Deep Learning Enhances Rheological Parameter Determination for Landslide Runout Prediction

This dataset contains numerical simulation results and deep-learning–based predictions used to investigate transferable rheological parameters for landslide runout modeling. The data were generated using a physics-based shallow water equation (SWE) framework coupled with a deep neural network (DNN) to calibrate μ(I) friction law parameters. Included files provide optimized rheological parameters, model performance metrics, and sensitivity analyses for multiple landslide scenarios under varying topographic and material conditions. These data support the analysis of parameter generalizability and model transferability across different landslide events and catchments.