Abstract
We estimated soil moisture in and around Sarobetsu mire, in which desiccation worsens using data of PALSAR and AVNIR-2, which were put on Advanced Land Observing Satellite (ALOS). The backscattering coefficients that are obtained as PALSAR data include information of vegetation and surface roughness besides soil moisture. Therefore we incorporated the temporal differential NDVI as a variable about vegetation, and removed effect of surface roughness by using difference between the obtained satellite data and the data of the reference day to estimate soil moisture from backscattering coefficients. Then we made a model from the processed data and measured volumetric soil moisture. The estimated volumetric soil moisture shows a linear relationship with the measured one, with high correlation coefficient and the validation of the method was shown.
