| Microwave remote sensing is independent of both weather conditions and the time of day, and hence widely used in the dynamic monitoring of forests, crops or other vegetation, getting a large amount of data about vegetation. Lots of useful information can be obtained from these data by the theoretical research of electromagnetic scattering characteristics of vegetation. This dissertation will study the electromagnetic scattering characteristics of vegetation through the electromagnetic scattering theory of random media. The main contributions of the dissertation include:1. A Monte-Carlo coherent scattering model for the electromagnetic scattering of vegetation canopy is developed. According to the characteristics of vegetation, a simplified discrete particle model (ball, ellipsoid, disk and cylinder, etc.) and Monte-Carlo technique is used to model different vegetation. Based on multi-path scattering mechanism, the scattering field of particle and volume-surface interaction can be calculated. Then the total scattering field of vegetation is obtained by coherent superposition. In this process, the effects of attenuation caused by the vegetation are also taken into account by introducing the wave propagation constant of inhomogeneous dielectric layer. The backscattering enhancement is analyzed on basis of this model.2. In order to study the electromagnetic scattering characteristics of vegetation, the statistical average of a large number of random samples is required in the above model. However, the efficiency of this process is unsatisfactory so that the application for large-scale vegetation scene is seriously affected. For this reason, the first-order iterative solution of the vector radiative transfer theory (VRT) for multi-layer vegetation is studied and then a multi-layer vegetation bistatic scattering analytical model is established. Starting from the scattering amplitude function of Rayleigh ellipsoidal scatterer and thin cylinder, the analytical form of one-order solution about the scattering contribution is derived by assuming uniform distribution of scatterers. This analytical form avoids the random sampling process and greatly improves the computational efficiency. In addition, the effectiveness of the analytical model is further verified through comparison with the measured data and other model. |
PDF Full Text Request