Seismic Landslide Exposure Analysis for Türkiye

Earthquake-induced landslides are a major secondary hazard in tectonically active regions like Türkiye, often disrupting infrastructure and emergency response. However, national-scale assessments remain limited. This study addresses that gap by developing a data-driven model to simulate landslide susceptibility and estimate infrastructure exposure under different seismic hazard scenarios. Using a Generalized Additive Model (GAM) trained on inventories from the 2011 Van and 2023 Kahramanmaraş earthquakes, the model incorporates six key variables: slope, local relief, PGA, proximity to faults, lithology, and land cover. Susceptibility was simulated across Türkiye under four PGA exceedance probabilities (2%, 10%, 50%, and 68% in 50 years), showing increasing landslide likelihood with stronger but rarer earthquakes. Exposure analysis revealed that infrastructure overlap shifts from lower to higher susceptibility zones as earthquake intensity increases. While the model does not simulate runout or dynamic conditions, it offers a practical, scenario-based framework for national risk assessment. Future work should integrate higher-resolution data, physical models, and open data practices to improve robustness and enable broader application.