Uncertainty Analysis of Water Cloud Model Calibration for Soil Moisture Retrieval from SAR Data

The Water Cloud Model (WCM) is a semi-empirical model widely used to simulate microwave backscattering from vegetated soil surfaces. Combined with a soil scattering model, the WCM is also employed to retrieve soil moisture content. However, the relatively simple conceptual structure of the WCM with a few model parameters may result in ill-posed calibration and poorly transferable soil moisture retrievals. In order to investigate the optimal parameter space and calibration uncertainty, we calibrated the WCM combined with a linear soil scattering model using a Markov Chain Monte Carlo (MCMC) approach and data from the SMAPVEX12 field campaign. With the validation dataset, we then retrieved the soil moisture using the optimal parameter values for VV, HH and VH polarisations. The results showed superior accuracy of VH- and HH-pol soil moisture retrievals over VV-pol retrievals in terms of the root mean squared error (RMSE) and Pearson’s correlation coefficient (R).