Reference-Based Terrain Coordination Assessment of Abandoned Sand Mining Sites in an Arid Region Using UAV-Derived DEMs

Abandoned sand mining sites in arid regions are characterized by slow vegetation recovery and extensive bare-surface exposure, making it difficult to evaluate restoration outcomes using vegetation-based or conventional engineering indicators alone. More importantly, reference-based terrain assessment requires reliable natural benchmarks; otherwise, terrain similarity may be misinterpreted when candidate reference areas have been modified by human activities. This study developed a reference-based terrain coordination evaluation framework using high-resolution UAV-derived DEMs for three historical abandoned sand mining sites in the Aksu region, northwestern China. Candidate reference units were first screened using a 30 m DEM based on area, mean slope, and slope variability, and were then verified using 0.1 m UAV orthomosaics and DEM-derived hillshade maps. Surface roughness, local topographic relief, and mean absolute curvature were extracted to quantify terrain similarity between restored areas and verified natural reference units, and DEM-resolution sensitivity was further assessed.