A Sense of Sea Ice

In the bestselling Danish novel Smilla's Sense of Snow, the sea ice is primeval, beautiful but deadly, intertwined in human lives in ways mysterious yet real.

Looking over the new ice in Copenhagen harbor, the half- Greenlandic heroine says, I'm happy because I know that now the frost has gained momentum; now the ice will stay, now the crystals have formed bridges and enclosed the salt water in pockets that have a structure like the veins of a tree through which the liquid slowly seeps; not many who look over toward Holmen think about this, but it's one reason for believing that ice and life are related in many ways.

map of polar ice cap

The ice-life relationship is more than a novelist's symbol, according to Penn State geographer Robert Crane. "When you run the big global climate models in a climate-change mode — simulating, say, a doubling of CO2 — you find that a lot of the change occurs in the high latitudes," Crane says. "The Arctic may be more important to climate than people have thought."

Sea ice affects the earth system in two ways: as a mirror, reflecting the sun's heat back into space; and as a cap, holding in the warmth of the seas. "If you cool the climate, you increase the ice," says Crane, "and the increased ice will cool the climate even more. Or, in the two-times-CO2 scenario, the reverse happens." The positive feedback in this sea-ice loop is key to the currently accepted global warming prediction, Crane notes, of a two- to four-degree average increase in temperature if the CO2 concentration in the atmosphere doubles.

But to capture the sea ice in a computer's set of algorithms is not easy. "The Arctic ice doubles from 7.5 million square kilometers to 15 million every year," says Crane. "It would be more than that, but most of the Arctic is landlocked." The unconstrained Antarctic ice balloons yearly from 2.5 million square kilometers to 20 million.

And, while most of the Antarctic ice floats north into warmer waters and melts each spring, Arctic ice lives from 5 to 20 years, circling round and round the gyre that is the Arctic Ocean, until it escapes with the current down the Greenland coast. As the oldest floes crash into each other, bending or clambering on top of one other, the ice can build to a thickness of 90 meters (a 20-meter-high freeboard and 70-meter-deep keel). Their swirling passage always leaves a small percentage of the Arctic seas un-iced.

"The way the climate models deal with the polar regions is not very accurate," says Crane. "They certainly don't simulate the ice cover. They do a good job of evaluating the energy balance, but not the distribution of the ice."

For sea ice — like wind and clouds — is always dynamic, always on the move; its location and extent can't be predicted simply by the time of year and the latitude. Continues Crane, "Genesis, the model we're running experiments with here at Penn State, is one of the first models that has dynamic ice. Unfortunately, the Genesis model doesn't predict wind circulation in the Arctic very well, so the ice is moving — but moving in the wrong direction."

Since 1978, satellites flying microwave sensors have been tracking the sea ice. "Ice emits more microwave radiation than ocean does," says Crane (strangely, to the microwave sensor ice looks warmer than water), "so we can see the contrast between them." Better, ice that has lasted all summer can be distinguished from new ice — for, as the fictional Smilla wisecracks, "In some ways ice is so transparent. It carries its history on its surface." Explains Crane, "As ice forms, there are pockets of brine. As the ice warms over the summer, the pockets drain, leaving air cavities. Microwaves are scattered by these pockets of air, so there is less microwave radiation emitted by the multi-year ice."

By plugging 15 years of microwave data into the Genesis climate model, Crane and his colleagues at Penn State's Earth System Science Center hope to fine tune the model's sense of sea ice, helping it, for instance, predict the onset of spring, that fleeting moment when the ice first begins to recede and the cycle of seasons begins to turn.

By studying the microwave data, Crane has found that the thaw "is not a simple seasonal melt," not a response to the sun moving up to its summer latitudes. Instead, it is controlled "largely by low pressure systems and cloud cover," or, in other words, precisely what the global climate models hope to predict: the weather.

Ice does, it seems, have a strong tie to life.

Robert G. Crane, Ph.D., is associate professor of geography and a member of the Earth System Science Center in the College of Earth and Mineral Sciences, 103 Deike, University Park, PA 16802; 814-865-7482. This research was funded by NASA.

Last Updated June 01, 1995