Despite the persistent rainfall deficit observed from more than forty years, the West African Sahel exhibits a significant increase in runoff, in rivers discharge and in water erosion. This fact is attributed to land use changes, particularly to soil overexploitation, characterized by the shortening of fallow periods. This leads to the degradation of soil structure, and to soil crusting. The soil crusting is the first factor of the increase in runoff. The purpose of this research is to determine the land use evolution in the western part of the Niger, to underline the factors and processes leading to the significant changes in surface water regime. The increase in discharges and runoff coefficients is accompanied with a reduction in the lag time of the slopes and basins when occurs a rainy event. This is directly linked to the decrease in soil water holding capacity, due to soil crusting. Hence, land use/land cover evolution is the main explaining factor of water cycle changes. Remote sensing is used in order to determine land uses changes. This implies to improve the knowledge about the “regreening” of the Sahel, observed and measured by several research teams: is it a myth or a reality? In what regions? During what periods? What is the role of classification methods of remote sensing in the highlighting of the regreening? We need to ensure that regions of interest used in this study are representative of the regional area. Methods of “hydro-geomorphometry” are needed to divide the study area into two differents areas, “plateau” and “non-plateau”, in order to avoid errors in the pixel classification of some different vegetation types which have neighbouring NDVI. A study by interpretation of aerial photographs is attempted over the Square degree of Niamey, with the aim of improving the separation between the respective role of the drought and the human activities in the evolution of land cover. After a comparison of our research with numerous previous studies realised on the same region, we perform an analyse of the impact of land use changes on infiltration and runoff, from the plot scale to the great basin scale, with a special focus on the observed evolution of rivers discharges and regimes and on the increase in flood hazard, detected during the last years. It appears that 2 to 5% of the total area exhibiting soil crusting are sufficient to increase and concentrate enough the discharges and to provoke a significant rise in the flooding, without avoiding a certain global “regreening” such as it is observed using the evolution of vegetation indices.
