Abstract
Using the radar backscattering coefficient (σ0) measurements over the ocean surface by a spaceborne scatterometer, one can estimate the near-surface wind by using a geophysical model that relates σ to winds and a wind estimation algorithm. The so-called SOS algorithm, which is basically an algorithm with weighted least squares in the log domain (WLSL), was used to process the Seasat SASS data. In this paper, we compare the performances of seven wind estimation algorithms, including the WLSL, maximum-likelihood (ML), least squares (LS), weighted least squares (WLS), adjustable weighted least squares (AWLS), Ll norm, and least wind speed squares (LWSS) algorithms, for wind retrieval. For each algorithm, we present performance simulation results for the NASA scatterometer (NSCAT) [4] system planned to be launched in the 1990's. A relative performance merit based on the root mean square value of wind vector error is devised for this comparison study. According to this merit, performances for all algorithms are quite comparable. However, the results do indicate that the ML algorithm performs best for the 50-km wind resolution cell case and the Ll norm algorithm performs best for the 25-km wind resolution cell case. We have also considered preaveraging the σ0 measurements obtained from each antenna beam for the 50-km resolution wind cell case. The performances of all algorithms are even more similar in these cases with preaveraging, although the Ll norm algorithm performs best. Finally, the issue of using a two-stage wind estimation method in order to reduce the computational load and its impact on algorithm performance are discussed. © 1988 IEEE