A Regional-Scale Framework for Landslide Prediction Combining Three-Dimensional Hydrological Modeling and the Local Field Factor of Safety

Landslides represent a major threat to human safety and infrastructure, particularly in mountainous regions. Accurately predicting landslide susceptibility in a physically based deterministic manner requires an integrated, multidisciplinary approach that combines geology, geomorphology, and hydrology. In this work, a hydromechanical modeling framework is developed to forecast the initiation of large-scale shallow landslides by computing the local factor of safety (LFS) as a measure of slope instability. The framework couples (1) a finite element method (FEM) solver for hydromechanically coupled landslide processes implemented within a Java-based, object-oriented modeling environment, with (2) an external hydrologic model, allowing for detailed three dimensional simulations of slope response to transient rainfall events across extensive hillslope domains. The proposed framework is first validated using a benchmark test on a homogeneous hillslope with constant inclination and is subsequently applied to a real-world large-scale case study in the Braies Alpine Catchment, Alto Adige, Northern Italy. In the benchmark scenario, the model successfully reproduces shallow landslide triggering under prolonged rainfall, while in the real-case application it reliably captures the initiation of multiple landslides during an intense summer storm. These results highlight the framework’s robustness and accuracy in predicting landslide initiation in complex terrain, demonstrating its potential as a cost-effective tool for landslide hazard and risk assessment.