期刊名称：Science et changements planétaires / Sécheresse
印刷版ISSN：1147-7806
电子版ISSN：1777-5922
出版年度：2007
卷号：18
期号：4
页码：359-362
DOI：10.1684/sec.2007.0112
出版社：John Libbey Eurotext
摘要：Figures See all figures Authors Elmar Csaplovics , Catherine Mering , Yveline Poncet Université de Dresde, Helmholtzstrasse 10, D-01062 Dresde, Allemagne, UMR Prodig, Université Paris Diderot, 75013 Paris, Institut de recherche pour le développement (IRD), Centre d’Orléans, 5, rue du carbone, 45072 Orléans cedex 2 Key words: Niger, radar image, remote sensing, satellite imaging, soil surface condition, wetland DOI : 10.1684/sec.2007.0112 Page(s) : 359-62 Published in: 2007 Synthetic Aperture Radar (SAR) images produce pixel values which are closely related to soil physical parameters such as topography, roughness, moisture and vegetation. These characteristics are particularly interesting in the study area of the spatial and temporal variability of the ground surface in a wetland such as the inner delta of the Niger in Mali. The exploitation of successive SAR images thus allows good detection and quantification of the ground surface changes. Classic techniques for identifying changes are based on the additive synthesis in RGB trichromy and on simple arithmetic comparisons between the two images. However, the “salt and pepper” features of ERS images, due to the speckle, reduce the efficiency of such techniques. Hereafter, we show experimentation of a morphologic filter intended to reduce the noise of the original images before using the classic technique of detection and mapping of changes. The interpretation of these images at general scales produces significant results on the spatial and temporal organization of moisture in the delta area. These results are directly exploitable for the study of local climatic variability.
