GEOSTROPHIC DRAG COEFFICIENTS FOR THE CENTRAL ARCTIC DERIVED FROM SOVIET DRIFTING STATION DATA

Based on 4 years of central arctic atmospheric sounding data and 10 ye ars of surface wind data from Soviet drifting stations, combined with geostrophic winds from the arctic buoy program, the following relation between air-ice stress, tau, and the surface geostrophic wind speed, g, can be recommended tau = rhoC(g)2gamma2g2, rohC(g)2 = 0.85 x 10(-3) (nt m-2) (m2 s-2)-2, C(g) = 0.024, November-March, rhoC(g)2 = 1.10 x 10(-3) (nt m-2)(m2 s-2)-1, C(g) = 0.029, June-August. If the spacing b etween sea level pressure values from buoys is greater than 400 km, su ch as in the arctic buoy array, a speed enhancement factor of gamma = 1.3 should be applied to correct for insufficient sampling and smoothi ng in generating the geostrophic wind field. If the spacing is of orde r 100 km or less, then gamma = 1.0. An inflow angle alpha, the angle b etween the geostrophic wind and the surface wind, of 33-degrees- is re commended for winter and 23-degrees- for summer. Values for the transi tion months April, May, September and October can be interpolated betw een winter and summer values. The winter value of Cg is calculated 3 w ays: from the Soviet station surface wind-geostrophic wind speed ratio using suitable 10 m drag coefficients, from regression equations base d on surface-900 mb stability, and from the AIDJEX analyses. The summe r values are based on the surface wind-geostrophic wind ratio, model, and AIDJEX derived values. There is considerable day-to-day variabilit y in atmospheric stability and geostrophic coefficients, but no statis tically significant variation in the within-season monthly mean and me dian values; month-to-month variability is within 5 % for C(g) and 5-d egrees- for alpha for the winter and summer seasons. Neglect of stabil ity variations for daily cases can contribute an error of +/- 40 % in the relation between surface stress and geostrophic wind speed squared compared with using constant winter values. Use of atmospheric temper ature profiles from satellites may increase the accuracy of C(g) for d aily cases by providing an estimate of lower-atmospheric inversion str ength.