Landmark study casts doubt on controversial theory linking melting Arctic to severe winter weather
By Paul Voosen
Every time severe winter weather strikes the United States or Europe, reporters are fond of saying that global warming may be to blame. The paradox goes like this: As Arctic sea ice melts and the polar atmosphere warms, the swirling winds that confine cold Arctic air weaken, letting it spill farther south. But this idea, popularized a decade ago, has long faced skepticism from many atmospheric scientists, who found the proposed linkage unconvincing and saw little evidence of it in simulations of the climate.
Now, the most comprehensive modeling investigation into this link has delivered the heaviest blow yet: Even after the massive sea ice loss expected by midcentury, the polar jet stream will only weaken by tiny amounts—at most only 10% of its natural swings. And in today’s world, the influence of ice loss on winter weather is negligible, says James Screen, a climate scientist at the University of Exeter and co-leader of the investigation, which presented its results last month at the annual meeting of the European Geosciences Union. “To say the loss of sea ice has an effect over a particular extreme event, or even over the last 20 years, is a stretch.”
The idea that Arctic sea ice loss could influence midlatitude winter weather first gained traction in 2012, in a paper by two climate scientists, Jennifer Francis, now at the Woodwell Climate Research Center, and Stephen Vavrus at the University of Wisconsin, Madison. It started with a simple observation: The Arctic is warming nearly three times faster than the rest of the world. At the time, sea ice loss was thought to be the primary accelerant for this amplification: As bright, reflective ice is replaced by dark, sunlight-absorbing water, the Arctic heats up, causing more ice loss, and more warming in turn.
The warming, Francis and Vavrus proposed, would inflate the height of the polar troposphere—the lowest layer of the atmosphere and home to its weather. That would decrease the pressure differences between polar and midlatitude air that drive the polar jet stream, which separates the air masses and keeps cold air collared around the pole. The jet would grow weaker and wavier, allowing cold air to intrude farther south. In their paper, Francis and Vavrus argued such a trend was visible in weather records and worsening with Arctic warming and ice loss.
A lot has changed since then, Francis now says. “Like all things, as you dig into them, they become more complicated.” Most significantly, the 25-year trend that she and others had identified in observations from the late 1980s to early 2010s has weakened after another decade of observations. Although sea ice loss has continued, there are few signs of colder winters in Eurasia or North America, more cold extremes, or more frequent weakening or waviness in the jet stream. The new computer modeling matches the observations, says Doug Smith, a climate scientist at the United Kingdom’s Met Office and another co-leader of the modeling effort. “There’s not an inconsistency.”
In the yearslong investigation, called the Polar Amplification Model Intercomparison Project (PAMIP), researchers ran more than a dozen climate models 100 times each. One set of model runs simulated the Arctic atmosphere without pronounced sea ice loss, using ocean temperatures and sea ice extent from 2000. The other kept the ocean temperatures the same, but reduced the ice coverage to the extent expected decades from now, after 2°C of global warming, when the Arctic could be ice free in the summer. Keeping the oceans the same should highlight the influence—if any—of sea ice loss.
In addition to finding only a tiny effect of sea ice loss on the polar jet stream, the models also found no coherent sign of a second proposed effect of reduced sea ice: more frequent disruptions of the stratospheric polar vortex—a second set of swirling winds, much higher up. Such disruptions, which occur every 2 years on average, ultimately allow cold air lower in the atmosphere to spill southward, causing extreme winter storms, including the cold that gripped Texas this past winter.
Judah Cohen, director of seasonal forecasting at Atmospheric and Environmental Research, has long argued that increased snow cover and diminished sea ice in Siberia favor weather patterns that propagate energy into the stratosphere, making the high-altitude disruptions more frequent. But although some model runs show this happening, on average “there is no clear response,” says Yannick Peings, a climate scientist at the University of California (UC), Irvine.
Cohen isn’t convinced, noting that the models also forecast unrealistically warm winter weather in the midlatitudes, making other predictions suspect. “There’s clearly something missing.” And Francis says the PAMIP experiment may be too simplistic, now that “we know there’s a lot more to Arctic amplification than sea ice loss.” Satellites and weather balloons have shown that the high troposphere in the tropics is warming fast because of tremendous storms that shoot hot, humid air upward. The Arctic is much less stormy, but many scientists now believe so-called atmospheric rivers regularly deliver this warm tropical air to the Arctic—a mechanism that PAMIP ignored.
Several PAMIP scientists, including Peings, tried to address this shortcoming in a paper last year in Geophysical Research Letters. They compared simulations that accounted just for sea ice loss, which tended to warm only the surface, with models in which tropical air warmed the whole Arctic atmosphere. Those models showed a striking effect of a warmer Arctic: At lower latitudes in Siberia, temperatures dropped 2°C by 2060. “That was a big eye opener for everybody,” Francis says. It made the focus on just sea ice seem like “kind of a waste of time,” she says. Gudrun Magnusdottir, a climate scientist at UC Irvine and co-author of the study, agrees. “It’s dangerous to emphasize just one area and one point,” she says.
The debate is far from over. Indeed, new evidence from weather records, published last month in the Journal of Geophysical Research, suggests the jet stream actually has gotten slightly wavier since the 1950s. The true cause of it—and the influence of global warming—remains to be seen.