Understanding the triggering mechanism and deformation characteristics of a reactivated landslide in the Baihetan Reservoir

Reservoir impoundment can significantly alter the hydrogeological conditions of landslides, potentially leading to the reactivation of ancient landslides and the formation of new landslides within the reservoir region. Understanding the deformation mechanism and triggering factors of these landslides can provide valuable insights for evaluating their stability and mitigating the associated hazards. This study considers the Tuandigou (TDG) landslide as an example of a massive reactivated ancient landslide in the Baihetan Reservoir region. We first present the geological setting, material composition, and structural characteristics of the TDG landslide. Subsequently, the surface displacement of the TDG landslide before and after impoundment was acquired through time-series Synthetic Aperture Radar Interferometry (InSAR) analysis of two overlapping C-band Sentinel-1 satellite tracks. Additionally, the digital elevation model (DEM) data derived from light detection and ranging (LiDAR) obtained through UAV-based surveys were employed to utilize various visualization methods for the identification of deformation cracks and micro-topographical features associated with the landslide movement. High-resolution Global Navigation Satellite System (GNSS) monitoring data were utilized to determine the critical reservoir water level that triggers the onset of rapid movement in the landslide. Furthermore, the reactivation mechanism of the TDG landslide was proposed based on the accumulated evidence. The results indicate that the TDG landslide exhibited relative stability prior to reservoir impoundment; subsequent reservoir filling is identified as the primary trigger for landslide reactivation. During the initial phases of reservoir filling, bank collapse occurred at the leading edge of the landslide, resulting in deformation that gradually propagated toward the rear and concentrated on the southern flank. Multi-source monitoring data revealed a significant increase in landslide movement as the reservoir water level rose to higher elevations. These observations suggest that the TDG landslide can be classified as a buoyancy-driven landslide, whereby buoyancy effects reduce the slip resistance of the highly permeable geotechnical materials submerged in water, thereby accelerating landslide movement.