The Measurement of Temperature in Ice Stream B, Antarctica

Keith Echelmeyer surverying the margin.

The western third of the Antarctic Ice Sheet is grounded below sea level. As a result, it may be vulnerable to rapid mechanical disintegration in response to minor climate or ocean perturbations; the present breakup of Columbia Glacier in Alaska may be a good analogue. A key role in the dynamics of the ice sheet is played by several great streams of fast-moving ice, 35 to 50 km wide, which drain into the Ross Ice Shelf. These mysterious streams seem to die, and new ones to be born in different locations, on the time scale of a few centuries. Our program is studying the role of the margins in the flow of Ice Stream B, at a location where the stream is 1 km thick and 35 km wide. The first goal is to determine if the margins are significant in the "force balance" of the ice stream. In other words, could they be exerting as much force on the ice stream (against the down-slope component of gravity) as the bed itself, despite the 35 to 2 ratio of bed to margin contributions to the perimeter? The second goal is to determine whether the positions of the margins are changing.

Temperature data were obtained from sensors placed in holes across one of the highly crevassed margins of the ice stream (Figures 1 and 2). Figure 3 illustrates what happens to the ice as it is fed into the side of the ice stream at a rate of 1 or 2 meters per year. "Stage" shows the temperature before it is fed into the ice stream. By the time the ice has progressed to "Lost Love", about 500 m into the ice stream, two things have happened; it has warmed at the bottom and cooled from the top. These processes have continued at "Chaos", 1000 m in. The heating comes from straining of the ice, and the implied side drag indicates that the margins are indeed of about the same importance as the bed in the balance of forces. The cooling from the top is due to the ponding of cold air in the crevasses that permeate the margin; the depth of penetration of the cold into the ice below the crevasses implies that the ice at "Lost Love" and "Chaos" has been incorporated in the stream for about 50 and 100 years respectively. Given the 1 or 2 meters per year motion of ice into the ice stream, this can only be possible if the margin is unstable, moving outward at roughly 10 meters per year.

NSF Grant DPP91-17911: Scientific personnel; W. D. Harrison, K. A. Echelmeyer and C. Larsen.