Classification of polarimetric SAR data over wet and arid regions of India

Polarimetric SAR data from ALOS PALSAR, SIR-C and ENVISAT ASAR over wet and arid regions were processed for classification and soil moisture estimation. HH and HV dual polarized ALOS PALSAR could classify wetlands of Mumbai coastal area with an accuracy of 96%, whereas fully polarized SIR-C data over Kolkata gave 92% accuracy. The accuracies are based on selected training areas and not based on test areas. ALOS PALSAR could clearly discriminate water, mangrove forest and ocean water. With Dual polarized data, discrimination between Ocean water and wetlands is not possible. Several features in arid data can also be classified using PALSAR data in addition to the estimation of soil moisture.