Controls on the deposition of extremely large post-earthquake debris flows in Wenchuan

Debris flows are the dominant process delivering sediment from hillslopes into channels following the 2008 Wenchuan earthquake. Post-earthquake debris flows continue to pose a significant hazard to the recovering local communities. In 2019, a period of intense rainfall triggered several extremely large debris flows. The flows bulked to volumes in excess of 100 000 m³,  much larger than their initiation volumes, and transited catchments to be deposited in the Min Jiang river. The scale of these flows highlights our limited understanding of why and where large debris flows deposit. Previous studies have shown that topography (notably bed slope and channel width), flow composition (grain size), and flow characteristics (velocity and depth) can all control debris flow runout. Yet, there is limited understanding of how these interrelate. For example, whether abrupt changes in topography, such as increased channel width, lead to the deposition of certain grain size fractions and subsequently encourage further deposition. Alternatively, whether changes in bed slope affect flow velocity and this results in the entrainment of specific grain size fractions by the flow. An understanding of these relationships will help to better constrain where and how post-earthquake debris flows are more likely to deposit.

In this study, we determine how debris flow characteristics (velocity and depth) and the grain size distribution (GSD) deposited by the debris flow evolve with changes in topography and distance from the initial debris flow source. To achieve this, we simulated two post-earthquake debris flow events in the Liusha and Luoquan catchments, China, using the 2D dynamic debris flow model, Massflow. GSDs were collected by sampling and sieving pits located equidistantly along the centre of each 2019 debris flow deposit. Bed topography data was recorded both in the field and using a 30 m resolution DEM. We compared changes in the flow characteristics and GSDs deposited for each debris flow with the data for bed topography to explore how controls on debris flow runout interrelate. Preliminary findings for the Luoquan debris flow suggest a relationship between negative changes in curvature and the deposition of fine-grained material. This work will help to better understand controls on debris flow runout, subsequently aiding future studies of post-earthquake debris flow hazard prediction.