Lithological controls on clustered landslides: a case study of landslides triggered by Typhoon Gaemi (2024) in Zixing, Hunan Province, China
摘要
Short-duration, high-intensity rainfall events are becoming more frequently associated with catastrophic geohazards, such as large debris flows and clustered landslides worldwide. Between 26 and 28 July 2024, exceptional rainfall during Typhoon Gaemi triggered clustered landslides in Zixing, Hunan Province, China. This study, based on a landslide inventory created through remote sensing interpretation and field validation, analyzes the spatial distribution, controlling factors, and mobility characteristics of the landslides using statistical analysis, machine learning, and experimental testing. Results show that both landslide clustering and mobility characteristics are strongly conditioned by geological setting, with lithology exerting the primary control. Landslide mobility is primarily controlled by slope gradient and exhibits a negative correlation with it. In granite areas, landslides predominantly occur within shallow residual soil. This region exhibits the highest landslide number and area density, as well as relatively high mobility, an average H/L ratio of 0.2, with an average slope angle of 16.8°. In contrast, landslides in sandstone areas primarily initiate along the interface between residual soil and weathered bedrock. The steeper topography in these regions has an average slope angle of 24.3°. In these settings, mobility is relatively uniform, an average H/L ratio of 0.3, indicating lower mobility than in the granite areas with gentler slopes. The synergy between the erosion-prone nature of granite residual soil and the channeling topography of the hills enabled landslides to rapidly mobilize into extensive debris flows, amplifying their mobility and destruction, which led to the most concentrated damage in the granite areas. These findings indicate that hazard assessment should explicitly consider lithology and landslide mobility to improve runout prediction and enable targeted mitigation under intensifying rainfall extremes.