Soil moisture estimates from AMSR-E brightness temperatures by using a dual-frequency algorithm

This paper investigates the possibility of estimating the soil moisture content on a global scale from dual-frequency (C- and X-bands) microwave data of the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E). Because some anomalous behavior was occasionally found in AMSR-E C and X band data, a calibration check compared the AMSR-E data with measurements from the SSM/I sensor over two reference targets, namely a Russian evergreen forest and the sea surface, both of which have already been studied in the past. The algorithm for retrieving soil moisture uses both the brightness temperature at C-band, in H polarization, and the Polarization Index at X-band for correcting the effects of vegetation. This algorithm is based on a simplified Radiative Transfer (Tau-Omega) model, which has been inverted by using the Nelder-Mead iterative minimization method. The algorithm was validated with microwave data collected on two sites during the MASMEx02 (Microwave Alpine Soil Moisture Experiment) and the SMEX02 (Soil Moisture Experiment) experiments, respectively. The first site, in Italy, was characterized by natural vegetation covers, while the second, in Iowa (US), was covered primarily in agricultural crops. In general, the soil moisture estimated by the algorithm from AMSR-E data and the SMC measured on ground were in good agreement with each other in both sites, and five classes of soil moisture were easily identified. Keywords – AMSR-E brightness temperatures; soil moisture content retrieval; C- and X-band emission; Polarization Index; Tau-Omega model.