Probabilistic identification of debris flow source areas in the Wenchuan earthquake-affected region of China based on Bayesian geomorphology.

After the 2008 Wenchuan earthquake (M_S 8.0), a large number of debris flow source areas formed and further affected the occurrence of debris flows. Identifying the source area that may participate in debris flow activities requires understanding its dynamically changing conditions of location and probability. Based on this, the Bayesian geomorphology method was used to analyze the variation probability in the sediment supply from coseismic landslide deposits to subsequent debris flows and to undertake a quantitative assessment as well. The segment of the Min River, stretching from the urban area of the Wenchuan County to the Yingxiu town, China, was selected as the study area. Geomorphologic characteristics, including elevation, distance to faults, seismic intensity, land cover, gradient, monthly precipitation, and lithostratigraphy, were determined as assessment factors. The method was applied and verified in three clusters of debris flow events from 2010 to 2019 in 37 debris flow catchments. The results indicate that the method is applicable for identifying the dynamic debris flow source areas in earthquake-affected regions with a high accuracy of over 87% and a refined resolution of 30 m. This study finds that the sediment supply types have changed since 2019, that is, the sediment supply from Wenchuan earthquake-induced landslide deposit zones occupied by the debris flow-prone area has dropped from an average of 95.9% in 2013 to approximately 44.7% in 2019. Still, a large number of coseismic landslide deposits exist in the debris flow basins; the probability of debris flow source areas could thus be used as an important criterion for evaluating the long-term evolution of landslides and subsequent debris flow activities in earthquake-affected regions. [ABSTRACT FROM AUTHOR]