Potential of palsar data in retrieving spatial variability of soil moisture in tropical catchment

Abstract

The potential of PALSAR data for soil moisture retrieval over four land use type of tropical area in Malaysia was evaluated using backscattering regression model. The catchment is divided into four land use types which are rubber, shrub, oil palm and grass land area. Without having any knowledge on surface roughness, the regression model was inversed and applied to the data. The regression coefficient of the model is analyzed to overall catchment and each type of the land use. The final output showed the calibration between backscattering coefficient and measured soil moisture is R = 0.52. The validation result between derived and measured soil moisture is found to decrease in R = 0.47. However, the division of each land uses are found to give a different result with one to another due to the type of land used based on the observation during the fields. The higher coefficient is found in rubber area and followed by oil palm, grassland and shrub area with R = 0.89, 0.79, 0.38 and 0.13 respectively. The result showed the potential of PALSAR data is reliable to be used in tropical catchment.