Two-dimensional full-wave scattering from discrete random media in layered rough surfaces

Modeling of electromagnetic scattering from discrete random media in layered rough surfaces finds various applications, including depth retrieval of layered snow-covered ice and remote sensing of vegetation canopy. In this paper, a coherent technique for solving scattering from discrete random media in layered rough surfaces is presented. The significance of the development of a full-wave solution to this problem stems from the fact that both co-polarized phase difference and polarized scattering coefficients can only be accurately determined using a coherent approach. Therefore, the objective of this paper is to formulate a full-wave solution for scattering from discrete random media in layered rough surfaces as well as to demonstrate the potential in the retrieval of subsurface parameters pertaining to the physical properties of rough surfaces and discrete random media using polarized scattering coefficients and co- polarized phase difference. The core of our technique lies in the use of plane wave decomposition. Plane wave solution for the scattered field due to a rough surface is obtained using extended boundary condition method (EBCM). The recursive T-matrix algorithm together with cylindrical-waves-to-plane-waves transformation matrices is employed to deal with scattering from discrete random media. Subsequently, plane wave solutions for the scattered fields due to rough surfaces and discrete random media are then cast into reflection and transmission matrices. These reflection and transmission matrices facilitate the application of scattering matrix technique which coherently accounts for electromagnetic interactions between layered rough surfaces and discrete random media. Various numerical results are examined and it is shown that the subsurface parameters may significantly impact backscattering coefficients and co-polarized phase difference even when the subsurface ground is covered by a rough layer of discrete random media.