Effects of wind variability on scatterometry at low wind speeds

Donelan and Pierson [1987] predicted that the microwave cross section of the sea at moderate incidence angles should show a dramatic increase in level at a certain, low value of wind speed, the threshold wind speed. This occurs because the wind speed must be high enough to overcome viscous damping of the short waves that scatter microwaves from the sea surface. Recent measurements in a wind-wave tank have demonstrated the validity of this prediction. Nevertheless, the present author shows that probability distributions of cross sections measured by NSCAT during its 10 month lifetime are not consistent with those expected if such a low-wind-speed threshold were present in the cross sections measured by the spaceborne scatterometer. The author shows that this is to be expected if sufficient variability of the wind exists within the very large footprint of the scatterometer on the sea surface. If this is the case, then the behavior of cross sections at low wind speeds, when taken over the whole data set, are more consistent with the model function that was used to process the NSCAT data. This model function was extrapolated linearly to very low wind speeds on a scale where the log of the cross section is plotted against the log of the wind speed. The author shows, however, that areas of the globe exist where this model function, which has a single form, cannot explain the observations and that a model function with a dependence on the wind variability may be required