Facet-based electromagnetic scattering modeling of island–sea composite scenes using DEM and land-cover classification

Abstract

Accurate modeling of electromagnetic scattering characteristics in island-sea composite scenarios is of great significance for improving target detection and identification in marine environments. This paper proposes an island-sea composite scene modeling method by integrating Digital Elevation Models (DEM) with remote sensing imagery. Remote sensing imagery is utilized for land-cover classification to identify different media types, including sea, forest, rock, and city. The corresponding electromagnetic and roughness parameters are then assigned to each discrete facet according to its land-cover category. Based on this framework, an improved two-scale model combining the Advanced Integral Equation Model (AIEM) and Kirchhoff Approximation (KA) is employed to calculate the scattering response. The simulated backscattering coefficients are validated against ALOS-1 PALSAR measurements, and the results demonstrate that the prediction errors for all four land-cover classes are within 1.79 dB.