| Numerous disciplines require knowledge of the wind field at the ocean surface, such as weather forecasting, global climate and ocean wave modeling, and ship routing. Recently, satellite-borne radar scatterometers have made it possible to estimate the wind or the friction velocity at the ocean surface. Most techniques use scatterometer measurements to estimate the wind speed at the ocean surface. However, the friction velocity, and not the wind speed, is more closely coupled to ocean surface roughness. Most wind retrieval techniques are empirical and do not use physical models of either the waves at the ocean surface or of the microwave scattering from the air-sea interface.; Here, we present a technique for estimating the magnitude and the direction of the friction velocity at the ocean surface from radar scatterometer measurements that is based primarily on the use of in situ measurements and physical models. This technique uses in situ measurements to characterize the ocean waves at low wave numbers, and the Phillips 1985 equilibrium range spectral model to characterize the ocean waves at mid and high wave numbers. The directional spreading of ocean waves is treated using models derived from both in situ and remote measurements of the sea surface. Estimates of the radar cross section are determined by applying the composite surface model for electromagnetic scattering from a rough surface to these two-dimensional ocean wave spectra.; Using a data assimilation technique, radar cross section predictions derived from these physical models are combined with scatterometer measurements obtained by the ERS-1 system, to derive friction velocity estimates. In this work we successfully implement such a scheme, and demonstrate that it is capable of obtaining accurate friction velocity retrievals under a variety of wind and wave conditions. We also use this technique to characterize the impact of long wavelength waves on the ocean surface radar cross section. This technique was also used as a tool for investigating ocean wave model parameters, and their potential impact on radar cross section prediction. |
