Title :
Sahelian-Grassland Parameter Estimation from Backscattered Radar Response
Author :
Monsivais-Huertero, Alejandro ; Chenerie, Isabelle ; Sarabandi, Kamal
Author_Institution :
LERISM, Univ. Paul Sabatier, Toulouse
Abstract :
In recent years a special emphasis has been placed on the retrieval of physical parameters from polarimetric radar at microwave frequencies in many research programs. In this paper, we adapted a technique based on an empirical model and a genetic algorithm, and verify its applicability for a complex class of vegetation within a wide temporal interval. This complex class of vegetation is Sahelian grassland which is mainly composed of annual grass and shrubs. The proposed retrieval algorithm is conformed of 3 main steps: (1) Identification of sensitive parameters, (2) Development of the empirical model, and (3) Implementation of a genetic algorithm for the inverse process. For this class of vegetation the sensitive parameters are: the soil moisture content ms, the grass density D, and the grass moisture content mv. When applying the retrieval algorithm to simulated radar responses, a great agreement (an error of 6% when estimating the soil moisture content, 13% for the grass density, and 18% for the grass moisture content in the adult-plant stage) is observed between input parameters and estimated ones.
Keywords :
backscatter; genetic algorithms; geophysical techniques; microwaves; moisture; radar polarimetry; remote sensing; soil; vegetation; Africa; Sahelian-grassland parameter estimation; backscattered radar; empirical model development; genetic algorithm implementation; grass density; grass moisture content; microwave frequency; polarimetric radar; sensitive parameters identification; shrub; soil moisture content; vegetation; Backscatter; Blades; Genetic algorithms; Parameter estimation; Predictive models; Radar remote sensing; Radar scattering; Scattering parameters; Soil moisture; Vegetation; Inversion algorithm; Sahelian grassland; radar remote sensing;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. IEEE International
Conference_Location :
Boston, MA
Print_ISBN :
978-1-4244-2807-6
Electronic_ISBN :
978-1-4244-2808-3
DOI :
10.1109/IGARSS.2008.4779551