New methodology for soil surface moisture estimation and its application to ENVISAT-ASAR multi-incidence data inversion

This paper presents an original methodology to retrieve surface soil moisture based on the use of the ENVISAT-ASAR multi-incidence angle sensor. Previous studies using ERS and RADARSAT SAR have shown the potential of radar signals to monitor surface soil moisture with one incidence-angle data and a simple linear relationship. This work aims at developing this linear approach to estimate mean soil surface moisture of a small watershed using ASAR multi-incidence angle data. A method of normalisation of all radar data acquired at different incidence angles is first described. Secondly, the effects of roughness and soil texture on the linear relationship between moisture and radar signals are discussed. Finally, the proposed methodology to reduce these effects is developed. The validation of our approach is based on eight experimental campaigns with different ASAR incidence angles, and different surface moisture conditions. The processed radar signal shows a linear relationship with the mean measured soil surface moisture, with very high correlation (R-2 = 0.97) and a slope of 0.28. These results illustrate the high potential of the developed approach and ENVISATASAR to monitor surface moisture. (c) 2005 Elsevier Inc. All rights reserved.