The equilibrium drift of Ice Station Alpha

The problem of the equilibrium (unaccelerated) drift of an ice floe (no internal resistance) is solved by a treatment similar to Shuleikin's. Curves of drift angle, drift speed, and wind factor as functions of wind speed are computed for two different sets of parameters and compared with Shuleikin's curves for one of the sets. As a test of the theory, drift angles, drift speeds, and wind factors were plotted against wind speed for 107 periods between July 1, 1957, and August 31, 1958, in which the drift of station Alpha was essentially unaccelerated. The data exhibited considerable scatter, and it was necessary to compute mean drifts in various wind categories in order to bring out the empirical relationships. The observed drift angles were smaller than the theoretical in all wind categories, though the departures were not as great for our curves as for Shuleikin's curve. The graphs of drift speed and wind factors also exhibited systematic deviations from the theoretical curves, the ice drifting faster than the theoretical speed in light winds and slower in strong winds. In explaining the deviations between observation and theory six sources of error are considered: navigation errors, errors due to residual accelerations, errors due to use of averaged data, errors due to variations in parameters, errors caused by gradient currents, and errors attributable to internal stresses. Although part of the scatter is found to be the result of the navigation errors, it is concluded that gradient currents and internal stresses are probably the main causes of drift anomalies. Evidence from the present study and from other sources suggests that a permanent current of about 3 cm sec−1 exists in the region of the drift of Alpha. This would account for the above-normal drift in light winds. The below-normal drift in strong winds is believed to be the result of internal resistance, though tests failed to show this conclusively.