Something was strange about the smog. Usually smog will dissipate when sources of air pollution—like cars or factories—shut down for a time. But this cloud was stubborn. As part of solar new year celebrations in February 2013, millions of families drove their cars out of Beijing to go on vacation, and the government ordered all factories to cease operations. The smog didn’t subside much, and less than a week later the full-on “airpocalypse” returned as bad as before.

What made the winter smog so bad that year—and in the winters since, which have also been stubbornly smoggy? Two new studies revisit the episode. Both of them argue that climate change will make this kind of smog event much more common. And, remarkably, one of them asserts that the Chinese smog of January 2013 was worsened by two weather phenomena thousands of miles away. Because the Arctic Ocean froze less than it usually does, and because higher-than-usual snowdrifts piled up across the boreal forests of Russia, millions of Chinese people were subjected to some of the worst air pollution ever measured.

That’s because the smog of January 2013 wasn’t the result of emissions alone: Weather played the accomplice. For most of that winter, air over eastern China barely circulated. Trade winds went dormant, so smog could not ventilate to the east; and vertical circulation slowed, meaning particulate matter could not float up into the higher atmosphere. And as is typical for Chinese winters, rain never arrived to wash air pollution out.

The first paper, published Monday in Nature Climate Change, uses an ensemble of 15 climate models to suggest that those wind conditions—which lead to a stagnation of air across China’s most populated regions—are more likely to occur in the century to come. The haze-inducing conditions are 50 percent more likely to form between 2050 and 2099 than they were from 1950 to 1999. And once those patterns fall into place in any one episode, they are 80 percent more likely to persist.

These calculations assumed a worst-case emissions scenario, so any reduction in projected carbon-dioxide emissions (to meet the goals of the Paris Agreement, for instance) would likely reduce the frequency and persistence of haze episodes.

A second paper goes further. A team at Georgia Tech asked: What in the world, specifically, shaped those wind patterns? They applied a rudimentary form of artificial intelligence to historical climate and pollution data. Ultimately, they found two big influences: the extent that the Arctic Ocean has frozen over, and the amount of Siberia and the Russian far east covered in snow. These shape the climate of China mostly through their differing temperatures. And when the Arctic is especially warm, and Siberia is covered with snow and especially cold, they combine to reduce the atmospheric-pressure gradient across Asia. This weakens the pull of the East Asian monsoon wind and leaves air stagnating across eastern China.