The Washington PostDemocracy Dies in Darkness

Arctic climate change may not be making winter jet stream weird after all

New research is pouring cold water on once-hot theory

Jet stream visualization. (NASA)

An influential, highly publicized theory — that a warming Arctic is causing more intense winter outbreaks of cold and snow in midlatitudes — is hitting resistance from an ongoing sequence of studies, including the most comprehensive polar modeling to date.

The idea, first put forth in a 2012 paper by Jennifer Francis, now at the Woodwell Climate Research Center, and Stephen Vavrus, at the University of Wisconsin at Madison, is that two well-established trends — Arctic amplification (intensified global warming at higher latitudes) and depleted sea ice — can force the polar jet stream to dip farther south, thus causing more intense bouts of winter weather than might have otherwise occurred.

From 2012: Shrinking Arctic ice and the wicked backlash on our weather

Over the past decade, this hypothesis sparked widespread public interest and scientific debate, as various high-profile cold waves and snow onslaughts hit North America and Eurasia, including a deadly, prolonged cold wave in Texas last February.

The Texas power grid failure that left millions without power during the week of Feb. 15 prompted calls for a system more resilient to extreme weather. (Video: The Washington Post)

Winter temperatures over the past three decades have shown cooling in some areas of the northern midlatitudes, especially eastern Asia.

But the cooling has been far from ubiquitous and the Arctic-midlatitude link has been difficult to detect in simulations by global computer models. Instead, the models point more strongly toward the gradual, longer-term trend of milder midlatitude winters that one would expect in a human-warmed climate. (A separate line of research is addressing extremes during the summer, such as the unprecedented heat wave that struck the Pacific Northwest in June; see below.)

Some emerging work, not yet peer-reviewed, does reveal faint fingerprints of the Francis-Vavrus hypothesis in new simulations of Arctic and midlatitude winter climates, part of the Polar Amplification Model Intercomparison Project (PAMIP).

“It does look like there’s something there, but it looks fairly weak,” James Screen, a climate scientist at the University of Exeter, said in an interview. “That doesn’t mean it’s not one part of the overall jigsaw.”

“There’s genuine reason to be concerned about climate change and its effects,” Screen added, “but an increase in the frequency in cold events would not be at the top of my list.”

Screen has published several related studies with Russell Blackport, now at Environment and Climate Change Canada, over the past two years. The title of one of their latest, published in April in the Journal of Climate, is itself an assertion: “Observed Statistical Connections Overestimate the Causal Effects of Arctic Sea Ice Changes on Midlatitude Winter Climate.”

In this study and others, Screen and Blackport suggest that the connection between Arctic sea ice loss and extreme midlatitude events is real, but not necessarily causal. Instead, they argue, a third factor — most likely large-scale changes in atmospheric circulation that may not be permanent — is probably driving both the sea ice loss and the extreme winter events.

Francis asserts that the new PAMIP work actually lends support to some of the building blocks of her hypothesis, even if the model findings aren’t as strong as recent observations: “The jet stream shifted southward, just like we expect to see as the Arctic warms,” Francis said in an interview. “The zonal [west-to-east] winds were weaker when there was less sea ice. All these things we’ve been talking about for almost 10 years, they found in their models.”

At the same time, Francis maintains that global models can’t fully replicate the Arctic-midlatitude connection simply by adding or removing sea ice, especially if the results lump together conditions across a whole winter or the entire midlatitudes.

“If you’re interested in things like whether the jet stream is going to get wavier, whether you’re going to get more extreme weather events, more blocking — these are all very regional, short-term responses,” Francis said.

Satellite-based images from NASA show the recession of Arctic sea ice from March 2020 to September 2020. (Video: NASA)

Francis also points to work led by Zachary Labe of the University of California at Irvine, which published in August 2020 in Geophysical Research Letters. Labe and colleagues found that when they modeled not only sea ice loss but also the deep warming observed in the Arctic atmosphere, then a distinct pattern emerged, with a stronger Siberian High and colder winter temperatures across eastern Asia.

As it happens, eastern Asia is one of the few regions on Earth where winters have trended cooler over the past several decades. The winter of 2020-21 was colder than the 1981-2010 average across most of Russia and northern Asia.

Like Francis, Screen is intrigued by the Labe paper: “It’s interesting, but I’m not entirely convinced — I’d like to see that result confirmed in other studies.”

The appeal of a counterintuitive idea

With mountains of research pointing to so many dire consequences of human-caused climate change — some of them already taking shape, such as the ominous growth of “hot droughts” in the U.S. Southwest — it’s easy to see how winter storms and cold blasts could be accepted as just another negative outcome, albeit a freakish one.

“The generalization that climate change makes extreme weather worse causes people to assume it makes all extreme weather worse,” said Texas state climatologist John Nielsen-Gammon. He’s now analyzing the extreme cold and snow that paralyzed his state in February. No studies have yet tried to directly attribute any or all of the Texas event to climate change.

While the cold blast was on par with some of the worst in Texas history, the catastrophic consequences — close to 150 deaths and $20 billion in damage, according to reinsurance broker Aon — were widely considered to be the result of the state’s main power grid operator, ERCOT, being unprepared for such high-end events in a fast-growing region.

Texas cold snap was not ‘unprecedented,’ and it was inexcusable to be unprepared

Nielsen-Gammon found that the statewide average temperature, weighted by population within the ERCOT area, was the third coldest on record on Feb. 15, just behind Feb. 12, 1899, and Dec. 23, 1989. The coldest nonoverlapping seven-day period of the cold wave was in a virtual tie with three other historical periods for second place, with a week in December 1983 outdoing all of them.

On the whole, according to Nielsen-Gammon, extreme cold in Texas is getting less rather than more common.

“No matter how they’re measured, cold-air outbreaks have decreased in frequency and severity in Texas and in most locations in the Northern Hemisphere midlatitudes,” Nielsen-Gammon said.

Across Texas, he noted, the coldest readings in each winter month have been on the rise over the past several decades — in fact, rising twice as quickly as the average winter temperature.

Blackport and Screen found much the same for midlatitude winters across the Northern Hemisphere. In a letter published last December in Nature Climate Change, they found that as a group, the coldest readings per winter at each point were rising at a more rapid pace than average winter temperatures at each point.

Stay tuned

The lead players in this decade-long debate agree on one thing: The case isn’t fully settled. “A lot of progress has been made, but there’s still a lot of open questions,” Screen said.

Francis and colleagues, including Vavrus and Judah Cohen, who directs seasonal forecasting at Atmospheric and Environmental Research, are now working on a paper that will employ self-organizing maps, an increasingly popular tool for teasing out relationships that might otherwise go unanalyzed.

Meanwhile, research by Michael Mann at Pennsylvania State University and colleagues suggests that the “stuckness” of some recent summertime patterns, often leading to prolonged heat waves or flooding, may be related to resonant jet stream modes induced by global warming that lead to long-lasting, nearly stationary upper lows and highs. The last generation of global climate models couldn’t capture this resonance, Mann and colleagues found.

A quick-response study coordinated by the World Weather Attribution group concluded that the Pacific Northwest heat wave of late June would have been “virtually impossible without human-caused climate change.” Although the report did not analyze the jet stream hypothesis directly, it said that nonlinear interactions not captured by climate models might have boosted the odds of an event this extreme. The other possibility, they said, is that the heat wave was an exceedingly low-probability event made more probable by overall rising temperatures — “the statistical equivalent of really bad luck, albeit aggravated by climate change.”

Pacific Northwest heat wave was ‘virtually impossible’ without climate change, scientists find

Because there are different physics at work in the climate extremes of summer vs. winter, Mann isn’t sure if the same kind of modeling issues might be playing out with winter events, “but my gut instinct is that they may well be.”

Overall, Mann said in an email, “I think the jury is still very much out. And I applaud both ‘camps’ for continuing to examine the best available data and models to get to the bottom of this. Eventually, this work will converge toward a scientific consensus. It’s not easy, but it’s how the gears of science turn, and it’s how we move forward toward a better physical understanding of the world.”

Bob Henson is a meteorologist and journalist based in Boulder, Colo. His books include “The Thinking Person’s Guide to Climate Change” and “Weather on the Air: A History of Broadcast Meteorology.”