Regressive Retrieval of Soil Moisture in Corn Fields using a Full-Wave Microwave Scattering Model

Estimating soil moisture using radar is a well-studied, but difficult application, especially when vegetation is present. A common type of agricultural vegetation is corn. Retrieving soil moisture from corn fields would be beneficial because of corn’s wide cultivation. Recent work has used full-wave scattering models to predict the radar backscatter from corn fields. A retrieval algorithm based on full-wave scattering models could have important benefits for achieving accurate results. A macromodel based on an L-band, full-wave model is developed in this work that uses second-order, multivariate polynomials to approximate the radar backscatter coefficients in terms of vegetation and soil parameters. The macromodel is used to provide an inverse model for soil moisture retrieval. The macromodel backscatter results are accurate to within ± 2 dB, and the inverse model can retrieve soil moisture to within ±2.1%.