Full-wave Modeling of Microwave Radar Scattering from Maize and Inversion Methods for Biophysical Parameter Estimation

Passcode: eecskaleor

Abstract:

Soil moisture and biomass are two important quantities that can affect the climate, weather, and agriculture. There is interest in measuring them from space because of the global, frequent, and repeated observations that could be achieved. Synthetic aperture radar is an attractive imaging sensor for this because microwaves can readily penetrate foliage, and SAR can take high resolution images in all weather conditions. Translating radar backscatter to biomass or soil moisture is challenging, though, because it does not solely depend on these variables.

This study explores novel methods for translating radar backscatter to soil moisture and biomass for corn fields. This is done by making forward and inverse models, which predict measurements from relevant variables and vice versa. This paradigm is first applied to a novel device for making in-line measurements of a liquid’s permittivity. It is then applied to corn fields with a novel forward model that uses full-wave analyses to predict the backscatter. The forward model is then inverted by combining regularized inversion with an artificial neural network. The forward and inverse models are also evaluated with measured data. Key findings include the feasibility of using full-wave models and better inversion accuracy for biomass than soil moisture when polarimetric SAR systems are used at L-band (1.25 GHz).

Chair: Professor Kamal Sarabandi