Land-Based Scatterometer Measurements and Retrieval of Surface Parameters Using Neural Networks

In this study, a multi-band FM-CW land- based scatterometer is used to measure the backscattering coefficient of bare soil surface. And then combined with the neural network which can be simulated in any nonlinear problem in theory, and integrated equation model (IEM) which has a wide range of surface roughness, has realized all of the surface parameters inversion through the theoretical simulation and the experimental data, including dielectric constant, surface rms high and correlation length. The results have a better consistency with the actual measurement parameters, and the difference among different angles inversion results reflects the uneven characteristics of reality surface. Final shows that retrieval of multi-parameter using neural networks to be effective, at the same time that simultaneous measurement of backscattering coefficient and surface parameter is a effective means that study features of microwave scattering.