Emulator-based global sensitivity analysis for flow-like landslide run-out models

Landslide run-out modeling involves various uncertainties originating from model input data. It is therefore desirable to assess the model's sensitivity. A global sensitivity analysis that capable of exploring entire space and accounts for all interactions, often remains limited due computational challenges resulting a large number necessary runs. We address this research gap by integrating Gaussian process emulation into landslide apply it open-source simulation tool r.avaflow. The feasibility efficiency our approach illustrated based on 2017 Bondo event. aggregated outputs, such as apparent friction angle, impact area, well spatially resolved maximum flow height velocity, dry-Coulomb coefficient, turbulent coefficient release volume are studied. results first-order effects consistent with previous common one-at-a-time analyses. In addition that, allows rigorously investigate interactions. Strong interactions detected margins path where expectation variation velocity small. generally become weak increasing velocity. Besides, there stronger between two coefficients than each coefficient. future, promising extend other computationally expensive tasks like uncertainty quantification, calibration, smart early warning.