Right now Greenland is no threat to beachfront property. Steven Nerem of the University of Colorado at Boulder, who monitors sea level by satellite, says the oceans have been rising an eighth of an inch (0.3 centimeter) a year. At that rate the sea would go up a foot (0.3 meters) by 2100, roughly what a United Nations panel on climate change predicted earlier this year. “But that’s nothing compared to what we expect if Greenland really starts to go,” Nerem says.

The latest signs from Greenland have persuaded many ice researchers that sea level could rise three feet (one meter) by 2100. Rignot, who has measured the rush of glaciers to the sea, says even that figure may turn out to be an underestimate. Greenland, he notes, could ultimately add ten feet (three meters) to global sea level, “and if this happens in the next hundred years instead of the next several hundred years, that’s a very big deal.”

The fog has lifted, though a drizzle lingers. The helicopter has unloaded Steffen’s team onto the pitted, eroding ice sheet, about five miles (eight kilometers) in from the edge. This is the ice sheet’s ablation zone, its warm, low margin, where ice from the interior comes to die. It’s August, the height of the melting season. Washbasin-size pools of meltwater dapple the surface, and blue streams thread across the icescape.

All that water—already a loss to the ice sheet—could be causing the ice to dwindle even faster. The first hints of this feedback came ten years ago at Steffen’s springtime ice camp. NASA scientist Jay Zwally set out some GPS beacons to measure how fast the ice was creeping outward from the heart of Greenland to the ablation zone. Within a couple of years an odd correlation had turned up: the more the surface melted, the faster the great pile of ice moved.

At first the effect was subtle: For a few days in high summer, the ice sped up by 10 percent over its usual pace of about a foot a day. But year by year, as the climate warmed and the melting season lengthened, the lurching increased. Zwally and Steffen proposed that the cause was meltwater pooling at the base of the ice and lubricating its contact with the bedrock. Like a car skidding on a wet road, they suggested, the ice was hydroplaning across its bed.

Somehow, water must be percolating all the way from the surface of the ice down to its base. Here and there in the ablation zone, summer melt collects in azure lakes hundreds of yards across. Sometimes, a lake vanishes from one day to the next down some invisible drain. Not far from where Steffen’s helicopter touched down, a cleft in the ice, called a moulin, swallows a foaming torrent of meltwater.