CO₂ TRACES: Deposits of the mineral nahcolite (brown) mixed in with salt crystals (clear) suggest that the earth's last major warm period was high in CO₂.
DAVID TUTTLE, Binghamton University

Those trying to get a glimpse of what the future holds for the global climate may want to dig into the past. The earth's last major warm period probably witnessed relatively high concentrations of carbon dioxide (CO₂) in the atmosphere--similar to the upper limit of concentrations we might expect in the coming century--according to an analysis of ancient mineral deposits.

The clues come from the period of time between 49 and 56 million years ago, called the Eocene. Global temperatures jumped at the start of this era, judging from marine sediments as well as fossilized palm fronds and crocodiles found near the North Pole. The various indirect estimates of CO₂ concentration for the Eocene vary widely, however, from 100 parts per million (ppm), which is about one fourth of today's concentration, to more than 3,000 ppm. Researchers would like to narrow down that value so they can test their models of how CO₂ levels influence global temperature and thus better predict the planet's future warming.

The new estimate is based on the discovery of the rare sodium carbonate mineral nahcolite in two large deposits that date to the Eocene. Crucially, nahcolite is only expected to crystallize from salty water when the atmosphere contains 1,125 ppm or more of carbon dioxide. "That's the lowest it could have possibly been," says geologist Tim Lowenstein of Binghamton University, who reports the discovery with his colleague Robert Demicco in the September 29 Science. At lower CO₂ levels a mineral called trona usually forms instead. The Intergovernmental Panel on Climate Change has estimated that global CO₂ will reach roughly 500 to 1,000 ppm by 2100, depending on what humans do to control emissions.

"It's great that they found this," says Robert Berner of Yale University, who has calculated that CO₂ concentrations during the Eocene should have been about 1,000 ppm. "This is a period of intense warming, and a lot of people believe it was due to greenhouse effects," he says. "This adds more credence to this idea." All proxies for past carbon dioxide have different strengths and limitations, which makes incorporating a new one tricky, says Dana Royer of Wesleyan University. "Even including this paper the data tends to fall out between 500 and 1,000 ppm," he argues. In his view, that could still leave CO₂ levels somewhat too low during this era to fully explain its warming, leaving room for additional mechanisms on top of the greenhouse effect.