Exploring the spatial heterogeneity of deep soil drought by using electrical resistivity tomography (ERT) at multiple scales

The formation of dried soil layer (DSL) significantly threatens the sustainability of ecological restoration. Yet capturing the high-resolution spatial heterogeneity of DSL remains a challenge. This study utilized Electrical Resistivity Tomography (ERT) to conduct a wide-scale assessment of DSL dynamics across 82 transects on the Chinese Loess Plateau, covering precipitation gradients (300–600 mm), diverse vegetation types, stand ages, and topographic units within the two catchments. The reliability of ERT-derived quantitative index (QI) was validated against neutron probe moisture measurements ( n = 58 pairs). ERT demonstrated high fidelity in capturing DSL dynamics ( r = 0.73, p < 0.001, RMSE = 0.25). At the scale of transects, the mean QI was 0.50 ± 0.26, with soil water content (SWC) within the DSL reaching a critically low 12.1 ± 2.2%. DSL severity increased northward with decreasing precipitation, while QI rose from 0.34 to 0.62 with decreasing precipitation along the gradient of 600–300 mm. The vegetation type significantly modulated DSL intensity, with shrubland and grassland experiencing relatively severe drought (QI >0.7). Interestingly, mixed shrub-grass planting showed potential in mitigating desiccation (QI = 0.47). Stand age in orchards induced an increasing then decreasing DSL trend, peaking at ∼35 years (QI = 0.67). Spatially, tablelands (QI = 0.61) exhibited the most severe DSL, being characterized by “deep-blocky” patterns in forests compared to the “carpet-like” continuous patterns in grasslands. Gullies (QI = 0.092) remained the most resilient zones. Regarding of these findings, we propose a three-tier classification framework of permanent land-use type DSL (tablelands), permanent regional type DSL (slopes), and temporary land-use type DSL (gullies). This study validates ERT as a powerful minor soil disturbance tool for visualizing deep-soil hydrological processes, providing a scientific basis for topography-based and species-specific ecological management on the Loess Plateau and similar regions around the world.