Many of us share some dim apprehension that the world is flying out of control, that the center cannot hold. Raging wildfires, once-in-1,000-year storms, and lethal heat waves have become fixtures of the evening news—and all this after the planet has warmed by less than 1 degree Celsius above preindustrial temperatures. But here’s where it gets really scary.

If humanity burns through all its fossil fuel reserves, there is the potential to warm the planet by perhaps more than 10 degrees Celsius and raise sea levels by hundreds of feet. This is a warming spike comparable in magnitude to that so far measured for the End-Permian mass extinction. If the worst-case scenarios come to pass, today’s modestly menacing ocean-climate system will seem quaint. Even warming to half of that amount would create a planet that would have nothing to do with the one on which humans evolved, or on which civilization has been built. The last time it was 4 degrees warmer there was no ice at either pole and sea level was hundreds of feet higher than it is today.

I met University of New Hampshire paleoclimatologist Matthew Huber at a diner near campus in Durham, New Hampshire. Huber has spent a sizable portion of his research career studying the hothouse of the early mammals, and he thinks that in the coming centuries it’s not impossible that we might be headed back to the Eocene climate of 50 million years ago, when there were Alaskan palm trees and alligators splashed in the Arctic Circle.

“The modern world will be much more of a killing field than the Paleocene–Eocene Thermal Maximum was,” he said. “Habitat fragmentation today will make it much more difficult to migrate. But if we limit it below 10 degrees of warming, at least you don’t have widespread heat death.” In 2010, Huber and coauthor Steven Sherwood published one of the most ominous science papers in recent memory: “An Adaptability Limit to Climate Change Due to Heat Stress.”

“Lizards will be fine, birds will be fine,” Huber said, noting that life has thrived in hotter climates than even the most catastrophic projections for anthropogenic global warming. This is one reason to suspect that the collapse of civilization might come long before we reach a proper biological mass extinction. Life has endured conditions that would be unthinkable for a highly networked global society partitioned by political borders. Of course, we’re understandably concerned about the fate of civilization, and Huber says that, mass extinction or not, it’s our tenuous reliance on an aging and inadequate infrastructure—perhaps, most ominously, on power grids—coupled with the limits of human physiology that may well bring down our world.

In 1977, when power went out for only one summer day in New York, whole swaths of the city devolved into something like Hobbes’s man in a state of nature. Riots swept across the city, thousands of businesses were destroyed by looters, and arsonists lit more than a thousand fires. In 2012, when the monsoon failed in India (as it’s expected to do in a warmer world), 670 million people—that is, 10 percent of the global population—lost access to power when the grid was crippled by unusually high demand from farmers struggling to irrigate their fields, while the high temperatures sent many Indians seeking kilowatt-chugging air conditioning.

“The problem is that humans can’t even handle a hot week today without the power grid failing on a regular basis,” he said, noting that the aging patchwork power grid in the United States is built with components that are allowed to languish for more than a century before being replaced. “What makes people think it’s going to be any better when the [average summer temperature] will be what, today, is the hottest week of the year in a five-year period, and the hottest temperatures will be in the range that no one has ever experienced before in the United States? That’s 2050.”

By the year 2050, according to a 2014 MIT study, there will also be 5 billion people living in water-stressed areas. “Thirty to fifty years from now, more or less, the water wars are going to start,” Huber said. In their book Dire Predictions, Penn State’s Lee Kump and Michael Mann describe just one local example of how drought, sea level rise, and overpopulation may combine to pop the rivets of civilization: Increasingly severe drought in West Africa will generate a mass migration from the highly populous interior of Nigeria to its coastal mega-city, Lagos. Already threatened by rising sea levels, Lagos will be unable to accommodate this massive influx of people. Squabbling over the dwindling oil reserves in the Niger River Delta combined with potential for state corruption will add to the factors contributing to massive social unrest.

“Massive social unrest” here being, of course, a rather bloodless phrase masking the utter chaos coming to a country already riven by corruption and religious violence. “It’s sort of the nightmare scenario,” said Huber. “None of the economists are modeling what happens to a country’s GDP if 10 percent of the population is refugees sitting in refugee camps. But look at the real world. What happens if one person who was doing labor in China has to move to Kazakhstan, where they aren’t working? In an economic model, they’d be immediately put to work. But in the real world they’d just sit there and get pissed. If people don’t have economic hope and they’re displaced, they tend to get mad and blow things up. It’s the kind of world in which the major institutions, including nations as a whole, have their existence threatened by mass migration. That’s where I see things heading by midcentury.”

And it doesn’t get any better after 2050. But forecasts about the disintegration of society are social and political speculations and have nothing to do with mass extinctions. Huber is more interested in the hard limits of biology. He wants to know when humans themselves will actually start to disintegrate. His 2010 paper on the subject was inspired by a chance meeting with a colleague.

“I presented a paper at a conference about how hot tropical temperatures were in the geological past and [University of New South Wales climate scientist] Steve Sherwood was in the audience. He heard my talk, and he started asking himself the very basic question, ‘How hot and humid can it get before things start dying?’ It was literally just an order of magnitude kind of question.

I guess he thought about it and realized that he didn’t know the answer and wasn’t sure anyone else did either. . . . Our paper really wasn’t motivated by the future climate per se, because when we started we didn’t know if there was any kind of realistic future climate state that would fall within this habitability limit. When we started, it was just like, ‘We don’t know. Maybe you have to go to, like, 50 degrees Celsius global mean temperature.’

Then we ran a whole set of model results, and it was rather alarming to us.” Sherwood and Huber calculated their temperature thresholds using the so-called wet-bulb temperature, which basically measures how much you can cool off at a given temperature. If humidity is high, for instance, things like sweat and wind are less effective at cooling you down, and the wet-bulb temperature accounts for this.

“If you take a meteorology class, the wet-bulb temperature is calculated by basically taking a glass thermometer, putting it in a tight wet sock, and swinging it around your head,” he said. “So when you assume that this temperature limit applies to a human, you’re really kind of imagining a gale force wind, blowing on a naked human being, who’s doused in water, and there’s no sunlight, and they’re immobile, and actually not doing anything other than basal metabolism.”

Today the most common maximums for wet-bulb temperatures around the world are 26 to 27 degrees Celsius. Wet-bulb temperatures of 35 degrees Celsius or higher are lethal to humanity.

Above this limit, it is impossible for humans to dissipate the heat they generate indefinitely and they die of overheating in a matter of hours, no matter how hard they try to cool off. “So we were trying to get across the point that physiology and adaptation and these other things will have nothing to do with this limit. It’s the E-Z Bake Oven limit,” he said. “You cook yourself, very slowly.”

What that means is that this limit is likely far too generous for human survivability.

“When you do real modeling, you hit a limit much sooner, because human beings aren’t wet socks,” he said. According to Huber and Sherwood’s modeling, 7 degrees Celsius of warming would begin to render large parts of the globe lethally hot to mammals.

Continue warming past that, and truly huge swaths of the planet currently inhabited by humans would exceed 35 degrees Celsius wet-bulb temperatures and would have to be abandoned. Otherwise, the people who live there would be literally cooked to death.

“People are always like, ‘Oh, well, can’t we adapt?’ and you can, to a point,” he said. “It’s just after that point that I’m talking about.”

Already in today’s world, heated less than 1 degree Celsius above preindustrial times, heat waves have assumed a new deadly demeanor. In 2003, two hot weeks killed 35,000 people in Europe. It was called a once-in-500-year event. It happened again three years later (497 years ahead of schedule). In 2010, a heat wave killed 15,000 people in Russia. In 2015, nearly 700 people died in Karachi alone from a heat wave that struck Pakistan while many were fasting for Ramadan. But these tragic episodes are barely a shade of what’s projected.

“In the near term—2050 or 2070—the Midwest United States is going to be one of the hardest hit,” said Huber. “There’s a plume of warm, moist air that heads up through the central interior of the US during just the right season, and man, is it hot and sticky. You just add a couple of degrees and it gets really hot and sticky. These are thresholds, right? These aren’t just like smooth functions. It gets above a certain number and you hurt yourself very badly.”

China, Brazil, and Africa face similarly infernal forecasts, while the already sweltering Middle East has what Huber calls “existential problems.” The first flickers of this slow-motion catastrophe might be familiar to Europeans struggling to accommodate the tens of thousands of refugees at their borders: the collapse and mass migration of Syrian society came after a punishing four-year drought. Still others have noted that the Hajj, which brings 2 million religious pilgrims to Mecca each year, will be a physically impossible religious obligation to fulfill due to the limits of heat stress in the region in just a few decades.

But for the very worst-case emissions scenarios, heat waves would not merely be a public health crisis, or a “threat multiplier,” as the US Pentagon calls global warming. Humanity would have to abandon most of the earth it now inhabits. In their paper, Huber and Sherwood write: “If warmings of 10 degrees C were really to occur in the next three centuries, the area of land likely rendered uninhabitable by heat stress would dwarf that affected by rising sea level.”

Huber said, “If you ask any schoolchild, ‘What were mammals doing in the age of the dinosaurs?’ they’d say they were living underground and coming out at night. Why? Well, heat stress is a very simple explanation. Interestingly, birds have a higher set point temperature—ours is 37 degrees Celsius, birds’ is more like 41. So I actually think that’s a very deep evolutionary relic right there. Because that wet-bulb temperature was probably maxing out around 41 degrees Celsius in the Cretaceous, not 37.”

The Motivated Point Man

Huber told me about his “favorite story”: the US Army’s real-life parable of the so-called Motivated Point Man. In 1996, a platoon of light infantry spent days in the Puerto Rican jungle acclimatizing to stifling heat and humidity, cautiously monitoring their water intake before simulating a nighttime raid. The platoon included “some of the most fit and motivated soldiers in the battalion.” When the evening of the raid came, the platoon leader began leading his troops through the jungle, machete-ing a path through the brush. Before long, he was felled by fatigue and delegated his leadership to an underling.

When the second private failed to advance the platoon quickly enough, the platoon leader demanded to lead again. But soon he found himself hyperthermic and unable to walk. His soldiers had to douse him in cold water and supply him with intravenous infusions. Eventually four soldiers had to carry him. Before long, the extra demands vitiated the entire platoon, all of whom began to fall prey to heat stress. The exercise had to be called off before it became a massacre.

“So I look at that as, if it’s nighttime and acclimatized, fit people can just disintegrate into a pool of useless people on stretchers. That’s what I see happening to society, to cultures,” Huber said.

“If you want to know how mass extinctions happen, that’s how. So when people talk about the Pleistocene megafauna extinctions and Clovis people, sometimes they act like it’s a mystery how these things happen. But it happens in exactly the same way. You have something tearing apart the strongest members, the weaker ones try to fill in the gaps, they’re really not strong enough to take it, and the whole thing collapses.

“You want to know how societies collapse?” Huber said.

“That’s how.”