Uncertainty of debris flow mobility relationships and its influence on the prediction of inundated areas

Empirical mobility relationships can be used for delineating debris flow inundated areas. A database of documented events in alpine mountain catchments of northeastern Italy is used to test the model DFlowz (Berti and Simoni, 2007). We back-analyzed 25 DF events, ranging in volume from 3000 to 350,000 m3, with the support of high resolution topographic information derived from LiDAR. The analysis makes use of an objective methodology for evaluating the accuracy of the predictions and involves the calibration of the model based on factors describing the uncertainty associated with the empirical relationships.Results indicate that the model is capable of reproducing the observed behavior with a maximum uncertainty of a factor of 3. The most relevant source of error lies in the estimation of the deposited volumes which affects the results of back-calculation and is mainly responsible also for the scatter associated with the empirical mobility relationships. On the contrary, the influence of different flow properties on the depositional process appears to play a minor role as the mutual relations between the three main scaling parameters (volume, inundated area, and cross-sectional area) are respected in the vast majority of cases and calibrated mobility coefficients show no significant relationship with the angle of reach of the deposit.

► 25 debris flow events are back-analyzed by means of the model DFlowz.
► The dataset confirms the validity of the empirical scaling relationships.
► Prediction uncertainty is not primarily attributable to differences in flow mobility.
► DFlowz can be used for prediction of inundated areas with specific tolerance levels.