Air drag coefficient and roughness length of a cover of sea ice

Detailed wind and temperature profiles up to a height of 4 meters have been obtained at 30-sec intervals for a period of 20 days in March 1970 from an ice field drifting in the Gulf of St. Lawrence in Canada. Wind speeds were measured with fast-response cup anemometers at heights of 37.5, 50, 100, 200, and 400 cm, and temperatures were measured with shielded copper constantan thermopiles at heights of 12.5, 25, 50, 100, 200, and 400 cm above the ice surface. Some 250 values of roughness length z0 have been calculated from wind profiles averaged over periods of 15 min when conditions of dynamic near-neutral stability were observed (i.e., the Richardson number at a height of 1.22 meters was less than 0.03). The roughness length was found to be a sensitive function of surface conditions. Changes were caused by rain and snow and by fluctuations of near-surface temperature through 0°C. During the first week, March 7–14, the median value of z0 was 0.084 cm, and the probability that 0.002 cm < z0 < 0.15 cm was 0.71. After moving the instrument masts to an area of visibly rougher ice, on March 19–27, z0 was found to have a median value of 0.34 cm with a probability of 0.80 that 0.05 cm < z0 < 0.40 cm. The corresponding drag coefficient C10, calculated from the wind profiles by extrapolation to the standard height of 10 meters had a median value of 1.7 × 10−3, with a probability of 0.92 that 1.0 × 10−3 < C10 < 2.5 × 10−3 and, for the second period, a value of 2.2 × 10−3 for the median, with a probability of 0.96 that 1.5 × 10−3 < C10 < 3.0 × 10−3.