Remote Sensing of Soil Moisture based on Dynamic Vegetation Scattering Properties for AMSR sensors

Abstract:

Accurate mapping of soil moisture and its spatial-temporal variations are of great significance to scientific studies on global water, energy and carbon cycles as well as operational applications including flood and drought monitoring, water resources management and crop yield forecasts. An approach for deriving volumetric soil moisture using satellite passive microwave radiometry from the Advanced Microwave Scanning Radiometers AMSR-E and AMSR2 was developed in this study. The algorithm adopts a weighted averaging strategy for soil moisture estimation based on a dynamic selection of empirically determined vegetation single-scatter albedo values. The resulting soil moisture retrievals demonstrate more realistic global patterns and seasonal dynamics relative to the baseline University of Montana (UMT) soil moisture product. Quantitative analysis of the new approach against in situ soil moisture measurements over four global study regions also indicates significant improvement over the baseline algorithm, with coefficients of determination (R²) between the retrievals and in-situ measurements increasing by approximately 16.9% and 41.5% respectively; and bias-corrected RMSEs decreasing by about 25.0% and 38.2% for respective ascending and descending orbital data records. Initial comparisons between soil moisture retrievals from AMSR2 and SMAP indicate coherent global and seasonal patterns.