Summer temperatures rarely rise above 0 degrees Celsius in Antarctica. There are no trees or flowers, no cars or cable TV–just perpetual daylight, a hunk of ice the size of the continental U.S., and glaciers as big as cities, moving hundreds of meters a year. Deep within those glaciers, under millions of pounds of pressure, the history of the atmosphere lies buried.

“I don’t actually like the cold,” admits Kurt Cuffey. “I get no end of flack from my colleagues for that.” Nevertheless, Cuffey, 34, spends two months a year in Antarctica, using high-resolution GPS receivers to measure ice-flow rates and zipping around in a snowmobile to collect humidity and wind records. The data is fodder for numerical models he uses to interpret past climate changes and predict future ones. Cuffey is helping reframe the debate about global warming and the speed at which it can happen. His research has revealed a skittish, sensitive Earth where ice sheets can melt more–raising sea levels faster–than anyone had previously imagined.

Cuffey’s first big discovery came when he was working on ice cores as a doctoral student at the University of Washington. Ice cores–long cylinders extracted from an ice sheet by drilling down a mile or farther–provide a physical record of annual snowfalls going back thousands of years. By analyzing the chemical composition of the ice, researchers determine when past temperature swings occurred. Cuffey wanted to know how big those swings were. After a thermometer was lowered into a borehole left by the removal of an ice core, he analyzed the temperatures of the ice layers within, adjusting for variables such as the diffusion of surface temperatures through the ice over time. He calculated that since the last ice age, about 10,000 years ago, the Greenland ice sheet has warmed a full 15C–and 10 of that warming occurred in just a decade. “Our impression of how volatile the climate is has changed quite dramatically,” says Bernard Hallet, Cuffey’s graduate adviser. “Kurt’s work had a profound impact on the whole community.” No one is prophesying tsunamis hitting the Statue of Liberty, as depicted in Hollywood’s recent thriller The Day after Tomorrow, but coastal cities would be inundated, and droughts would descend on key agricultural areas.

While politicians debate, Cuffey is content to spend most of the year in his sunny California office, crunching numbers to figure out how and when the icy places of the world will melt.