The integral equation model and surface roughness signatures in soil moisture and tillage type determination

A proposed application of RADARSAT, C-band, HH-polarized, satellite-based, synthetic aperture radar (SAR) identifies agricultural tillages and soil moisture through multi-incidence angle observations. A procedure to accomplish this has been developed using the integral equation (IE) model and a roughness signature for each anticipated tillage type. Through a comparison of the IE model, using each roughness signature over a range of moisture levels, to the measured backscatter one can select the case with the least error. This procedure yields little confusion between rough and smooth tillages using scatterometer measurements of test plots where extensive moisture and roughness ground truth were obtained. There ist however, significant confusion in classification of tillages with similar roughness characteristics. The bare-field moisture level was determined with a mean error of 1.8% and a standard deviation of 5.8%, when the initial moisture level was constrained to +/-10% of the actual level.