Scatterometer Wind-Speed Exponents: A Review and SAXON 35-GHz Measurements

Wind exponents Gamma are very important in the study of radar scattering cross-section under ocean-surface conditions. Most experiments apply the ordinary least-square regression OLR. However, OLR accounts for error in only one variable in the regression, and therefore the results of OLR may be not statistically correct. In contrast, the orthogonal regression OR provides a good way to estimate two correlation variables x,y when both of them have errors. The best results are obtained if x and y have similar scales. Our study shows that OR has agreement with OLR when x is perfectly known without errors. Otherwise, OR and OLR have different results. For example, SAXON-FPN experiment 35 GHz data demonstrates that orthogonal regression OR and ordinary linear regression OLR methods have similar wind exponent, Gamma if the wave and the wind directions are same. If however, the directions are not the same, the estimates disagree. In this case, the estimate of Gamma from OR and OLR methods. Thus, it is seen that the wind exponent Gamm, has a highest value in both OR and OLR methods. Thus, it is seen that the wind exponent, Gamma has a highest value when the wind direction and the wave direction are normal to the radar.