A critical review of UAV structure from motion photogrammetry for measuring soil surfaces and inferring soil loss

Abstract

Soil erosion remains difficult to measure reliably because detachment, transport, temporary storage, deposition, and export are expressed at different spatial and temporal scales, so even precise observations may still be weakly tied to process. UAV-based structure-from-motion photogrammetry (UAV-SfM) has made repeated, high-resolution topographic surveying practical, but it measures surface geometry and, in repeated surveys evaluated against uncertainty, detectable net surface change rather than erosion, soil loss, sediment export, or conservation effect directly. This structured critical review draws on peer-reviewed soil- and erosion-relevant photogrammetry studies published since 2005 that were identified through database searching and citation tracking, interpreted alongside a smaller contextual set of methodological papers, and synthesized through structured claim-level coding and evidential appraisal. Across the evidence base, UAV-SfM is strongest for direct reconstruction of soil microtopography, roughness, rills, and many exposed gully forms, but it becomes progressively more conditional when used to infer diffuse erosion, redistribution, connectivity, or conservation performance. Accordingly, the decisive issue is not image resolution alone, but whether change detection is supported by stable control, spatially explicit uncertainty, explicit treatment of non-erosional surface change, and independent process validation. UAV-SfM should therefore be understood first as a method for measuring soil-surface geometry and detectable net surface change; it becomes defensible evidence of erosion, deposition, redistribution, or conservation effect only when detectability, mechanism, storage context, and scale are explicitly constrained.