Calibration of IEM Model for the Soil Moisture Mapping of Non-Inundated Paddy Fields Using ALOS/PALSAR Data

This paper reports calibration of model parameters to estimate soil moisture of non-inundated paddy fields using SAR (synthetic aperture radar) data. The IEM (Integral Equation Method) model, theoretical model to represent surface scattering in the microwave region, was examined. Surface roughness parameters (standard deviation of surface height, autocorrelation function of the surface height, and autocorrelation length) are important parameters for microwave backscattering. However, the surface roughness of the paddy fields was found to change, and the autocorrelation length is quite difficult to accurately measure on the surface. Therefore, the two parameters, standard deviation of surface height, and autocorrelation length, were calibrated with actual soil moisture. As a result, it was found that the objective function to minimize the sum of residuals between estimated soil moisture and actual soil moisture produced the most optimal estimation results. In this case, the root mean square of errors between the estimated volumetric soil moisture and the actual soil moisture used for the calibration was 9.4%. This can be acceptable to consider the fluctuation of the surface roughness parameters during the non-inundated season.