New research raises the uncomfortable possibility that climate change will not only make weather more severe but also harder to predict, potentially giving us less time to prepare for extreme floods, storms and heat waves in the years to come.
“It seems that colder climates are just inherently more predictable than warmer ones,” said Aditi Sheshadri, an atmospheric scientist at Stanford University and lead author of the paper.
In the current climate, the practical extent of reliable weather forecasting is about 10 days, owing both to the limits of forecasting technology and to the complexity of weather itself. Even the tiniest unaccounted-for detail in a weather model — the proverbial flap of a butterfly’s wing — can introduce error.
Small errors grow into larger errors, causing the model to diverge from the actual weather over time. Knowing more about the initial conditions can make forecasts marginally more accurate, but weather is fundamentally chaotic, and beyond a certain point, the future is unknowable, at least in any great detail.
To account for errors, meteorologists will run an ensemble of weather models, inputting slightly different initial conditions into each model and then watching where the models agree and disagree. As they look deeper into the future, the models will branch further and further apart, until they bear as much resemblance to each other as to a model based on completely different inputs. This is the point of “error saturation,” when the models “lose memory” of the initial conditions, as Sheshadri wrote.
In the paper, she and her co-authors asked whether weather models would reach this point more quickly in a warmer climate. They ran a weather ensemble in computer simulations of both warmer and cooler climates, measuring the time to error saturation. The study focused on the middle latitudes — which include the United States, Europe and China — finding that, in warmer climates, storms grow more quickly and errors propagate faster.
“Its time to error saturation, which is associated with the accurate window of predictability, that time scale is just shorter in a warmer climate,” Sheshadri said.
At 5.4 degrees (3 degrees Celsius) of warming, the window for accurately predicting precipitation shrinks by about one day in the mid-latitudes, according to the study. The window closes more slowly for predictions of wind and temperature, shortening by about half a day.
“It’s an interesting idea to ask the question, ‘Are weather forecasts going to be more or less predictable in the future under climate change?’ ” said Timothy Palmer, an Oxford physicist specializing in the predictability of weather who was not affiliated with the research. He said the study perhaps raises more questions than it answers, relying, as it does, on highly simplified simulations of climate change.
In the simulations, for instance, the equator heats up faster than the poles, but, in reality, the picture is more complicated: At high altitudes, the tropics are warming faster than the poles; at the Earth’s surface, the poles are warming faster than the tropics.
Simplified, or “idealized,” simulations are useful for allowing researchers to home in on a few key variables, but in glossing over some of the complexity of the actual climate, they may not reveal all the changes that scientists would expect to see in real life.
“I think they’ve revealed some interesting sensitivities, but it would be nice to then go and test this in a way that’s less idealized,” said Isla Simpson, a climate scientist at the National Center for Atmospheric Research, who was unaffiliated with the study.
“It’s a good starting point,” she added. “And I think it will motivate other studies that look at this from a more comprehensive viewpoint.”
Simpson said that, in addition to deploying a more realistic representation of climate change, future studies should investigate whether predictability will vary by season or region. Although weather forecasts may grow less reliable in some parts of the world, they could become more reliable in others.
Future research should also investigate the different ways that climate change could frustrate weather forecasting, Palmer said. Extreme rainfall and longer dry spells are already difficult to predict several days in advance, and rising temperatures will yield more of both phenomena.
One saving grace is that improved weather forecasting could allow meteorologists to continue to produce accurate predictions even if climate change introduces more chaos into the system.
“It’s unclear whether this effect would be big enough to counteract any advances we may make as computing power becomes better or our models advance,” Simpson said. “Climate change is not the only factor that’s going to affect our skill at weather forecasts in the future.”