Soil erosion is a major threat in drylands, where sparse vegetation and intense rainfall accelerate land degradation. Traditional erosion models, such as RUSLE, capture surface processes and overlook the role of subsurface deformation, leaving a critical gap in understanding erosion dynamics. In this study, we present a novel framework that couples RUSLE with Persistent Scatterer Interferometric SAR (PSInSAR) and Transport-Limited Sediment Delivery (TLSD) modelling, thereby integrating surface and subsurface controls on erosion. We applied this framework in two contrasting Israeli basins: the arid Habsor and the semi-arid Yarkon–Ayalon. RUSLE factors were derived from IMS rainfall (R), Sentinel-2 NDVI (C), and terrain-based LS factors, while Sentinel-1 time series provided PSInSAR velocities to refine LS estimates and highlight zones of active subsidence or uplift. TLSD routing further transformed gross soil loss into spatially explicit maps of net erosion and deposition. Results revealed extreme variability, with gross erosion exceeding 420 t ha⁻¹ yr⁻¹ in Habsor and up to 41.6 t ha⁻¹ yr⁻¹ in Yarkon–Ayalon. Hotspots were concentrated on steep, sparsely vegetated slopes, while depositional sinks aligned with concave terrain and areas of positive Persistent Scatterer Interferometry (PSI) velocity. ANOVA confirmed significant differences in PSI velocities across erosion classes (p < 0.001), demonstrating the strong coupling between ground deformation and erosion dynamics. This framework is the first to integrate satellite-based subsurface deformation with surface erosion modelling in drylands, identifying critical erosion–subsidence zones and providing a realistic decision-support tool for soil conservation and dryland ecosystem resilience.
