Abstract: Landslides are common geological hazards that are primarily influenced by rainfall, seismic activity, and changes in the geological structure of soil and rock. In particular, heavy rainfall significantly increases both the frequency and destructiveness of landslides. The dynamic variation of pore water pressure is considered one of the key driving factors behind the occurrence and development of landslides. This study focuses on the impact of pore water pressure variation on landslide disasters, using the 2018 Baige landslide as a typical case. Sentinel-1A ascending orbit D-InSAR technology was employed to monitor potential surface displacement and deformation in the region following the landslide, providing insight into how pore water pressure changes drive landslide dynamics. Additionally, a three-dimensional geomechanical model was developed, incorporating parameters such as the permeability coefficient, shear strength, and geological structure of the rock and soil mass. FLAC^3D was used to conduct dynamic simulations of the landslide, quantitatively characterizing the relationships between unsaturated seepage, pore water pressure, and shear strength under heavy rainfall infiltration conditions. The results show that an abnormal increase in pore water pressure in the toe resistance zone at the foot of the slope weakens the shear strength of the slip surface, triggering the expansion of the deformation zone and a sudden increase in displacement. This process shows strong correlation with the deformation patterns obtained from InSAR monitoring, and ultimately leads to a chain reaction of dynamic instability in the landslide. This study provides a quantitative analytical model and theoretical support for dynamic early warning and risk assessment of high-altitude landslide disasters in areas with heavy rainfall.

Key words: 2018 Baige landslide, Pore pressure, D-InSAR, Dynamic simulation