Utilisation d'une approche multi-incidence pour l'estimation de l'humidité du sol et de la rugosité de surface à partir d'images RSO

Abstract

Applications in the field of agriculture as well as studies carried out on soil erosion and runoff have shown the potential of SAR type radar sensors for measuring the two principal parameters characterizing soil status: roughness and surface wetness. Many studies have tried to develop models for analyzing the direct link between these two parameters and the backscattering coefficient. Present scientific work is focused on the development of empirical models combining the physical approach and ground data. The objective for using such models is to produce precise cartographic documents permitting the detailed analysis of the complex processes of soil runoff and erosion mechanisms. The semi-empirical model of Dubois et al (1995) permits the simultaneous extraction of the information concerning roughness and wetness from airborne multipolarisation data. Since RADARSAT works only in HH polarisation while still providing the capability of acquiring images under different incidence angles, the principal mandate of this study was to demonstrate the validity of this model for spatial data following a multi-incidence approach. The second objective was to create thematic maps of the surface parameter information derived from the satellite data and aimed at improving hydrological models such as HYDROTEL."--Résumé abrégé par UMI