Landslide susceptibility mapping under cropland expansion: considering root cohesion variability and land use effects on saturated hydraulic conductivity

Cropland expansion in hilly regions reduces root cohesion and alters land use, increasing landslide susceptibility. Existing physically based models often assume uniform root cohesion and ignore land use impacts on saturated hydraulic conductivity, limiting their accuracy. This study advances landslide susceptibility modeling by incorporating both the spatial heterogeneity of root cohesion and the land use-driven variability of hydraulic conductivity within the infinite slope stability framework. Root cohesion is mapped using satellite-derived tree height and above-ground biomass products. Land use effects on saturated hydraulic conductivity are quantified by assigning land use-specific hydraulic conductivity values through an adjustment method. Incorporating these factors in the infinite slope stability model significantly enhances the accuracy of landslide susceptibility assessments, with AUC improving by 6%–17%. Results indicate deforestation caused by agricultural expansion declines in root cohesion and hydraulic conductivity markedly elevate rainfall-induced landslide risks. This approach enhances the infinite slope stability model, emphasizing the environmental trade-offs of cropland development. It provides a practical tool for identifying high-risk zones, setting cropland expansion limits, and optimizing disaster mitigation strategies in landslide-prone terrains undergoing significant land use changes, such as agricultural expansion.