No one can predict the future. But sometimes we can get a solid idea of what’s coming by looking at the past. In his new book, Our Fragile Moment: How Lessons from Earth’s Past Can Help Us Survive the Climate Crisis, renowned climate scientist Michael Mann describes the world climate change is creating based on what we know from specific times in Earth’s four-billion-year history when the planet was extremely hot or extremely cold.

Scientific American asked Mann, director of the Penn Center for Science, Sustainability and the Media at the University of Pennsylvania, to give us the main lessons from each era and to explain the warning, and the hope, they provide for today and the future. As Mann says in his book’s intro, “the collective evidence from … the paleoclimate record of Earth’s past climatic changes … actually provides a blueprint for what we need to do to preserve our fragile moment” on a planet that has survived much more than what we humans could.

[An edited transcript of the interview follows.]

Let’s start with the first two eras: the Faint Young Sun era was three billion years ago, and then Snowball Earth occurred 800 to 550 million years ago. What happened, and what did we learn?

Early on, the sun was 30 percent less bright, but the planet wasn’t frozen; the oceans were teeming with life already. As the sun gradually got brighter and brighter, the concentration of carbon dioxide in the atmosphere got lower and lower during a couple billion years. As living organisms spread, they moderated the atmosphere and temperature. It suggests that there are restorative mechanisms—that life itself helps keep the planet within livable bounds. But only to a point!

Cyanobacteria loaded Earth’s atmosphere with oxygen, which had previously been largely anoxic [deficient in oxygen]. Oxygen scavenges methane, so there was a rapid disappearance of methane; Earth lost that early methane greenhouse effect. Positive feedback loops occurred. The planet spun out of control into a snowball.

Life can help keep the planet within habitable bounds, but it can also push the planet beyond those boundaries. Today we are the living things that are impacting our climate. Is our future one of resilience or instability? The paleoclimate record tells us we’re somewhere in between. We can still achieve stability, but if we continue burning fossil fuels, we will have instability.

A massive buildup of carbon dioxide in the atmosphere 250 million years ago, during the Permian period, led to the Great Dying, when most life on Earth was wiped out. What does it tell us about the so-called sixth extinction we’re in right now?

The Permian has the greatest documented extinction—something like 90 percent of all life went extinct. There was great natural warming driven by unusually active volcanism that loaded the atmosphere with carbon dioxide. It warmed the planet rapidly on a geological timescale, although it was nowhere near the rapidity of what we’re doing today.

Some people cite this era as a reason for believing that we are experiencing runaway warming and that our extinction is now ensured. They say we’re experiencing runaway methane-driven warming from thawing permafrost—and that it’s too late to do anything about it; we’ll all be extinct. But I spent quite a bit of time going through the literature, and it doesn’t hold up. There’s no evidence that there was any major release of methane at that time. There are a whole bunch of things that make it a bad analogue for today. I go through them in the book. For example, there was a massive continent that was very dry with very tenuous, early forests that were very vulnerable to wildfire and to collapse. So there was a much greater potential for massive deforestation and therefore a massive lowering of oxygen. There was also a huge increase in sulfur in the ocean that probably extinguished quite a bit of sea life.

There are all these things that contributed to that particular catastrophe that aren’t analogous today. There’s no evidence that we’re going to see substantial lowering of oxygen concentrations from anything that we’re doing. There’s no evidence that we’re seeing massive releases of sulfur—although deoxygenation like the Black Sea has experienced, with a larger anoxic zone and die-offs, is a bit of a warning.

About 56 million years ago Earth became very hot again—as hot as it ever has been. This was the so-called Paleocene-Eocene Thermal Maximum, the PETM. Are we headed for that instead?

This is the same as the Great Dying. Scientists no longer think that methane played a major role in the PETM. But there is a different lesson. The PETM is notable for the rapid warmup—it happened not across millions of years but in as short as 10,000 or 20,000 years. This is very rapid from a geological standpoint, although, again, it’s 100 times slower than today. The warming spike happened on top of an already warm planet; it took the planet to temperatures higher than anything that’s documented in the geological record.

The PETM reached levels of heat that would be dangerous for human beings, and we are already encountering wet-bulb temperatures [an estimation of the effect of temperature and humidity] that are deadly in some parts of the world. The PETM would have been a world where large parts of the planet were too hot for humans. So people say, “Oh, look, life adapted.” There was a massive miniaturization of some species. Horses shrunk 30 percent in order to adapt [smaller bodies, with a higher ratio of surface area to volume, have less trouble shedding heat]. The reality is that when you see something so dramatic as horses shrinking by 30 percent, that means there would have been very large amounts of maladaptive species; there would be a massive loss of life along the way. The idea is that human beings can just adapt, but those selective pressures don’t favor anyone.

Let’s jump back 10 million years before the PETM to 65 million years ago. An enormous asteroid struck the Earth, shrouding the planet in dust, which rapidly cooled its surface, killing the land-based dinosaurs (not the avian ones). That’s very different from earlier events and from climate change today. What can that episode tell us?

The dust very rapidly cooled the planet, so any animal that couldn’t burrow into the ground or find shelter—everything larger than a dog, basically—died out. The climate story is that even though it’s a scenario of global cooling rather than global warming, it was rapid. [The event is also known as the K–Pg boundary, the transition between the Cretaceous period and the Paleogene period.]

This also relates to societal fragility. In the height of the cold war, we were focused on nuclear winter. An all-out nuclear war would shroud the planet with dust, smoke and ash. The fate that befell the dinosaurs could be our fate. Carl Sagan, of course, was the one who really raised awareness. He and his colleagues published a paper in late 1983 that said it isn’t just the physical destruction that’ll get us; what will really get us is the rapid cooling of the planet.

As the cold war ended, the world felt that that particular threat had waned. But with recent tensions with Russia’s invasion of Ukraine and the threat by Putin to use tactical nuclear weapons, all of a sudden this threat has reemerged. The point that applies from the dinosaurs it that it isn’t the absolute levels of warmth that matter today; it’s the planet we are evolved for. The dinosaurs had evolved for a certain climate, and when it cooled rapidly, they perished. Other animals were able to exploit the niches that emerged. Ironically, it was our ancestors, the early mammals. In one sense, we’re here because the dinosaurs perished. If we have eight billion people adapted to a climate that is disappearing as we continue to warm the planet, that’s a real danger.

Much more recently we’ve had several ice ages; the Last Glacial Maximum was about 20,000 years ago. What did these cold periods reveal about our increasingly hot period now?

The K–Pg event was a punctuated interval of cooling during an otherwise warm era. About three million years ago, CO2 levels dropped to near what they are today. To some extent, the Pleistocene [which started about 2.6 million years ago] is a better analogue for our climate today. There was no Greenland ice sheet. Sea levels were 10 feet higher at least, maybe 20. The planet was warmer than it is today. Is that the future that we are now committed to? The answer isn’t so clear-cut because of hysteresis [when a physical change lags the force that created it]. The behavior of things when you’re on a cooling scenario is different from the behavior of things when you’re on a warming scenario. You can reach the same point, and the climate can look very different depending on how you got there. It’s probably not the case that we have committed yet to the melting the Greenland ice sheet. That hysteresis effect buys us a little bit of a margin of error but not a big one. Maybe it buys us a half a degree more warming. Once again it shows us the fragile nature of this moment. We could soon exceed that range of resilience if we continue on the path we’re on.

The last timeframe in the book is the Common Era, the past 2,000 years, when humans have dominated life on Earth. You address questions we are confronting today: How will warming affect El Nino or the Asian summer monsoons? Will the North Atlantic Ocean’s conveyor-belt circulation change? Are our climate models underestimating the pace and extent of changes underway? Given all that, what worries you the most? What surprises you?

What worries me the most is beyond the hockey stick. [The “hockey stick” was a graph published by Mann and others in 1999. It showed that the global average temperature was the same or slightly decreasing for more than 900 years and then turned sharply upward from the mid-1900s through 1999. It looked like a hockey stick laying on its side, with the blade at the far right pointing up in the air.] The obvious difference from past events is that we’ve warmed the climate so much faster during this timeframe. It turns out that El Nino, sea-level rise and Arctic sea ice levels can all follow the hockey stick pattern. There’s a theme: changes to some of these things are happening sooner than we expected.

One of these is the Atlantic meridional overturning circulation, or AMOC—the ocean conveyor belt. That’s one of the surprises: the dramatic slowdown that we already see. There has been a dramatic slowdown in this circulation in the past century, even though the models say any slowdown should only occur during the upcoming next century. The blade of that hockey stick is coming about a century too early. One of the reasons is probably that we’re losing Greenland ice faster, so we’ve got more fresh water already running off into the North Atlantic earlier than we expected.

What gives you the most hope?

We don’t know precisely how close we are to triggering some devastating tipping point that could threaten human civilization. The collective evidence from the past tells us that we’ve still got a safety margin. Science tells us that if we act quickly, if we act dramatically, we can avoid warming that will bring far worse consequences. That’s the fragility of this moment: we have a little bit of a safety margin, but it’s not a large safety margin. The phrase I use often these days, a phrase that characterizes the message of this book, is the pairing of urgency and agency. Yes, it’s bad, and we face far worse consequences if we don’t act. We can see devastating climate consequences already. That’s the urgency. But the paleoclimate record tells us we haven’t triggered runaway warming yet. We can avoid that point of no return if we act quickly and dramatically. That’s the agency. We’ve got 4 billion years of Earth history. Let’s try to learn from it.