COSMO-SkyMed and radarsat image integration for soil moisture and vegetation biomass monitoring

Contribution of Polarimetry and Multi-Incidence to Soil Moisture Estimation Over Agricultural Fields Based on Time Series of L-Band SAR Data

“…Over the last four decades, considerable research efforts have been devoted to soil moisture estimation by means of synthetic aperture radar (SAR) and proved the potential of SAR data at L, C, and X bands for estimating SSM over bare and vegetated soils [2], [3]. The use of physical models (IEM, AIEM), semi-empirical models (Oh, Dubois, and Shi) [4], decomposition theorems (Freeman-Durden, Yamaguchi, and Cloude-Pottier) [5], [6], [7], change detection techniques [8], [9], [10], and statistics-based methods [11], [12], [13] applied to multiple SAR observations have improved the capability to obtain SSM information at high spatial resolution (less than 10 m).…”

“…Notably, the complexity of the volume scattering model [4], [5] and the number of modeled scattering components are important aspects in the description of the scattering properties of this type of scenes. Another strategy to take into account the presence of vegetation on the ground consists in adopting statistical approaches, for instance based on Bayes theorem or on machine learning methods [11], [12], [21]. Artificial Neural Networks (ANNs) and Supported Vector Regressions (SVR) have reported some capability to minimize the retrieval uncertainties induced by multiple combinations of surface parameters and improve the inversion of SSM from SAR data [13], [22], [23].…”