Correlation of microwave sensor returns with soil moisture

Abstract

Microwave sensor soil data were collected by aircraft over agricultural croplands. Multiple incident angle scatterometer data (13.3, 4.75, 1.6 and 0.4 GHz), passive radiometer data (L and C-band), and soil moisture ground truth measurements were collected coincidentally. Each sensor and angle of incidence was linearly analyzed against the measured soil moisture. For bare agricultural soils, the optimal single sensor for soil moisture preduction is the L-band passive radiometer. The effects of vegetation and differing surface roughness prove significant. When both bare and vegetated surfaces were studied, the masking due to the vegetation renders the single sensor approach ineffective in soil moisture prediction. Multisensor techniques are necessary to remotely measure soil moisture when a priori knowledge of vegetation is not available.