Azimuthal harmonic coefficients of the microwave backscattering from a non-Gaussian ocean surface with the first-order SSA model

In this paper, the first-order small slope approximation is applied to a rough sea surface with non-Gaussian statistics, for which the third- and the fourth-order statistics are taken into account in the calculation of the radar cross section. From the Cox and Munk slope distribution, the higher order statistic moments are derived, and behaviors of the corresponding correlation functions are assumed. We show that the fourth order (related to the peakedness or kurtosis) is isotropic, whereas the third order (related to the skewness) has a behavior as cos(/spl psi/), where /spl psi/ is the wave direction along the wind direction. Thus, using the Elfouhaily et al. sea height spectrum, related to the second-order statistics, we show that the normalized radar backscattering cross section (NRBCS) can be expanded as an even Fourier series in cos(n/spl phi/) (where n is a positive integer), for which the harmonic coefficients require only a single integration over the radial distance. This result is consistent with experimental data done for microwave frequencies. In addition, we show for microwave frequencies (like C- and Ku-bands) that the Fourier series can be truncated up to the second order, since the higher order harmonic coefficients vanish. The NRBCS is also compared with empirical backscattering models CMOD2-I3 and SASS-II, valid in C- and Ku-bands, according to the scattering angle and the wind direction. The first-order harmonic coefficient predicts the surface asymmetry along the upwind and downwind directions, whereas the second-order harmonic coefficient describes the surface asymmetry along the upwind and crosswind directions.