Assessing the Impact of Ground Acceleration during Earthquake on Landslide Triggering Using a Simple Physic-Based Model : Application to the 2015 Gorkha Earthquake

Landslides pose significant hazard in mountain regions, driving hillslope erosion and mobilizing large amounts of sediment to rivers. Earthquake-triggered landslides are commonly clustered near ridges and steep slopes, influenced in part by the topographic amplification of seismic waves. Understanding the spatial distribution of these landslides is critical for evaluating sediment supply to river and connectivity. While several complex physical-based models have been developed to explore the spatial distribution and river connectivity of earthquake-triggered landslides, challenges remain in accurately modeling the influence of earthquake-induced ground acceleration.

Here we test Slipos, a simple physic-based model accounting for landslide source and a runout, to study the impact of ground acceleration from the 2015 M_w 7.8 Gorkha earthquake on the spatial distribution of landslides and their connection to rivers. The landslide source component of Slipos is calibrated by varying rock strength parameters, while the runout component is refined by exploring transport-deposition parameter spaces.

Preliminary results show some discrepancies between modeled and observed landslides, in terms of location and source volume. We infer that the noise affecting post-event DEM lead to unrealistic landslides. Integrating peak ground acceleration leads to an increase in the area and volume of each individual landslide. However, the runout component accurately reproduces observed landslide locations when parameter spaces are appropriately adjusted. Initial findings on landslide connectivity indicate that up to 70% of modeled landslides deposit material in proximity of a river channel, consistent with observations. Our preliminary results highlight the need to use high-quality and high-resolution DEM when modeling earthquake-triggered landslides. In addition, the Slipos model, particularly its runout component, has the potential to accurately reproduce landslides connectivity.