Assessment of Sand Encroachment in Arid Regions: A Case Study of El Hajeb Municipality, Biskra Province, Southeastern Sahara, Algeria

Abstract

This study examines sand encroachment in El Hajeb, Algeria, from 2000 to 2023—a region highly vulnerable to desertification. A multi-source remote sensing framework was used, combining optical and radar satellite data (Landsat, MODIS, Sentinel-1 SAR) with spectral indices (NDVI, MSAVI, BSI, NDESI) to monitor dune dynamics, vegetation health, and sand distribution. Pre-processing steps included atmospheric correction, cloud masking, and normalization. Advanced geospatial analysis and supervised classification were conducted using machine-learning algorithms—Random Forest (RF) and Support Vector Machine (SVM)—, which improved classification accuracy and land, cover discrimination. Results show an 11.90% reduction in stabilized dunes and a 6.05% increase in sparse sand areas. These shifts are linked to vegetation exploitation, soil moisture status, and intensified aeolian activity driven by prolonged droughts and prevailing winds. Sentinel-1 SAR data contributed to understanding surface roughness and moisture variability, reinforcing the analysis. Spatial maps revealed high-risk zones along the periphery of vegetated areas. To combat land degradation, the study recommends windbreaks, sand fences, and afforestation efforts. Overall, the research highlights the value of integrating remote sensing and machine learning for environmental monitoring and supports the development of adaptive land management strategies to enhance resilience in arid landscapes.