Back to the Miocene: What the climate 13.8 million years ago could tell us about our future world

Projections of our future under climate change paint a picture of extreme weather and acidified oceans, a world many of today’s animals — including humans  —  may struggle, or fail, to survive. Yet despite overwhelming scientific evidence screaming at us to reverse course, humanity just keeps burning more fossil fuels, skidding past climate goals, and ramping up consumption. We don't really know what to expect from the climate we're hurtling towards. What might the world be like in 50 or 100 years? Maybe looking to the past can give us clues.

For prehistory fans, the Miocene, with its fantastic mammal life, is an immensely attractive period. From Dryopithecus, a lineage of extinct primates that included forerunners of humans, to the toxodonts, large-hoofed mammals with long, curved incisors, to mammals similar to sloths, armadillos and anteaters, to marsupial carnivores, this epoch that stretched from 23.03 to 5.33 million years ago was a glorious time of the weird and oddly familiar. The Miocene is even named using the Greek words for “less” and “new.”

In the oceans, molluscs and crocodilians radiated into diverse new forms. Continents were arranged quite similarly to their configuration today, while plate tectonics pushed the Andes up from the Earth, and the Himalayas began a slow rise towards their current extravagant height. Over almost 18 million years, the mostly steamy climate gradually cooled and carbon dioxide levels declined from an early high. What a time to be alive!

Paleoecology, the study of past ecosystems, employs geology and the study of fossil evidence from plants, animals or other kingdoms of life to reconstruct what it would’ve been like back then. Most often, paleoecologists use modern understanding of how organisms interact with their environment and apply them to what we know about the past in order to bring it to life, as it were.

Increasingly, though, it’s of great interest to reverse that process: that is, to use the past to imagine the world we may face in our not-too-distant future. Of the hundreds of millions of years of history at our disposal, the Miocene epoch stands out as the closest thing to a proxy for conditions we may expect to face within the next 50 or 100 years, as human activities have raised our atmosphere’s CO2 content by 50% since the Industrial Revolution. These and other greenhouse gas emissions are heating up the planet and causing sea levels to rise while contributing to erratic weather patterns that are decimating agriculture and intensifying natural disasters.

A proxy for our near future

Before we put ourselves on a clear path to overshooting the maximum global heating target we set for ourselves, a different epoch, the Pliocene (which took over from the Miocene and ended just 2.4 million years ago), seemed like a reasonable proxy for our coming world. With its CO2 levels that maxed out at 425 ppm (roughly where we’re at already), the Pliocene offered a mostly warm and pleasant climate preceding the ice ages of the Pleistocene. Not such a bad vision for our future. But that was then.

“Not too long ago, a lot of people would have been talking about the Pliocene as a good proxy for modern [climates] but we’ve actually, by all reasonable estimates, surpassed the Pliocene CO2 levels, and that’s why the focus is mainly on the Miocene,” explained Tammo Reichgelt, a geologist and paleobotanist at the University of Connecticut, in a video interview with Salon.

Under the legally-binding 2015 Paris Agreement, signatories committed to holding the now-inevitable increase in the global average temperature “to well below 2° C above pre-industrial levels” while still working to keep it to no more than 1.5° C — although more recently it’s been understood that we really need to avoid exceeding 1.5° C by the end of the century, which according to the United Nations means that this year at the latest should see a peak in greenhouse gas emissions, followed by a steep 43% reduction by 2030.

Of the different best to worst case scenarios outlined by the Intergovernmental Panel on Climate Change, the relatively recent Pliocene epoch is indeed a fair proxy for the very optimistic scenario for the near-term future we face, Reichgelt explained. By contrast, the earlier Eocene, when C02 reached 1000 ppm and the global mean temperatures were 13º C warmer than today, could represent the worst case scenario. But it’s frankly too late to put much faith in that wildly optimistic scenario.

“And as we are currently not in the best case scenario, but also not the worst case scenario, we’re sort of on a middle trajectory, and that is what the Miocene best represents,” Reichgelt said.

Hyaenodon, extinct carnivorous mammals, late Eocene to middle Miocene. (Florilegius/Universal Images Group via Getty Images)Between the nightmare of an Eocene-like world and the no-longer realistic Pliocene as a proxy lies a frightening but quite plausible span of possible global CO2 concentrations between around 540 ppm and something under the apocalyptically worst-case scenario of 900 or higher ppm by the end of the century.

Along with CO2 levels similar to what we might expect under that not-great, could-be-worse scenario, the Miocene boasted a relatively similar continental configuration to our own, which ensures a more similar climate than otherwise, as well as its biomes.

“The ecosystems would be recognizable to what they are today. You know, like flowering plants are the dominant plants,” Reichgelt said. “You would have grasslands, which didn’t really exist that much in the Eocene.”

Significantly, the Miocene was a nearly 18 million year epoch full of change, albeit far slower change than ours. It started with a period of glaciation that must have been a chilly change from the greenhouse-like Oligocene, and ended with a prolonged period of glaciation, too. But through much of the Miocene, it was a warm world compared to today’s, a high CO2 planet that gradually cooled over millions of years until ice sheets developed in the Northern Hemisphere and Antarctica. 

Around the middle of the epoch, we reached what is called the Miocene Climate Optimum (MCO), a roughly two million year-long greenhouse period when the world experienced its last period of sustained warmth, and the CO2 level was at least 500 ppm. This is the period we’re talking about, most specifically, when we talk about the Miocene as a proxy for our future, although changes throughout the Miocene are relevant: basically, from the middle Miocene Earth went through a process roughly opposite the one we are experiencing (and causing) today.

Want more health and science stories in your inbox? Subscribe to Salon's weekly newsletter Lab Notes.

“We go from a globally warm world with an Antarctic ice sheet [in the middle Miocene] to a globally cool world with a larger Antarctic ice sheet and an Arctic Greenland ice sheet. Contemporary climate change will likely mean the reverse process over the next 100 or 200 years,” Reichgelt said.

Reichgelt is part of a thriving community of paleoecologists interested in what we can learn, and infer about the future, from accurate modeling of conditions in deep time, tens of millions of years ago. They put their different models and data together then analyze the resulting ensembles, an approach called a Model Intercomparison Project that aims to understand the differences and similarities between geophysical models. Since the paleoecologists focus on Deep Time, or epochs from tens of millions of years ago, they call their ensembles DeepMIPs. The Miocene-focused DeepMIP builds on a 2021 synthesis of various existing Miocene climate models along with available reconstructions of the ground and ocean surface temperatures during this period.

The DeepMIP-Miocene researchers found that the global mean surface temperatures during the Miocene ranged from 5.3º C to 11.5º C higher than the temperatures on earth just before the start of industrialization a couple of hundred years ago. Only about 2 degrees of that difference from today can be explained by factors other than CO2.

So what can we learn from the Miocene, its climate and CO2 levels and transitions, that might help us understand how life on Earth will be affected by our own, speedy transition to a warmer world?

Under the sea

Plankton, the abundant tiny creatures that exist in many different types today, were far less diverse during the Miocene. Katherine Crichton, a paleoclimatologist at the University of Exeter, explained to Salon in an email interview how her team explains their findings of reduced plankton biodiversity during the Miocene compared to now.

“The carbon cycle describes how carbon moves and is distributed in the ocean — this carbon includes organic and inorganic carbon. Phytoplankton living in surface waters photosynthesize, and take up carbon from the waters around them (they convert inorganic carbon into organic carbon by growing!). These plankton are predated by other plankton, and so on, and form a whole load of organic carbon near the surface,” Crichton said.

What about lower down? Well, due to gravity, we get what Crichton calls marine snow — the continuous shower of organic matter made up of everything from dead plankton to fecal pellets, that falls from the upper ocean to deeper waters. And that marine snow becomes food for the creatures of the deep. 

“If you imagine that lots of organic matter is leaving the surface waters and sinking, this is effectively removing carbon from the surface and moving it to deeper water. Between the surface waters and the atmosphere, air-sea gas exchange occurs,” Crichton told Salon. “If surface ocean waters have less carbon dissolved (inorganic), then they absorb more carbon from the atmosphere (it is more complicated than and depends on other things too). But this is how the ocean carbon pump works. So, a strong ocean biological carbon pump can draw down atmospheric carbon levels. The biological carbon pump is a result of the marine ecological system.”

Then there’s temperature, which was higher back in the Miocene.

“A fridge works because it preserves food by keeping it at lower temperature — bacterial degradation of the food happens slower at lower temperatures. If ocean waters are colder, that organic matter that is sinking degrades slower, it can get deeper, and provide a source of food for deeper-living fish, creatures, etc.,” Crichton explained. “In the warmer Miocene, less food at depth was probably why there were fewer plankton at depth, and during the cooling since the Miocene, the number of plankton species at depth greatly increased. This we attribute principally to increasing food availability.”

In the world we can expect under most scenarios of global heating over the coming decades, the biological carbon pump will become weaker. And a weaker biological carbon pump is less able to draw down atmospheric carbon into the ocean, “which means we may be underestimating future temperatures,” Crichton said, as this is a positive feedback to warming.

And on land

In a paper published January 2025, Reichgelt and collaborator Christopher West compared the shapes of fossil leaves of the Miocene gathered and analyzed in many previous studies with a dataset of modern leaf morphologies. They were able to draw inferences about the global climate on land, about productivity of different regions, and about characteristics of the various biomes of the Miocene compared to today. Among other things, they found that precipitation was much more varied from today than temperature.

“You would expect that in a high CO2 world, the temperature is the main signal that you get. But instead, we found that precipitation was the largest signal. I don’t know what to make of it yet,” Reichgelt told Salon.

What is clear is that this was a very moist world, with 89% of the reconstructed Miocene biomes being wetter than today while only 66% were warmer, despite our usual focus, when we think of the ways in which we are setting out on the Miocene trajectory only in reverse, on the warmer world we are entering, rather than the sogginess we should probably expect in many regions.

Carbon dioxide levels affect plants by allowing for greater photosynthesis rates, and by increasing water use efficiency, in that plants can achieve the same amount of photosynthesis with less loss of water through the pores in their leaves, because higher availability of CO2 absorbed through open pores means they can keep them closed more of the time. Thanks to all this, it was also “a globally greener Miocene world,” as Reichgelt and West write in the 2025 paper. Various forms of evidence suggest that the biosphere was more productive during the Miocene compared to now, and that at higher latitudes, this effect was more pronounced.

So what would it be like?

There is a huge amount of research that looks at very specific aspects of life in different parts of the Miocene, in different areas of the Earth, in different biomes on land and under the sea. Beyond Reichgelt’s general description of a globally greener, wetter, warmer world, it’s hard to say how we personally, or our personal favorite forms of life, are likely to fare under Miocene-like CO2 and temperature levels — nor should we, since life exists in dependent relationships, or food webs. What we can say is what did well in the past.

Take palm trees. Although plants don’t get up and migrate like herds of wildebeest, they do shift over time as climate zones change. In fact, we’re already seeing a northward shift in the distribution of palm trees, those tropical icons. Since palms are one type of plant that really can’t propagate in freezing temperatures, Reichgelt wanted to know just how far north they might spread. As he and his colleagues reported a few years ago in Scientific Reports, many palm species are able to do well in fairly temperate, rather than tropical, climes, and paleoclimate reconstructions suggest they can do just fine at a minimum temperature of 2º to 8º C (35.6º F to 46.4 Fº.)

Fifty million years ago, in fact, palm trees grew in Antarctica, as palm fossils attest, and in the more recent but still mostly toasty Miocene, they were still found in North America, Europe, Asia and in South America far south of the tropics. Back in the present day, it’s been reported that exotic palms are already displacing native species in the Swiss Alps, posing a fire risk. More fires and more palms in our future? Perhaps.

And what enjoys a good lush plant to chew on? Herbivores. In a 2020 research article that provides a rich overview of what we know about conditions and biology during the Miocene, Margret Steinthorsdottir and colleagues outline the radiation of mammals, both plant-eaters and the carnivores that feed on them, that occurred thanks to the warming climate around the middle Miocene. There were radiations of hoofed mammals; of primates including our ancestors; of whales and dolphins; of the all-time biggest sharks. Oh, and 10-foot tall, fast-running, meat-eating terror birds. ​​

By the end of the Miocene, though, as the Earth followed the opposite path to the one we’re on now, grasslands expanded at the expense of forests, and we saw the development of tundra and cactus-filled deserts, with desert-specialist animals like snakes and rats soon to follow.

Limitations on the Miocene as a proxy for our future world

Sadly, the taxodonts will not grace our future world. The long-armed, horsey Chalicotheriidae, reminiscent of Bojack Horseman, won’t be joining us at the bar. Smilodon, the catty predator whose ancestors emerged in the early Miocene, will not smile on us again. Nor the “bizarrely specialized” family of carnivorous marsupials, Malleodectidae, which used their massive ball peen-like third premolars to crush snails. Not the dog bears, Hemicyoninae, who emerged before and lived through the Miocene, nor the bear dogs, Amphicyonidae, which died out by the late Miocene. Evolution doesn’t work like that. Barring the odd de-extinction attempt, what’s lost is gone forever (that includes, thank goodness, the terror birds.)

Crichton explains that climate proxies help us to understand if our best Earth system models capture the main drivers of importance for the climate system. Past climate data helps to improve such models so that we can make predictions about the impacts we might expect on ecological and human systems, and have confidence about our results and our ability to make decisions based on model projections. Findings like Crichton’s might also provide a warning that a future climate could lack the deep-sea plankton necessary to sustain large deep-sea fish.

But there are limits to how far we can take the Miocene analogy. Reichgelt, for example, notes that while all the paleoclimate proxy models that rely on comparisons of fossil plants to modern ones use leaf surface area as a measure of precipitation, increased water use efficiency resulting from higher atmospheric CO2, rather than more rain, could actually explain the observed larger leaf sizes during the Miocene. This would mean that the models could be overestimating just how wet the world was.

It’s also possible that we’re being seduced by the idea that the Miocene might represent a “happy medium.” As Steinthorsdottir and colleagues write, “More pessimistic scenarios of unmitigated greenhouse gas emissions quickly move us beyond the Pliocene state, pushing Earth's systems into a potentially vulnerable position where many of its ‘tippable’ subsystems such as glaciers, sea ice, forest biomes, deserts and coral reefs will be permanently destabilized […] an ‘intermediate’ deep-time climate analog, where boundary conditions are close to modern but extreme climate changes occurred, is therefore of great interest.”

As humans we have a notorious tendency to believe that whatever’s in the middle of two given extremes is moderate, cozy, all around OK. (In politics, this results in the Overton Window.) But Miocene-style hydrological or water cycles favor high altitude wind events, like cyclones and hurricanes, that transport heat and moisture evaporating from the tropics to higher latitudes, or California’s intense seasonal rainstorms. The future may be lush, sure, but it’ll also be erratic and dangerous for us. And the “tippable” subsystems Steinthorsdottir mentions may have tipping points that occur well within a Miocene-like context, as scientists have warned.

It’s also important, from the perspective of the world humanity may find ourselves living in, to understand that the most significant limit in using the Miocene as a proxy for tomorrow’s world has to do with the drastic difference in the rate of change.

“We have to remember scale, right?” Reichgelt said. “We were talking about millions of years of evolution versus hundreds of years or maybe even decades of change [now]. So really different pace of change, and we’re also going in the opposite direction. So whether that will play out like that in the short term, where you have a greener world, is really a much more open question.”

He returned to the issue of water use efficiency, which should be greater in a warmer, higher CO2 world. Turns out, it’s not that simple. 

“That would in theory favor the expansion of forests, but forests need a really long time to establish, and over a shorter period of time, human disturbance or drought [are] much more effective at killing whatever is capable of establishing itself,” Reichgelt said. 

Crichton echoed the same sobering warning.

“The warming we are currently imposing is happening much much faster than changes seen during the Miocene cooling. It took say 10 million years for the Earth System to change temperature naturally by around 5 degrees — now it could happen in 100 years,” Crichton explained. If, she added, we are unlucky — and don’t take necessary steps to avoid such extreme and rapid temperature change.