干旱区生态环境知识资源中心

The purpose of this work was to establish a multitemporal methodology for monitoring and detecting changes of arid rangeland by remote sensing. Three steps compose this method. In the first step, visual interpretation of Spot images allows description of the distribution of vegetative biomass in the study area. A "color intensity" criterion indicates changes of each landscape unit. In the second step, radiometric analysis revealed the dynamic of each formation by the comparison of the coefficient of correlation between the Near Infrared (NIR) and Red (R) channels. The rate of vegetation development was correlated with climatic and socioeconomic data to determine its principal causes. In the final step, the changes occurring in each formation in relation to time were mapped according to a simple vegetation index (ratio NIR/R) and the coefficient of variation (standard deviation/mean) for each class. This latter coefficient indicated the change dynamics of each class. It corroborated the results obtained with the correlation coefficient and was directly linked to the relative heterogeneity of each class. The proposed temporal monitoring approach is based on simple criteria that are easy to apply and that demonstrate the differential change of each landscape unit. The degraded areas where change is regressive or those little frequented by animals where change is progressive can, thus, be rapidly identified. This is essential information which, in the hands of the decision-makers, will be able to ensure a rational and sustained management of these rangelands.