Evaluation of radar backscatter models IEM, OH and Dubois using experimental observations

Abstract

Synthetic aperture radar (SAR) observations have been used in numerous studies to estimate soil moisture and surface roughness, because of their important roles in many domains. The objective of this paper is to validate, on bare soils, the most common surface backscattering models: the semi‐empirical models of Oh et al. and Dubois et al., and the theoretical integral equation model (IEM) of Fung et al. This evaluation is based on a large database consisting of C‐band SAR images (ERS‐2, RADARSAT‐1 and ASAR) and ground measurements (roughness and moisture). The discrepancies observed between the radar signals measured by the SAR sensors and those predicted by the models have been correlated with radar incidence angle, soil moisture (mv), and rms surface height. The results show that the models frequently tend to over‐estimate the radar response. The errors indicate that the models tested are dependent on the rms surface height, the soil moisture, and/or the radar incidence.

Acknowledgements

This work was supported by the BRGM (the French Geological Survey) and by the PNRH (the French National Hydrology Programme, RIDES project). The authors wish to thank the ESA (European Space Agency) for kindly providing the ASAR images under project no. 351 ENVISAT/ASAR. Some RADARSAT images were provided by the Canadian Space Agency under the RADARSAT User Development Programme (RUDP) and the Application Development and Research Opportunity Programme (ADRO). The database was compiled from work carried out by the Bureau de Recherches Géologiques et Minières (BRGM, French Geological Survey), the Centre d'étude des Environnements Terrestres et Planétaires (CETP), and the Centre d'applications et de recherches en télédétection (CARTEL) of the Université de Sherbrooke, Canada.