Abstract
Low‐level cloud motion vectors determined by the National Oceanic and Atmospheric Administration (NOAA) and the University of Wisconsin in a 5‐degree square centred at 0°, 152°30’ W were intercompared with moored buoy wind measurements made at 0°, 152°W during April 1979‐February 1980. At the site of the intercomparison test the prevailing wind direction was easterly; monthly mean values of the meridional wind speed were less than 20% of the zonal component. The surface winds measured during the observational period were similar to the climatological‐mean wind conditions. Although the satellite wind speeds were larger than buoy wind speeds, as a priori expected, because of the vertical separation between the measurements, the comparison indicated that in the case of the zonal component there was a maximum usable frequency (muf) below which cloud drifts and surface wind vectors were 95% significantly correlated and the correlation was greater than the 50% noise level. The muf were 0.17 and 0.3 cpd for the NOAA and Wisconsin cloud motion vectors, respectively. At frequencies below the muf, an algorithm describing the frequency‐dependent differences of the rms zonal wind speed amplitudes was developed. Coherences involving the meridional wind speeds were too low for the estimation of a muf.