The occurrence of strong earthquakes in highly undulating terrain around the eastern boundary of the Tibetan Plateau can trigger numerous coseismic landslides. Identifying the primary types of these landslides is crucial for assessing the scale and risk of secondary effects. This study utilized multi-source remote sensing images, comparing pre- and post-earthquake images of the September 5th, 2022, Luding Ms 6.8 earthquake, to extract an inventory with a total number of 8,385 coseismic landslides covering an area of 40.46 km². The majority of these landslides were distributed along the valleys of the Daduhe River and predominantly occurring on the middle and upper sections of slopes. The long and narrow shape indicates that the provenance area is limited, and the circulation and accumulation area are controlled by the steep slope morphology. The small scale of volumes of the inventory with the type of mountain peeling caused significant damage to mountain zig-zag roads and local buildings but less impact on the reservoir near the epicenter. This research enhances the understanding of spatiotemporal distribution characteristics and classification standards of coseismic landslides in high mountain canyon areas. We offer valuable insights into secondary disasters resulting from earthquakes in regions with highly variable terrain.

Supplementary material:https://doi.org/10.6084/m9.figshare.c.7774642