Spatial and temporal trends in oasis ecosystems and their response to prolonged drought in an arid zone of Algeria

Algerian oasis ecosystems are among the few resilient ecosystems, that manage to maintain a delicate balance between resource availability and the needs of societies. They constitute a unique agroforestry system, providing essential services to the ecosystem and local communities, such as provisioning, regulation, and cultural services. Following climate change and repeated droughts, these ecosystems are undergoing significant environmental modifications, marked by changes in the nature and behavior of vegetation cover, a fundamental element of ecological stability. Therefore, in order to maintain the ecological balance of oasis ecosystems and combat their degradation, it is necessary to understand the relationship between climate change and its impacts on the transformation of these oasis ecosystems. For this study, the pre-Saharan oasis zone of ASLA (south of Naâma), in Algeria, was chosen because of the significant environmental changes it has undergone following a prolonged decrease in rainfall. Our study is based on exploring spatio-temporal variations and identifying changes in land cover. To do this, we extract classification categories representing soil conditions, which we subdivide into classical classification categories corresponding to plant groups. To this end, we classified land cover by combining several spectral indices, calculable from satellite data for each spectral band, to create multiband input data for a supervised classification approach based on a support vector method (SVM). This method was applied to Landsat 8 OLI images with 30-meter resolution, combined with images from the Algerian satellite Alsat-2B with 2.5-meter resolution for panchromatic images. Archive images from 2008, 2013, 2018, and 2023, at five-year intervals, were used to detect changes. We then studied the relationship between variations in land surface parameters and changes in land surface temperature (LST) and normalized difference vegetation index (NDVI) before and after drought. Using GIS, we integrated climatic parameters (precipitation and land surface temperature) combined with land cover and NDVI results. Then followed expert recommendations to determine the weights to be assigned to each parameter in the model. This method allowed us to clarify the situation regarding the degradation of the oasis ecosystem, to classify the study area according to its degree of vulnerability, and to determine the spatiotemporal changes that occurred over the 15-year period.