American Scientist

If ever there was an inanimate cousin to the snail or tortoise, it is the glacier. Like its fleshy counterparts, the glacier's sluggishness has etched it a special place in both legend and lexicon. Congress, baseball games, awards ceremonies—all creep along at a glacial pace.

The Columbia Glacier crushes that stereotype. The ice behemoth—60 kilometers long, 5 kilometers wide and up to 1,000 meters thick—has been calving icebergs into Prince William Sound, site of the 1989 Exxon oil spill, at speeds that would make a filibustering senator's head spin. Pushing 35 meters a day, it is the world's fastest glacier.

The Columbia Glacier is among 51 Alaska tidewater glaciers, so-called because they end in the sea. It first became an object of curiosity and concern in the 1970s, when prescient scientists with the U.S. Geological Survey noted that an increase in its calving rate could pose a risk to the sound's shipping lanes. Every summer since then, the Columbia site has been the Indy 500 of glaciology, attracting time-trial-conducting scientists.

"Tidewater glaciers are fairly fast creatures under any circumstances," says the latest of those timers, Tad Pfeffer, associate director of the Institute of Arctic and Alpine Research at the University of Colorado at Boulder, where he is also an associate professor of civil engineering. "Parts of the Columbia Glacier were moving as fast as a kilometer a year even before its retreat started. Now it's moving at more than 10 kilometers a year." The fastest-melting mountain glaciers move 100 meters any given year. Antarctica glaciers barely budge, 10 to 20 centimeters a year.

Tidewater glaciers don't respond to climate change as directly as other glaciers—the melting that has led to the Columbia Glacier's accelerated disintegration began 50 years ago—but once they start chugging, watch out. As the glacier slides ever more quickly into deeper water, iceberg production picks up. The race is on when a glacier reaches what Pfeffer calls "critical thinness." Melting reduces the ice mass and increases the flow of water beneath it, buoying the glacier and reducing friction; the glacier channel behaves like a giant flume carrying a too-big ice log that becomes thinner as it moves. If the glacier's terminus, its leading edge, speeds up, "the ice stretches toward the flow, which causes the ice to further thin vertically. That means more loss of weight and basal traction, more speed-up, more thinning and so on. It's a feedback loop." Once this cruise-ship-wowing iceberg-birthing show winds up, "the terminus will have retreated back up its channel, creating a 25-kilometer-long fjord."

Pfeffer and colleagues will continue timing the Columbia Glacier this summer, enlisting a technique that combines the best attributes of two standard methods: surveying and aerial photography. "Both involve locating a survey target or unique feature such as a crevasse edge on the glacier surface at two different times," Pfeffer says. "The difference in location gives the average velocity over the interval." Surveying is more accurate than aerial photography because the instrument is close to its target, beside it as opposed to between 8,000 and 29,000 feet above. But from the high vantage point, it is possible to track many more points on the ice.

Pfeffer will set up a specially designed camera beside the glacier as if surveying. "But instead of shooting angles and distances to unique points on the glacier, I make high-resolution photographs from two locations." This will imbue the results with the close-up accuracy of surveying but also mimic the images derived from aerial photography, over intervals of several hours.

"I can't wait for too long between photographs," he says, "because the surface is changing so fast that I can't easily recognize the same object in both photos. I have to be able to re-identify points!"—William J. Cannon