UC Berkeley

Rainfall-induced landslides are driven by intense storms in mountainous regions. However, drainage basin-scale modeling faces challenges due to uncertainties in input parameters and computational limitations. Particularly, the Digital Elevation Model (DEM) resolution used in the hydrological and slope stability models can impact the prediction of landslides. Hence, understanding the influence of the domain’s resolution on the results is critical. This study uses a pseudo-three-dimensional slope stability approach coupled with the Parallel Integrated Hydrological model (ParFlow) to predict failures by Hurricane Maria in a watershed in Utuado, Puerto Rico. A 1-m resolution pre-hurricane DEM is used as the base scenario for both ParFlow and the slope stability analyses. The DEM resolution was then varied, and the results were compared to the base scenario with a focus on the landslide area density, timing of failures, and associated failure mechanisms. Results demonstrated that a coarser DEM for the slope stability analysis significantly affected the predicted landslides, whereas it had minimal impact on the hydrologic model. Resampling the DEM for the hydrological model while maintaining a high-resolution DEM for the slope stability model reduced the required computational resources significantly without compromising the validity of the results in regional-scale applications.