A SIMPLE-MODEL FOR RETRIEVING BARE SOIL-MOISTURE FROM RADAR-SCATTERING COEFFICIENTS

A simple algorithm based on a rough surface scattering model teas developed to invert the bare soil moisture content from active microwave remote sensing data. In the algorithm development, a frequency mixing model was used to relate soil moisture to the dielectric constant. In particular, the Integral Equation Model (IEM) was used over a wide range of surface roughness and radar frequencies. To derive the algorithm, a sensitivity analysis was performed using a Monte Carlo simulation to study the effects of surface parameters, including height variance, correlation length, and dielectric constant. Because radar return is inherently dependent on both moisture content and surface roughness, the purpose of the sensitivity testing was to select the proper radar parameters so as to optimal by decouple these two factors, in an attempt to minimize the effects of one while the other was observed. As a result, the optimal radar parameter ranges can be chosen for the purpose of soil moisture content inversion. One thousand samples were then generated with the IEM model followed by multivariate linear regression analysis to obtain an empirical soil moisture model. Numerical comparisons were made to illustrate the inversion performance using experimental measurements. Results indicate that the present algorithm is simple and accurate, and can be a useful tool for the remote sensing of bare soil surfaces.