When Ted Shepherd calls me on his way to the airport, he has a predicament. He’s planning on flying to Japan and, even though there was a forecast for a typhoon in the region, he’s weighed up his chances and figures he has a strong chance of making it there safely. “You know, it’s probabilistic,” he explains. “But I have to make a decision.”
As an expert in risk and climate modelling at the University of Reading, he was probably in a better position than the average individual to make that judgement. But he acknowledges that uncertainty is difficult to communicate. “We can present them as probabilistic, but when people pick up a newspaper, they’re looking at them as deterministic,” he says.
Extreme weather forecasting – with near-apocalyptic headlines – tends to be the domain of the tabloid press, some of which have a track record of climate change denial. And if you put any stock in tabloid headlines, it might seem that the age of extreme weather is on us: “-3C Arctic FREEZE TONIGHT before bizarre tropical ‘battle’ next week”, reads one from The Daily Express. “Hottest October day in SEVEN years hits today as temperatures reach 77F but a month’s worth of rain is on the way to fall in just three days”, suggests MailOnline. Maybe we’ll all have to make those decisions soon.
These predictions – which look at the upcoming season and declare it will be the worst one on record – are spurious and sensationalised. Weather conditions are becoming more extreme with each passing year, but it remains incredibly difficult for seasoned weather forecasters to accurately predict what this winter has in store for us. Why? It’s complicated.
“Tabloids don’t really like to talk about climate change itself, but they talk about the weather instead,” Shepherd says. “When we talk about climate change at a local level, we’re talking about the weather – those stories can be exciting, but they’re human stories too.”
In order to understand why we can trust scientists on long-term research looking at climate change, or weather forecasters on if it will rain tomorrow, but not what lies between, the first thing to clarify is the relationship between climate and weather. “When we talk about weather, we’re talking about climate at the average level, so what it feels like to live in London in October,” says Colin Cotter, who works in numerical analysis and computing at Imperial College London. “The weather is changing all the time, but we can view it as some variations around the average, which is set by the climate.”
That distinction is what causes the split – so while you can check the BBC forecast for the weekend weather and find it useful, after a certain amount of time it becomes harder to state with any real certainty that a certain kind of weather condition will prevail. Since we don’t even fully know the current state of our atmosphere, we can never have perfect predictions.
This is down to what’s known as the butterfly effect. Coined by Edward Lorenz in the 1980s, it essentially means that small changes in one area of the atmosphere can have a huge effect elsewhere – so, slight changes in atmospheric wind pressure can lead to huge changes in wave height. It’s also well documented that the butterfly effect has a very pronounced impact on the mid-latitudes, where the UK is located. So, Cotter suggests, it’s only possible to obtain a useful forecast for around seven days in the future.
“It is very difficult because seasonal forecasting sits in the difficult regime between weather and climate,” he adds. “The current state of the system still has significant impact on the forecast, but is very entangled with the chaotic behaviour of the atmospheric system.”
But Adam Scaife, the head of long range prediction at the Met Office, says that it doesn’t mean we have to give up on weather forecasting altogether. “To account for the butterfly effect we have to make dozens of forecasts,” he says. What tends to happen is that after the dozens of forecasts are made, they are averaged out (using advanced computing models) and they can give a picture of the most likely scenario. This is why weather forecasters can say that winter will be cold, but they can’t tell you what the temperature will be on December 15 until maybe a few days before. As models become more advanced, those impressions can become more accurate, but it’s a long, complicated process.
Even so, the Met Office has stopped issuing seasonal forecasts. This is because it’s hard to predict with any level of local detail what the conditions of a season will be in the future. The Met Office now issues long-range weather forecasts, about a month into the future. These look at the UK as a whole, rather than focusing on specific regions.
Read more: Could we reverse global warming by redesigning the atmosphere?
Long term weather forecasts are often useful for businesses who will find themselves impacted by small changes in business conditions, such as crop prices in energy markets. But these businesses have the capabilities to use those weather forecasts in the ways that make sense for them, which individuals generally lack the skill set to do.
“It’s a question of communication. How do you communicate that risk, and do it in a way that’s useful and honest?” says Shepherd. He suggests that a storytelling paradigm – where you present several probable weather forecasts to people, and they look at the evidence and weigh up what outcome is likely to happen – might be more useful than out-and-out predictions of doom.
What’s bizarre is that newspapers fixated on sensationalist weather headlines often also deny climate change – and have spent years pushing back on the issue from scientific bodies, including the Met Office. By continuing to push ridiculous claims with a tenuous link to reality, we risk completely eroding the basis of belief in weather models, which will have a harmful effect in the long run.
This article was originally published by WIRED UK
