Read our COVID-19 research and news.

Up in smoke. A sudden global warming 55 million years ago might have been fueled by carbon dioxide released from permafrost, as happened in Alaskan fires in 2007.

Bureau of Land Management/Alaska Fire Service

Did Melting Antarctic Permafrost Drive Ancient Global Warming?

Fifty-five million years ago, the world abruptly warmed by a scorching 5 degrees Celsius, the oceans turned acidic, and life ran a gantlet of extinction. Sound familiar? That was the closest nature ever came to foreshadowing the fossil-fuel-fed warming that is just getting underway now. But for 15 years, climate scientists have puzzled over how that ancient global warming could have happened. Now a group of researchers is offering a new explanation: The needed greenhouse gases could have gushed from thawing permafrost on an ice-free Antarctica.

The problem with the warming 55 million years ago—called the Paleocene-Eocene Thermal Maximum (PETM)—has been its magnitude. Somehow, some sort of carbon-rich deposit had to release several trillion tons of carbon over a few millennia in the form of the greenhouse gases carbon dioxide and methane. Absent humans burning fossil fuels, the best bet has been the methane locked in ice—so-called methane hydrates—beneath the sea floor. But judging by today's stores of methane hydrates, there doesn't seem to have been enough methane stored at the time of the PETM to drive that much warming.

Climate scientist Robert DeConto of the University of Massachusetts, Amherst, and his colleagues think they have found an ancient carbon storehouse big enough to do the job. Carbon in the form of rotted plants is stored in coal, the mud of oceans and lakes, and soil, but the rotted-plant carbon in permanently frozen ground is both abundant and easily released once carbon-laden permafrost has thawed. So DeConto and colleagues set out to calculate how much permafrost carbon might have been around at the time of the PETM and how fast it might have been released. That required some educated guesses: how much land was at high latitudes, what kind of plants grew there, where permafrost could form under the then-warmer climate, and how much carbon dioxide and methane would be released when permafrost thawed as the world warmed slightly, among others.

In the team's modeling, at least, the PETM world held about 3.7 trillion metric tons of carbon in permafrost, nearly half of it on Antarctica (where there is nearly none today). Once the warming that had been going on for millions of years reached a threshold and that permafrost began to thaw, decomposing organic matter yielded 1.2 trillion tons of carbon as carbon dioxide and methane. The resulting greenhouse warming thawed more permafrost, releasing more greenhouse gases. After 10,000 years, almost all of the permafrost stores would have been released, according to the modeling, raising the global temperature just as much as actually happened.

In a field that's long lacked an explanatory paradigm, one might expect a possible solution to get a warm reception. Not when the PETM is involved. "It's another idea to explain this really big problem," says Gerald Dickens of Rice University in Houston, Texas, a founding member of the PETM-studying community, "but by no means do I believe it's the right answer." He sees the same fundamental problem with permafrost as with methane hydrates-the likely amount of available carbon falls short of that needed by a factor of two or three. And it gets worse. "How you test any of these ideas," he says, "is really problematic. How do you prove there's a whole bunch of permafrost" 55 million years ago?