D. Halpern, VALIDATION OF SPECIAL SENSOR MICROWAVE IMAGER MONTHLY-MEAN WIND-SPEEDFROM JULY 1987 TO DECEMBER 1989, IEEE transactions on geoscience and remote sensing, 31(3), 1993, pp. 692-699
Since July 1987 the Remote Sensing Systems, Santa Rosa, California, ro
utinely computes from first principles the wind speed from Special Sen
sor Microwave Imager (SSM/I) measurements of the intensity of microwav
e radiation emitted at the ocean surface. The accuracy of monthly-mean
SSM/I wind speeds is determined by comparisons with moored-buoy wind
measurements, which were recorded by four different institutions at 44
sites in the Atlantic and Pacific Oceans during July 1987 to December
1989. All results for 1988 were virtually identical with 1989. The ra
nge of monthly mean moored-buoy wind speeds was 2-10 ms-1. During 1987
, the equatorial matchups were not equivalent with 1988 and 1989, and
the cause remains unknown. The root-mean-square (rms) difference of 69
7 monthly-mean matchups of the composite 1988 and 1989 data set was 1.
2 ms-1. The rms differences were smaller in the equatorial zone and hi
gher in middle latitudes, in accord with the monthly standard deviatio
n. At middle latitudes the time series of rms differences displayed an
annual cycle. In the equatorial zone the agreement between SSM/I and
in situ data was better in regions with a lesser amount of clouds, and
vice versa. For SSM/I monthly standard deviations of 1-2, 2-3, and 3-
4 ms-1, the average absolute values of the monthly-mean difference bet
ween SSM/I and moored-buoy wind speeds were 0.6, 0.9, and 1.4 ms-1, re
spectively.