Flying over Yakutia, in northeastern Russia, I watched the dark shades of the boreal forest blend with patches of soft, lightly colored grass. I was strapped to a hard metal seat inside the cabin of an Antonov-2, a single-­engine biplane, known in the Soviet era as a kukuruznik, or corn-crop duster. The plane rumbled upward, climbing above a horizon of larch and pine, and lakes the color of mud. It was impossible to tell through the Antonov’s dusty porthole, but below me the ground was breathing, or, rather, exhaling.

Three million years ago, as continent-­size glaciers pulsed down from the poles, temperatures in Siberia plunged to minus eighty degrees Fahrenheit and vast stretches of soil froze underground. As the planet cycled between glacial and interglacial periods, much of that frozen ground thawed, only to freeze again, dozens of times. Around eleven and a half millennia ago, the last ice age gave way to the current interglacial period, and temperatures began to rise. The soil that remained frozen year-round came to be known as permafrost. It now lies beneath nine million square miles of Earth’s surface, a quarter of the landmass of the Northern Hemisphere. Russia has the world’s largest share: two-thirds of the country’s territory sits on permafrost.

In Yakutia, where the permafrost can be nearly a mile deep, annual temperatures have risen by more than two degrees Celsius since the Industrial Revolution, twice the global average. As the air gets hotter, so does the soil. Deforestation and wildfire—both acute problems in Yakutia—remove the protective top layer of vegetation and raise temperatures underground even more.

Over thousands of years, the frozen earth swallowed up all manner of organic material, from tree stumps to woolly mammoths. As the permafrost thaws, microbes in the soil awaken and begin to feast on the defrosting biomass. It’s a funky, organic process, akin to unplugging your freezer and leaving the door open, only to return a day later to see that the chicken breasts in the back have begun to rot. In the case of permafrost, this microbial digestion releases a constant belch of carbon dioxide and methane. Scientific models suggest that the permafrost contains one and a half trillion tons of carbon, twice as much as is currently held in Earth’s atmosphere.

Trofim Maximov, a scientist who studies permafrost’s contribution to climate change, was seated next to me in the Antonov, shouting directions to the pilot in the cockpit. Once a month, Maximov charters the plane in order to measure the concentration of greenhouse gases in the atmosphere above Yakutia. He described the thawing permafrost as a kind of feedback loop: the release of greenhouse gases causes warmer temperatures, which, in turn, melt the permafrost further. “It’s a natural process,” he told me. “Which means that, unlike purely anthropogenic processes”—say, emissions from factories or automobiles—“once it starts, you can’t really stop it.”

A hose attached to the plane’s wing sucked air into a dozen glass cylinders arrayed on the floor of the cabin. By comparing the greenhouse-gas numbers over time, and at various altitudes, Maximov can estimate how permafrost is both affected by a warmer climate and contributing to it. When he started taking airborne measurements, half a decade ago, he found that the concentration of carbon dioxide in the air above Yakutia was increasing at double the rate of historical averages. Methane has a shorter life in the atmosphere than carbon dioxide, but it is more than twenty-­five times as effective at trapping heat. According to Maximov’s data, methane is also being released at an accelerated rate: it is now accumulating fifty per cent faster than it was a generation ago.

At the moment, though, I was mainly concerned with the stomach-turning lurches the plane was making as it descended in a tight spiral. We had dropped to a few hundred feet above the ground so that Maximov’s colleague, a thirty-three-year-old researcher named Roman Petrov, could take the final sample, a low-altitude carbon snapshot. The plane shook like a souped-up go-kart. Petrov held his stomach and buried his face in a plastic bag. Then I did the same. When we finally landed, on a grass-covered airstrip, I staggered out of the cabin, still queasy. Maximov poured some Cognac into a plastic cup. A long sip later, I found that the spinning in my head had slowed, and the ground under me again took on the feeling of reassuring firmness—even though, as I knew, what seemed like terra firma was closer to a big squishy piece of rotting chicken.

Throughout the seventeenth and eighteenth centuries, as the Russian Empire expanded eastward, reports filtered back to the capital of a “firm body of ice” in the ground, in the words of one explorer, that “was never heard of before.” In Yakutsk, the capital of Yakutia, early settlers struggled to grow crops and find sources of fresh groundwater. In the summer of 1827, a merchant named Fedor Shergin, whom the tsar had dispatched to Yakutia as a representative of the Russian-­American Company, tried to dig a well. Shergin’s team of laborers spent the next decade chiselling a shaft, reaching three hundred feet down, only to find yet more frozen earth. Finally, in 1844, Alexander von Middendorff, a prominent scientist and explorer, made his way from St. Petersburg to Yakutsk and estimated, correctly, that the soil under the shaft was frozen to a depth of at least six hundred feet. His findings jolted the Russian scientific academy, and eventually reached the salons of Europe.

Today, the entrance to Shergin’s shaft, as it is known, is housed in a log cabin in the center of Yakutsk, wedged between a concrete apartment block and the burned-out shell of a former military academy. One afternoon last summer, I visited the site with Yuri Murzin, a scientist from the Melnikov Permafrost Institute, based in Yakutsk. “The study of permafrost began here,” he said. “Before Shergin’s shaft, practically no one outside of Yakutia had any idea such a thing existed.” Murzin and I wanted to have a look inside the shaft, which required lifting a series of heavy wooden lids. A column of cold air rushed upward. I looked down but saw only a wall of black. A musty aroma of dirt and ice wafted into the cabin. “It smells of antiquity, of time gone by,” Murzin said.

In a widely read monograph published in the nineteen-twenties, a Soviet scientist named Mikhail Sumgin called the country’s frozen earth vechnaya merzlota, literally “eternal frost,” a neo­logism that was later rendered into English as “permafrost.” Sumgin was something of a permafrost romantic, writing that “vechnaya merzlota astounds the human intellect and imagination.” He likened it to a “Russian Sphinx”—inexplicable, alluring, a riddle to be solved.

For others, permafrost posed a confounding engineering problem. Soviet ideology contained a strong Prome­thean impulse, encapsulated by Maxim Gorky’s axiom, paraphrasing Marx, that “in transforming nature, man transforms himself.” The construction of the Trans-Polar Railroad was one of many infrastructure projects under Stalin that had to contend with the particularities of land that might sink by several inches in the summer or heave upward in the winter. As one scientist declared in the thirties, “It is necessary to defeat the enemy—vechnaya merzlota—and not surrender.”

Fewer than two hundred thousand people live in the Arctic reaches of Alaska and Canada, and there are no large towns; the Soviet Union, by contrast, sought to populate its north­eastern territories. With the influx of inhabitants, and the construction projects that followed, a new problem arose: buildings create their own heat, warming the permafrost and causing the ground to buckle and squirm. In 1941, the Yakutsk headquarters of the N.K.V.D., the Stalin-­era secret police, sank into the earth, leading one of its walls to split open, spraying plaster over a room of operatives.

Yakutsk is one of two large cities in the world built in areas of continuous permafrost—that is, where the frozen soil forms an unbroken, below-zero sheet. The other is Norilsk, in Krasnoyarsk Krai, Russia, where Gulag prisoners were sent in the nineteen-­thirties to construct a new settlement. Norilsk is home to some of the largest nickel deposits on Earth. To service the mining and smelting industries, the city needed factories, apartment blocks, schools, hospitals, and auditoriums. Many of these early structures didn’t last long. Valery Grebenets, a professor of engineering at Moscow State University, worked in Norilsk in the eighties. Some of his colleagues there recounted stories of engineers facing severe consequences when their projects collapsed. “When your neighbors start getting shot, you begin to think a bit more vividly,” Grebenets grimly remarked. As advances were made in the study of permafrost, he continued, “people started to understand its properties, to come up with new ideas.”

One of the more outlandish proposals came from a Soviet scientist named Mikhail Gorodsky, who called for positioning an artificial dust ring—similar to Saturn’s rings—around Earth, to create a heat dome over the poles that would raise temperatures to the point that the permafrost would vanish entirely. In the mid-fifties, Mikhail Kim, an engineer who had first arrived in Norilsk as a Gulag prisoner, devised a more practical solution. His idea was to build on top of cement piles driven as far as forty feet into the permafrost. The piles would elevate a building’s foundation, keeping it from warming the ground below and allowing cold air to penetrate deep into the soil. An Arctic construction boom followed.

Soviet engineers came to treat vechnaya merzlota as exactly that: eternal, stable, unchanging. “They believed they had conquered permafrost,” Dmitry Streletskiy, a professor at George Washington University, said. “You could construct a five- or nine-story building on top of piles and nothing happened. Everyone was happy.” But, Streletskiy went on, “that infrastructure was meant to serve thirty to fifty years, and no one could imagine that the climate would change so dramatically within that span.”

By 2016, a regional official had declared that sixty per cent of the buildings in Norilsk were compromised as a result of permafrost thaw. On May 29, 2020, a fuel-storage tank belonging to Norilsk Nickel, one of Russia’s largest mining companies, cracked open, spilling twenty-one thousand tons of diesel into nearby waterways and turning the Ambarnaya River a metallic red. Executives at the company said that the damage had been contained. But Georgy Kavanosyan, a hydrogeologist based in Moscow, who has a popular YouTube channel, travelled to Norilsk and took samples farther north, from the Pyasina River, which empties into the Kara Sea. He found pollutant concentrations two and a half times permitted levels, threatening fish stocks and ecosystems for thousands of miles.

The Kremlin could not ignore the scale of the disaster, which Greenpeace compared to the Exxon Valdez oil spill. In February, 2021, the state ordered Norilsk Nickel to pay a two-billion-­dollar fine, the largest penalty for environmental damage in Russian history. The company had said that the piles supporting the tank failed as the permafrost thawed. An outside scientific review found that those piles had been improperly installed, and that the temperature of the soil was not regularly monitored. In other words, human negligence had compounded the effects of climate change. “What happened in Norilsk was a kind of demonstration of how severe the problem can be,” Vladimir Romanovsky, a professor of geophysics at the University of Alaska Fairbanks, said. “But it’s far from the only case. Lots of other accidents are happening on a smaller scale, and will continue to.”

To get a sense of how permafrost thaw is changing the landscape, I took a drive out of Yakutsk with Nikolay Basharin, a thirty-two-year-old researcher at the Permafrost Institute. Our destination was Usun-Kyuyol, the village where Basharin grew up, eighty miles away. His family, like many in Yakutia, had a cellar dug into the permafrost, where they stored meat and jam and lake ice, which they melted for drinking water. “You live on it for all these years but never really fully understand it,” Basharin told me, explaining his decision to study permafrost science. We set off at dawn to catch the first ferry across the Lena River; because of the ever-changing effects of permafrost on soil structure, building a bridge has thus far proved unfeasible.

The area on the Lena’s right bank, a valley of some twenty thousand square miles, is known for its large deposits of yedoma, a type of permafrost that is especially rich in ice. Whereas some permafrost is nearly all frozen soil, yedoma contains as much as eighty per cent ice, forming solid wedges, invisible from the surface, that can extend multiple stories underground. This is problematic for several reasons. Water is an efficient conductor of heat, soaking up atmospheric temperatures and warming the ­permafrost below. As yedoma thaws, it can create depressions in the land that fill with water, a process known as thermokarst.

Yedoma is also a very absorbent carbon trap, accumulating organic matter in silt and sediment that, at a certain point in the past tens of thousands of years, froze underground. When it thaws, it can release ten times more greenhouse gases than other, sandier types of permafrost. Yedoma is found in parts of Alaska and Canada, but it is most prevalent in northeastern Siberia; in Yakutia, it makes up a tenth of the region’s territory.

Basharin and I drove past the pooling remains of thawing yedoma. Some areas were the size of small ponds, others were effectively lakes. We stopped at the edge of a large alas—a thermokarst lake that has dried up, becoming a kind of scooped-out crater. This alas had likely taken more than five thousand years to form. Basharin told me that fragments of hundred-and-fifty-year-old birch trees had recently been found at the bottom of a smaller alas nearby, suggesting that a process which once took thousands of years is now happening in little more than a century. “In geological terms, that’s no more than a milli­second,” he said.

We drove on to Usun-Kyuyol, where Basharin lived until he was twelve. Cows grazed in front of wooden houses, their chimneys puffing out dark wisps of smoke. One stretch of road was pockmarked with oval mounds several feet high. Patches of yedoma had thawed, leaving steep pits where the tops of the ice wedges had once been. It started, Basharin said, around twenty years ago, following a silkworm infestation in a nearby birch forest. The trees died, leaving the permafrost vulnerable to sunlight and rising temperatures. “At first, people were happy—the next year was a good one for berries,” Basharin told me. But, as the permafrost thawed, the road became so bumpy as to be impassable, a mogul skiing course turned horizontal. A number of houses cracked as the ground beneath them gave way. A few stood abandoned.

We stopped at the home of Basharin’s aunt and uncle, who invited us in for lunch. “We watch television, we hear about warming,” his uncle, Prokhor Makarov, told me. “But we live in a village. Our main problem is making sure we have enough hay for the winter.” Their house wasn’t in imminent danger of collapse, but the earth around it was craggy and dotted with small indentations. The fence around their property had the lurching quality of a person at the bar who’s had one too many. Makarov told me that, in the summer, he shovels dirt around to keep things level. “We’re used to it,” he said.

After we left, Basharin told me, “People don’t understand the end of this story.” Try as they may to adapt, he went on, “the thaw will reach them all the same.”

Three days later, I caught a flight on a propeller plane leaving Yakutsk for Chersky, a speck of a town on the Kolyma River, near the delta where it empties into the East Siberian Sea. In the nineteen-thirties, Chersky was a transit hub for the Gulag camps; later, it served as a base for the planes that ferried Soviet explorers on Arctic expeditions. These days, in late summer, residents who have spent their vacations on the “mainland,” as they call it, return for the start of the new school year, bringing with them items that are rare and expensive in the northernmost reaches of Siberia. The plane was packed, not only with people but with trays of eggs, bouquets of flowers, and boxes containing newly purchased televisions and blenders.

On arrival, I walked out of the Chersky airport—which is not much more than a small waiting room—and saw a Land Rover parked on a dusty road. A man with a flowing silver beard and a black beret sat behind the wheel. I immediately recognized him as Sergey Zimov, who is something of a permafrost soothsayer. “Get in,” he said.

We sped off toward the Northeast Science Station, his research center, on the outskirts of town. Zimov, who is sixty-six, studied geophysics in Vladivostok and, in the waning years of the Soviet Union, moved to Chersky, along with his wife, Galina; a son, Nikita, was born shortly afterward. The Soviet collapse is but one of many events, past and future, that Zimov claims to have foretold. “When you know the history of civilization, it is very easy to make predictions, and, so far, I have not been wrong,” he told me. During the next week, I heard Zimov hold forth on global population trends, Russian military logistics, and the gold standard. (“My rule is simple: if you get a dollar, use it to buy gold.”)

But it was Zimov’s ideas on permafrost that had brought him scientific renown. In the early nineties, he was among the first to come to several related realizations: permafrost holds immense quantities of carbon; much of that carbon is released as methane from thermokarst lakes (the presence of water and the absence of oxygen produce methane, as opposed to carbon dioxide, which is released from upper layers of soil); and a sizable portion of those emissions comes in the fall and the winter, cold periods that Arctic scientists had previously considered unimportant from a climate perspective.

In the spring of 2001, an American Ph.D. student named Katey Walter Anthony, who had met Zimov at an academic gathering in Alaska, arrived in Chersky to help collect data on methane emissions. “When I first saw him in Alaska, I thought he looked so wild, with these big eyebrows and crazy eyes,” Walter Anthony told me. “But when I got to Chersky I realized that, though nothing about him had changed, in that setting he looked totally normal.”

Walter Anthony positioned methane traps, which she’d fashioned out of sheets of plastic, around Chersky’s thermokarst lakes. “Sergey had thought up these really excellent ideas,” she said. “But he had collected just as much data as he thought he needed to prove his point, which was much less than what Western scientists would like to see.” Walter Anthony returned the following year; this time, she stayed until the fall and the onset of the first frost.

One morning after breakfast, Zimov suggested that they visit one of the lakes. The ice was still thin and brittle, and Walter Anthony was nervous about walking on it. “Don’t worry,” Zimov told her. “Autumn ice is friendly—it tells you before it breaks.” He pointed down. Walter Anthony saw thousands of tiny air bubbles, giving the frozen surface the look of a starry night. “The ice was essentially a map pointing to where the methane was coming up,” she said. She was able to place her traps precisely where methane was being emitted, rather than, as she put it, “shooting an arrow into the sky.”

Walter Anthony found methane emissions five times higher than Zimov’s initial estimate. Radiocarbon dating showed that the gas was emitted from organic matter that formed between twenty and forty thousand years ago, during the Pleistocene era, indicating that permafrost thaw had reached layers that were deep and ancient. The research was published in a paper in Nature, in 2006, which immediately became a foundational text in establishing the impact of permafrost thaw on climate change.

When I was in Chersky, Zimov took me out to the lake. We walked through shrubs and felt the crunch of bright-red cloudberries under our feet. At the water’s edge, Zimov asked, “You see the bubbles?” Once I knew to look for them, they were impossible to miss. It was as if the lake were a giant cauldron on the brink of a very slow, barely perceptible boil, with a pop of air here and there. Methane.

Zimov explained that, even during Chersky’s frigid winters, temperatures under the lake’s surface remain above freezing. Unfrozen water allows microbes to keep digesting organic matter long after the surrounding landscape is covered in snow. Water also has a powerful erosion effect. “The bank is slowly thawing and collapsing, taking with it fresh pieces of permafrost into the lake,” Zimov said—more fuel for the release of methane. As Walter Anthony, who is now a professor at the University of Alaska Fairbanks, put it to me, “Once permafrost thaws to the point where it creates depressions filled with water, the thaw starts to go deep and fast and expands laterally—you can’t really stop it.”

The mean annual temperature in Chersky has risen by three degrees Celsius in the past fifty years. An equally pressing problem is snow cover. “Snow is like a warm blanket—it doesn’t allow the wintertime cold to penetrate all the way into soil,” Zimov said. One of the effects of climate change is more precipitation in the Arctic ecosystem around Chersky. Yearly snowfall has increased by as much as twenty centimetres since the early eighties, adding two more degrees of warming effect. As a result, Zimov explained, permafrost that used to be minus seven degrees Celsius is now on the verge of thawing, if it hasn’t already.

Adecade ago, a paper about emissions from undersea permafrost led to a moment of hysteria over a so-called methane bomb in the Arctic, poised to release a devastating amount of warming gas all at once. In the years since, much of the scientific community has come to see permafrost thaw more as a slow-motion disaster. “The permafrost isn’t going to release a catastrophic explosion of carbon that would, say, double overnight the amount of carbon dioxide in the atmosphere,” Ted Schuur, who leads a project on permafrost thaw and climate change at the University of Northern Arizona, told me. “Instead, this carbon is going to leak out from all over the Arctic and, over time, add a substantial amount to the carbon humans have already added by burning fossil fuels.”

In 2018, a report prepared by the U.N.’s Intergovernmental Panel on Climate Change gave humans a maximum carbon budget of some five hundred and eighty billion tons in order to have an even chance of limiting warming to one and a half degrees ­Celsius. The panel’s models have only recently started factoring in various permafrost-thaw scenarios, but they offer such a wide range of possible outcomes that permafrost has become, as Schuur put it, the “wild card” of climate science. He and his colleagues estimate that permafrost emissions might make up five to fifteen per cent of the I.P.C.C.’s allotment.

The I.P.C.C.’s models also miss a significant cause of greenhouse-gas emissions from permafrost. Its estimates presume that all thaw will be gradual, caused by rising air temperatures, and do not take into account thermokarst, or “abrupt thaw,” as Schuur prefers to call it, which can trigger nonlinear events like rapid erosion or landslides. “Those events are essentially irreversible on human time scales,” Susan Natali, a scientist at the Woodwell Climate Research Center, in Falmouth, Massachusetts, said.

Average global temperatures are on track to rise by nearly two and a half degrees Celsius this century. At the latest U.N. climate-change conference, held in Glasgow in November, participating countries reaffirmed the goal of holding warming to one and a half degrees, even as plans for doing so remain vague. Most models presume that temperatures will surpass that limit, and that a successful global effort to keep warming at a manageable level will involve measures to bring them down again. “The problem is, you can’t just turn off, let alone reverse, permafrost thaw,” Natali said. At a certain point, nature takes over. Even the most forward-thinking legislature in the world can’t pass a law banning emissions from permafrost. As Natali put it, “It won’t be possible to refreeze the ground and have it go back to how it was.”

All across the Arctic, ecosystems are shifting from carbon sinks—which absorb more greenhouse gases than they release—to carbon sources. One day in Chersky, I visited a site along the river managed by a German research team from the Max Planck Institute for Biogeochemistry. I was shown around by Mathias Göckede, the project’s lead scientist. We jumped between grassy tussocks sprouting up from the tundra and came to a spot where, seventeen years earlier, his colleagues had purposely degraded the upper layer of yedoma. The idea was to mimic permafrost thaw in order to see how the landscape would react and how the local carbon budget would change.

In the first year of the experiment, Göckede explained, the soil released more carbon dioxide than the vegetation could absorb, and the site switched from a sink to a source. Then larger shrubs and trees appeared, which sucked up emissions. The site settled into a new equilibrium, at a higher level of both emissions and absorption than before. “I find that encouraging,” Göckede told me.

But trees can grow only so much. And, in the Arctic, light is limited to a few months in the summer, forming a narrow window in which photosynthesis can remove greenhouse gases from the atmosphere. Microbes in the soil, meanwhile, can digest organic material in the thawed permafrost for a much longer season and, given the deep stores of carbon, with seemingly no end. “There is a limit to how much the vegetation can grow and absorb carbon,” Göckede said. “But there is virtually no limit to how much the soil can heat up and release more carbon.”

Earlier in the summer, I visited Yamal, a peninsula that juts into the Kara Sea like a crooked finger. Yamal is home to the Nenets, an ethnic group native to the Russian north, and one of the largest remaining nomadic populations. Nenets live in chums—the local version of yurts—and drive herds of reindeer up and down the peninsula, in search of seasonal pastures. In the Nenets language, Yamal means “the edge of the world.”

After taking a passenger ferry up the Ob River, I stopped to spend a night in the chum of a Nenets family. I slept under a reindeer hide and, following a breakfast of fresh fish, headed farther upriver to Yar-Sale, a settlement that functions as an administrative center for the nomad camps in the tundra. There, I met Vitaly Laptander, a reindeer herder.

In July, 2016, a heat wave hit Yamal, with temperatures reaching a hundred degrees Fahrenheit. Laptander was with his flock of two thousand animals near Lake Yaroto, in the middle of the peninsula. “I hadn’t felt such heat before,” he told me. One morning, he came across a horrifying sight: fifty of his reindeer lay dead in the tundra. There was no power or cellular service. Lap­tander walked for ten hours to call for help, finally coming across a Nenets encampment with a satellite phone. By the time he had trekked back to his herd, two hundred more reindeer were dead. “I didn’t know what to do,” he said. “Things were clearly really bad, and I was scared.”

A helicopter arrived, and discharged a team of medics and veterinarians in hazmat suits. They took samples from the dead reindeer and flew off, delivering them to laboratories in Moscow and Siberia. Two days later, the helicopter returned, and officials told Lap­tander that his animals had likely been infected by anthrax.

Within days, specialists from the Army’s Radiological, Chemical, and Biological Defense forces had arrived in Yamal. They searched for reindeer carcasses, and burned them where they lay. After two weeks, quarantine measures and an accelerated vaccination campaign brought the outbreak under control. By then, more than twenty-five hundred reindeer had been lost on the peninsula; Laptan­der’s herd was cut in half. The contagion had also spread from animals to humans. Dozens of people were hospitalized; a twelve-year-old boy died.

The outbreak represented the first anthrax cases on Yamal since 1941. Just about everyone, from scientists to herders, had believed that the bacteria-borne disease was eradicated long ago. Two hundred thousand soil samples taken during the previous decade showed no evidence of anthrax spores. But in a normal summer the upper layer of permafrost in Yamal thaws to a depth of twenty inches or so; in 2016, it had reached nearly three feet in some places. In a subsequent report on the causes of the outbreak, a panel of Russian experts wrote, “The emergence of anthrax was triggered by the activation of ‘old’ infection sites following anomalously high air temperature and the thawing of the sites to a depth beyond normal levels.”

Permafrost thaw has brought to the surface all sorts of mysteries from millennia past. In 2015, scientists from a Russian biology institute in Pushchino, a Soviet-era research cluster outside Moscow, extracted a sample of yedoma from a borehole in Yakutia. Back at their lab, they placed the piece of frozen sediment in a sterilized culture box. A month later, a microscopic, wormlike invertebrate known as a bdelloid rotifer was crawling around inside. Radiocarbon dating revealed the rotifer to be twenty-four thousand years old. In August, I drove out to Pushchino, where I was met by Stas Malavin, a researcher at the laboratory. “It’s one thing for a simple bacterium to come back to life after being buried in the permafrost,” he said. “But this creature has intestines, a brain, nervous cells, reproductive organs. We’re clearly dealing with a higher order.”

The rotifer had survived the intervening years in a state of “cryptobiosis,” Malavin explained, “a kind of hidden life, where metabolism effectively slows down to zero.” The animal emerged from this geological “time machine,” as he put it, not just alive but able to reproduce. A rotifer lives for only a few weeks, but replicates itself multiple times through parthenogenesis, a type of asexual reproduction. Malavin removed from the lab fridge a direct descendant of the rotifer that had crawled out of the permafrost and placed it under a microscope. An oval-shaped plankton squirmed around; I imagined this blob, two-tenths of a millimetre in size, as a nervous explorer who awoke to find itself in a strange and unexpected future.

“Why be modest?” Malavin asked. Unlocking the secret of how an animal with a complex anatomy was able to shut down for tens of thousands of years and then turn itself back on might, for example, offer hints for using cryogenic conditions to store organs for donation. Neuroscientists at M.I.T. have been in touch. “I’m obviously not saying our findings will lead to people being put into long-term cryogenic slumber tomorrow,” Malavin said. “But it’s a step in that direction.”

Perhaps the most exciting biological specimens to come out of the permafrost are mammoth remains, many of which, thanks to millennia of natural cold storage, are remarkably well preserved. In Yakutsk, I visited the Mammoth Museum, a two-story facility full of bones and tusks and teeth. The mammoth appeared a hundred and fifty thousand years ago, roaming over grassland steppe that stretched from the Iberian Peninsula to the Bering Strait.

The species began to die out near the end of the Pleistocene era, around twelve thousand years ago, for reasons that were long the subject of debate. One camp held that the mammoth was among the first victims of anthropogenic extinction. “Mammoths didn’t have any natural predators—except for humans,” Sergey Fedorov, the head of the museum’s exhibitions, told me. But in October an international team of scientists published a study in Nature that purported to settle the case. By analyzing ancient environmental DNA, they determined that rapidly warming temperatures melted the glaciers and inundated the tundra, wiping out the mammoth’s food supply. “Our results suggest that their extinction came when the last pockets of the steppe-tundra vegetation finally disappeared,” the authors wrote.

Yakutia is the world leader when it comes to mammoth finds. These remains, the first of which were recovered by Russian scientists in 1806, have taught us a great deal about the Pleistocene in general: the gastrointestinal tract of one mammoth, found in 1971, was so well preserved that scientists were able to analyze its last meal. Fedorov told me about an expedition, in 2013, to Maly Lya­khovsky Island, off the northern coast of Yakutia; when researchers there dug up a frozen mammoth carcass, its flesh started to bleed. A British paleobiologist at the site later described the specimen as “really juicy,” like a “piece of steak.”

The prospect of forty-thousand-year-old hemoglobin was exciting for a coterie of scientists who have dreamed of using gene-editing techniques to reproduce a living mammoth. (In the end, the tissue samples from the Maly Lyakhovsky mammoth did not produce enough usable DNA to reconstruct the animal’s genome.) George Church, a prominent geneticist at Harvard Medical School, has co-founded a startup dedicated to the mammoth de-extinction effort, and hopes that his team will be ready to produce embryos of neo-mammoths within the next few years.

Fedorov brought me to a large walk-in freezer, where lumps of flesh and fur were piled on metal shelves; the crescent bend of a tusk was unmistakable. As Fedorov explained, these mammoth remains, dug up across Yakutia, were being stored at zero degrees Fahrenheit, awaiting further scientific study. The space was cramped and frigid—so this is what it’s like to be locked in the permafrost, I thought. I picked up a leg that once belonged to the Maly Lyakhovsky mammoth, a thick stump with reddish-brown hair. “Look, its footpad is very well traced,” Fedorov said. “You can see its toenails.”

One clue to how permafrost will survive this current era of warming is how it fared during the previous one. Five years ago, Julian Murton, a scientist and professor at the University of Sussex, led a team of researchers to the Batagaika Crater, a permafrost thaw slump in central Yakutia. A thaw slump is essentially a drawn-out landslide set off by thawing yedoma; the Batagaika Crater is the largest in the world, a half-mile-long gash in the earth with walls as high as two hundred and eighty feet. The crater is constantly thawing and collapsing, growing by as much as a hundred feet a year. Locals call it a “gateway to Hell.” A more apt metaphor may be a geological layer cake, whose exposed walls allow a rare opportunity to look at hundreds of thousands of years of permafrost all at once.

Murton told me that the first thing that struck him during his time at the crater was the sound. “It’s like an orchestral piece,” he said. “In the summer, when the head wall is thawing quickly, you hear the constant trickle of water, like first violins. And then you have these massive chunks of permafrost, up to half a ton, that fall to the bottom with a big thud. That’s the percussion.”

Murton and his team drilled boreholes down the crater’s walls, and used a method called luminescence dating to estimate the age of the sediment that they extracted. The bottom layer of permafrost turned out to be at least six hundred and fifty thousand years old. As Murton explained, that means it survived the previous interglacial period, which began some hundred and thirty thousand years ago, when parts of the Arctic were as much as four or five degrees Celsius warmer than they are today. “The oldest permafrost in Eurasia has been kicking around for over half a million years,” Murton told me. “Seeing as it survived intense global-warming events in the past, it must be pretty resilient.”

That’s the good news. “If you like permafrost, as I do, we’re not going to be short on it in our lifetimes,” Murton said. But his hypothesis on the resilience of permafrost applies to frozen earth that extends hundreds of feet below the surface. “The top several metres are certainly under threat,” he said. That is exactly where the carbon is: the upper three metres of permafrost hold half as much carbon as similar soil depths in the rest of the planet’s ecosystems combined. Moreover, as Murton put it, “even as it appears that the ecosystem can protect permafrost from high air temperatures, if that ecosystem is disturbed, permafrost suddenly becomes very vulnerable.” The Batagaika Crater itself formed after a large patch of forest was clear-cut, in the nineteen-sixties.

These days, fire is the biggest threat to the landscape. Last summer was Yakutia’s worst fire season in history, with eight million hectares ablaze—an area about the size of Maine—releasing the equivalent of more than five hundred megatons of carbon dioxide. It is hard to predict what sort of long-term effect fire will have on the permafrost. In some parts of Yakutia, the boreal forest has been able to regenerate itself, bringing new trees and underbrush that sequester carbon, and the situation has returned to equilibrium. But in other places—especially those full of ice-rich yedoma—fires have caused irreversible changes in the landscape, such as a thermokarst lake or a crater like Batagaika. Sander Veraverbeke, a climate scientist at Vrije Universiteit Amsterdam, who has done extensive field work in Yakutia, told me, “In that scenario, the permafrost never recovers.”

One day in Chersky, Zimov showed me a site where he had tried to mimic the result of a fire on the permafrost. He drove us in a motorboat down the river, the wind slicing through my jacket and chafing my face. We tied the boat to some bushes, and set off through the spongy moss of the tundra. “I actually hate terrain like this,” Zimov said. “Everything is soft and squishy, with mosquitoes everywhere.”

Half an hour later, we came to a clearing that had the same bumpy features that I had seen in the village of Usun-Kyuyol. In 2003, Zimov had used a “very, very large bulldozer,” which he borrowed from a nearby gold mine, to uproot shrubs and moss and remove the topsoil, much the way a fire might. (“This is the kind of experiment Sergey likes,” Göckede had told me. “For him, a bulldozer is a scientific instrument.”) Within a year, the ice in the yedoma began to melt, collapsing the ground and leading the permafrost to thaw at ever greater depths.

Zimov and I were each carrying a long metal probe, the permafrost scientist’s classic field tool. The point at which the tip hits hard ice reveals the depth of permafrost thaw. Zimov has an ear for frozen soil, able to judge its consistency by the sound it makes when struck by metal. “It’s loose, ready to crumble,” he declared. Thirty years ago, during an average summer, the permafrost thawed to a depth of less than a metre. Now, at the bulldozed site, Zimov had to fasten two probes together, finally hitting solid ice at a depth of three and a half metres. All that thawed soil was producing carbon dioxide and, at deeper levels, where there is less oxygen, methane. “You’d need five very cold, raw winters in a row to freeze it again,” Zimov said. “And I don’t quite believe we’ll see that again.”

In May, Russia’s environmental minister proposed a nationwide system to monitor climate-induced changes in the permafrost, noting that its thaw could cause more than sixty billion dollars’ worth of damage to the country’s infrastructure by 2050. The next month, Vladimir Putin, who in 2003 had remarked that global warming simply means “we’ll spend less on fur coats,” said of the country’s permafrost zone, “We have entire cities built on permafrost in the Arctic. If it all starts to thaw, what consequences will Russia face? Of course, we are concerned.”

It’s possible to imagine technical solutions to avoid the worst effects of permafrost thaw on buildings, industrial facilities, or even whole settlements. In Yakutsk, I passed apartment blocks with large metal tubes installed near their foundations, filled with a cooling agent that, during the winter, condenses and flows belowground to keep the soil frozen. In Salekhard, the capital of Yamal, temperature sensors have been lowered into boreholes under the foundations of certain buildings—if the soil is at risk of thawing, scientists will get an alarm signal, presumably in time to make engineering fixes. Yaroslav Kamnev, the director of an initiative launched by the regional government to study the warming of the soil, told me, “You simply have to understand what is going on inside the permafrost, and everything will stay standing just fine.”

But what to do with the huge reserves of carbon in the ground, waiting to be turned into greenhouse gas? You can’t effectively monitor, let alone cool, millions of square miles of uninhabited tundra. “Technological fixes are impossible,” Merritt Turetsky, the director of the Institute of Arctic and Alpine Research at the University of Colorado Boulder, said. The most obvious answer, tragic in both its banality and its unlikelihood, is for humans to quickly and dramatically limit the burning of fossil fuels. “There is one way to keep permafrost frozen that we know is proven and demonstrated—reducing human emissions,”  Turetsky said. “A focus on other solutions might be intriguing, but it’s ultimately a distraction.”

Zimov has his own idea. As a ­graduate student, during field visits to the Arctic, he was struck by the bones and other assorted remains he found: mammoths, horses, bison, elk, and wolves. On a walk around an eroding hillside by the river outside Chersky, I stumbled across the dark-brown skull of a wild horse. Zimov’s son, Nikita, who now runs the day-to-day operations at the research station, estimated that it was between twenty and forty thousand years old.

During the Pleistocene era, the Arctic was covered by grassy steppe, which acted as a natural buffer for the permafrost. The mammals that roamed this lost savanna depended on it for food and also perpetuated its existence. Zimov wants to re-create that ecosystem. “We must return nature to order,” he said. “It will then take care of the climate.”

The theory rests on the warming effect of snow. As Zimov explained, there isn’t much hope of quickly cooling air temperatures. But lessening the snow cover during the winter would allow more cold air to reach the permafrost. “You could do this mechanically, by sending three hundred million workers with shovels across Siberia,” he said. “Or you can do the same, for free, with horses, musk ox, bison, sheep, reindeer.” Those animals would break down shrubs and churn the soil, allowing grasslands to reappear. In summer, owing to the albedo effect—light surfaces reflect heat, dark ones absorb it—the pale grass would stay cooler than the brown shrubs that currently blanket the tundra.

In 1998, Zimov brought the first horses to what he called Pleistocene Park, a fenced tract of land an hour’s boat ride from the research station. Since then, the park has grown to eight square miles, and it is now home to a hundred and fifty animals, not just horses but bison, sheep, yaks, and camels. To give them a head start, Nikita sped about the territory in the family’s “tank”—a hefty, all-terrain transport vehicle on treads—knocking down trees and under­growth.

Two years ago, Zimov and Nikita completed a study with a team of researchers from the University of Hamburg, which showed that the animals reduced average snow density by half, and lowered the average temperature of the permafrost by nearly two degrees Celsius. The researchers theorized that thirty-seven per cent of Arctic permafrost could be saved from thawing by the wide-scale introduction of large herbivores. (Not all scientists are so enthusiastic: Duane Froese, a professor of geology at the University of Alberta, who has done extensive research on the Pleistocene ecosystem, told me, “The kind of animal density you’d need in order to impact vegetation in the way Sergey is envisioning greatly exceeds anything that could be maintained naturally.”)

Nikita, who is thirty-eight, has a degree in applied mathematics, but he is not exactly a scientist. His fluency in the world of permafrost came from years spent with Zimov around the station, an informal education that has made him an energetic steward of his father’s vision. For much of the time that I was in Chersky, he was tracking a shipment of a dozen bison that had begun their journey on a farm in Denmark, nearly five thousand miles away. They were on a container ship sailing on the Arctic Ocean, but because of storms at sea the journey was taking longer than planned. One morning, he announced that he was headed to the park to install a new greenhouse-gas flux sensor, which a group of scientists at the University of Alaska Fairbanks had sent to measure emission levels. I volunteered to go along.

It was a clear fall day on the river, with the golden leaves of the bushes and stunted trees of the tundra giving the scene the feel of a New England autumn in miniature. An hour later, we pulled up to the entrance of the park, marked by a few wooden steps built into the muddy riverbank. Nikita lugged the sensor in a backpack up a hundred-­foot tower and tinkered with it for a while, without success. After he came down, we walked through the territory, with pockets of knee-high grasses rising out of the flat expanse. “We’re not reinventing the wheel here,” he said. “This all existed at one point, we know that. How to re-create it now, though? That’s the question.”

We came to a caravan of camels, munching on grass and craning their necks in wary avoidance of us. They looked out of place this far north, but the fossil record shows that camels once grazed all over the high Arctic, their fatty humps providing stores of energy during the long winters. Like the mammoth, the Arctic camel disappeared during the late Pleistocene era, along with giant beavers and sloths, horses and cave lions—a Noah’s ark of lost Arctic species.

The permafrost, sealed underground, has managed to survive a while longer. But it couldn’t stay out of harm’s way forever. Neither could humans, for that matter. Whether we are thawing the permafrost or fighting to keep it frozen, its presence, like that of so much on this planet, is far less eternal than we once convinced ourselves. “People didn’t start acting as gods fifty or a hundred years ago, or even one thousand, but ten thousand years ago,” Nikita said. “The point isn’t whether it’s O.K. to act like a god but whether you’re acting like a benevolent or wise one.”