Preliminary calibration of a numerical runout model for debris flows in southwestern British Columbia

Debris-flow hazard and risk assessments require reliable estimates of inundation area, velocity and flow depth to evaluate spatial impact and impact intensity. Semi-empirical numerical runout models that simulate bulk flow behavior with simple rheological models are useful in forecasting these parameters, however, they require calibration by back-analyzing past events. This paper presents the back-analysis of six debris flows in southwestern British Columbia using a novel automated calibration approach that systematically optimizes the Dan3D runout model to fit field observations. The calibration method yielded good simulations of runout length, but under-predicted flow depths in some cases, and over-predicted velocities in all cases. The best-fit Voellmy rheology parameters for the studied cases ranged from 46 to 531 m/s2 for the turbulence coefficient and 0.08 to 0.18 for the friction coefficient. There is a potential inverse correlation between friction coefficient and event volume. Calibrated parameters were compared to morphometric parameters for the study sites, which may be useful for guiding parameter selection once a larger dataset is calibrated. Ongoing work is focused on refining the calibration technique, including standardization of input parameters more relevant to debris flows. The long-term goal is to apply the technique to a larger dataset of debris-flow cases and provide practitioners with better guidance on the selection of model input parameters for forecasting purposes.