Comparison of asymptotic backscattering models (SSA, WCA, and LCA) from one-dimensional Gaussian ocean-like surfaces

In this paper, recent asymptotic backscattering models are compared for one-dimensional multiscale dielectric sea surfaces with Gaussian statistics and by considering the Elfouhaily et al. height spectrum. We focus on the calculations of the normalized radar cross sections (NRCS) obtained from the weighted and local curvature approximations (WCA and LCA), recently published by Elfouhaily, and of the first- and second-order small slope approximation (SSA(1) plus SSA(2) denoted as SSA), developed by Voronovich. Voronovich et al. (2002 Waves Random Media 11 247-269) presented simulations of the SSA by making an assumption over the SSA(2) contribution (the model is referred to as SSAM). The NRCS computation is then similar to SSA(1), where the sea spectrum is substituted by a modified spectrum defined as the product of the sea spectrum by the second-order polarization term. The second-order statistical moment of WCA is calculated rigorously for any two-dimensional height correlation of the surface with Gaussian statistics. The NRCS of the WCA, WCAQ (obtained from a quadratic approximation of WCA), LCA, SSA, and SSAM backscattering models are compared for moderate wind speeds, for microwave frequencies and for backscattering angles ranging from 0 (nadir) to 70/spl deg/.