Real-Time Dynamic Hazard Maps for Shallow Colluvial Landslides in Eastern Kentucky

Landslide hazards pose a persistent threat to communities and infrastructure in Eastern Kentucky, where steep slopes, shallow colluvial soils, and variable hydrological conditions drive frequent slope failures. This work advances landslide hazard mapping (LHM) through the development of dynamic, spatiotemporal models for shallow colluvial landslides. Two studies refine the use of a limit equilibrium framework to enhance predictive capability. The first study establishes a novel LHM workflow that integrates Hydrus-1D simulations of soil moisture dynamics, driven by precipitation and evapotranspiration data, into slope stability analysis. Factor of Safety (FS) parameters are applied to statistics-based landslide susceptibility maps (LSM), producing accurate, predictive hazard maps validated against documented landslides in Pike and Breathitt Counties. This approach demonstrates the feasibility of generating robust LHMs using publicly accessible datasets. The second study extends this workflow by improving slope stability assessments through refinement of the depth-to-bedrock parameter and incorporation of soil root cohesion into FS calculations. By automating data extraction and map generation in a Python environment, this study enables efficient, near-real-time forecast capabilities, including 72-hour landslide hazard predictions. Results underscore the critical role of vegetation and root cohesion in slope stability while highlighting the potential for proactive hazard mitigation. Together, these studies establish a comprehensive, adaptable methodology that leverages both physical process modeling and machine learning, providing a pathway toward accurate, operational landslide hazard forecasting in data-limited regions.