ANNOUNCEMENTS
Sand dune encroachment poses a significant and escalating environmental threat, particularly in arid and semi-arid regions where shifting sands increasingly impact human infrastructure, agriculture, and natural ecosystems. This study focuses on the Riyadh region of Saudi Arabia, where rapid urban expansion and climatic variability have made the region particularly susceptible to dune migration. The research aims to comprehensively assess current sand encroachment patterns and estimate the vulnerability of critical areas using a multi-factor geospatial model called the Sand Dune Encroachment Vulnerability Index (SDEVI).
The project is structured around four key components: area estimation, volume calculation, encroachment mapping, and vulnerability analysis. A machine learning–based classification approach (Random Forest) was applied to high-resolution Sentinel-2 optical imagery to isolate and map active sand dune areas. These classified outputs were converted to polygons for calculating total sand-covered area using GIS tools. For volumetric analysis, digital elevation models (DEMs) were integrated with the classified sand extent to estimate dune volume based on height differentials.Encroachment was assessed using change detection techniques over multi-temporal classified rasters. This step identified shifts in dune positions and quantified their intrusion onto nearby infrastructure, particularly roads. Additionally, the analysis incorporated movement vectors inferred from ancillary wind data to validate sand movement patterns.
The potential risk zones were evaluated using the SDEVI framework, which integrates several environmental and climatic layers including land use/land cover, NDVI, slope, soil moisture, wind speed, and wind direction. Each layer was preprocessed, reclassified into vulnerability classes, and spatially aligned before summation into the final index. The result is a continuous vulnerability surface that highlights areas most at risk of future encroachment. The study reveals that areas characterized by low vegetation, gentle slopes, dry soil, and strong prevailing winds are particularly vulnerable. The findings underscore the importance of integrating multi-source satellite data with environmental parameters for proactive dune management. This approach not only facilitates precise risk mapping but also aids in prioritizing mitigation strategies for infrastructure planning and desertification control.
