Atmospheric CO2 was the primary driver of a 400,000-year global warming event, known as the middle Eocene climatic optimum (MECO), according to a new study. The finding, which could help climatologists better understand the precise relationship between CO2 concentration and climate change today, is described in the November 5 issue of Science.

The climate trend across the entire Eocene, an epoch between 55 million and 34 million years ago, was actually characterized by long and gradual cooling. Much of that activity took place during the middle Eocene, when the planet transitioned from a warmer climate to a cooler one. But the MECO (about 40 million years ago) interrupted that trend, representing the last major temperature increase before the end of the epoch, which was marked by Antarctic glaciation. In order to investigate the role of CO2 during this warming episode, researchers analyzed sediment taken from deep beneath the ocean floor off the eastern coast of Tasmania. The core contained a record of fossils that spanned the relevant time interval.

The researchers relied on two separate organic proxies, also called paleothermometers, to reconstruct changes in sea-surface temperature during the MECO. Both are based on variation—due to temperature change—in structural characteristics of the molecular remains of specific microorganisms. In other words, the team analyzed specific molecules in which certain variations are known to be reflective of temperature change.

The MECO has for a long time been enigmatic to climate researchers, says study co-author Alexander Houben, a paleoecologist at Utrecht University in the Netherlands. Although there is lots of evidence for a large-scale temperature increase during this time, so far it's all been based on a different proxy: the occurrence of a specific oxygen isotope in ancient carbonate samples. Oxygen isotope values, however, are influenced not only by temperature, but also by seawater composition and changes in ocean ice flow. This study for the first time employs two independent proxies that "can clearly differentiate between temperature changes and other factors,” Houben says. The authors report that the paleothermometers indicated a warming of the sea surface during the MECO, at least in the southwestern Pacific, by 3 to 6 degrees Celsius.

To figure out how much atmospheric CO2 increased during the MECO, the researchers reconstructed the changes in its concentration by determining the ratio of stable carbon isotopes in organic molecules called alkenones. Variation in the composition of these molecules, produced by algae, can serve as an indicator of atmospheric CO2. Isotopic proportions are also influenced by the growth rate of the algae, which is proportional to the available food and nutrients in the water. So the authors considered multiple scenarios to account for variations in nutrient availability, concluding that CO2 concentration increased by a factor of two to three during the MECO.

James Zachos, a professor of Earth and planetary sciences at the University of California, Santa Cruz, who specializes in the time period and methodology used the study, said in an e-mail to Scientific American that apart from what may be an overestimation of the mean partial pressure of CO2, and the magnitude of its rise, "the data look reasonable." The study adds value, he says, because "the application of multiple proxy records in one core to reconstruct climate is unique, at least for this time interval."

The MECO makes for a useful "historical laboratory," Houben says, within which climate dynamics can be studied over a much longer timescale compared with other historical warming events of interest to climate researchers. By investigating CO2 variations and temperature dynamics over longer periods, researchers can obtain a clearer understanding of the still partly unknown role played by long-term climate feedbacks, such as changes in the ocean's carbon chemistry and/or large-scale changes in vegetation.

Currently, climatologists have a much better understanding of the role short-term feedbacks, such as changes in water vapor or sea ice. Incidentally, there were no glaciers at the time of the MECO, so Hoeben's group could look exclusively at the relationship between CO2 and temperature without having to account for the fact that changes in the amount of sea ice can increase temperature, too. "We've shown that if you include those long-term factors, then CO2 might very well be the leading factor for temperature increase, especially in a world without a major ice sheet," Houben says.

The result, he says, will help climatologists get a better grip on the concept of climate sensitivity—the degree to which a global temperature increase is entirely dependent on an accompanying rise in CO2. The authors conclude that the climate sensitivity during the MECO led to a 2- to 5-degree C increase per doubling of atmospheric CO2.

The study does leave one big question outstanding: Where did all the MECO CO2 come from? This remains an area of speculation, Houben says, although scientists are fairly sure the source was not organic. However the CO2 got there, the takeaway from this study is simple: "In the past," Zachos says, "whenever atmospheric carbon dioxide levels rise, the climate warms."