Spatial and temporal evolution of long-term debris flow activity and the dynamic influence of condition factors in the Wenchuan earthquake-affected area, Sichuan, China

Abstract

In recent decades, there have been several extreme earthquakes around the world, with the 2008 Wenchuan earthquake resulting in the highest number of landslides and substantial deposition of loose material on steep terrain and deep gullies, thereby promoting post-earthquake debris flows. Our monitoring indicates that debris flows have become long-term threats for over a decade after the earthquake, and it is crucial to determine the long-term spatial and temporal changes of debris flows and the evolving influence of condition factors in the earthquake-affected area. Thus, we established a database of 1628 debris flow events occurring between 2000 and 2020 within the 31,627 km2 area affected by the Wenchuan earthquake. The database included information on the date of occurrence, GNSS (Global Navigation Satellite System) coordinate at the gully outlet, event washout volume, and scale level obtained from various sources. Post-seismic debris flow activity witnessed a significant increase, followed by oscillation with rainfall and gradual weakening. Using the average washout volume for each scale in different periods, we applied a grid calculation method to characterize the debris flow intensity (DFI) using the total debris flow volume of cells in different epochs to study the spatial and temporal evolution of debris flow activity. Thus, based on the DFI calculation, we introduced the certainty factor and Geodetector to determine the dynamic evolution of condition factors. Our results indicate that lithology, rainfall, and slope degree primarily controlled debris flows before the earthquake. However, coseismic landslides became the main promoting factor after the earthquake. As loose materials were consumed, hydrodynamic and topographic factors began to play a more significant role in controlling debris flows. During this process, the requirements for rainfall, slope degree, and terrain relief amplitude increased over time. Based on the annual volume of debris flows from 2000 to 2020, we developed a preliminary conceptual model, which indicated that post-earthquake debris flow activity can be categorized into three phases: active (2008–2013), unstable (2014–2033), and recession (2034–2045), with the debris flow activity anticipated to return to pre-earthquake levels by 2045.