On the first day of the new year, eight countries and hundreds of cities and towns across Europe recorded their all-time highest January temperatures, in some cases by almost 10 degrees Fahrenheit. With records falling from the Netherlands to Belarus, experts called it “the most extreme heat wave in European history.”
This may sound familiar. The number of unprecedented climate and weather disasters has exploded in the past several years, suggesting that scientists underestimated the growing potential for extreme outlier events — such as 2022′s deadly and damaging heat wave in India and Pakistan, the destructive flooding in Yellowstone National Park and the U.K.’s all-time highest temperatures last July. Recent experience suggests that as the world warms, these events might not be such outliers.
“What we’re being faced with … [is] an increasing occurrence of events that are more extreme than any that that we’ve encountered in our historical experience,” said Noah Diffenbaugh, a professor of earth system science at Stanford University. Even as the range of future warming estimates has somewhat narrowed in recent years, the ferocity of individual heat waves, floods and fires continue to catch the world by surprise.
Can’t rely on history
Soaring temperatures across the Pacific Northwest in the summer of 2021 killed hundreds of people. The heat wave was so extreme, and so far outside the range of what was considered likely or even possible, that it caused some scientists to question the computer models used to simulate the future climate.
In fact, a study on that heat wave determined that while the models could theoretically account for an event that extreme, it would only occur around once every 100,000 years. Either the region got remarkably unlucky, or the models — and our historical climatological record — aren’t properly capturing the odds of such extremes.
Diffenbaugh told Grid that climate models have done a reasonable job in capturing the potential for outlier events, but relying on past events and statistics as harbingers of the future is particularly fraught since the climate is changing. He published a study in 2020 that demonstrated the problem. If researchers examine the historical record and calculate the likelihood of extreme events based on that record, “that analysis shows that there’s a huge underestimation for extreme hot events and extreme wet events,” he said.
The climate of the present is no longer like the climate of the past — so looking backward isn’t a good way to gauge what the future might bring.
This past summer, another “unprecedented” event struck outside of most predictions when monsoon rains in Pakistan submerged one-third of the country and displaced tens of millions of people. Regional weather authorities did predict the region would have “normal to above normal” monsoon rains, but what eventually fell was almost a four-fold increase over average.
Again, recent research suggests that such underestimation of precipitation extremes is likely. Last summer, researchers at Yale and the University of Oslo in Norway showed that models could underestimate the risks of future flooding events by as much as twofold, a difference the authors called “staggering.”
Short-changing the potential for extreme events has wide-ranging implications for everything from emergency management to building and highway codes.
“If we can’t quantify the impacts, we don’t know how much they will cost us,” said Jacob Schewe, a scientist at the Potsdam Institute for Climate Impact Research in Germany. “It’s hard to plan and to adapt to them. ... To make those decisions, you need to be able to quantify at least roughly what the risks are.”
Schewe led a 2019 study that found a variety of other types of computer models — agricultural output models, hydrology and water impacts, and excess mortality models, among others — can underestimate the impacts of extreme events. They used the devastating 2003 European heat wave, which killed tens of thousands of people, particularly in France, as a test case. They found a wide range of impacts that outstripped what models would predict — including deaths, which in some cities like Paris were underestimated by half.
In some sense, this means the world is doubly unprepared: The extreme events themselves may exceed the predictions of both historical records and climate models, and the impacts on humans and other systems may leap past what computer programs warn about.
Diffenbaugh pointed out that the impacts of extreme events depend on how humans have prepared for them and that there are often there are tipping points beyond which that preparation fails. “There are lots of examples where you’re safe below the threshold, and there’s literal catastrophe beyond the threshold,” he said. “Flooding is a great example.” If a levee can handle a certain amount of water, everything is mostly fine up until that amount is exceeded — and then everything is very much not fine.
Not just a short-term problem
Outside of acute disasters, some longer-term climate change impacts are also coming into clearer focus. A new paper published Wednesday in Science found that previous estimates of glacier ice loss are also likely underestimated — but every fraction of a degree of warming avoided could cut those losses sharply.
Compared to data included in the most recent Intergovernmental Panel on Climate Change report as well as other glacier computer modeling, “we’re projecting more glacier mass loss and a greater glacier contribution to sea level rise than they did,” said David Rounce, an assistant professor of civil and environmental engineering at Carnegie Mellon University.
Using data that examines the world’s 215,000 glaciers individually, Rounce and his colleagues found that even at 1.5 degrees Celsius (2.7 degrees Fahrenheit) warming — the low end of Paris Agreement targets that is still achievable but slipping out of reach with each passing day — 26 percent of the ice present in 2015 will disappear by 2100. With more warming, that number could balloon to 41 percent. Globally, that’s as much as an 8 percent jump from previous best estimates, and in some regions — such as Alaska and the southern Andes — the difference is upward of 20 percent.
Rounce said that losing glaciers can have wide-ranging cultural impacts, and the melt will also contribute more to sea level rise than previously anticipated — 115 millimeters (4.3 inches) by 2100 if the current warming trajectory is maintained, or around half the amount of rise observed since 1880. And he added that while the new study’s projections are on the scale of decades, acute extreme events like heat waves can play a huge role in what actually happens to the world’s glaciers.
“You can have these extreme summers, extreme heat waves like Europe experienced last summer, that can really decimate a large fraction of the glacier volume,” he told Grid. “It can kind of cause this additional mass loss than just a regular model would predict.” There are further connections as well. The additional sea level rise glaciers may contribute can worsen coastal flooding when extreme events like hurricanes strike.
In general, there is widespread interest in an improved picture of climate extremes as adaptation efforts across the world ramp up. “That’s a very important field of research, to understand better the tail of the distribution … the kind of the freak cases,” Schewe said from Germany, where nearly 1,000 local heat records fell in the first few days of the year. “It will become the new normal, or at least much more frequent in the future.”
Thanks to Brett Zach for copy editing this article.