Remote sensing assessment of dam impact on arid basins in Southern Saudi Arabia: A machine learning and space-for-time approach

This study focuses on four basins in southern Saudi Arabia: Hali, Baish, Yiba, and Reem. These regions are characterized by arid conditions and are significantly impacted by dam construction. The research investigates the environmental impacts of dam construction using a space-for-time substitution approach, remote sensing, and machine learning techniques. A key focus is analyzing non-linear environmental impacts, particularly in data-limited, arid regions where traditional methodologies fall short. The study introduces a novel framework that combines space-for-time substitution and Dynamic Time Warping (DTW) to assess temporal and spatiotemporal changes in key environmental factors such as NDVI, soil salinity, groundwater, and runoff. The results reveal significant changes post-dam construction. In the Yiba-Hali basins, DTW values increased across several parameters: NDVI (0.08–0.25), soil salinity (0.09–0.25), and runoff (0.45–0.90), indicating reduced similarity between pre- and post-dam conditions. In the Reem-Baish basins, the Baish dam caused notable increases in DTW values for NDVI (0.16–0.31), soil salinity (0.15–0.30), groundwater (0.52–1.19), and runoff (0.53–1.33), with the most significant changes observed in groundwater and runoff. Additionally, regression models showed a decrease in predictive accuracy from 2010 to 2020, as evidenced by lower R² values for NDVI (0.82–0.37), soil salinity (0.77–0.38), groundwater (0.98–0.34), and soil moisture (0.96–0.24). • A novel framework integrates space-for-time substitution, remote sensing, and machine learning to assess dam-induced environmental changes. • DTW captures non-linear shifts, quantifies changes in water flow and vegetation from dam construction, and aids in basin response assessment. • Long-term monitoring tracks NDVI, soil salinity, groundwater, moisture, and runoff, highlighting dam-induced hydrological changes.