The objective of this study is to detect land cover changes in fact of flood risks, with using a new method, fast and efficient. This technique needs only a diachronic satellite images in input data. The variability of reflectance spectral indexes is a characteristic of land cover changes. For applying this theory, we were including four spectral indexes those look on the principal indicator for land cover changes which are (Brightness index, Color index, Salinity index and clay index). To applying this method, we choose the Souss Massa basin as a study area. In 2014, this basin especially rounded of Guelmim city was affected by floods which caused a lot of damages in different ecosystems. Another step is used in the laboratory to validate our procedure; we talk about the physic-chemical analysis of land cover taken in two different dates, before and after floods.
…………………………………………………………………………………………………….... Introduction:-Availability increased free download data of remote sensing has allowed to make environmental studies much more temporal multi easy. A huge advantage has summer for the study of the degradation of soils under the effect dare floods and erosion of the soil. This decline in the quality of the soil leads to a reduction of productivity land that is more uniform, the same landscape lies in a deteriorated condition. The space images provide an effective way for the management costs of this degradation direct on the ground then the methods are expensive and tedious (Mumby et al, 1999). Thus, this information extracted on the land cover, using remote sensing and geographic information systems tools, provide data on a large scale and to better identify the areas exposed to the risk of change with their geographical distribution (Yuan et al.., 2005).Previous studies on the detection of land cover changes, using remote sensing and geographic information systems tools showed that there are several approaches to detect them.