Skip to main content
U.S. flag

An official website of the United States government

Want Clues to Future Climate? Let's Look Back 3 Million Years.

Right-click and save to download

Detailed Description

Scientists have reconstructed what the climate conditions were 3 million years ago, and are using these data as one of the closest analogs to estimate future climate conditions.

USGS scientists Harry Dowsett and Marci Robinson discuss some of their findings regarding carbon dioxide’s impact, Arctic conditions, and the deep ocean’s system.




Public Domain.


[Intro Music]

Jessica Robertson: Hello and welcome to the USGS CoreCast. I'm Jessica Robertson.

Scientists have reconstructed what climate conditions were 3.3 to 3 million years ago and are using these data as one of the closest analogs to estimate future climate conditions. This past period was characterized by warm temperatures similar to those projected for the end of this century.

In looking to the past, scientists have discovered that the climate may be 30-50 percent more sensitive to atmospheric CO2 in the long term than previously thought. They also found increased evidence that the Arctic could face seasonally ice-free conditions and much warmer temperatures in the future. And they gained new insight on the deep ocean’s temperature variability and circulation system.

This new understanding of past climate conditions will help in projecting what the future may hold.

Today, we are joined by USGS scientists Harry Dowsett and Marci Robinson to discuss some of their recent findings. Thank you both for joining us today.

Harry Dowsett: It’s great to be here.

Marci Robinson: Thank you for having us.

Jessica Robertson: Harry, first, can you tell us what the benefit is of looking at climate conditions in the past and how that’s related to our current and future climate? Also, why did you choose to examine the mid-Pliocene period 3 million years ago?

Harry Dowsett: Sure. Looking at climate conditions in the past allows for a true understanding of how the Earth’s climate system really functions. The data we generate help refine climate models to build more precise long-term projections and help prepare for the impacts to our world.

We chose the environment 3 million years ago, that’s the mid-Pliocene, because global average surface temperatures during that time were about three degrees Celsius greater than today which is within the range projected for the 21st century by the IPCC.

So while the warming 3 million years ago was due to natural variability and the warming we see today is clearly human-induced, studying this time period serves as a plausible scenario for what the future may hold.

I'd also like to note that our research in the mid-Pliocene is the most comprehensive global reconstruction for any warm period. We created this reconstruction by analyzing fossils dating back to the mid-Pliocene.

Jessica Robertson: Thanks, Harry. Can you now tell us what discoveries you found regarding the sensitivity of our climate to carbon dioxide in the atmosphere?

Harry Dowsett: We found that the climate may be 30-50 percent more sensitive to atmospheric CO2 in the long term than previously thought. The temperatures we found during the mid-Pliocene were higher than what you would expect from the CO2 present at that time.

And this means that projections over the next hundreds of years of climate may need to be adjusted. This research was led by the University of Bristol. The USGS provided the reconstruction of environmental conditions during this timeframe and that allowed our collaborators to test the impact of CO2.

Jessica Robertson: Now, Harry, can you tell us why the impact of carbon dioxide was previously underestimated?

Harry Dowsett: Well, the underestimates occurred because current projections don’t accurately take into account the long term sensitivity of the Earth's system. Over the long term or during the mid-Pliocene, the Earth had more time to adjust to some of the slower impacts of climate change.

For example, warming leads to a reduction in ice and that allows more sunlight to be absorbed and warming continues into the future. With this new understanding, we can now update the projections of future climate change.

Jessica Robertson: Marci, the USGS also made some discoveries regarding conditions in the Arctic. Can you tell us about those new insights?

Marci Robinson: Sure. We found increased evidence that the Arctic could face seasonally ice-free conditions and much warmer temperatures in the near future. We documented the first quantitative evidence that the Arctic Ocean and Nordic Seas were too warm to support summer sea ice during the mid-Pliocene warm period.

Other researchers have suggested this warming in the Arctic but we have the first actual sea surface temperatures. We found temperatures that were between 10 and 18 degrees Celsius, which is about 50-64 degrees Fahrenheit, during the mid-Pliocene, which was very warm considering that current temperatures are around or below zero degrees Celsius or freezing.

This suggests that the record-setting melting of Arctic sea ice that we've seen over the past few years could be an early warning of more significant changes to come. In addition, this adds to our previous finding that the mid-Pliocene had a very different pattern of heat distribution than today with much warmer waters in the high latitudes.

Our increased understanding of how the Arctic responds to warmth will help refine climate models, which currently underestimate the rate of sea ice loss in the Arctic.

Jessica Robertson: What are some of the impacts that could result from the loss of sea ice?

Marci Robinson: There are several impacts. There are more than you may initially think. First, the loss of sea ice could contribute to continued Arctic warming by reducing the reflective properties of the ice. With no sea ice, the ocean is there to absorb the heat from the sun.

We may also see accelerated coastal erosion due to increased wave activity. Other researchers have found impacts to polar bears and seals that depend on sea ice cover. We may also see changes in weather patterns, such as intensified mid-latitude storm tracks and increased winter precipitation in western and southern Europe and less rainfall in the American west.

Jessica Robertson: Thanks, Marci. Harry, the USGS also made some conclusions regarding the deep ocean’s temperature variability and circulation system. Can you tell us about that?

Harry Dowsett: Sure. We created a three-dimensional reconstruction of the ocean during the mid-Pliocene. Based on conditions during that time, we found the deep ocean was affected more by surface warming than we previously thought. The average temperature of the entire ocean during the mid-Pliocene was maybe one degree warmer than current conditions, and that’s a lot of heat. It wasn’t just at the surface but it was at all depths. Understanding ocean temperature variability allows for more accurate predictions of factors such as sea level rise and ice volume change.

High ocean surface temperatures have also been found to result in a more vigorous deep ocean circulation system. This results in a faster transport of large amounts of warm water, with possible impacts like reduction of sea ice and overall warming of the Arctic.

Our findings are significant because it was previously thought that the deep ocean stayed relatively stable and cold, and that the deep ocean circulation system would slow down as surface temperatures increased.

Jessica Robertson: Thank you, Harry. Marci, is there anything else you’d like to share with us about your research on the mid-Pliocene?

Marci Robinson: I would like to mention some research that we are conducting right now. We’re testing the hypothesis that the increased transport of heat to the high northern latitudes may be partly related to the changes in the depth of ocean ridges in the North Atlantic. So look out for that. That will be coming soon.

Jessica Robertson: And Harry, is there anything else you’d like to add?

Harry Dowsett: Well, I should recognize our collaborators on this research. USGS leads the research through the Pliocene Research, Interpretation and Synoptic Mapping group known as PRISM. Our primary collaborators in PRISM are the University of Leeds, University of Bristol, Columbia University, Brown University, and the British Geological Survey.

Jessica Robertson: Thank you both for joining us. And thank you to all of our listeners.

To learn more about PRISM research and for links to read the full articles on each of the discussed topics, visit and view the transcript for episode 115.

CoreCast is a product of the US Geological Survey, Department of the Interior.

[Ending Music]

More Information:

PRISM research

Nature Geoscience article on impacts of atmospheric CO2

Stratigraphy article on Arctic conditions during the mid-Pliocene

Climate of the Past article on the deep ocean’s temperature variability and circulation system during the mid-Pliocene

Show Transcript