Read our COVID-19 research and news.

Paleobotanist Scott Wing stands in wintry Wyoming badlands, where alligators swam 56 million years ago.


A 500-million-year survey of Earth's climate reveals dire warning for humanity

When it opens next month, the revamped fossil hall of the Smithsonian Institution's National Museum of Natural History in Washington, D.C., will be more than a vault of dinosaur bones. It will show how Earth's climate has shifted over the eons, driving radical changes in life, and how, in the modern age, one form of life—humans—is, in turn, transforming the climate.

To tell that story, Scott Wing and Brian Huber, a paleobotanist and paleontologist, respectively, at the museum, wanted to chart swings in Earth's average surface temperature over the past 500 million years or so. The two researchers also thought a temperature curve could counter climate contrarians' claim that global warming is no concern because Earth was much hotter millions of years ago. Wing and Huber wanted to show the reality of ancient temperature extremes—and how rapid shifts between them have led to mass extinctions. Abrupt climate changes, Wing says, "have catastrophic side effects that are really hard to adapt to."

But actually making the chart was unexpectedly challenging—and triggered a major effort to reconstruct the record. Although far from complete, the research is already showing that some ancient climates were even more extreme than was thought.

Ancient glaciations are easy enough to trace, as are hothouse periods when palms grew near the poles. But otherwise little is certain, especially early in the Phanerozoic, which spans the past 541 million years. Paleoclimate scientists study their own slices of time and use their own specialized temperature proxies—leaf shape, say, or growth bands in fossilized corals—which often conflict. "We don't talk to each other all that much," says Dana Royer, a paleoclimatologist at Wesleyan University in Middletown, Connecticut. So at a meeting last year, Wing and Huber assembled a loose-knit collaboration, dubbed Phantastic, dedicated to putting together a rigorous record. "Most people came away quite inspired to do something about this," says Dan Lunt, a paleoclimate modeler at the University of Bristol in the United Kingdom.

The value of a deep-time temperature curve extends beyond the exhibit. Similar curves exist for atmospheric carbon dioxide (CO2). Combine the two and you can see how much warming CO2 caused in the past, says Jessica Tierney, a paleoclimatologist at the University of Arizona in Tucson. Because the latest climate models seem to forecast more warming than earlier ones, "using paleoclimate to constrain the models is becoming much more important," she says. "We feel we have to step up."

Fever line

A preliminary global temperature curve shows that marine life diversified in extreme heat (1) before land-based plants absorbed carbon dioxide (CO2) and polar ice caps formed (2). Volcanoes and erosion swung CO2 levels up and down (3), but mammals evolved in a warm period (4). Now, humans are rapidly warming the climate again (5).

500 Millions of years ago 10˚ 15.5˚ 21.1˚ 26.7˚ 32.2˚C 50˚ 60˚ 70˚ 80˚ 90˚F 450 400 350 300 250 200 150 100 50 Today World without polar caps World with polar caps 2 3 5 4 1

But ancient global temperatures are elusive because they varied with location and season, and because the gauges drop away as you move deeper in time: Tree rings go back only thousands of years and ice cores only a million years or so. Still, oxygen isotopes in tiny fossilized shells on the ocean floor give a fairly reliable longer-term record. Because water molecules with lighter oxygen variants evaporate faster and end up locked in ice sheets, the ratio of light to heavy isotopes in the fossils indicates the volume of global ice, a rough guide to temperatures.

However, ocean floor older than 100 million years or so is scarce, devoured by the constant churn of plate tectonics. To go deeper in time, Ethan Grossman, a geochemist at Texas A&M University in College Station, looks for marine fossils found on land—mostly teeth and once-common bivalves called brachiopods. They tend to be from the shallow, isolated seas that formed inside ancient supercontinents. To glean temperatures from those fossils, scientists have to assume those seas had a balance of oxygen isotopes similar to the ocean today.

This "water problem" is decades old. But scientists in Phantastic are attacking it with a second thermometer, based on a new technique, called clumped isotopes, that measures the abundance of two or more rare isotopes. Using sensitive mass spectrometers, they analyze the fossil shells for carbonate molecules that contain a heavy isotope of oxygen bound to a heavy carbon, which form more frequently at lower temperatures. The results will be misleading if the fossil has been exposed to heat and pressure during its burial, but researchers have learned how to identify altered specimens. "We've moved into the place where we can apply it," says Kristin Bergmann, a geobiologist at the Massachusetts Institute of Technology in Cambridge, who is using clumped isotopes to prepare a temperature record of the past billion years.

In collaboration with Gregory Henkes, a geochemist at the State University of New York in Stony Brook, and others, Grossman has gone through his samples, tossing out those that show signs of alteration, and analyzed their clumped isotopes. The results match his existing oxygen-isotope measures, and they tell a startling story, he and Henkes reported last year in Earth & Planetary Science Letters. Some 450 million years ago, ocean waters averaged 35°C to 40°C, more than 20°C warmer than today. Yet marine life thrived, even diversified. "It's unsettling for the biologists, these warm temperatures we're proposing," Grossman says. "These are extreme for modern organisms."

To turn such data into a global temperature curve, researchers need to fill gaps in geography and time. One Phantastic collaborator, Christopher Scotese, a geologist at Northwestern University in Evanston, Illinois, has come up with a simple way to spread limited data into a global picture. He uses the presence of polar ice caps to indicate whether the world had a steep temperature differential between the equator and its poles.

Other collaborators are using the sparse data to calibrate computer simulations of the ancient climate, the way weather models use satellite data as a reality check. Lunt and Paul Valdes, also at Bristol, are ground-truthing a suite of several hundred paleoclimate simulations. They've been able to extrapolate temperatures across the planet for broad stretches of the Phanerozoic.

Although Wing and Huber are pleased with the work they've seeded, they also ran out of time. The temperature curve they're presenting at the museum opening is a beta, Wing says. "It's sort of jamming together different kinds of observations, different kinds of models, different kinds of procedures, and probably different assumptions." The plan is to replace it once the Phantastic team's efforts reach maturity. But even the draft version should open eyes, Grossman says. "This kind of work gives people a sense of how easy it is to tip into a warm period. Because the world has been warm."

*Correction, 24 May, 11:40 a.m.: A previous version of this story incorrectly stated that brachiopods are extinct.