Nature | News

Longest historic temperature record stretches back 2 million years

Suggests greenhouse gases may warm planet more than previously thought.

Article tools

Rights & Permissions

Carmen Jaspersen/picture-alliance/dpa/AP

Researchers use tools such as ocean sediment cores, seen here at a German storage facility, to help reconstruct past climates.

A global temperature record published on 26 September in Nature1 extends 2 million years into the past — the longest continuous log yet published — and has sparked debate about how Earth's climate will change in the future.

The study harnesses data from dozens of ocean-sediment cores, as well as climate models, to provide estimates of global average surface temperatures. Prior reconstructions have gone back further — in some cases back to 3 million years — but were less comprehensive or focused only on particular time periods. The longest comprehensive temperature record available before this study went back 22,000 years.

“It’s a useful starting place,” says lead author Carolyn Snyder, director of the Climate Protection Partnerships Division of the US Environmental Protection Agency in Washington DC. “People can take this and improve upon it as more records become available in the future.”

Using a subset of the reconstructed temperature data, Snyder, who began the study while at Stanford University in Palo Alto, California, analysed the relationship between past temperatures and carbon dioxide (CO2) levels estimated from Antarctic ice cores covering the past 800,000 years. Based on that analysis, she found that future long-term warming induced by greenhouse gases could be more severe than many previous estimates.

Even if the amount of atmospheric CO2 were to stabilize at current levels, the study suggests that average temperatures may increase by roughly 5 °C over the next few millennia as a result of the effects of the greenhouse gas on glaciers, ecosystems and other factors. A doubling of the pre-industrial levels of atmospheric CO2 of roughly 280 parts per million, which could occur within decades unless people curb greenhouse-gas emissions, could eventually boost global average temperatures by around 9 °C. This is on the high end of existing estimates.

Proceed with caution

And this is where the study has encountered scepticism. “The key part of this paper is the temperature reconstruction, which is really valuable,” says Eelco Rohling, a palaeoclimatologist at the Australian National University in Canberra. But he adds that the question of how the planet will respond to atmospheric CO2 over the long term requires more detailed analysis. 

In particular, the study fails to account for subtle changes in Earth’s orbit that affected global temperatures and helped to drive the expansion and retreat of glaciers throughout the time period covered by the analysis, says Gavin Schmidt, director of NASA’s Goddard Institute for Space Studies in New York City. He says the effects of such orbital variations must be considered when comparing the glacial era to the present.

Schmidt's research into Earth's long-term sensitivity to CO2 has focused on a warmer era before the glacial ages, about 3 million years ago. That work suggests that the millennial-scale warming that would be expected from rising greenhouse gases is around 4.5 °C for a doubling of pre-industrial atmospheric CO2 levels2.

The new analysis has little bearing on the short-term climate sensitivity, which is what humanity might expect to see over the next century or two. The Intergovernmental Panel on Climate Change estimates that short-term climate sensitivity is around 3 °C of warming for a doubling of atmospheric CO23

Bare necessities

Snyder isn't convinced that the orbital effects are that important in this case. She says her study provides a single measure of the relationship between historic temperatures and CO2 levels. If that relationship were to hold up, it would suggest that the Earth is in store for even more warming in the future. But she stresses that conclusions about future climate change should be viewed with caution. “This is not an exact prediction or a forecast,” she says. “The experiment we as humans are doing is very different than what we saw in the past.”

Although Snyder’s temperature record will probably be improved as new data are incorporated in the future, the study fills a crucial gap in the archives of climate science, says Jeremy Shakun, a geologist at Boston College in Chestnut Hill, Massachusetts, who developed the previous temperature record out to 22,000 years.

“This is something that just needed to be done,” Shakun says. As for the debate about long-term climate sensitivity, he says scientists can debate the details, but it’s clear that heat will continue to build up in the Earth system for a very long time.

Journal name:


  1. Snyder, C. W. Nature (2016).

  2. Lunt, D. J. et al. Nature Geosci. 3, 6064 (2010).

  3. Intergovernmental Panel on Climate Change. Climate Change 2013: The Physical Science Basis. Working Group I contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge Univ. Press, 2013).

For the best commenting experience, please login or register as a user and agree to our Community Guidelines. You will be re-directed back to this page where you will see comments updating in real-time and have the ability to recommend comments to other users.


Commenting is currently unavailable.

sign up to Nature briefing

What matters in science — and why — free in your inbox every weekday.

Sign up



Nature Podcast

Our award-winning show features highlights from the week's edition of Nature, interviews with the people behind the science, and in-depth commentary and analysis from journalists around the world.