Modeling Tools for Backscattering from Rough Surfaces

In this study we show that backscattering models are available for a variety of surface conditions including surfaces with a narrow or a broad surface spectrum. In particular, we shall consider the application of three backscattering models in algebraic form to field measurements and numerical simulation results. All three models have a wider range of applicability than most existing models currently available in the literature: (1) a scattering model for surfaces whose high spatial frequency components are Gaussian correlated and may shift the relative amount of the contributions between high and low frequency components as specified by the user so that when there is a larger contribution from the higher spectral components the contribution from the lower spectral components will decline and vice versa, (2) a scattering model for nearly exponentially correlated surfaces which may contain a varying amount of spectral components smaller than or comparable to the correlation length of the surface as specified by the user and (3) a scattering model with a two-scale correlation function, an algebraic sum of two exponential-like functions, the relative weighting of which can be adjusted by the user. It is believed that most surface backscattering problems can be explained with one of the three models. A general conclusion from model applications is that the roughness components of a given surface is represented more explicitly by its spectrum than its correlation function. It is also found that the exponential correlation tends to overestimate backscattering at large angles of incidence.