocked, at 

least during tsunami inundation of the Basin of Flokas-Pelopio. 

One of the high-energy flood deposits 

encountered near Olympia was dated to 585-647 cal 

AD which fits well with the earthquake in 551 AD 

during which Olympia is reported to have been destroyed. 

However, there are no historic accounts on 

catastrophic flooding of Olympia. 

finally covered completely by sediments. (ii) Similar 

high-energy flood deposits were also encountered at 

the eastern fringe of the adjacent Basin of Flokas- 

Pelopio; due to their fossil content, their geochemical 

fingerprint, their geomorphological position and their 

stratigraphical pattern, they are interpreted as of 

tsunamigenic origin. (iii) In both areas, the highenergy 

facies is associated to sedimentary structures 

known from historic to recent tsunami events (basal 

unconformity, fining upward sequences, mud caps, 

ripped up intraclasts etc.) and is characterized by 

abundant fragments of a marine fauna. (iv) Geomorphological 

features such as marks of scouring and 

undercutting let us assume that tsunami overflow 

occurred across the comparatively shallow saddles 

between the villages of Flokas and Platanos wheras 

backflow was accomplished along the river valley as 

soon as blocking of the entrance of the Alpheios 

River into the Basin of Flokas-Pelopio by tsunami 

waters had ceased. We call this scenario “The Olympia 

Tsunami Hypothesis”. (v) Our results, together 

with manifold geoarchaeological destruction patterns 

at Olympia, rather indicate catastrophic event-related 

flooding by tsunami than by the River Kladeos itself. 

Acknowledgements: Sincere thanks are due to our cooperating 

partners G. Chatzi-Spiliopoulou (Olympia), I. Fountoulis 

(Athens), H.-J. Gehrke (Freiburg), A. Hoppe, R. Lehné 

(Darmstadt), K. Reicherter (Aachen), D. Sakellariou (Athens) 

and R. Senff (Athens/Olympia). Work permits were 

kindly issued by the Greek Institute for Geology and Mineral 

Exploration (IGME). We gratefully acknowledge funding by 

the German research Foundation (DFG, VO 938/3-1). 

References 

Fig. 3: Ca/Ti ratios for vibracore ALP 9A drilled at the western 

fringe of the Basin of Flokas-Pelopio at the foot of the 

homonymous ridge nearby Olympia. Allochthonous siltdominated 

deposits of a quiescent limnic environment are 

characterized by a Ca/Ti base level around 50. Episodic 

interferences from the sea side by tsunami waves left behind 

marine sand deposits with Ca/Ti ratios up to 200. 

Similar Ca/Ti profiles and stratigraphies were found for 

vibracores in the Kladeos River valley and nearby Olympia. 

CONCLUSIONS 

Detailed geomorphological, sedimentological, geophysical, 

geochemical and microfaunal studies of the 

Olympia terrace in the Kladeos and the lower Alpheios 

River valleys allow to draw the following conclusions. 

(i) The Kladeos valley and the environs of 

Olympia were affected by at least four distinct phases 

of high-energy flood events by which the site was 

Büdel, J. (1981). Klima-Geomorphologie. 2 nd 

edition. 

Bornträger. Berlin, Stuttgart. 304 p. 

Christaras, B., Mariolakos, I., Dimitriou, A., Moraiti, E. & D. 

Mariolakos (2002). Slope instability at Olympia archaeological 

site in southern Greece. Intern. Symp. “Landslides 

Risk Mitigation and Protection of Cultural and Natural 

Heritage”, Abstract Volume. Kyoto. 339-342. 

Dufaure, J.-J. & E. Fouache (1988). Variabilité des crises 

d’âge historique le long des vallées d’Elide (Ouest du Peloponnèse). 

Cahier Inter-universitaire d’Etudes Méditerranéennes, 

12: 259-278. Poitiers. 

Fouache, E. & K. Pavlopoulos (2011). The interplay between 

environment and people from Neolithic to Classical 

times in Greece and Albania. In: Landscapes and societies. 

Selected cases (Martini, I.P., Chesworth, W. eds.). 

Springer. Dordrecht. 155-166. 

Fountoulis, I., Mariolakos, I., Mavroulis, S. & I. Ladas 

(2008). Flood periods during prehistoric and Roman 

times in the Kladeos torrent basin – ancient Olympia 

(Greece). In: 8 th 

Intern. Hydrogeol. Congr. Greece, 3 rd 

MEM Workshop Fissured Rocks Hydrology, Abstract 

Volume (Migiros, G., Stamatis, G., Stournaras, G. eds.). 

Athens. 809-818. 

Institute for Geology and Mineral Exploration (IGME, 1982). 

Geological map of Greece, 1:50,000, Olympia Sheet. 

Athens. 

Vött, A., Bareth, G., Brückner, H., Lang, F., Sakellariou, D., 

Hadler, H., Ntageretzis, K. & T. Willershäuser (2011). 

Olympia’s harbour site Pheia (Elis, western Peloponnese, 

Greece) destroyed by tsunami impact. Die Erde (in 

press). 

262


EARTHQUAKE 

ARCHAEOLOGY 



LANDSLIDE MAPPING TO ANALYSE EARTHQUAKE ENVIRONMENTAL EFFECTS (EEE) 

IN CARMONA, SPAIN – RELATION TO THE 1504 EVENT? 

Vollmert, Andre (1, Klaus Reicherter (2), Pablo G. Silva (3), Tomas M. Fernandez-Steeger (4) 

(1) Lehrstuhl für Ingenieur- und Hydrogeologie, RWTH Aachen University, Lochnerstr. 4-20, 52056 Aachen. GERMANY. Email: 

andre.vollmert@rwth-aachen.de 

(2) Lehr- und Forschungsgebiet Neotektonik und Georisiken, RWTH Aachen University, Lochnerstr. 4-20, 52056 Aachen. 

GERMANY. Email: k.reicherter@nug.rwth-aachen.de 


SPAIN. Email: pgsilva@usal.es 

(4) Lehrstuhl für Ingenieur- und Hydrogeologie, RWTH Aachen University, Lochnerstr. 4-20, 52056 Aachen. GERMANY. Email: 

fernandez-steeger@lih.rwth-aachen.de 

Abstract (Landslide mapping to analyse earthquake environmental effects in Carmona, Spain – relation to the 1504 

event?): The 1504 Carmona earthquake (intensity IX EMS) claimed the loss of human life and caused a number of Earthquake 

Environmental Effects. On the basis of historical data reported by George Bonsor (1918) this study is intended to estimate 

coseismic slope performance. The aim is to combine field investigations, geotechnical parameters and computerized models to 

generate digital probabilistic seismic landslide hazard maps on a local scale. GIS-based simulations of mass movements driven by 

hydrodynamical and gravitational processes are performed by means of the factor of safety, which is calculated for dry and fully 

water saturated conditions. Following Newmark´s sliding block model these approaches are extended to assess the potential of 

earthquake-triggered slope movements. Assuming a Peak Ground Acceleration of 0.3 g, representing the 1504 event, the most 

affected areas show a failure probability of 33.5 %. 

Key words: 1504 Carmona Earthquake, Seismic Landslide Hazard Assessment, South Spain 

INTRODUCTION 

In 1918 the archaeologist George Bonsor was the 

first scientist who published the effects of a strong 

earthquake near Carmona (South Spain) in 1504 

(Bonsor, 1918). Based on the ESI-2007 Intensity 

Scale, Silva et al. (2009) attract notice again on this 

event in order to update Bonsor´s data. They focus 

on ground cracks, liquefaction, anomalous waves, 

flooding in rivers, temporary turbidity changes in 

wells and, especially, on mass movements, since 

landslides and rock falls belong to the most relevant 

phenomena of all EEE being observed in Carmona. 

This study provides different approaches to calculate 

the site scaled slope instability in terms of the 1504 

earthquake (IX EMS) as a potential triggering factor 

for a number of observed landslides. Each of the 

methods combines geotechnical results and slope 

angles derived from a Digital Elevation Model (DEM). 

Figure 1 points out the sequential steps leading to 

the hazard-mapping procedure of the study. All 

simulations have been performed under dry and fully 

water saturated conditions. 

LANDSLIDES IN CARMONA 

Within the southern margin of the Guadalquivir river 

valley Carmona is founded on a small NE-SW 

trending ridge (Los Alcores). It consists of Miocene 

blue marls and grey clays coming from the southeast 

located Betic Cordillera front. 

Fig. 1: Flow chart showing the steps involved in producing a seismic landslide hazard map; white: input parameters; blue: 

results (modified after Jibson et al., 2000). 

263


EARTHQUAKE 

ARCHAEOLOGY 



This substratum is covered by a Late Neogene 

calcarenite unit, which outcrops in a steep cliff, 

surrounding large parts of the city. Both units can be 

subjected to massive landslides. In order to 

distinguish seismically triggered slope movements 

from others driven by hydrodynamical and 

gravitational processes, all landslide phenomena are 

classified according to type of movement, material 

and size of the displaced mass. Furthermore, all 

possible causes including geological, morphological, 

physical and human influences are determined as an 

important aim. 

All important investigation sites are indicated in 

figure 2 showing the studied area around Carmona. 

The map also includes joint diagrams from 

Fig. 2: Map of Carmona showing the sampling points for later geotechnical investigations, joint diagrams and the location of 

landslide phenomena illustrated in figure 3 (Gauss-Krüger coordinates). 

264


EARTHQUAKE 

ARCHAEOLOGY 



calcarenite strata and the location of observed rock 

falls, topples and slides as well as earth slides in the 

unit of blue marls and grey clays. Typical examples 

of these types are illustrated in Figure 3. 

66 %) are supposed to be essential for the 

occurrence of massive earth slides on steep slopes 

along the courses of streams. 

Apart from these invariant parameters, both, water 

saturated soils caused by intensive rainfalls and 

earthquake shaking can be seen to be the most 

relevant causal factors for landslides in Carmona. 

Therefore, these triggering factors are considered in 

the following simulations. 

Fig. 3: Huge rock fall underneath the Picacho (A); 

Toppling process of a column of calcarenite rock (B); 

Calcarenitic boulders are transported downslope on 

softer clayey materials (C); Earth slides due to water 

saturated conditions and steep slope angles (D). 

Landslides in Carmona are related to a number of 

preparative and triggering factors. Observed rock 

falls and topples in calcarenite strata can be mainly 

subjected to SSE- and ENE-striking tension cracks 

(fig. 2) and relatively low shearing parameters 

(c´ = 17 MPa; ´ = 41°). Laboratory results of 

analysed loose material samples indicate effective 

cohesions between 13.04 and 20.16 kPa and angles 

of internal friction between 17.35 and 26.23°. This 

data and the high contents of clay minerals (35 – 

265 

SIMULATION OF SLOPE STABILITY 

First slope stability has been simulated be means of 

the factor of safety, which is calculated by the ratio of 

the sum of the resisting forces that act to inhibit a 

slope failure to the sum of the driving forces that tend 

to cause a failure. The application of a Geographical 

Information System (ArcGIS 9.3) allowed a 

differentiated calculation for every grid cell (2 x 2 m), 

where input parameters vary due to different slope 

angles. 

Based on the factor of safety the site-specific critical 

acceleration was calculated in a second step. 

According to Newmark´s sliding block analogy 

(Newmark, 1965) the critical acceleration is defined 

as the minimum horizontal seismic acceleration that 

is necessary to overcome the shear resistance of a 

friction block, resting on an inclined plane. That 

means, the higher the degree of slope stability, the 

higher the critical acceleration, which is needed to 

cause a failure. 

To estimate the cumulative slope displacement 

during an earthquake, Wilson & Keefer (1983) 

developed a double integration approach based on 

numerically cumbersome calculations performed by 

Newmark (1965). Thereby, those sections of an 

earthquake accelerogram that exceed the critical 

acceleration of a slope are integrated two times to 

obtain the velocity and the cumulative displacement 

of the sliding block. Considering also the PGA of the 

Carmona earthquake (0.3 g) it was possible to 

determine the Newmark Displacement. 

Newmark displacement rates are not directly 

correlated to the potential of earthquake-triggered 

landslides. For this reason, Jibson et al. (2000) 

developed a probabilistic empirical model, which 

allows the estimation of the probability of a failure for 

every grid cell (Eq. 1): 

P 

 

 

 

1.565 

f 0.335 

1exp0.048D 

N [1] 

They have calibrated these parameters with data 

from Southern California and anticipate that the 

mapping procedure is applicable in any areas 

susceptible to seismic slope failure. Therefore the 

model was used to compile digital probabilistic 

landslide hazard maps for dry and fully water 

saturated conditions in the study area (fig. 4).


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 4: Probabilistic landslide hazard map showing the probability of failure in case of an intensity-IX (EMS) earthquake 

under fully water saturated conditions (Gauss-Krüger coordinates). 

The most affected areas show a failure probability of 

33.5%. They are generally related to the same slopes 

indicating a higher potential of landslides under nonseismic 

conditions, however, they are extended. 

CONCLUSION 

The performed assessment of earthquake-triggered 

landslides provides useful information to estimate 

potential damages during future earthquakes. In this 

sense, the designation of vulnerable areas can be 

used to predict interruptions of access roads, gas 

and water pipes or electrical lines in case of another 

strong earthquake in Carmona when landslides of 

large volumes will be triggered with high probability. 

Acknowledgements: This work would not have been 

possible without the support of many persons. It is a 

pleasure for me to thank those who always helped and 

inspired me during this project. 

References 

Bonsor, J. (1918): El terremoto de 1504 en Carmona y en 

Los Alcores. Boletín de la Real Sociedad Espan ola de 

Historia Natural, 18, 115-123. 

Jibson, R. W., Harp, E. L. & Michael, J. A. (2000): A method 

for producing digital probabilistic seismic landslide hazard 

maps. Engineering Geology, 58, 271-289. 

Newmark, N. M. (1965): Effects of earthquakes on dams 

and embankments. Géotechnique, 15, 139-160. 

Silva, P.G., M.A Rodríguez-Pascua, R. Pérez-López, J.L. 

Giner-Robles, J. Lario, T. Bardají, J.L. Goy & C. Zazo, 

(2009). Geological and archaeological record of the 1504 

AD Carmona earthquake (Guadalquivir Basin, South 

Spain): a review after Bonsor, 1918. In: 

Archaeoseismology and Palaeoseismology in the Alpine- 

Himalayan Collisional Zone (Pérez-López, R., Grützner, 

C., Lario, J., Reicherter, K., Silva, P.G. eds). Baelo 

Claudia, Spain, 139-142. 

Wilson, R. C. & Keefer, D. K. (1983): Dynamic analysis of a 

slope failure from the 6 August 1979 Coyote Lake, 

California, earthquake. Bulletin of the Seismological 

Society of America, 73, 863-877. 

266


EARTHQUAKE 

ARCHAEOLOGY 



STORM SURGE LAYERS WITHIN A CHANGEFUL HOLOCENE ENVIRONMENT OR 

SEDIMENTARY TRACES OF PALAEO-TSUNAMIGENIC EVENTS. PROS AND CONS OF 

ON-SITE FINDINGS, JADE BAY, SOUTHERN NORTH SEA, GERMANY 

Wartenberg, Wolfram (1), Andreas Vött (2), Hanna Hadler (2), Timo Willershäuser (2), Holger Freund (1), Stefanie Schnaidt (1) 

(1) Institute for Chemistry and Biology of the Marine environment (ICBM), Carl von Ossietzky Universität Oldenburg, 26382 

Wilhelmshaven. Germany. Email: wolfram.wartenberg@uni-oldenburg.de 

(2) Institute for Geography, Johannes Gutenberg-Universität, 55099 Mainz. Germany. Email: voett@uni-mainz.de 

Abstract (Storm surge layers within a changeful Holocene environment or sedimentary traces of palaeo-tsunamigenic 

events. Pros and cons of on-site findings, Jade Bay, Southern North Sea, Germany): Tsunamigenic events are well known to 

occur within the Mediterranean and along many seismo-tectonically active coasts all over the world. Their incidence is almost 

excluded for the southern North Sea, a shore line well known to be prone to storm surges that in parts occur at a regional scale. 

Accordant chronicles describe floods of sheer enormity, for instance affecting the Dutch and the German coastal sections at the 

same time (e.g. Cosmas and Damian flood, 1509). 

Key words: tsunami, storm, North Sea, Germany 

INTRODUCTION 

The Jade Bay in Lower Saxony, northwest Germany, 

is the largest tidal inlet of the German North Sea 

coast. The modern embayed tidal flat system shows 

a changeful Holocene sedimentary record from 

terrestrial-driven to seaward-influenced environments 

(Streif, 2004). The Jade area has been investigated 

interpreting sedimentary markers, pollen and macro 

remains taken from 45 cores. Probing was carried 

out between 2009 and 2011. The sedentary 

chronology is based on 36 radiocarbon and pollen 

datings. Direct age determination of clastic sediments 

will be complemented in late 2011 using optical 

dating. 

The palaeo-geomorphology of the Jade area is 

influenced by a major north-south trending channel 

(Sindowski, 1972), representing the structural 

rudiment for an early stage of the advancing sea. 

From approximately 4500 cal BC onwards, the 

palaeo-coastline must have been close, starting to 

increase the groundwater level. Alder carr to 

Cyperaceae fen peat started to develop extensively 

before marine conditions became dominant from 

~3000 to ~2800 cal BC (Wartenberg & Freund, in 

press). Two different palaeo-environments are 

related to the present-day Jade Bay, each identifying 

a distinctive local depositional development 

(Wartenberg & Freund, in press). From the west to 

the centre, the equivalent early Holocene landscape 

morphology is drainage-driven, feeding the 

associated pronounced basal peat with minerogenic 

water but being autonomous from isochronic relative 

sea-level. To the east, basal peat is absent within the 

sedimentary succession. Here, the facies zone is 

dominated by tidal flat to brackish-lagoonal 

267 

sediments, in places intercalated by fen peat layers 

dating back to minimum 4490 cal BC. 

Comparing the western and the eastern palaeoenvironments 

reveals different sedimentary signals 

identifying coincident event-stratigraphic markers of 

early to late Holocene age (Fig 1). Their eventstratigraphic 

signal may be linked to distinct 

sedimentary horizons at a regional scale. The clastic 

material deposited in parts shows rhythmic layers of 

coarser material within one event horizon (Fig. 2). 

Fig. 1(above): Description of core PR 321, tidal flat Arngast 

Sand, central Jade Bay. Drawing: Stefanie Schnaidt. 

Fig. 2 (below): Event layer of core PR 321 at 1.77 to 1.64 m. 

The poster presents detailed 

discussion on the origin of these 

high-energy event layers with 

respect to major storm or tsunami 

influence. Feasible Tsunami 

triggers may have been landslides 

in the northern North Sea area 

(alike the Storegga event at ~8000 

BP) or earthquakes along active 

faults at the coastal sections of


EARTHQUAKE 

ARCHAEOLOGY 



southern Spain and Portugal. 

Acknowledgements: The geological field work was 

accomplished as part of an interdisciplinary study of the 

Jade Bay. We gratefully acknowledge project funding by the 

Ministry for Science and Culture of Lower Saxony 

(Sponsorship program: “Niedersächsisches Vorab der 

Volkswagen Stiftung”). 

Wartenberg, W., Freund, H. (2011): Late Pleistocene 

and Holocene sedimentary record within the Jade 

Bay, Lower Saxony, Northwest Germany – new 

aspects for the palaeo-ecological record. 

Quaternary International (in press). 

References 

Sindowski, K.-H. (1972): Zur Geologie des 

Jadebusen-Gebietes. Oldenburger Jahrbuch 72, 

175-181. 

Streif, H. (2004): Sedimentary record of Pleistocene 

and Holocene marine inundations along the North 

Sea coast of Lower Saxony, Germany. Quaternary 

International 112, 3-28. 

268


EARTHQUAKE 

ARCHAEOLOGY 



TESTING EARTHQUAKE RECURRENCE MODELS WITH 3D TRENCHING ALONG THE 

DEAD-SEA TRANSFORM 

Wechsler, Neta (1, 2), Thomas K. Rockwell (2), Yann Klinger (1), Amotz Agnon (3), Shmulik Marco (4) 

(1) Equipe de Tectonique, Institut de Physique du Globe de Paris, CNRS, Paris, France. Email: wechsler@ipgp.fr 

(2) Department of Geological Sciences, San Diego State University, San Diego, CA, United States. 

(3) Institute of Earth Sciences, Hebrew University of Jerusalem, Jerusalem, Israel. 

(4) The Department of Geophysics and Planetary Sciences, Tel Aviv University, Tel Aviv, Israel 

Abstract (Testing earthquake recurrence models with 3D trenching along the Dead-Sea Transform): We propose to test 

earthquake recurrence and slip models via high-resolution three-dimensional trenching of the Betieha site on the Dead Sea 

Transform (DST) in northern Israel. We extend the earthquake history of this simple plate boundary fault to establish how often 

earthquakes have occurred in the Holocene (past 50-100 centuries), to determine the amount of slip per event, and to test 

competing rupture models (characteristic, slip-patch, slip-loading, and Gutenberg Richter type distribution). We do this by 3D 

trenching and documentation of offset buried streams of various ages across the DST. This information is critical for improving 

seismic hazard analysis and earthquake forecast models in general, and for establishing the current earthquake risk in Israel and 

the surrounding Middle East. 

Key words: Dead Sea Transform, 3D Paleoseismology, slip per event, historical earthquakes 

INTRODUCTION 

Understanding earthquake production along major 

plate-boundary faults is critical for improving seismic 

hazard assessment and earthquake forecast models. 

Models used to forecast future seismicity make 

fundamental assumptions about fault behavior, 

whether it ruptures in a random, quasi-periodic, or 

clustered pattern. Those models are based on limited 

observations of recurrent slip at a point along a fault, 

or variations in recurrence times at multiple 

paleoseismic sites along individual faults. Models 

such as the “characteristic earthquake” model 

(Schwartz and Coppersmith, 1984) or the “slip-patch” 

model (Sieh, 1996) rely on assessments of fault 

segmentation and assume that large ruptures 

terminating at invariant segment boundaries 

(Wesnousky, 2008). However the self-similarity of 

large ruptures, both in terms of magnitude and of slip 

distribution, is not clearly established thus far, as 

there is yet to be a recorded repeated large event 

since the advent of modern instrumental 

measurements. In order to better understand longterm 

earthquake recurrence there is need for 

comprehensive event records that includes 

magnitude, location and displacement data. The 

Beteiha (Bet-Zayda) site, located on the Dead-Sea 

Transform fault (DST), provides us with an 

opportunity for constructing long term record for an 

active plate-boundary via high-resolution threedimensional 

trenching. The data can be used to test 

earthquake recurrence and slip models, as well as to 

address other key issues such as slip-rate variation 

with time, or GPS-geological slip-rate disparity. 

Geological Background 

The DST is a major plate boundary and a source of 

significant hazard in the Middle East, accommodating 

the relative motion between the African and Arabian 

269 

Fig. 1: a) Generalized tectonic framework of the Middle 

East. b) The DST (on land) from the Red Sea northward to 

the East Anatolian fault (EAF) in Turkey. Slip rate data are 

shown in black (Daeron et al., 2004, Ferry et al., 2007, Le 

Beon et al., 2010, Meghraoui et al., 2003).


EARTHQUAKE 

ARCHAEOLOGY 



plates. It transfers slip northward from the oblique 

opening at the Red Sea to the Taurus-Zagros 

collision zone, and a consequence of this northward 

motion of Arabia and collision with Asia is the 

westward extrusion of Anatolia along the North 

Anatolian Fault system (Fig. 1a). The cumulative 

offset of the DST is ~105 km, representing the total 

motion between the Arabian plate and Sinai subplate 

since the mid Miocene (e.g., Freund et al. 1968; 

Garfunkel 1981). The rate of sinistral motion 

measured across the fault is estimated to be between 

3 and 7 mm/year in northern Israel (Le Beon et al. 

2008 and references therein, Figure 1b). Despite the 

rich record of historical seismicity in the Middle-East, 

not enough data exists to constrain surface rupture 

extent for historical earthquakes. Previous work 

(Ellenblum et al., 1998; Marco et al., 2005) on this 

segment of the DST obtained slip per event for the 

last 2 historical earthquakes (1759 and 1202 CE) and 

established that they indeed ruptured this far south. 

Results 

We present results from a 2 year trenching campaign 

in the Beteiha valley, on the northern shore of Lake 

Tiberias. Previous trenches in the same locale by 

Marco et al. (2005) exposed 5 buried paleo-channels 

which provided slip-per-event data for the last two 

earthquakes, as well as slip-rate estimate for the last 

5k years. In the first year, we established the 

feasibility for an extended 3D trenching project in that 

locale by excavating a 300m long fault parallel trench 

(Figure 2) and exposing at least 7 additional buried 

channels that can be used as offset markers, in 

addition to the 5 channels that were originally 

mapped by Marco et al. (2005). In order to constrain 

each paleo-channel’s age, 30 14 C samples were 

dated. The obtained dates span the range 915 years 

BP to more than 4000 years BP. A re-dating of 

charcoal from the channel (CH1) previously dated to 

5kyr using bulk soil 14 C, found it to be younger by at 

least 1kyr (4k and not 5k). Based on the new dates, 

the slip-rate estimate of 3 mm/yr based on 15m offset 

measured for CH1 is revised to 3.7 mm/yr. The ages 

of the buried channels cover the period between the 

last large event on this segment (1202 CE) and over 

4000 years ago. 

The second year’s trenching season will commence 

in May 2011, with the main goal to follow the paleochannels 

across the fault and measure the amount of 

offset for each, thus refining the slip-rate and slip-perevent 

history. We plan to have several of the 

channels excavated and mapped by July, and we will 

present our results in Corinth, 2011. 

References 

Ellenblum, R., Marco, S., Agnon, A., Rockwell, T.K., and 

Boas, A., 1998, Crusader castle torn apart by earthquake 

at dawn, 20 May 1202, Geology 26, 303–306. 

Daeron, M., Benedetti, L., Tapponnier, P., Sursock, A., and 

Finkel, R.C., 2004, Constraints on the post 25-ka slip rate 

of the Yammouneh fault (Lebanon) using in situ 

cosmogenic 36 Cl dating of offset limestone-clast fans, 

Earth Planet. Sci. Lett., 227, 105–119. 

Ferry, M., M. Meghraoui, N. Abou Karaki, M. Al-Taj, H. 

Amoush, S. Al-Dhaisat, and M. O. Barjous, 2007, A 48- 

kyr-long slip rate history for the Jordan Valley segment of 

the Dead Sea Fault, Earth Planet. Sci. Lett., 260, 394– 

406. 

Freund, R., Zak, I., and Garfunkel, Z., 1968, Age and rate of 

the sinistral movement along the Dead Sea Rift, Nature, 

220, 253– 255. 

Garfunkel, Z., Zak, I., and Freund, R., 1981, Active faulting 

in the Dead Sea rift, Tectonophysics, 80, 1-26. 

Le Béon, M., Klinger, Y., Amrat, A. Q., Agnon, A., Dorbath, 

L., Baer, G., Ruegg, J. C., Charade, O., and Mayyas, O., 

2008, Slip rate and locking depth from GPS profiles 

across the southern Dead Sea Transform, J. Geophys. 

Res., 113, B11403, doi:10.1029/2007JB005280. 

Le Beon, M., Klinger, Y., Al Qaryouti, M., Mériaux, A-S, 

Finkel, R. C., Elias, A., Mayyas, O., Ryerson, F. J., and 

Tapponnier, P., 2010, Early Holocene and Late 

Pleistocene slip rates of the southern Dead Sea Fault 

determined from 10 Be cosmogenic dating of offset alluvial 

deposits, J. Geophys. Res., 115, B11414, 

doi:10.1029/2009JB007198. 

Marco, S., Rockwell, T.K., Heimann, A., Frieslander, U., 

Agnon, A., 2005, Late Holocene activity of the Dead Sea 

Transform revealed in 3D paleoseismic trenches on the 

Jordan Gorge segment. Earth Planet. Sci. Lett., 234, 

189–205. 

Meghraoui, M., et al., 2003, Evidence for 830 years of 

seismic quiescence from palaeoseismology, 

archaeoseismology and historical seismicity along the 

Dead Sea Fault in Syria, Earth Planet. Sci. Lett., 210, 35– 

52. 

Schwartz, D. P., and K. J. Coppersmith, 1984, Fault 

behavior and characteristic earthquakes: Examples from 

the Wasatch and San Andreas fault zones, J. Geophys. 

Res., 89, 5681–5698. 

Sieh, K., 1996, The repetition of large-earthquake ruptures, 

Proc. Natl. Acad. Sci. U.S.A. 93, 3764-3771. 

Wesnousky, S.G., 2008, Displacement and geometrical 

characteristics of earthquake surface ruptures: Issues 

and implications for seismic-hazard analysis and the 

process of earthquake rupture, Bull. Seismol. Soc. Am. 

98(4), 1609-1632. 

Acknowledgements: Thanks are due to all the trenching 

assistants, especially to K. Farrington, J. B. Salisbury and 

E. Bowles-Martinez. We thank the village of Almagor for 

letting us trench in their fields, with special thanks to Avshi 

Herzog for his assistance. This project is funded by NSF 

grant EAR-1019871 to T. K. Rockwell and by a grant from 

the city of Paris to N. Wechsler and Y. Klinger. 

270


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 2: Location of first year trenches, on an aerial photograph of the field from summer 2009. Approximate location of previous 

trenches is marked with a blue rectangle. The buried paleo-channel locations are marked in the trench by letters, and the youngest 

14C calibrated date is given for each. Marco et al. (2005) followed sub-units of channel E to constrain slip for the 1202 and 1759 

events. Their CH1, used to estimate the last 5kyr slip-rate, was exposed in T1, and its approximate location is marked on the 

photo, just south of our paleo-channel F. 

271


EARTHQUAKE 

ARCHAEOLOGY 



A TERRESTRIAL CLOSE RANGE VIEW OF THE NORMAL FAULT ZONE NEAR 

ARCHANES (EAST YIOUCHTAS MT., HERAKLION BASIN, CRETE) 

Wiatr, Thomas (1), Ioannis D. Papanikolau (2, 4, Klaus Reicherter (1), Tomás Fernández-Steeger (3) 

(1) RWTH Aachen University, Neotectonics and Natural Hazards Group, Lochnerstr. 4-20, 52056 Aachen, Germany 

(2) Laboratory of Mineralogy & Geology, Department of Sciences, Agricultural University of Athens, 75 Iera Odos Str., 11855 

Athens, Greece 

(3) RWTH Aachen University, Department of Engineering Geology and Hydrogeology, Lochnerstr. 4-20, 52064 Aachen, 

Germany 


London UK 

Abstract (A terrestrial close range remote sensing view of the normal fault zone near Archanes (East Yiouchtas Mt., 

Heraklion basin, Crete): The focus of investigation in this paper is the reconstruction of different fault plane conditions and the 

primary interpretation of them, based on terrestrial close range LiDAR (Light Detection and Ranging) data. For reconstruction the 

slip per event along the bedrock scarp and where possible (e.g. postglacial scarp) an estimate regarding the slip rate of the 

individual faults we used the backscattered signal of the laser beam, the geomorphological geometry and the fault plane 

conditions. In this paper we discuss the N-S striking normal fault near Archanes in the Heraklion basin. To the north of this fault 

zone the ancient Minoan temple Anemospillia (caves of the wind) is located, which was destroyed by an earthquake around 1700 

BC and forms the nearest seismic source to the site. 

Key words: close range LiDAR, normal fault, fault scarp morphology, Crete 

INTRODUCTION 

Crete is the largest Greek island with an area of ~ 

8300 km² and approx. 900 km coastline. The 

development of a multidirectional tectonic regime on 

Crete is interpreted as a result of the Hellenic 

subduction zone in the south and the westward 

extrusion of the Anatolian plate in the north. The 

island is forms a horst structure in the Hellenic fore 

arc zone, which is also influenced by the roll back of 

the African plate. Rapid uplift of ~ 1.2 mm/yr can be 

observed on the entire island (Meulenkamp et al., 

1994). Crete has been uplifted since the Middle 

Miocene from 1 up to 2 km depending on the 

influence of different tectonic blocks. The island of 

Crete lies on top of the active subduction zone for 

about 30 ma years, implying that it experiences high 

strain rates and constant deformation processes 

(Papanikolaou, 1993). Crete is characterised by a 

complex geological and tectonic structure that results 

from: i) the successive thrusting of the alpine 

geotectonic units on top of each other (Bonneau, 

1984), ii) the activity of major detachment faults 

(Fasoulas et al., 1994, Papanikolaou and Vassilakis, 

2010, Zachariasse et al., 2011), iii) by the intense 

neotectonic and active faulting (Monaco and 

Tortorici, 2004, Peterek and Schwarze, 2004, Caputo 

et al., 2010). Crete is located in a high seismicity 

area. Over the last 40 years the active Hellenic 

subduction zone produced earthquakes in a depth 

range from 18 to 162 km with magnitudes of M = 4.9 

up to 6.1 (Benetatos et al., 2004). Hypocentral 

depths of earthquakes showed that the north dipping 

Wadati-Benioff seismic zone close to the low angle 

subduction along the convex side of the Hellenic arc 

trench is located in a depth of around 60 to 90 km 

near Crete (Papazachos et al., 2000). But this region 

has also experienced strong thrusting 

paleoearthquakes with magnitudes up to M > 7.5 - 

8.0 and hence one of the most intense seismic 

activity area of the Aegean region (Papazachos & 

Papazachou, 1997). 

Decoding paleoearthquakes in fault bedrock scarps 

is important for seismic hazard assessment. Shallow 

earthquakes greater than M S 6 can produce an 

imprint in the landscape named fault scarps (Stewart 

& Hancock, 1990). Bedrock fault scarps are 

indicators of large surface faulting events and may 

provide not only slip rates, but also information on 

slip per events when they are analysed with 

cosmogenic isotope dating (Benedetti et al., 2003). 

Fault scarps are preserved in the landscape when 

the slip rate is greater than the erosion rate. 

Therefore, these are regarded as postglacial scarps 

that were formed since the last glaciation (Benedetti 

et al., 2002). The Neogene fault plane solutions and 

the seismic activities indicate large earthquakes and 

a rapid uplift with complex tectonic settings of Crete 

(Dewey & Sengör, 1979; Papazachos et al., 1987). 

This paper is focused on the N-S striking fault zone in 

the Heraklion basin to the south of Knossos (Fig.1). 

This basin is characterized by block tectonics and the 

Yiouchtas Mt. represent a neotectonic horst structure 

which is subdivides the Heraklion Basin in a western 

and an eastern subbasin (Papanikolaou and 

Nomikou 1998). Major aims of the investigation were 

to find quantitative and qualitative data for the 

reconstruction of surfaces and to analyze the tectonic 

geomorphology and paleoseismicity of active faults 

with terrestrial laser scanning (TLS) to reconstruct 

fault history and activity along surface rupturing 

scarps. 

The varying scale of structural heterogeneity and 

discontinuous geometry of the exhumed foot wall slip 

plane along a fault zone and the complexity of the 

surface features like the subslip-plane breccia sheet, 

brecciated colluvium or frictional water-wear 

striations on the rupture plane, makes it difficult to 

recognize the paleoevents on the fresh fault scarp 

272


EARTHQUAKE 

ARCHAEOLOGY 



above the level of exhumation (Stewart & Hancock, 

1991; Roberts, 1996). Even, more the detail 

geometrical characterisation of the slip surface 

depends on the view direction to the strike-slip 

direction (parallel or perpendicular), the calculation 

approaches and the scaling size of the analysis, 

because the anisotropy properties and fractal 

dimensions of fault morphology are decisive 

(Mandelbrot, 1985; Fardin et al., 2001, 2004; 

Rahman et al., 2006; Renard et al., 2006; Sagy et al., 

2007, 2009; Candela et al., 2009). 

Additionally, several time-dependent and overlapping 

processes influence the condition of the free face 

fault plane that become degraded. These processes 

involve weathering, pedogenesis of the unbrecciated 

colluvium, vegetation, karstification and erosion of 

the colluviums and the fault outcrop. 

PRIMARY RESULTS OF THE LiDAR 

INVESTIGATION ON THE NORMAL FAULT ZONE 

NEAR ANEMOSPILIA "CAVES OF THE WIND" 

(1900-1700 BC MM II/III) IN THE HERAKLION 

BASIN 

In the northern part of the Mount Juktas about 7 km 

south of Knossos, the legendary area and tomb of 

Cretan Zeus is situated, which is considered to be 

one of the earliest Minoan temples: Anemospilia 

(Fig.1). 

The three small rooms, each of them opened into a 

corridor, were described and discovered by 

Sakellarakis during the 1979 expedition. They 

assumed that the temple was destroyed by a 

sequence of large earthquakes around 1700 BC and 

based this conclusion on pottery and artefacts (Nur, 

2008). Furthermore, they found a skeleton with 

broken legs under an ash layer implying that the 

earthquake was strong enough to damage the 

massive temple and was followed by a fire. Five 

different close range LiDAR scans were made in the 

middle of the N-S striking normal fault 2 km south of 

the Anemospilia temple (Fig.2). The free natural fault 

plane in the interesting area is around 6 m high and 

the scanned area is around 30 m wide. The whole 

outcrop is by this location around 70 m continuous 

wide. Our primary goal in this study was to use the 

TLS for fault tectonomorphology reconstruction. The 

TLS data have a point to point range between 3 and 

7 mm. The examples in this paper (Fig.3/4) had 

around 2.8 and 3.8 million numbers of shots (points) 

and the average range between LiDAR and defined 

scan window was 10 m. This allowed a spatial 

reconstruction of the scan sequence without gaps 

and ensured a good data quality and spatial 

resolution for the interpolation between the points 

and for the analysis of the plane morphology. The 

detailed structural analysis of rock surfaces has 

shown that the surface conditions are changing from 

base to top (Fig.4). The hillshades in figure 3b and 4b 

illustrates the plane morphology with different karstic 

features and degradation in the upper part of the 

scarp plane (variance of rougher surface conditions 

in section III, IV, V, VI in Fig.3). 

Fig. 1: Investigation area near Archanes in the Heraklion 

basin of Crete including earthquakes recorded since 

1979 (a) (SRTM, USGS, www.usgs.gov) and the main 

faults, the LiDAR position, location and looking direction 

of photos (see Fig.2), dip direction of the scanned fault 

plane, recent stress field and ancient sites (b) (modified 

from Fassoulas, 2001; ten Veen & Meijer, 1998). 

273 

Fig. 2: Photo of the east dipping continuous normal fault zone 

western Archanes and eastern Mount Yiouchtas within the 

ancient Minoan site Anemospilia in the northern part of 

Mount Juktas (a). b) Zoom in photo of the fault plane. 

Parts I and II of this section have no significant 

karstification. The sections are dominated by 

striation, slickensides and small fractures (compare 

Fig.4). The combination of the plane morphology with 

the detected backscattered signal of the LiDAR is 

shown in figure 3d/3e and 4d. By using this 

technique, a different point of view allowed that the 

detected near infrared laser signal can be used for 

the classification of different functions of weathering, 

morphological and erosion features on the fault plane


EARTHQUAKE 

ARCHAEOLOGY 



and also reveals the exhumation history. The signal 

intensity can be used to identify vegetation (including 

lichen) and its influence on the fault plane. The 

results shown the influence of the colluvium on the 

base of the scarp (I in Fig.3e), which can be detected 

by a change of the backscattered intensity. 

Furthermore it is possible to distinguish between the 

different parts of the fault plane, which are 

characterised and dominated by 1) degradation; 2) 

karstification; 3) slickensides; and 4) the influence of 

the colluvium. Following Giaccio (2002) and the 

model of natural free normal bedrock scarp, we 

identified different weathering mircomorphologies 

depending on the scarp height (Fig.4e). The different 

sections with determinates features on the fault plane 

represent the time-dependent fault scarp evolution 

(Stewart, 1996). 

Fig.4: a) photo of the fault plane includes the LiDAR scan 

window, b) hillshade of the fault plane for geometrical and 

morphological analysis, c) dip direction of the plane, d) 

distribution of backscattered signal with primary surface 

interpretation, e) brown, red, green and yellow boxes are 

demonstrate the different surface conditions on the fault 

plane in depending on the scarp height. 

Fig. 3: Primary results of the LiDAR investigation based 

on the model in Fig.1. a) photo of the fault plane includes 

the LiDAR scan window, b) hillshade of the fault plane for 

geometrical and morphological analysis, c) dip direction of 

the plane, d) distribution of the backscattered signal with 

primary surface interpretation, e) primary interpretation of 

all fault plane conditions. 

The time- and height-dependent features of bedrock 

fault scarps are shown figure 4e. The boxes illustrate 

examples of the fault plane morphology from bottom 

to top (young to old; brown, red, green, yellow). 

Conspicuous is the increasing roughness from young 

to old (brown to yellow) and the specific surface 

features in different heights. The reason could be the 

different bio-karstic, bio-erosional, physical and 

biochemical processes which depend on time 

(Giaccio et al., 2002). The brown box shows the 

striation in the lower part of the fault. The red box 

illustrates small fractures and a rougher surface than 

in the brown box. The green box the gradation of the 

karstification and the yellow box show the rillen karst 

in an advanced stage. Close range LiDAR 

investigation on postglacial natural normal fault 

scarps has shown that reconstruction of the spatial 

distribution of different plane evolution indicators is 

possible. 

These fundamental phenomena can be realized by 

and imaged with a high resolution digital elevation 

model (HRDEM) in combination with the 

backscattered laser impulse. The primary 

interpretations of the fault surface conditions and 

their interaction are described in figure 3. Based on 

the LiDAR results and the field survey we created a 

principal model of fault scarp alteration for the East 

Yiouchtas fault, following the general model of 

Giaccio et al. 2002 (Fig. 5). 

Fig. 5: The model of a normal fault with fault plane 

evolution indicators (modified from Giaccio et al. 

2002). 

274


EARTHQUAKE 

ARCHAEOLOGY 



The investigation with LiDAR and the field survey has 

shown that the eastern Yiouchtas fault is active with 

several events in the post-glacial time period. 

References 

Benedetti, L., Finkel, R., King, G., Armijo, R., 

Papanastassiou, D., Ryerson, F., Flerit, F., Farber, D. & 

Stavrakakis, G., 2003: Motion on the Kaparelli fault 

(Greece) prior to the 1981 earthquake sequence 

determined from 36Cl cosmogenic dating. Terra Nova, 

15:118-124. 

Benedetti, L., Finkel, R., Papanastassiou, D., King, G., 

Armijo, R., Ryerson, F., Farber, D. & Flerit, F., 2002: 

Post-glacial slip history of the Sparta Fault (Greece) 

determined by 36 Cl cosmogenic dating: evidence for nonperiodic 

earthquakes. Geophys. Res. Lett., 29: 87-1–87- 

4. 

Bonneau, M., 1984: Correlation of the Hellinides nappes in 

the south-eastern Aegean and their tectonic 

reconstruction. Special Publication of the Geological 

Society of London 17, 517-527 

Benetatos, C., Kiratzi, A., Papazachos, C., & Karakaisis G., 

2004: Focal mechanisms of shallow and intermediate 

depth earthquakes along the Hellenic Arc. J. Geodyn., 

37, 253-296, doi:10.1016/j.jog.2004.02.002. 

Candela, T., Renard, F., Bouchon, M., Brouste, A., Marsan, 

D., Schmittbuhl, J. & Voisin, C., 2009: Characterization of 

fault roughness at various scales: implications of threedimensional 

high resolution topography measurements, 

Pure appl. geophys. doi: 10.1007/s00024-009-0521-2. 

Caputo, R., Catalano, S., Monaco, C., Romagnoli, G., 

Tortorici G. & Tortorici, L., 2010: Active faulting on the 

island of Crete (Greece). Geophys. J. Int., 183, 111–126, 

doi: 10.1111/j.1365-246X.2010.04749.x. 

Dewey, J.F. & Sengör, A.M.C., 1979: Aegean and 

surrounding regions: Complex multiplate and continuum 

tectonics in a convergent zone. Bull. geol. Soc. Am., 90, 

84-92. 

Fardin, N., Stephansson, O. & Jing, L., 2001: The scale 

dependence of rock joint surface roughness. Int. J. Rock 

Mech. & Min. Sci., 38, p. 659-669. 

Fardin, N., Feng, Q. & Stephansson, O., 2004: Application 

of a new in situ 3D laser scanner to study the scale effect 

on the rock joint surface roughness. Int. J. Rock Mech. & 

Min. Sci., 41, p. 329-335. 

Fassoulas, C., 2001: The tectonic development of a 

Neogene graben at the leading edge of the active 

European margin, The Heraklion basin, Crete, Greece, J. 

Geodyn., 31, 49–70, doi:10.1016/S0264-3707(00)00017- 

X. 

Giaccio, B., Galadini, F., Spasato, A., Messina, P., Moro, 

M., Zreda, M., Cittadini, A., Salvi, S. & Todero, A., 2002: 

Image processing and roughness analysis of exposed 

bedrock fault planes as a tool for paleoseismological 

analysis: results from the Campo Felice fault (central 

Apennines, Italy). Geomorphology, 49, 281-301. 

Mandelbrot, B.B., 1985: Self-affine fractals and fractal 

dimension, Phys. Scripta, v. 32, p. 257-260. 

Meulenkamp, J.E., van der Zwaan, G.J., van Wamel, W.A., 

1994. On late Miocene to recent vertical motions in the 

Cretan segment of the Hellenic arc. Tectonophysics, 234, 

53-72. 

Monaco, C. & Tortorici, L., 2004: Faulting and effects of 

earthquakes on Minoan archaeological sites in Crete 

(Greece). Tectonophysics, 382, 103-116. 

Nur, A., Burgess, D., 2008: Apocalypse: Earthquakes, 

Archaeology and the Wrath of God. Princeton University 

Press. Princeton and Oxford. 309 p. 

Papanikolaou, D. (1993). Geotectonic evolution of the 

Aegean. Bull. Geol. Soc. Greece, 28/1, 33-48. 

Papnikolaou, D.J. & Nomikou, P.V., 1998: Neotectonic 

blocks and planation surfaces in Iraklion basin, Crete, 

Greece. Bull. Geol. Soc. Greece, XXXII/1, 231-239. 

Papanikolaou, D. and Vassilakis, E. (2010). Thrust faults 

and extensional detachment faults in Cretan tectonostratigraphy: 

Implications for Middle Miocene extension. 

Tectonophysics 488, 233-247 

Papazachos, B.C., Papadimitriou, E.E., Kiratzi, A.A., 

Papaioannou, C.A. & Karakaisis, G.F., 1987: Probabilities 

of Occurrence of Large Earthquakes in the Aegean and 

Surrounding Area During the Period 1986-2006. 

Pageoph., 125, 4, 597-612. 

Papazachos, B. & Papazachou, C., 1997: The Earthquakes 

of Greece. Zitis, Thessaloniki. 

Papazachos, B.C., Karakostas, V.G., Papazachos, C.B. & 

Scordilis, E.M., 2000: The geometry of the Wadati– 

Benioff zone and lithospheric kinematics in the Hellenic 

arc. Tectonophysics, 319, 275-300. 

Peterek, A. & Schwarze, J., 2004: Architecture and Late 

Pliocene to recent evolution of outer-arc basins of 

Hellenic subduction zone (south-central Crete, Greece). 

J. Geodyn., 38, 19-55. 

Rahman, Z., Slob, S. & Hack, R., 2006: Deriving roughness 

characteristics of rock mass discontinuities from 

terrestrial laser scan data. Proceedings of the 10th IAEG 

Congress, Engineering geology for tomorrow's cities, 

Nottingham, United Kingdom, 6-10 September 2006. 

Engineering geology for tomorrow's cities. 

Renard, F., Voisin, C., Marsan, D. & Schmittbuhl, J., 2006: 

High resolution 3D laser scanner measurements of a 

strike-slip fault quantify its morphological anisotropy at all 

scales, Geophys. Res. Lett. 33, L04305, doi: 

10.1029/2005GL025038. 

Roberts, G.P., 1996: Variation in fault-slip directions along 

active and segmented normal fault systems. J. Struct. 

Geol., Vol. 18, No. 6, 835-845. 

Sagy, A., Brodsky, E.E. & Axen, G.J., 2007: Evolution of 

fault-surface roughness with slip. Geology, 35: 283-286. 

Sagy, A. & Brodsky, E.E., 2009: Geometric and rheological 

asperities in an exposed fault zone. J. Geophys. Res., 

114, B02301, doi: 10.1029/2008JB005701. 

Sakellarakis, Y. & Sapouna-Sakellaraki, E., 1981: Drama of 

death in a Minoan temple. National Geographic, February 

1981. 

Stewart, I.S. & Hancock, P.L., 1990: Brecciation and 

fracturing within neotectonic normal fault zones in the 

Aegean region. In: Knipe, R.J. and Rutter, E.H., (eds.) 

Deformation Mechanisms, Rheology and Tectonics. 

Geol. Soc. Spec. Publ., 54: 105-112. 

Stewart, I.S. & Hancock, P.L., 1991: Scales of structural 

heterogeneity within neotectonic normal fault zones in the 

Aegean region. J. Struct. Geol., Vol. 13, No. 2, 191-204. 

Stewart, I., 1996: A rough guide to limestone fault scarps. J. 

Struct. Geol., 18,1259-1264. 

ten Veen, J.H. & Meijer, P.T., 1998: Late Miocene to Recent 

tectonic evolution of Crete (Greece): geological 

observations and model analysis. Tectonophysics, 298, 

191-208. 

Zachariasse, W.J., van Hinsbergen, D.J.J., & Fortuin, A.R., 

2011: Formation and fragmentation of a late Miocene 

supradetachment basin in central Crete: implications for 

exhumation mechanisms of high-pressure rocks in the 

Aegean forearc. Basin Research, doi: 10.1111/j.1365- 

2117.2011.00507.x. 

275


EARTHQUAKE 

ARCHAEOLOGY 



THE DISCONTINUITY OF A CONTINUOUS FAULT: DELPHI (GREECE) 

Wiatr, Thomas (1), Klaus Reicherter (1), Ioannis D. Papanikolau (2, 3) & Tomás Fernández-Steeger (4) 

(1) RWTH Aachen University, Neotectonics and Natural Hazards Group, Lochnerstr. 4-20, 52056 Aachen, Germany 

(2) Laboratory of Mineralogy & Geology, Department of Sciences, Agricultural University of Athens, 75 Iera Odos Str., 11855 

Athens, Greece. papanikolaou@ucl.ac.uk 


London UK. papanikolaou@ucl.ac.uk 

(4) RWTH Aachen University, Department of Engineering Geology and Hydrogeology, Lochnerstr. 4-20, 52064 Aachen, 

Germany. fernandez-steeger@lih.rwth-aachen.de 

Abstract (The discontinuity of a continuous fault: Delphi (Greece): We used a terrestrial laser scanning system for the 

reconstruction and analysis of the morphotectonic features on a fault segment near the ancient Delphi, Greece. Delphi is located 

on the northern part of the Gulf of Corinth and embedded in a seismic landscape and is the major onshore fault north of the 

Corinth Gulf. This paper concentrates on the LiDAR long range investigation of the fault for generating a digital elevation model. 

The model was used for estimating the natural fault plane height by using several vertical profiles. The results show a high 

horizontal and vertical variability on a 120 m long fault scarp. 

Key words: long range LiDAR, discontinuous fault, Corinth Gulf, Delphi 

INTRODUCTION 

The ancient Delphi with its oracle dedicated to the 

god Apollo, was the most popular place of worship 

around 700 B.C. to 400 A.D. in ancient Greece. 

Delphi is located at the southern flank of Mount 

Parnassus that is compiled by thick-bedded neritic 

Mesozoic limestones with significant bauxite 

deposits. The mountain range is situated on the 

northern coast of the Gulf of Corinth and bounded by 

an active fault zone that dips southwards (Fig.1). 

Ambraseys & Jackson, 1998; Papazachos & 

Papazachou, 2003 and Pavlides & Caputo, 2004). 

Piccardi (2000) described the 373 B.C. earthquake, 

which partly destroyed the ancient Delphi showing 

that there is a post-earthquake reconstruction phase 

at the shrine of Athena (located around 500 m east of 

the Temple of Apollo) (Fig.2). 

Fig. 2: Setting of the archaeological site of the shrine 

of Athena Pronaia (modified from Piccardi, 2000). 

Fig. 1: Tectonic and topographic overview map of the 

Delphi area. 

The Gulf of Corinth is a graben like tectonic structure 

with E-W trending normal faults and characterized 

one of the fastest extending regions worldwide with 

up to 20 mm/yr rate (Billiris et al., 1991; Briole et al., 

2000). Destructive historical earthquakes in the area 

are reported for 373 B.C. (Piccardi, 2000), 515 A.D. 

(Ambraseys & Jackson, 1998; Papazachos & 

Papazachou, 2003) and 1870 (1 st of August, Ms=6.7, 

Furthermore, he postulates that the Temple of 

Athena was relocated from its original position. 

Aim of our study was the reconstruction of the 

morphotectonic features of the Delphi fault scarp, 

located 2 km west of the Temple of Apollo by using a 

terrestrial remote sensing technique. We scanned a 

120 m long fault scarp with long range LiDAR. We 

used a terrestrial LiDAR (Light Detection and 

Ranging) system from Optech Inc. 

METHODS 

The ground-based LiDAR (Light Detection And 

Ranging) or TLS (terrestrial laser scanning) remote 

sensing method has been established as a versatile 

data acquisition tool in photogrammetry, engineering 

technologies, atmospheric studies and as a good 

data acquisition tool in geosciences and geological 

engineering in difficult accessible areas. 

276


EARTHQUAKE 

ARCHAEOLOGY 



As the TLS has a high spatial and temporal 

resolution it is an effective remote sensing 

technology for reconstruction, monitoring and 

observation of geosciences phenomena. 

The fundamental principle of ground-based LiDAR is 

to generate coherent a laser beam with little 

divergence by stimulated emission. LiDAR is a 

contact- and destructionless, non-penetrative active 

recording system which is stationary during the 

recording. The electromagnetic waves are reflected 

by surfaces and the receiver detects portions of the 

backscattered signal. All scan sequences were 

mapped with first pulse detection mode. The laser 

ranging system is based on measuring the time-offlight 

(two-way travel time) of the short wave laser 

signal. Advantages of the terrestrial method are the 

flexible handling, a relatively quick availability of an 

actual dataset, and a very high spatial resolution of 

the object with information about intensity, x-y-zcoordinates 

and range. The combination with a digital 

camera allows combining the point cloud with 

panchromatic information in order to achieve 

additionally the RGB colour-coding. 

For the morphological analysis twelve vertical profiles 

with 10 m distance to each other were generated 

from the LiDAR data. All profiles start on the same 

height level on top of the fault zone on the foot wall 

and are perpendicular to the fault plane towards the 

hanging wall. 

Results show a variation of the free face fault height 

with a difference of 5.4 m. The height data range of 

the natural bedrock fault plane is between 4 and 9.4 

m. Furthermore, horizontal variations of 3 m were 

detected (Fig.5). 

Laser scanning allows 3D surface data acquisition, 

which is specifically characterized by a digital data 

record and a computerized data analysis. 

Furthermore, an implementation of the dataset in a 

geographical information system (GIS) is 

uncomplicated with accurate digital elevation models 

(DEM) or digital terrain models (DTM) sourced 

directly from the raw dataset. 

The infrared laser scanner detected the 

monochromatic information of the backscattered 

intensity in 256 grey values. The information of the 

monochromatic wavelength, the detected 

backscattered intensity, reflexes the surface 

properties in the near infrared range. This 

wavelength is invisible for human eyes. Hence, the 

results show a different kind of view of the surface 

conditions. The quality of the reflection depends on 

the inclination angle of the laser beam, the range 

between the object and scanner, the material, the 

colour, the surface condition (weathering/roughness), 

and the spatial resolution. 

DISCONTINUOUS FAULT SEGMENT ON A 

CONTINUOUS FAULT NEAR DELPHI 

The scan position for the long range investigation 

was 250 m south of the fault plane (Fig.3). The raw 

data point cloud includes around 6.2 million points for 

an 130 m long and 60 m wide scan window. For this 

study a point resolution of 2 cm was chosen. 

Moreover, we recorded 9 close range scans with 

4 mm point resolution (Fig.4). 

After data validation and cleaning, the point cloud 

has been geo-referenced and imported into a GIS. In 

the GIS, the scans have been converted in a 

triangulated irregular network (TIN) and in a raster 

format. With the grid format it is possible to calculate 

the basic applications for morphologic specifications. 

Fig. 3: Panchromatic images of the investigation area. 

a) The geoeye satellite image shows the continuous 

fault by Delphi and includes the long range LiDAR 

study area (modified from Google earth). b) Photo of 

the south dipping continuous fault zone 2 km westward 

from the ancient shrine of Apollo by Delphi. 

We found that the Delphi fault has an oscillation in 

three dimensions (horizontal and vertical 

displacement), across the 120 m long scan 

sequence. The absolute elevation range of all profiles 

on the natural free face fault surface (vertical range 

of the fault plane) is between 574.5 m and 588.7 m 

above sea level (see Fig.5 right site number 1) and 

the horizontal variation is between 11.2 m and 21 m 

(see Fig.5 number 1 under the profiles). The 

horizontal and vertical value range (number 1) 

includes the variation of the knick point on top of the 

free face to the foot wall (number 2) and the knick 

point of the bottom of the free face to the hanging 

wall (number 3). Furthermore, the results had shown 

a variability of the alluvial deposits (vertical and 

horizontal displacement) of the hanging wall (number 

3). 

It turns out that the long term slip rate for postglacial 

scarps (last 19 ka ± 3 ka) in this case ranges 

between 0.21 ± 0.04 mm/yr and 0.49 ± 0.09 mm/yr 

277


EARTHQUAKE 

ARCHAEOLOGY 



depending on the fault segment analysed, and 

between 0.3 and 0.72 mm/yr for the last 13 ka (The 

13 ka pertains to Benedetti et al., 2002; 2003). 

Hence, the long term slip rate had a range between 

0.25 ± 0.05 mm/yr and 0.58 ± 0.09 mm/yr for the last 

16 ka ± 3ka. This implies for the long term slip rate 

estimation a variation of 0.28 mm/yr (19 ka ± 3 ka 

(0.4 mm/yr worst case scenario)) and 0.42 mm/yr (13 

ka), which is enormous. 

CONCLUSION 

Post-glacial throw variation along strike is evident, 

even over short distances, including locations that 

are apparently undisturbed by incision or deposition 

processes (Papanikolaou et al., 2005). This is the 

natural variation associated with coseismic surface 

slip and constitutes a major source of uncertainty that 

in central Apennines was measured at ± 20% 

(Papanikolaou et al., 2005). A fairly irregular surface 

slip distribution has been documented from several 

normal faulting events such as the 1981 Alkyonides 

earthquake sequence (Jackson et al., 1982). Herein, 

in Delphi we measured a higher variability of around 

± 28 %. Processes like erosion, deposits and debris 

as well as catchments analysis are not including in 

this research until now. 

But the LiDAR investigation on a fault scarp segment 

in Delphi has shown a massive variation in the 

vertical and horizontal displacement. The data 

collection for calculation of the long term slip rate with 

the profile method, it is necessary to produce a lot of 

profiles in the field, to get an impression of the fault 

variation. Due to the digital elevation model of the 

long range LiDAR data with 2cm spatial resolution 

and without gaps it is possible to get a high data 

quality for morphological analysis and ensured a 

documentation in a tectonic environment. 

Fig. 4: Close and long range terrestrial LiDAR 

investigation for morphological analysis on the 

southwards dipping continuous bedrock fault scarp 

2 km west of the archaeological site of Delphi (a). b) 

Illustrates the calculation of the largest postglacial 

vertical displacement in the scan window on one 

profile. c) Fault plane dip direction derived from 

structural mapping in the scan area. d) Fault plane 

reconstruction based on LiDAR data was realised by 

a hillshade in a GIS environment. e) Photo of the fault 

plane including the LiDAR scan window. f) Hillshade 

of the close range fault plane and g) distribution of 

backscattered signal with primary surface 

interpretations. 

Fig. 5: Twelve perpendicular profiles from the 120 m 

long Delphi scarp segment show the horizontal and 

vertical discontinuity of the fault plane. 1) Variations of 

the free face fault plane. 2) Top knick point of the free 

face to the foot wall. 3) Bottom knick point of the free 

face to the hanging wall. 

References 

Ambraseys, N.N. & Jackson, J.A., 1998: Faulting 

associated with historical and recent earthquakes in the 

Eastern Mediterranean region. Geophysical Journal 

International, 133, 390–406. 

278


EARTHQUAKE 

ARCHAEOLOGY 



Benedetti, L., Finkel, R., Papanastassiou, D., King, G., 

Armijo, R., Ryerson, F., Farber, D. & Flerit, F., 2002: 

Post-glacial slip history of the Sparta Fault (Greece) 

determined by 36 Cl cosmogenic dating: evidence for nonperiodic 

earthquakes. Geophys. Res. Lett., 29: 87-1–87- 

4. 

Benedetti, L., Finkel, R., King, G., Armijo, R., 

Papanastassiou, D., Ryerson, F., Flerit, F., Farber, D. & 

Stavrakakis, G., 2003: Motion on the Kaparelli fault 

(Greece) prior to the 1981 earthquake sequence 

determined from 36Cl cosmogenic dating. Terra Nova, 

15:118-124. 

Billiris, H., Paradissis, D., Veis, G., England, P., 

Featherstone, W., Parsons, B., Cross, P., Rands, P., 

Rayson, M., Sellers, P., Ashkenazi, V., Davison, M., 

Jackson, J. & Ambraseys, N., 1991: Geodetic 

determination of tectonic deformation in central Greece 

from 1900 to 1988. Nature, 350, 124– 129, 

doi:10.1038/350124a0. 

Briole, P., A. Rigo, H. Lyon-Caen, J. C. Ruegg, K. 

Papazissi, C. Mitsakaki, A. Balodimou, G. Veis, D. 

Hatzfeld, & Deschamps, A., 2000: Active deformation of 

the Corinth rift, Greece: Results from repeated Global 

Positioning System surveys between 1990 and 1995, J. 

Geophys. Res., Solid Earth, 21, 25, 605–625. 

Jackson, J., Gagnepain, A., Houseman, J., King, G., 

Papadimitriou, G.C.P., Soufleris, P. & Virieux, C., 1982: 

Seismicity, normal faulting, and the geomorphological 

development of the Gulf of Corinth (Greece): the Corinth 

earthquakes of February and March 1981. Earth and 

Planetary Science Letters 57, 377– 397. 

Papanikolaou, I.D., Roberts, G.P. & Michetti, A.M., 2005: 

Fault scarps and deformation rates in Lazio-Abruzzo, 

Central Italy: comparison between geological fault sliprate 

and GPS data. Tectonophysics 408, 147–176. 

Papazachos, C. & Papazachou, C., 2003: The Earthquakes 

of Greece. Ziti, Thesaloniki [in Greek]. 

Pavlides, S. & Caputo, R. 2004. Magnitude versus faults’ 

surface parameters: quantitative relationships from the 

Aegean Region. Tectonophysics, 380, 159–188. 

Piccardi, L., 2000: Active faulting at Delphi: seismotectonic 

remarks and a hypothesis for the geological environment 

of a myth. Geology, 28, 651–654. 

279


EARTHQUAKE 

ARCHAEOLOGY 



POSTSEISMIC DEFORMATION OF THE 2009 L’AQUILA EARTHQUAKE (M6.3) SURFACE 

RUPTURE MEASURED USING REPEAT TERRESTRIAL LASER SCANNING 

Wilkinson, Maxwell (1, Ken McCaffrey (1), Gerald Roberts (2), Patience Cowie (3), Richard Phillips (4) 

(1) Department of Earth Sciences, Durham Univerity, United Kingdom. Email: maxwell.wilkinson@durham.ac.uk 

(2) School of Earth Sciences, Birkbeck College, University of London, United Kingdom. 

(3) School of Geosciences, University of Edinburgh, United Kingdom. 

(4) Institute of Geophysics and Tectonics, University of Leeds, United Kingdom 

Abstract (Postseismic deformation of the 2009 L’ Aquila Earthquake (M6.3) surface rupture measured using repeat 

terrestrial laser scanning): We conducted an innovative survey using repeat terrestrial laser scan (TLS) technology at four sites 

on the surface rupture of the 2009 L’Aquila earthquake (M6.3), Central Italy. Between 8 – 126 days after the earthquake we 

repeatedly laser scanned four road sections cross-cut and vertically offset by the surface rupture. A method was developed to 

quantify the postseismic deformation. We modelled rupture afterslip and associated near-field postseismic deformation in the 

hangingwall at each site with millimetre to sub-centimetre precision. The observed postseismic deformation coincides with the 

coseismic slip deficit within the fault zone, which we suggest is the driving mechanism for afterslip and near-field postseismic 

deformation. Repeat TLS survey of actively deforming surface ruptures provide a new method to monitor and quantify postseismic 

deformation. 

Key words: Postseismic deformation, Laser scanning. 

INTRODUCTION 

We report the use of Terrestrial Laser Scan (TLS) 

technology to monitor near field postseismic 

deformation at four sites along the surface rupture of 

the 6th April 2009 L’Aquila earthquake, in the 

Abruzzo region, Apennines, central Italy. The main 

shock (Mw 6.3) occurred at 03:32 local time. The city 

of L’Aquila and its surrounding suburbs were 

subjected to the largest intensity shaking, resulting in 

308 deaths, 1,500 injuries and over 50,000 people 

made homeless. 

discontinuous surface rupture, ~ 12 km in length 

(Falcucci et al. 2009) along the Paganica fault with 

normal sense displacement, down thrown to the SE, 

with a maximum throw of 0.1 m as defined by the 

EMERGEO working group (2009) and Vittori et al. (in 

press). 

Fig. 1: Riegl LMS-z420i laser scanner set up at site PAG, 

one of the four surface rupture study sites. 

The source of the L’Aquila seismicity was identified 

as the Paganica fault, to the NE of the city of 

L’Aquila, defined by focal mechanisms, aftershock 

distribution (Chiarabba et al. 2009) and differential 

interferometry. The earthquake created a 

280 

Fig. 2: Modelled postseismic deformation and rupture 

afterslip at PAG for the various time intervals. (From 

Wilkinson et al. 2010) 

Between 8 – 124 days after the earthquake we 

repeatedly laser scanned four road sections crosscut 

and vertically offset by the surface rupture (Fig. 

1). By comparing each subsequently acquired 

dataset to the first at each site, we were able to


EARTHQUAKE 

ARCHAEOLOGY 



model the resultant near-field postseismic 

deformation and rupture afterslip over various time 

intervals with millimetre to decimetre precision. We 

used reflective targets to measure the horizontal 

component of deformation. 

DISCUSSION 

We observe progressive surface deformation in our 

modelled datasets whose rate decreases over time. 

We present the modelled data from site PAG in figure 

2 as this is site with the greatest magnitude of 

deformation. The three equations shown in figure 3 

are used to describe afterslip and postseismic 

deformation for three different studies of the 

Guatemala (1) and Superstition Hills (2 & 3) 

earthquake surface ruptures. The three equations 

differ slightly as they were derived using a mix of 

theoretical and empirical approaches, as well as 

datasets from different earthquakes or study sites. In 

figure 4 we compare these three equations with their 

earthquake specific parameters to our measurements 

for the L’Aquila earthquake derived from figure 2. We 

find that our data, with a decaying rate over time is 

consistent with previously published theoretical and 

empirical laws derived to explain afterslip 

phenomenon. 

At the study site PAG, we observe surface 

deformation of significant magnitude in the form of a 

developing syncline in the hangingwall, up to 30 m 

from the surface rupture. We interpret this 

deformation as the signal of shallow afterslip in the 

fault zone. Unpublished data from the three 

remaining study sites, plus a supplementary dataset 

of total station line of sight measurements from a fifth 

site show similar results, confirming postseismic 

deformation along the surface rupture is attributable 

to afterslip within the fault zone. 

We note our study sites experienced significant 

postseismic deformation, and are located above a 

zone of coseismic slip deficit within the fault zone 

(Fig. 5, after Cheloni et al. 2010). The correlation 

between a coseismic slip deficit in the fault zone and 

significant postseismic deformation at the surface 

suggests the coseismic slip deficit as the driving 

mechanism for afterslip and near-field deformation. 

Fig. 3: Theoretical and empirical afterslip models with 

parameters obtained from afterslip datasets of previous 

earthquakes. 

Fig. 4: Surface motion measurements of rupture throw 

(purple diamonds), hangingwall syncline subsidence (pink 

squares) and extension between reflective targets (green 

triangles), derived from the modelled data in Fig 2. Our 

measurements are compared to data for the three 

theoretical and empirical models for afterslip shown in figure 

3. (Figure from Wilkinson et al. 2010) 

Fig. 5: Coseismic fault slip map derived from coseismic 

GPS data. The contours represent coseismic slip (m). Note 

the location of study site PAG at the surface and the 

coseismic slip deficit (white patch) within the fault zone 

beneath it. (Figure adapted from Cheloni et al, 2010) 

CONCLUSIONS 

Repeat TLS survey of actively deforming surface 

ruptures provide a new method to monitor and 

quantify postseismic deformation. We have 

measured near-field postseismic deformation at four 

sites along the surface rupture of the L’Aquila 

earthquake. We observe rupture afterslip and 

hangingwall deformation whose decay in rate over 

time is consistent with field observations from other 

earthquakes, as well as theoretical and empirical 

laws derived to explain such phenomenon. We 

interpret the near-field postseismic deformation as 

the signal of afterslip within the fault zone. We note 

the correlation between the location of our sites 

experiencing significant postseismic deformation and 

a coseismic slip deficit within the fault zone beneath 

them. We suggest that the coseismic slip deficit is the 

driving mechanism for near-field postseismic 

deformation and afterslip within the fault zone. 

281


EARTHQUAKE 

ARCHAEOLOGY 



Acknowledgements: Funded by NERC grants 

NE/H003266/1 and NE/E016545/1 and Durham University 

Doctoral Fellowship (M. Wilkinson). Thanks to A. Michetti, 

E. Vittori, L. Guerrieri, A. M. Blumetti, N. De Paola, A. 

Yates, A. Bubeck and G. Sileo for assistance in the field. 

References 

Buckham, R.C., G. Plafker, and R.V. Sharp (1978), Fault 

movement (afterslip) following the Guatemala earthquake 

of February 4, 1976. Geology, 6, 170-173. 

Cheloni, D., et al. (2010), Coseismic and initial post-seismic 

slip of the 2009 M-w 6.3 L'Aquila earthquake, Italy, from 

GPS measurements. Geophys J Int, 181 (3), 1539-1546, 

doi: 10.1111/j.1365-246X.2010.04584.x 

Chiarabba, C. et al. (2009), The 2009 L’Aquila (central Italy) 

Mw6.3 earthquake: Main shock and aftershocks. 

Geophys Res Lett, 36, L18308, 

doi:10.1029/2009GL039627 

Emergeo Working Group (2010), Evidence for surface 

rupture associated with the Mw 6.3 L’Aquila earthquake 

sequence of April 2009 (central Italy). Terra Nova, 22, 43- 

51. 

Falcucci, E., et al. (2009), The Paganica Fault and Surface 

Coseismic Ruptures Caused by the 6 April 2009 

Earthquake (L’Aquila, Central Italy), Seismol. Res. Lett., 

80, 940–950, doi:10.1785/gssrl.80.6.940. 

Marone, C. J., S. H. Scholtz, and R. Bilham (1991), On the 

mechanics of Earthquake Afterslip, J. Geophys. Res., 96, 

8441–8452, doi:10.1029/91JB00275 

Scholz, C. H. (1990), The Mechanics of Earthquakes and 

Faulting, Cambridge Univ. Press, New York. 

Sharp, R.V. et al. (1989), Surface faulting along the 

Superstition Hills fault zone and nearby faults associated 

with the earthquakes of 24 November 1987, Bull. 

Seismol. Soc. Am., 79, 252-281. 

Vittori, E. et al. (2010), Surface faulting of the April 6th 2009 

Mw 6.3 L’Aquila earthquake in central Italy. Bulletin of the 

Seismological Society of America, in press. 

Wilkinson, M., et al. (2010), Partitioned postseismic 

deformation associated with the 2009 Mw 6.3 L’Aquila 

earthquake surface rupture measured using a terrestrial 

laser scanner, Geophys Res Lett, 37, L10309, doi: 

10.1029/2010GL043099 

Williams, P.L. and H.W. Magistrale (1989), Slip along the 

Superstition Hills fault associated with the 24 November 

1987 Superstition Hills, California, Earthquake. Bull. 

Seismol. Soc. Am., 79, 390-410. 

282


EARTHQUAKE 

ARCHAEOLOGY 



SEDIMENTARY EVIDENCE OF HOLOCENE TSUNAMI IMPACTS AT THE GIALOVA 

LAGOON (SOUTHWESTERN PELOPONNESE, GREECE) 

Willershäuser, Timo (1), Andreas Vött (1), Georg Bareth (2), Helmut Brückner (2), Hanna Hadler (1), Konstantin Ntageretzis (1) 

(1) Institute for Geography, Johannes Gutenberg-Universität Mainz, Johann-Joachim-Becher-Weg 21, 55099 Mainz. GERMANY. 

Email: Timo.Willershaeuser@uni-mainz.de 

(2) Institute for Geography, Universität zu Köln, Albertus-Magnus-Platz, 50923 Köln. GERMANY 

Abstract (Sedimentary evidence of Holocene tsunami impacts at the Gialova Lagoon (Southwestern Peloponnese, 

Greece): The coastal area around the Gialova Lagoon (southwestern Peloponnese, Greece), directly exposed to the tectonically 

highly active Hellenic Trench, was repeatedly affected by tsunamigenic impacts as known from historical sources. Detailed geoscientific 

studies were carried out in coastal environments in search of corresponding tsunami deposits using terrestrial 

vibracorings. Geomorphological, sedimentological and geochemical methods were applied to reconstruct the sedimentary 

fingerprints of Holocene tsunami events and the palaeogeographical evolution. Our results show that the palaeogeographical 

setting was strongly affected by high-energy tsunami impacts. Coarse-grained allochthonous sediments of marine origin were 

found intersecting muddy deposits of low-energy lagoonal and limnic environments. The Voidokilia washover fan and beachrock 

structures along the coastline also seem to be of tsunamigenic origin. Both sedimentological and geochronological criteria suggest 

multiple tsunami landfall since the mid-Holocene. 

Key words: Palaeotsunami, beachrock-type tsunami deposits, Holocene stratigraphy, washover fan. 

INTRODUCTION AND AIMS 

The eastern Mediterranean is a tectonically active 

region with a high tsunami risk (Papazachos & 

Dimitriou, 1991). The plate boundary of the Hellenic 

Arc, where the African Plate is being subducted 

beneath the Aegean microplate, is a hot spot for 

earthquakes and therefore highly capable for 

triggering tsunamis. Numerous historical accounts 

show that the surrounding coastal areas and their 

geomorphology were affected by multiple 

tsunamigenic impacts (Soloviev, 1990). Thus, 

palaeotsunami research in the eastern 

Mediterranean has been distinctly intensified in the 

last 20 years. Sedimentary characteristics of recent 

and subrecent tsunami deposits comprise e.g. (a) 

shell debris layers, (b) mixture of littoral and 

sublittoral material, (c) multi-modal grain size 

distribution, (d) rip up-clasts, (e) basal erosional 

unconformities, (f) fining upward and thinning 

landward tendencies, (g) lithified beachrock-type 

tsunamites and (h) washover deposits (Dominey- 

Howes et al., 2006; May, 2010; Vött et al., 2009a, 

2009b, 2010a, 2010b). 

The main objectives of this study are (i) to detect 

allochthonous high-energy deposits in the local 

stratigraphical record and (ii) to reconstruct 

palaeotsunami events against the background of the 

palaeogeographical evolution of the Gialova coastal 

area during the Holocene. 

REGIONAL SETTING AND METHODS 

The coastal area of Pylos and the Gialova Lagoon 

are located in the southwestern Peloponnese. The 

study area is directly exposed to the subduction zone 

283 

of the Hellenic Trench holding a high tsunami risk 

(Hollenstein et al., 2008, Sachpazi et al., 2000, 

Papazachos & Dimitriou, 1991). The Gialova Lagoon 

is located in the northern fringe of the Navarino Bay, 

a tectonic depression. The shallow lagoon is 

separated from the Bay of Navarino by a beach 

barrier system to the south and the semi-circular Bay 

of Voidokilia to the west (Fig. 1). 

Fig. 1: (a) Topographic and geomorphological 

overview of the study area and selected coring sites. 

General map based on Google Earth images (2003). 

(b) Bird’s eye view of Gialova Lagoon, view to the 

east. Photo taken by. T. Willershäuser, 2009. 

Vibracores in the environs of the Gialova Lagoon 

were retrieved by an Atlas Copco mk1 corer. In the


EARTHQUAKE 

ARCHAEOLOGY 



field, vibracores were analyzed by sedimentological 

and pedological methods. Laboratory studies 

comprised analyses of organic content (loss on 

ignition), concentration of calcium carbonate, pHvalue 

and electrical conductivity. All sediment 

samples were analysed for contents of Ca, Mn, Fe 

and more than 20 other elements using the XRF 

technique. 

SEDIMENTARY RECORD IN QUIESCENT NEAR- 

SHORE ENVIRONMENTS 

The Gialova geo-archive is dominated by quiescent 

low-energy conditions. For the Gialova Lagoon 

vibracoreprofiles PYL 2 and PYL 3 are considered to 

be most representative for the local coastal evolution 

and to detect potential high-energy signatures in the 

stratigraphical record. 

stratum of lagoonal mud documents that the preexisting 

quiescent conditions were again reestablished. 

Finally, the lagoonal mud is covered by 

sandy sediments of the recent dunes and marshy 

deposits. 

In summary, both Vibracores show distinct signatures 

of abruptly environmental changes in the 

sedimentary record. We detected a minimum of two 

distinct allochthonous layers of high-energetic 

conditions compared to the autochthonous, 

predominantly limnic pre-existing environments. 

Vibracoring site PYL 2 (N 36°57’46.9’’, 

E 21°41’29.3’’, ground surface at 0.47 m above sea 

level) and PYL 3 (N 36°57’51.6’’, E 21°39’51.3’’, 

ground surface at 0.22 m above sea level) are 

located at the eastern and western shores of the 

Gialova Lagoon, respectively (Fig. 1). 

At its base, the stratigraphy of PYL 2 (based on 

sedimentological and geochemical parameters), 

consists of homogenous silty sediments showing 

quiescent, most probably limnic conditions. This 

basal stratum is intersected by high-energy deposits 

of unsorted coarse-grained material of marine origin. 

This event layer is overlain by homogeneous and 

well sorted silty clay indicating an immediate reestablishment 

of pre-existing quiescent conditions. 

The limnic environment was again influenced by a 

second input of unsorted allochthonous grus, gravel 

and limestone fragments embedded in a loamy 

matrix. The associated sharp basal erosional contact 

again documents that sediment input occurred 

abruptly. Towards the top, the high-energy sediments 

are covered by a sequence of peat, organic mud, 

limnic mud and finally by recent marshy sediments. 

The base of PYL 3 is made out of well sorted fine 

sandy silt of a quiescent (fluvio-)limnic environment. 

This facies is separated by a sharp erosional 

unconformity from following allochthonous coarsegrained 

and unsorted sediments with distinct fining 

upward sequences and sublayers including wellrounded 

gravel. Subsequently, autochthonous limnic 

conditions were re-established. A second sharp 

erosional unconformity indicates another abrupt 

environmental change and the input of allochthonous 

marine material (Fig. 2). The fairly unsorted 

sediments consist of a mixture out of gravel, grus, 

sand and loam. This part of the profile is again 

characterized by distinct fining upward sequences 

and rip up-clasts of eroded underlying limnic 

sediments. After the event, quiescent limnic 

conditions were quickly re-established, subsequently 

turning into lagoonal conditions. Towards the top of 

vibracore PYL 3, the lagoon was influenced by 

another third distinct input of allochthonous marine 

sand. The intersecting event layer is characterised by 

rip up-clasts and several fining upward sequences. A 

284 

Fig. 2: Vibracore PYL 3 and facies interpretation. 

Details (a + b) show intersecting allochthonous 

marine sediments in between predominantly limnic 

deposits. 

BEACHROCK AS LITHIFIED TSUNAMI DEPOSIT 

First studies on the occurrence of beachrock along 

the coastline of Pylos were made by Kraft et al. 

(1980). These authors described that the beachrock 

includes sherds of probably Roman age, but no 

further information on the internal structure, 

geomorphological and sedimentary context was 

given. However, recent studies on beachrock-type 

calcarenitic deposits at adjacent coastal areas 

revealed that post-depositional pedogenetic 

decalcification and cementation of tsunami deposits 

must be assumed for their evolution (Vött et al., 

2010a & Scheffers et al., 2008). 

In the investigation area, several locations along the 

coastline show onshore and offshore occurrences of 

eroded and fragmented beachrock. The internal 

structure is characterized by typical sedimentological 

features of tsunami deposits. At Vromoneri (10 km to 

the north of Gialova Lagoon), we found beachrock 

sequences injected in between bedrock units and 

characterized by basal erosional discontinuities, 

partly well-laminated structure, embedded intra clasts 

and features of a distinct landward flow direction are 

detectable.


EARTHQUAKE 

ARCHAEOLOGY 



At Romanou (Fig. 3), the basal section of the 

beachrock is dominated by gravel, followed by 

coarse, medium and fine sand with distinct fining 

upward sequences. The sedimentary features are 

untypical of littoral environments and rather 

document high-energy flow dynamics. Thus, 

beachrock-type lithified sediments along the Pylos 

coastline are interpreted as deposits of high-energy 

impact. 

Fig. 3: Beachrock at Romanou (about 5 km to the 

north of the Gialova Lagoon - see Fig. 1). The present 

position of the fragmented beachrock is offshore (I) 

and onshore (II). 

DISCUSSION AND CONCLUSION 

Our case-study, based on sediment cores and 

geomorphological findings, show that tsunami 

signatures in coastal sedimentary environments are 

highly variable on large scales. 

The sediment profiles recovered from near-shore 

environments are characterised by allochthonous 

gravelly to sandy high-energy deposits intersecting 

fine-grained autochthonous sediments of a lower 

energetic potential. As the energetic potential as well 

as the landward extent of storm influence in the 

Mediterranean are known to be restricted to the 

immediate littoral zone (Vött et al., 2010b), 

allochthonous coarse-grained deposits found in the 

environs of Gialova Lagoon are interpreted as of 

tsunamigenic origin. Moreover, numerous findings of 

beachrock along the coastline must not be 

considered as consolidated littoral sediments. We 

assume that the origin of the beachrock in this case 

is attributed to sediment deposition by high-energy 

events followed by post sedimentary cementation 

and calcification (Vött et al., 2010a). Associated 

sedimentary characteristics, such as laminated 

structures, erosional contact at the base, fining 

upward sequences and embedded stones and 

ceramic fragments are characteristic for tsunami 

influence but cannot be described by recent littoral 

processes. The paleogeographical evolution of the 

Pylos area is characterised by predominant limnic 

and lagoonal conditions. The shifting and destruction 

of the palaeo-environment is mainly controlled by 

high-energy tsunami impacts induced by earthquake 

focal mechanisms. Gradual changes seem to be 

merely responsible for the re-arrangement of 

sediments after high-energy impacts. 

Acknowledgements: We acknowledge funding of the 

project by the German Research Foundation, DFG (Bonn, 

Gz. VO 938/3-1). Work permits were kindly issued by the 

Greek Institute of Geology and Mineral Exploration (IGME, 

Athens). 

References 

Dominey-Howes, D.T.M., G.S. Humphreys & P.P. Hesse 

(2006). Tsunami and palaeotsunami depositional 

signatures and their potential value in understanding the 

late-Holocene tsunami record. The Holocene 16 (8), 

1095-1107. 

Hollenstein, C., M.D. Müller, A. Geiger & H.G. Kahle (2008). 

Crustal motion and deformation in Greece from a decade 

of GPS measurements, 1993–2003. Tectonophysics 

449, 17-40. 

Kraft, J.C., G.R. Rapp, Jr. & S.E. Aschenbrenner (1980). 

Late Holocene Palaeogeographic Reconstructions in the 

Aerea of the Bay of Navarino: Sandy Pylos. Journal of 

Archaeological Science 7, 187-210. 

May, S.M. (2010). Sedimentological, geomorphological and 

geochronological studies on Holocene tsunamis in the 

Lefkada – Preveza area (NW Greece) and their 

implications for coastal evolution. PhD Thesis 159 pp., 

Cologne, Germany. 

Papazachos, B.C. & P.P. Dimitriu (1991). Tsunamis In and 

Near Greece and Their Relation to the Earthquake Focal 

Mechanisms. Natural Hazards 4, 161-170. 

Sachpazi, M., A. Hirn, C. Clément, F. Haslinger, M. Laigle, 

E. Kissling, P. Charvis, Y. Hello, J. C Lépine, M. Sapin& 

J. Ansorge (2000). Western Hellenic subduction and 

Cephalonia Transform: local earthquakes and plate 

transport and strain. Tectonophysics 319(4), 301–319. 

Scheffers, A., D. Kelletat, A. Vött, S.M. May & S. Scheffers 

(2008). Late Holocene tsunami traces on the western 

and southern coastlines of the Peloponnesus (Greece). 

Earth and Planetary Science Letters 269, 271-279. 

Soloviev, S.L. (1990). Tsunamigenic Zones in the 

Mediterranean Sea. Natural Hazards 3, 183-202. 

Vött, A., H. Brückner, S. Brockmüller, M. Handl, S.M. May, 

K. Gaki-Papanastassiou, R. Herd, F. Lang, H. 

Maroukian, O. Nelle & D. Papanastassiou (2009a). 

Traces of Holocene tsunamis across the Sound of 

Lefkada, NW Greece. Global and Planetary Change 66 

(1-2), 112-128. 



Fountoulis (2009b). The Lake Voulkaria (Akarnania, NW 

Greece) palaeoenvironmental archive - a sediment trap 



1-37. 



Lang, P. Masberg, S.M. May, K. Ntageretzis, D. 

Sakellariou & T. Willershäuser (2010a). Beachrock-type 



Akarnania, the Ionian Islands and the western 

Peloponnese. Zeitschrift für Geomorphologie N.F., 

Suppl- Issue 54 (3), 1-50. 

Vött, A., et al. (2010b), Sedimentological and 

geoarchaeological evidence of multiple tsunamigenic 

imprint on the Bay of Palairos-Pogonia (Akarnania, NW 

Greece), Quaternary International (2010),doi:10.1016/ 

j.quaint.2010.11.002 

285


EARTHQUAKE 

ARCHAEOLOGY 



IS THE RURRAND FAULT (LOWER RHINE GRABEN, GERMANY) RESPONSIBLE FOR 

THE 1756 DÜREN EARTHQUAKE SERIES? 

Winandy, Jonas (1, Christoph Grützner (1), Klaus Reicherter (1), Thomas Wiatr (1), Peter Fischer (2), Thomas Ibeling (3) 

(1) RWTH Aachen, Neotectonics and Natural Hazards Group, Aachen, Germany (jonas.winandy@rwth-aachen.de, +49 241 

8096358), 

(2) Köln University, Geographisches Institut, Köln, Germany 

(3) Archäologische Grabungen und Sondagen GBR, Köln, Germany 

Abstract (Is the Rurrand Fault (Lower Rhine Graben, Germany) responsible for the 1756 Düren earthquake series?): In 

1756, several strong earthquakes (M5-6.1) occurred close to Düren (Lower Rhine Graben, LRG) in Germany. The Rurrand Fault in 

the LRG located in the middle between Aachen and Cologne possibly indicates the Düren earthquake sequence. This fault is one 

of the most prominent NW-SE trending normal faults with a morphological expression in the area within the Lower Rhine Graben. 

Holocene sediments with significant offsets covered by thin colluvial sediments were found and a complex fault geometry was 

observed during archaeological excavations. DC geoelectrics and georadar were applied in order to image the deeper parts of the 

fault zone. Radiocarbon and luminescence dating of sediment samples are in progress, but the morphological expression of the 

fault, the shallow depths of the offset sediments, and geophysical data allow concluding on recent seismicity along this active fault 

with at least four surface-rupturing events. 

Key words: earthquake, geophysics, Rhine Graben, Rurrand Fault 

INTRODUCTION: THE DÜREN EARTHQUAKES 

1755/1756 

The area between Aachen and Cologne in western 

Germany was hit by a series of earthquakes in 

1755/1756. On 18 th February, 1756, the strongest 

event took place most likely west of the city of Düren, 

leaving two people dead (some reports claim three 

fatalities) and causing significant damages also in 

Aachen, Cologne, and nearby villages. Chimneys 

were destroyed in up to 70 km distance (Liège, 

Belgium), light damages were recorded in Brussels, 

Gießen and Osnabrück (200 km distance). A 

landslide was triggered 15 km SW from Düren. The 

shaking was felt as far as 400 km from the epicentre 

in London, Magdeburg, Halle, Paris, and Strasbourg 

(Meidow, 1995). Epicentral intensities of VIII are 

reported by Skupin et al. (2008) for the Eschweiler 

area (15 km W of Düren). A magnitude of 6.3 is 

assumed for the main event by Skupin et al. (2008), 

Meidow (1995) assumes M L =6.1. Our study shows 

that the Rurrand Fault was possibly activated during 

the Düren earthquake sequence. 

The Lower Rhine Embayment and especially its 

western part is one of the tectonically most active 

areas in Germany and dominated by the Lower 

Rhine Graben. NW-SE trending normal faults form a 

horst and graben structure with a number of single 

blocks (Krefeld-, Köln-, Venlo-, Erft-, and Rur blocks, 

from NE to SW). The faults show offsets of more than 

50 m in the Quaternary. Düren is situated in a NW- 

SE striking graben, which is flanked by the Rurrand 

Fault in the NE and the Stockheimer Sprung in the 

GEOLOGY AND TECTONIC SETTING 

The Lower Rhine Embayment (LRE) underwent 

subsidence from Miocene to recent, accompanied by 

uplift of the Rhenish Massif to the southwest and east 

of the study area. Tertiary and Quaternary sediments 

of more than 1 km thickness were deposited and also 

include lignites which are extracted in open pit mines. 

Fluvial and aeolian Pleistocene-Holocene sediments 

as well as loess cover wide areas. 

286 

Fig. 1: Neotectonics and historical earthquakes in the 

study area, the Lower Rhine Embayment. RRF: 

Rurrand Fault; SSF: Stockheimer Sprung


EARTHQUAKE 

ARCHAEOLOGY 



SW (Fig. 1). The Rurrand Fault is a NW-SE striking 

normal fault dipping to the SW and expressed by an 

escarpment. 

A large number of damaging earthquakes since 

Karolingian times has been reported for this area. 

The most recent one was the 1992 Roermond 

earthquake which reached M L =5.9. Recent studies 

reported that active faults in the study area are 

characterized by recurrence periods in the order of 

tens of ka (Skupin et al., 2008; Camelbeeck et al., 

2007), and that present day aseismic slip is assumed 

for the Rurrand Fault, resulting form the lowering of 

the groundwater level due to the nearby mining 

activities (Vanneste & Verbeeck, 2001). Active faults 

in Germany are often not visible in the field due to 

relatively high erosion rates. Therefore, the seismic 

hazard might be under-estimated. The Rurrand Fault 

was trenched already only approx. 2-3 km away, and 

only Pleistocene faulting evidence was found (Skupin 

et al., 2008). 

measurements. Data processing was done with 

ReflexW by Sandmeier Scientific Software. The 

geoelectrics data were gathered with the 4-point-light 

system (Lippmann), and 80 electrodes with 1.5 m 

spacing. Schlumberger, Dipole-dipole and Wenner 

configurations were applied. Soil samples were taken 

for radiocarbon and luminescence dating, which is 

currently in progress. 

RESULTS 

METHODS 

Fig. 2: Location of trenches, outcrops and geophysical 

profiles at the study area. 

Due to construction works for a new highway, 

extensive archeological excavations have proven 

findings from Roman times until recent. At this 

occasion, the Rurrand Fault was trenched in several 

places, where we mapped layer offsets, sediment 

deformation, and structural data (Fig. 2). The trench 

walls were sketched and photographed. Ground 

penetrating radar (GPR) and electric resistivity 

measurements have been applied in order to image 

deeper sediment structures and to map the fault 

trace. We used the GSSI 100, 270, and 400 MHz 

antennas with the SIR 3000 controller, a survey 

wheel and a GPS data logger for georadar 

Fig. 3: 270 MHz GPR profile (64) crossing the fault 

with an angle of 45°. 

The GPR data revealed sediment layers inclined 

towards the fault (Fig. 3). The fault cropped out 2 m 

away from the GPR profile, allowing a direct 

comparison. Down to a depth of 4 m several 

reflectors dip towards the NW. 

287


EARTHQUAKE 

ARCHAEOLOGY 



Fig. 4: Photograph of the fault, outcropped in an archeological trench (upper image), and reconstruction of the faulting 

history (lower image). S = sedimentation and erosional stages; E W - E Z = last earthquake events. 

A sharp contrast in reflection amplitudes marks the 

fault itself. On the footwall, only few layers appear to 

dip towards the Rurrand Fault. Similar observations 

were made at other GPR profiles that crossed the 

fault. The high-frequency antennas allowed 

identifying tilted sediments and the fault itself at 

various locations. The resolution of the 100 MHz 

288 

antenna was found too low for imaging fault features 

in this case. Geoelectrics data revealed low-resistivity 

anomalies (higher conductivities) at the fault zone, 

which are most likely related to an increase in water 

content at the fault zone.


EARTHQUAKE 

ARCHAEOLOGY 



Holocene surface-near sediments with significant 

offsets covered by thin colluvial sediments were 

found and a complex fault geometry was observed 

during the archaeological excavations (Figs. 4 - 6). 

Some deformation structures seem to be related to 

liquefaction. The offset of surface-near sediments is 

in the order of 5 cm (Fig. 5), deeper layers show 

greater offsets (Fig. 6). Growing displacement of the 

major fault downsection suggest more than one 

major, surface-rupturing earthquakes along the 

Rurrand Fault in the Holocene/Late Pleistocene. We 

developed a deformation model for the fault, 

assuming at least four surface-faulting events (E W - 

E Z ) that led to the present day geometry (Fig. 4) and 

seven stages of seismic quiescence (S 1 -S 7 ). Roots 

penetrated the soil at stage four, and liquefaction of 

fine grained material is likely to have occurred during 

the last earthquake event. Despite the results from 

dating are yet to come, we can assume four events 

since Late Pleistocene, which would result in slightly 

shorter recurrence periods than estimated by 

previous studies. The fault may also be responsible 

for the Düren 1756 events, as the evidence for 

surface-faulting earthquakes proves that it is capable 

for destructive events with magnitudes > 5.5. 

However, only the dating will allow associating the 

Düren earthquakes with the Rurrand Fault. 

We found evidence for four surface-faulting events at 

the active Rurrand Fault close to Düren, which might 

be responsible for the 1755/1756 earthquake series. 

Offset sediments clearly point to seismic activity 

since Late Pleistocene and at least four earthquake 

events. Geophysical data allowed mapping the fault 

trace where there were no outcrops (Geoelectrics 

and GPR) and revealed the fault geometry in depths 

of up to 6 m (GPR). 

Fig. 6: Offset surface-near layers have been found 

during the archeological excavations. 

References 

Fig. 5: A) Step-like offset clearly points to seismic 

deformation instead of soil creep. Displacement is 

about 5 cm. The offset reddish layer is made up of 

clayey-silty, loess material, 50 cm below the surface. 

B) The fault zone is clearly visible in the trenches, not 

only at the walls, but also on the floor. This enabled a 

very good correlation with the geophysical data and a 

precise analysis of subsurface features. 

CONCLUSION 

Camelbeeck, T., Vanneste, K., Alexandre, P., Verbeeck, K., 

Petermans, T., Rosset, P., Everaerts, M., Warnant, R. & 

Van Camp, M. (2007). Relevance of Active Faulting And 

Seismicity Studies To Assessments Of Long-Term 

Earthquake Activity and Maximum Magnitude In 

Intraplate Northwest Europe, between the Lower Rhine 

Embayment and the North Sea. In: Geological Society of 

America, Special Paper, vol. 425, pp. 193-224. 

Meidow, H. (1995). Rekonstruktion und Reinterpretation von 

historischen Erdbeben in den nördlichen Rheinlanden 

unter Berücksichtigung der Erfahrungen bei dem 

Erdbeben von Roermond am 13. April 1992. Dissertation, 

Universität Köln, Leverkusen, 305 p. 

Skupin, K., Buschhüter, K., Hopp, H., Lehmann, K., Pelzing, 

R., Prüfert, J., Salamon, M., Schollmayer, G., Techmer, 

A. & Wrede, V. (2008). Paläoseismische Untersuchungen 

im Bereich der Niederrheinischen Bucht. Scriptum 17, 

Krefeld (Geol. Dienst Nordrh.-Westf.), 72 p. 

Vanneste, K. & Verbeeck, K. (2001). Paleoseismological 

analysis of the Rurrand fault near Jülich, Roer Valley 

graben, Germany: Coseismic or aseismic faulting 

history? Netherlands Journal of Geosciences / Geologie 

en Mijnbouw, 80 (3-4): 155-169. 

289


EARTHQUAKE 

ARCHAEOLOGY 



FIND AND PRIMARY SEARCH OF AN ACTIVE FAULT AT THE GAIXIA SITE, GUZHEN 

COUNTY, ANHUI PROVINCE, P.R. CHINA 

YAO Da-quan (1), SHUO Zhi (2),TANG Ji-ping (3), WANG Zhi (2),SHEN Xiao-qi (1),CHEN An-guo (1) , LI Lin-li (1) 

(1)Seismological Administration of Anhui Province,Hefei,Anhui P.R.China,230031. email: daquany@aheq.gov.cn 

(2)Cultural Relic Archaeology Research Institute of Anhui Provine, Hefei,Anhui P.R. China,230001 

(3)Institute of Cultural Relic Archaeology of Tongling City,Anhui Province,Tongling, Anhui P.R. China,244001 

Abstract (Find and primary search of an active fault at the Gaixia Site, Guzhen County, Anhui Province, P.R. China): 

Recently a large number of ancient sites were excavated due to construction works in eastern China, which makes it possible to 

identify and trace thousands of years of natural deformation history. With this opportunity, one can benefit from the precise 

archaeological layer technology. In a collaboration of earthquake research institutions and archaeological departments, we 

analysed Quaternary tectonic deformation, in particular the deformation phenomena hosted in the prehistoric cultural layer. This 

paper reports the working progress of a special excavation at the Gaixia archaeological site in Haocheng town (Guzhen county, 

Anhui province, P.R. China). 

Key words: Quaternary deformation, prehistoric culture, Gaixia site, Anhui Province 

THE BASIC FEATURES OF THE VESTIGE 

The Gaixia site is located east of the Tancheng- 

Lujiang fault (TanLu Fault on Fig. 1). In the west 

there is the Guzhen-Fengtai fault, where the Ms 6.2 

earthquake event of 1831 took place in FengTai. The 

southern border is the NW-trending GuoHe fault, 

where a magnitude Ms 6 earthquake occurred in 

1481 in Guoyang (Fig.1). 

The Gaixia site lies in Bawangcheng village, 

Haocheng town, 24 kilometers eastward of Guzhen 

county, which is the centre battlefield of the 

competition between Chu and Han and the final 

Gaixia battle, where the ‘Gaixia Battle’ took place in 

202 B.C. The entire archaeological area is about 

150,000 km 2 , with an earth wall around the main site, 

which is 2-3 m higher than the outer surface level. 

The inner terrain is higher at the sides and lower than 

the surroundings in the middle. West and north of the 

archaeological site there is the Tuohe River. 

In order to clarify the specific age, the floor width and 

construction methods of the ancient city wall, 

archaeologists excavated a trench since March 2008 

(serial number: 2008 GGTG3, TG3 for short) with a 

length of 40 m and a width of 3 m, about 70 m from a 

former excavation in 2007. The trench extends from 

the city wall to the inner moat, which completely 

reveals the whole section of the city wall and 

excavates some of the cultural layers in the inner 

wall. Just in the Dawenkou Cultural layer in both 

sides and at the bottom of TG3, seismologists and 

archaeologists discovered an active fault (Fig.2). 

DISCOVERY AND ANALYSIS OF THE TENSION- 

SHEAR FAULT IN THE CULTURAL LAYER 

On May 12 th , 2008, when excavations reached greyblack 

ash layers, seismologists and archaeologists 

Fig. 1: A sketch of regional geology seismology of Gaixia ruins 

Fig. 2: TG3 prospecting trenchfault of Gaixia ruins 

290


EARTHQUAKE 

ARCHAEOLOGY 



found fault dislocation evidence. According to this 

clue, they carefully scraped the pace of the layer 

again for further identification, and confirmed the 

range of fault dislocation. 

point on the south wall is 1.68 meters below ground 

surface, and also extends to the raw soil layer. From 

the 18th layer downwards, 12 layers of the north wall 

had dislocation, not including the raw soil layer: the 

fault penetrated the wall with the layers 18, 33, 36, 

37, 39, 40, 41, 47, 54, 61, 68, 70, and also 

penetrated the narrow channel G6, which is S-N 

trending. The southern part of the prospecting trench 

also showed normal faulting (Fig.3). From the 18th 

layer downwards to the raw soil layer there are 10 

layers that were dislocated, not including the raw soil 

layer, and also showing normal faults (Figs.4,5). 

Fig.3: Tensile-shear fault on the northern wall. Left: 

Photo of the northern wall; Right: sketch of the 

northern wall. 

The dislocation was located in TG3, and the west 

segment is about 15.4 m to the west wall. According 

to the change of quality and color of the soil and the 

analysis of samples taken at the dislocation position, 

it could be confirmed that the dislocation is beneath 

the 17th layer of the city wall (serial number: “the 

17th layer wall”, following are the same), and the 

dislocation extends from the 18th layer, including raw 

soil layer (undisturbed soil). A crack with a width of 

Fig.5: Tensile-shear fault on the trough bottom. Left: 

photo towards the southern overlook; Right: photo 

towards the eastern overlook. 


In conclusion, active faulting was found in the culture 

layers of the vestige belong to the Late Dawenkou 

Dynasty (about 4300a BP), and the observation of 

the fault plane shows the characteristic of high speed 

deformation (He Yong-nian et al.,1985; Yao Daquan,2004). 

Fig.4: Tensile-shear fault on the southern wall. Left: 

photo of the southern wall; Right: sketch of the 

southern wall. 

2 mm - 4 mm can be seen at both the plane and the 

profile of the dislocation, which was very clear to see 

on the bottom of the soil layer, gradually becoming 

thinner upwards in the 17th layer. The crack was 

filled with grey clay and had a different colour than 

the surrounding soil. The soil’s thickness was 

symmetric and completely anastomosing on the two 

sides of the fault. 

The western part of the fractured plane was 3.8 cm 

higher than the eastern (= 3.8 cm vertical offset). The 

trend of the fracture plane is 353°, and the section 

dipped towards the E with an angle of ~60°. The fault 

plane is curved, with its highest point 1.79 meters 

below the ground surface at the western wall of the 

trench, extending to the raw soil layer. The highest 

The archaeological site is located right on the NNW 

trending Tancheng-Lujiang fault. According to the 

history records, several earthquakes of Ms~6 

occurred near this area. Our discovery of this 

earthquake in the ruins enriches the seismotectonic 

knowledge in this area and completes the earthquake 

catalogue for future hazard studies. 

Acknowledgements: This paper is a contribution to the 

Scientific Research Special Project of the Earthquake 

Calling200808064and Science and Technology Tackle 

Key Problem Plan Project of Anhui Province 

08010302204 

References 

He Yong-nian, Yang Zhu-en. 1985. Research and the 

significance of microscopic marks of ancient earthquakes 

[J]. Earthquake Research in China,1(3):76-81(in 

Chinese). 

Yao Da-quan 2004. Macroscopic and Microscopic Evidence 

of Periodical Stick-Slip Deformation in an Active Fault [J]. 

Recent Developments in World Seismology, (4):6-10 (in 

Chinese). 

291


EARTHQUAKE 

ARCHAEOLOGY 



THE MOVRI MOUNTAIN EARTHQUAKE: UNDERSTANDING ACTIVE DEFORMATION OF 

THE NW PELOPONNESE 

Zygouri, Vasiliki (1, Sotiris Kokkalas (1), Paris Xypolias (1), Ioannis Koukouvelas (1), Gerassimos Papadopoulos (2) 

(1). Department of Geology, Division of General, Marine Geology and Geodynamics, University of Patras, Patras 26500, Greece. 

Email: iannis@upatras.gr 

(2). Institute of Geodynamics, National Observatory of Athens, Lofos Nymfon, Thissio, 11810, Athens, Greece. Email: 

papadop@noa.gr 

Abstract (The Movri Mt earthquake: understanding active deformation of the NW Peloponnese): The Mw 6.4 June 8 th 

2008 Movri Mountain earthquake struck NW – Peloponnesus, Greece, caused widespread deformation and 

damage in buildings, as well as extensive ground hazards. Three surface ruptures were triggered by the 

earthquake, with the most promising for paleoseismology analysis lying near the epicenter of the event, attaining a 

maximum offset of 25 cm. In this surface rupture a paleoseismological trench was excavated. Based on seven 14 C 

samples, we identify two surface – rupturing earthquakes in the last 1Kyr prior the recent event. Thus, observations 

from paleoseismology suggest that the Nisi fault appear to be related to surface ruptures and events. In addition, 

our 14 C data support the view that the Nisi fault displays a slip rate in the order of 1.5 mm/yr during the last 1Kyr. 

Key words: paleoseismology, surface ruptures, colluvial wedge. 

INTRODUCTION 

On June 8 th 2008, a Mw 6.4 strike – slip earthquake 

(hypocentral depth of almost 20 km) occurred in the 

northwestern Peloponnese (Greece) without obvious 

relation to any mapped fault. Although strong 

earthquakes are common in western Greece, this 

event took place in a region previously considered as 

seismically quiet. This earthquake ruptured along an 

unknown dextral strike slip fault segment striking NE- 

SW, resulted in new geological conclusions for the 

area. During the Movri Mt earthquake, three main 

fault ruptures emerged on the epicentral area, the 4.5 

km long Vithoulkas surface rupture, the 5.0 km long 

Michoi surface rupture and the 6.0 km long Nisi 

surface rupture (Fig. 1). 

Fig. 1: Geological map of the study area showing lithology, active faults and surface ruptures during the Movri Mt 

earthquake. 

292


EARTHQUAKE 

ARCHAEOLOGY 



In order to investigate whether these ruptures and 

associated fractures record an individual or unique 

tectonic offset or a response to strong ground motion 

we performed a paleoseismological study across the 

Nisi surface rupture. The Nisi rupture trends NNW- 

SSE and has an almost straight, segmented trace 

and attained a maximum offset of 25 cm during the 

event. The trench has a length of 10 m, a width of 1.5 

m and a depth of 2.5 m. 

DESCRIPTION OF THE NISI TRENCH 

The sedimentary environment observed inside the 

trench suggests fine – grained coastal and fluvial to 

lagoonal sediments (fig. 2). It comprises sandy to 

clay horizons with generally yellow, brownish yellow 

and grey to dark brown sediments. Shallow coastal 

sediments are mostly found on the uplifted part of the 

trench. Shallow coastal sediments are overlain in the 

upthrown block of the fault either directly by the 

modern plowed soil, or by a succession of almost 0.5 

m thick yellow sandy and grey clay fluvial sediments 

in the downthrown block. The fault zone is almost 1 

m wide with open voids formed also during the latest 

event. The deeper excavated part of the fault zone 

includes rotated coastal sediments. Open voids allow 

the precipitation of meteoric water, and disturb the 

deposition of new material after every possible new 

event. The material inside the fault zone is 

characterized as an unconsolidated, unsorted and 

mixed assemblage of coastal sediments derived 

mainly from the fine grained footwall’s coastal 

sediments. Two small colluvial wedges (fig. 3) with a 

maximum thickness of 30-40 cm, though obscure, 

can be detected. These colluvials are characterized 

by wedge geometry, thickest at the fault surface and 

thinning away from it, and colour differences. 

Fig. 3: Detailed view of the upper colluvial wedge 

formed in the Nisi fault zone (The black box shows the 

Ps3 sample). 

Fig. 2: Trench stratigraphy of the Nisi Fault. he 

sampling sites within the trench are mentioned 

with Ps abbreviation . 

293 


Despite the flat and rather featureless landscape, 

certain evidences support the view that Nisi rupture is 

a newly emerged fault with recent seismic history. 

The colluvial wedges in the trench fulfil most of the 

parameters defining deposits in fault zones such as 

their shape and chaotic sedimentation (sensu 

Pavlides et al., 2004). The colluvial deposits are 

supplied by fine grained material of the upthrown 

block, controlled by the sedimentation environment 

and the ruptured sediments and therefore is bound to 

be formed by the finer deposits accumulating on the 

downthrown block’s free face. Open voids, filled 

cracks, the formation of at least one datable colluvial 

wedge and the increased deformation in upthrown 

horizons attests to possible episodic activity during 

the past 1000 years.


EARTHQUAKE 

ARCHAEOLOGY 



The seven samples (fig. 2) were collected from both 

the uplifted and subsided block of the fault as well as 

from the disturbed fault zone. The disturbed zone 

samples yielded two kind of ages, a younger one at 

the lowest part of the zone (sample Ps5) and an 

older one (sample Ps3) above the previous sample 

and at the base of the colluvial wedge (figs. 2, 3). 

This attests the effect of precipitation of meteoric 

water through the fault zone. After considering the full 

distribution of 14 C ages and taking into account the 

stratigraphy of the trench we interpret three events 

(with the more recent included) that can be 

evidenced. 

Our trenching study showed a recurrence interval 

between 300 to 600 years, we suggest that the Nisi 

fault shapes fine grained colluvial wedges without 

following a constant time pattern. Thus, we assume a 

Quaternary slip rate in the order of 1.5mm/yr and 

strong slip-rate similarities to the Egion and Schinos 

faults in Gulf of Corinth (Koukouvelas, 1998; 

Koukouvelas et al., 2005). 

Acknowledgements: We would like to express our 

appreciation to S. Pavlides, R. Caputo and A., Chatzipetros 

for fruitful discussion of the trench and S. Verroios and V. 

Chatzaras for their help during fieldwork. 

References 

Koukouvelas, I.K., (1998). The Egion fault, earthquakerelated 

and long term deformation, Gulf of Corinth, 

Greece. Journal of Geodynamics 26, 501-513. 

Koukouvelas, I.K., D. Katsonopoulou, S. Soter, & P. 

Xypolias, (2005). Slip rates on the Helike Fault, Gulf of 

Corinth, Greece: new evidence from geoarchaeology. 

Terra Nova 17, 158-164. 

Pavlides, S., I. Koukouvelas, S. Kokkalas, L. 

Stamatopoulos, D. Keramydas & I. Tsodoulos, (2004). 

Late Holocene evolution of the East Eliki fault, Gulf of 

Corinth (Central Greece), Quaternary International 115- 

116, 139-154. 

. 

294


Index 

Preface 

Baize, Stéphane, Laurence Audin, Thierry Winter, Alexandra Alvarado, Luis 

Pilatasig, Mercedes Taipe, Paul Kauffmann, Pedro Reyes 

First paleoseismic evidences in Ecuador: The Pallatanga Fault Record.............................................. 3 

Salvatore Barba and Debora Finocchio 

Some notes on earthquake and fault relationships for dip-slip events .............................................. 7 

Boulton, Sarah J. and I. S. Stewart 

Holocene coastal notches in the Mediterranean: palaeoseismic or palaeoclimatic 

indicators? .......................................................................................................................... 11 

Braun, Anika, Tomas M. Fernandez-Steeger, Hans-Balder Havenith, Almaz Torgoev, Romy 

Schlögel 

Analysing the landslide susceptibility with statistical methods in Maily-Say, 

Kyrgyzstan ......................................................................................................................... 14 

Burchfiel, B. C. and Royden, L. H. 

Tectonic interpretation of the 2008 Wenchuan Earthquake: Why it only 

propagated in one direction - the future? ................................................................................. 17 

Campos, Corina, Christian Beck, Christian Crouzet, Eduardo Carrillo 

C haracterization of Late Pleistocene-Holocene earthquake-induced 

“homogenites” in the Sea of Marmara through magnetic fabric. Implication for 

co-seismic offsets detection and measurements ........................................................................ 19 

Carmo, Rita, José Madeira, Ana Hipólito, Teresa Ferreira 

Paleoseismological evidence for historical surface rupture events in S. Miguel 

Island (Azores) .................................................................................................................... 22 

Čyžienė, Jolanta 

Fault tectonics regarding the Neotectonic period and influence of tectonic 

structures on glacial process in areas of thick Quaternary cover ................................................... 26 

Figueiredo, Paula M., João Cabral, Thomas K. Rockwell 

Plio–Pleistocene tectonic activity in the Southwest of Portugal...................................................... 30 

Foumelis Michael, Ioannis Fountoulis, Ioannis Papanikolaοu, Dimitrios Papanikolaou 

Geodetic evidence of the control of a major inactive tectonic boundary on the 

contemporary deformation field of Athens (Greece).................................................................... 34 

Fountoulis, Ioannis, Mavroulis Spyridon 

Neotectonics and comparison of the Environmental Seismic Intensity scale (ESI 

2007) and the traditional scales for earthquake intensities for the Kalamata 

(SW Greece) earthquake (Ms=6.2R, 13-09-1986) ...................................................................... 38 

Fountoulis, Ioannis D., Emmanuel Vassilakis, Mavroulis Spyridon, John Alexopoulos, 

Athanasia Erkeki 

Quantification of river valley major diversion impact at Kyllini coastal area (W. 

Peloponnesus, Greece) with remote sensing techniques .............................................................. 42 

Garduño-Monroy, Víctor Hugo, Raúl Pérez-López, Miguel Ángel Rodríguez-Pascua, Julián 

García Mayordomo, Isabel Israde-Alcántara and Jim Bischoff 

Could large palaeoearthquakes break giant stalactites in Cacahuamilpa? (Taxco, 

Central Mexico) ................................................................................................................... 46 

Gath Eldon and Tania Gonzalez 

Three-dimensional investigation of the AD 1621 Pedro Miguel fault rupture for 

design of the Panama Canal’s Boringquen dam.......................................................................... 50 

Georgiev, Ivan, Dimitar Dimitrov, Pierre Briole, Emil Botev 

Velocity field in Bulgaria and Northern Greece from GPS campaigns spanning 

1993-2008.......................................................................................................................... 54


Gielisch, Hartwig 

Acrocorinth - Geological history and the influence of paleoseismic events to 

recent archaeological research ............................................................................................... 57 

Goodman-Tchernov, Beverly N. 

Interpreting offshore submerged tsunami deposits: An incompletely complete 

record ................................................................................................................................ 60 

Guerrieri, Luca, Anna Maria Blumetti, Elisa Brustia, Eliana Esposito, Mauro Lucarini, 

Alessandro M. Michetti, Sabina Porfido, Leonello Serva & Eutizio Vittori, & the INQUA 

TERPRO Project #0811 Working Group 

Earthquake Environmental Effects, intensity and seismic hazard assessment: 

the EEE catalogue (INQUA Project #0418)................................................................................ 62 

Guzman, Oswaldo, Jean-Louis Mugnier, Rexhep Koçi, Riccardo Vassallo, Julien Carcaillet, 

Francois Jouanne, Eric Fouache 

Active tectonics of Albania inferred from fluvial terraces geometries .............................................. 66 

Hadler, Hanna, Andreas Vött, Benjamin Koster, Margret Mathes-Schmidt, Torsten Mattern, 

Konstantin Ntageretzis, Klaus Reicherter, Dimitris Sakellariou, Timo Willershäuser 

Lechaion, the ancient harbour of Corinth (Peloponnese, Greece) destroyed by 

tsunamigenic impact ................................................................................................................ 70 

Han, S.-R., M. Lee, J. Park, Y.-S., Kim 

Structural characteristics and evolution of the Yangsan-Ulsan Fault System, SE Korea ................................... 74 

Havenith, Hans-Balder 

Where landslides represent the most important earthquake-related hazards: the mountain 

areas of Central Asia ............................................................................................................... 77 

Hinzen, Klaus-G., Helen Kehmeier, Stephan Schreiber, Sharon K.Reamer 

A case study of earthquakes and rockfall - induced damage to a Roman mausoleum in 

Pinara, SW Turkey .................................................................................................................. 81 

Hipólito, Ana, José Madeira, Rita Carmo, João Luís Gaspar 

Neotectonics of Graciosa Island (Azores) – uncertainty in seismic hazard 

assessment in a volcanic area with variable slip-rates................................................................. 84 

Hoffmann, Gösta, Klaus Reicherter, Thomas Wiatr, Christoph Grützner 

Evidence for Holocene tsunami-impact along the shoreline of Oman ........................................................ 88 

Hoffmeister, Dirk, Konstantin Ntageretzis, Nora Tilly, Constanze Curdt, Georg Bareth, 

Helmut Brückner, Andreas Vött 

Monitoring coastal changes on the Ionian Islands (NW-Greece) by multi-temporal terrestrial 

laser scanning ....................................................................................................................... 92 

Jamšek, Petra, Lucilla Benedetti, Miloš Bavec, Jure Atanackov, Marko Vrabec, Andrej 

Gosar 

Preliminary report on the Vodice Fault activity and its potential for seismic hazard in the 

Ljubljana Basin, Slovenia........................................................................................................... 96 

Jankaew, Kruawun, Dominik Brill, Maria E. Martin, Yuki Sawai 

Distribution and sedimentary characteristics of tsunami deposits on Phra Thong 

Island, Thailand ................................................................................................................... 99 

Kázmér, M., Kamol Sanittham, Punya Charusiri, Santi Pailoplee 

Archaeoseismology of the AD 1545 Earthquake in Chiang Mai, Northern Thailand ...................................... 102 

Kinugasa, Yoshihiro 

Outline of the 3.11 Tohoku Earthquake in Japan.............................................................................. 106 

Koster, Benjamin, Klaus Reicherter, Andreas Vött, Christoph Grützner 

The evidence of tsunamigenic deposits in the Gulf of Corinth (Greece) w ith geophysical 

methods for spatial distribution .................................................................................................. 107 

Kübler, S., A. M. Friedrich, M. R. Strecker 

Coseismic surface rupturing in the epicentral area of Germany’s strongest historical 

earthquake ......................................................................................................................... 111


Lazauskiene, Jurga and Andrius Pacesa 

Seismotectonic and seismic hazard maps of Lithuania (Baltic region) – recent implications 

of intracratonic seismicity......................................................................................................... 114 

Lee, Minjung and Young-Seog Kim 

Preliminary study on damaged stone monuments in Gyeongju, SE Korea................................................. 118 

Malik, Javed N., Michio Morino, Mahendra S. Gadhavi, Khalid Ansari, Chiranjeeb Banerjee, 

B. K. Rastogi,Fumio Kaneko, Falguni Bhattacharjee, Ashok. K. Singhvi 

Earthquake geology and related hazard in Kachchh, Gujarat, Western India.............................................. 121 

Marco, Shmuel and G. Ian Alsop 

Seismogenic slumps in palaeo-dead sea sediments.......................................................................... 125 

McCalpin, James P. 

Mapping and measuring Holocene fault scarps in dense forests with LIDAR.............................................. 128 

Meilianda, Ella, Ben Maathuis, Marjolein Dohmen-Janssen 

Changes on the geomorphic settings of sand-poor environment coast of Banda Aceh, 

Indonesia subject to tectonic and tsunami events ............................................................................. 132 

Meskouris, Konstantin, Britta Holtschoppen, Christoph Butenweg, Julia Rosin 

Seismic analysis of liquid storage tanks ........................................................................................ 136 

Michetti, Alessandro M., Leonello Serva, Andrea Berlusconi, Livio Bonadeo, Fabio 

Brunamonte, Francesca FerrarioGianfranco Fioraso, Franz Livio, Giancanio Sileo, Eutizio 

Vittori 

Geological criteria for evaluating seismicity: Lessons learned from the Po Plain, Northern 

Italy ................................................................................................................................. 140 

Mishra, Anurag, D.C. Srivastava, Jyoti Shah 

Ancient seismites as geodynamical indicator: approach to construct a reactivation event on 

the main boundary thrust in the Himalayan region ............................................................................ 144 

Mouslopoulou Vasiliki, Andrew Nicol, John J. Walsh, John G. Begg, Dougal B. Townsend, 

Dionissios T. Hristopulos 

Sampling biases in the paleoseismological data .............................................................................. 148 

Niemi Tina M. 

Earthquakes in Aqaba, Jordan over the past 2,000 years: Evidence from historical, 

geological, and archaeological data............................................................................................. 152 

Nomikou, P., Alexandri M., Lykousis V., Sakellariou D., Ballas D. 

Sw ath bathymetry and morphological slope analysis of the Corinth Gulf................................................... 155 

Pantosti Daniela, Stefano Pucci, Paolo Marco De Martini, Alessandra Smedile 

Is the decadence of Leptis Magna (Lybia) the consequence of a destructive earthquake? .............................. 159 

Papageorgiou Elena and Paraskevi Nomikou 

On-shore prolongation of bathymetrically recognized fault zones based on geodetic GPS 

observations along Santorini Volcano .......................................................................................... 163 

Papaloizou Loizos and Petros Komodromos 

The dynamic analysis of multi-drum ancient structures under earthquake excitations.................................... 167 

Papanikolaοu Dimitrios, Royden Leigh, Vassilakis Emmanuel 

Neotectonic and active diverging rates of extension in the Northern and Southern 

Hellenides across the Central Hellenic Shear Zone........................................................................... 170 

Papanikolaοu Ioannis and Gerald Roberts 

Clustering and anticlustering in the Southern Apennines as evidenced from geological fault 

slip-rate seismic hazard maps and the historical record ...................................................................... 174 

Papanikolaοu Ioannis, Gerald Roberts, Georgios Deligiannakis, Athina Sakellariou, 

Emmanuel Vassilakis 

The Sparta Fault, Southern Greece: tectonic geomorphology, seismic hazard mapping and 

conditional probabilities........................................................................................................... 178


Papanikolaοu Ioannis, Maria Triantaphyllou, Aggelos Pallikarakis, Georgios Migiros 

Active faulting tow ards the Eastern tip of the Corinth Canal: Studied through surface 

observations, borehole data and paleoenvironmental interpretations....................................................... 182 

Passchier, Cees W., Gilbert Wiplinger, Gül Sürmelihindi, Paul Kessener, Talip Güngör 

Roman aqueducts as indicators of historically active faults in the Mediterranean Basin ................................. 186 

Pérez-López, R., J.L. Giner-Robles, M.A. Rodríguez-Pascua, F. Martín-González,J. García 

Mayordomo,J.A. Álvarez-Gómez, M.J. Rodríguez-Peces, J.M. Insua-Árévalo, J. J. 

Martínez-Díaz and P.G. Silva 

Testing archaeoseismological techniques w ith instrumental seismic data caused by the M 

5.1 Lorca Erthquake (5-11-2011, SE of Spain) ................................................................................ 190 

Reicherter, Klaus 

Frontiers of earthquake archaeology: the Olympia and Samicum cases (Peloponnese, 

Greece)............................................................................................................................. 194 

Roberts, Gerald, Joanna Faure Walker, Patience Cowie, Richard Phillips, Ken McCaffrey, 

Ioannis Papanikolaou, Max Wilkinson, Alessandro Michetti, Peter Sammonds 

Regional strain-rates on active normal faults and variability in the seismic cycle: an 

example from the Italian Apennines............................................................................................. 198 

Rockwell, Thomas K. and, Yann Klinger 

The variability of along-strike co-seismic slip: a new example from the Imperial Fault of 

Southern California................................................................................................................ 200 

Rodríguez-Pascua M.A., P.G. Silva, Perucha Atienza, M.A., J.L. Giner-Robles, R. Pérez- 

López 

Earthquake archaeological effects generated by the Lisbon Earthquake (first of November 

1755) in the Coria´s Cathedral (Cáceres, Western Spain) ................................................................... 204 

Rojas, Wilfredo, Nury Simfors-Morales, Luis Sanez, Åke Sivertun 

Neotectonic of the longitudinal fault system in Southern Costa Rica ....................................................... 207 

Ruano, Patricia, Antonio J. Gil, Jesús Galindo-Zaldívar, Gracia Rodríguez-Caderot, María 

Clara de Lacy, Antonio M. Ruiz, María Jesús Borque, Juan A. Armenteros, Antonio 

Herrera, Antonio Jabaloy, Angel C. López-Garrido, Antonio Pedrera, Carlos Sanz de 

Galdeano 

Geodetic studies in the Zafarraya Fault (Betic Cordilleras) .................................................................. 210 

Rudersdorf, Andreas, Jochen Hürtgen, Christoph Grützner, Klaus Reicherter 

Neotectonic activity of the Granada Basin – new evidence from the Padul-Nigüelas Fault 

Zone ................................................................................................................................ 214 

Sakellariou, Dimitris, Lykousis Vasilis, Rousakis Grigoris 

Holocene seafloor faulting in the Gulf of Corinth: the potential for underw ater 

paleoseismology................................................................................................................... 218 

Schreiber,Stephan and Klaus-G. Hinzen 

Damage assessment in archaeoseismology: methods and application to the archaeological 

zone Cologne, Germany ......................................................................................................... 222 

Scholz, Christopher H. 

Earthquake Triggering, Clustering, and the Synchronization of Faults ..................................................... 226 

Silva, Pablo G., Alex Ribó, Moises Martín Betancor, Pedro Huerta, M. Ángeles Perucha, 

Cari Zazo, Jose L. Goy, Cristino J. Dabrio, Teresa Bardají 

Relief production, uplift and active tectonics in the Gibraltar Arc (South Spain) from the Late 

Tortonian to the Present .......................................................................................................... 227 

Sintubin, Manuel, Simon Jusseret, Jan Driessen 

Reassessing ancient earthquakes on Minoan Crete getting rid of catastrophism ......................................... 231 

Smedile, Alessandra, Paolo Marco De Martini, Daniela Pantosti 

Paleotsunamis evidence from a combined inland and offshore study in the Augusta Bay 

area (Eastern Sicily, Italy) ........................................................................................................ 233 

Solakov D., L.Dimitrova, S.Nokolova, St. Stoyanov, S. Simeonova, L. Zimakov, L.Khaikin 

Bulgarian National Digital Seismological Netw ork............................................................................. 235


Štěpančíková Petra, Nývlt Daniel, Hók Jozef, Dohnal Jiří 

Paleoseismic study of the Sudetic marginal Fault at the locality Bílá Voda (Bohemian 

Massif) .............................................................................................................................. 239 

Tsang, Rebecca Y., Thomas K. Rockwell, Aron J. Meltzner, Paula M. Figueiredo 

Tow ard development of a long rupture history of the Imperial Fault in Mesquite Basin, 

Imperial Valley, Southern California............................................................................................. 243 

Vacchi, Matteo, Alessio Rovere, Nickolas Zouros, and Marco Firpo 

Mapping paleo shorelines in Lesvos Island: new contribution to the Late Quaternary relative 

sea level changes and to the neotectonics of the area ....................................................................... 247 

Vacchi, Matteo, Alessio Rovere, Marco Firpo and Nickolas Zouros 

Boulder deposits in Southern Lesvos: an evidence of the 1949’s Chios-Karaburum 

Tsunami?........................................................................................................................... 251 

Valkaniotis, Sotiris, George Papathanassiou, Spyros Pavlides 

Active faulting and earthquake-induced slope failures in archeological sites: case study of 

Delphi, Greece..................................................................................................................... 255 

Vött, Andreas, Peter Fischer, Hanna Hadler, Mathias Handl, Franziska Lang, Konstantin 

Ntageretzis,Timo Willershäuser 

Sedimentary burial of ancient Olympia (Peloponnese, Greece) by high-energy flood 

deposits – the Olympia tsunami hypothesis.................................................................................... 259 

Vollmert, Andre, Klaus Reicherter, Pablo G. Silva, Tomas M. Fernandez-Steeger 

Landslide mapping to analyse Earthquake Environmental Effects (EEE) in Carmona, Spain 

– relation to the 1504 event?..................................................................................................... 263 

Wartenberg, Wolfram, Andreas Vött, Hanna Hadler, Timo Willershäuser, Holger Freund, 

Stefanie Schnaidt 

Storm surge layers w ithin a changeful Holocene environment or sedimentary traces of 

palaeo-tsunamigenic events? Pros and Cons of on-site findings, Jade Bay, Southern North 

Sea, Germany ..................................................................................................................... 267 

Wechsler, Neta, Thomas K. Rockwell, Yann Klinger, Amotz Agnon, Shmulik Marco 

Testing earthquake recurrence models w ith 3D trenching along the Dead-Sea Transform .............................. 269 

Wiatr, Thomas, Ioannis D. Papanikolaοu, Klaus Reicherter, Tomás Fernández-Steeger 

A terrestrial close range view of the normal fault zone near Archanes (Heraklion Basin, 

Crete) ............................................................................................................................... 272 

Wiatr, Thomas, Klaus Reicherter, Ioannis D. Papanikolaοu & Tomás Fernández-Steeger 

The discontinuity of a continuous fault: Delphi (Greece) ..................................................................... 276 

Wilkinson, Maxwell, Ken McCaffre, Gerald Roberts, Patience Cowie, Richard Phillips 

Postseismic deformation of the 2009 L‘ Aquila earthquake surface rupture measured using 

repeat terrestrial laser scanning ................................................................................................ 280 

Willershäuser, Timo, Andreas Vött, Georg Bareth, Helmut Brückner, Hanna Hadler, 

Konstantin Ntageretzis 

Sedimentary evidence of Holocene tsunami impacts at the Gialova Lagoon (Southw estern 

Peloponnese, Greece) ........................................................................................................... 283 

Winandy, Jonas, Christoph Grützner, Klaus Reicherter, Thomas Wiatr, Peter Fischer, 

Thomas Ibeling 

Is the Rurrand Fault (Low er Rhine Graben, Germany) responsible for the 1756 Düren 

Earthquake Series? ............................................................................................................... 286 

YaoDa-quan, Shuo Zhi,Tang Ji-ping, Wang Zhi,Shen Xiao-qi,Chen An-guo, Li Lin-li 

Find and primary search of an active fault at the Gaixia Site, Guzhen County, Anhui 

Province, P.R. China.............................................................................................................. 290 

Zygouri, Vasiliki, Sotiris Kokkalas, Paris Xypolias, Ioannis Koukouvelas, Gerassimos 

Papadopoulos 

The Movri Mountain Earthquake: understanding active deformation of the NW Peloponnese........................... 292

M.A Rodríguez Pascua et al. (2011). Quaternary International, 242 (1), 20-30. Sp. Vol. on 

Earthquake Archaeology and Paleoseismology (P.G. Silva, M. Sintubin & K. Reicherter Eds.) 

INQUA 

TERPRO 

Paleoseismology 

and ActiveTectonics 

Proceedings INQUA - IGCP 567 CORINTH (GREECE), 2011

Proceedings - Johannes Gutenberg-Universität Mainz

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?