Landslides, Erosion and Landscape Evolution along the Eastern Margin of the Tibetan Plateau. (Invited)

Landslides are a prominent component of hillslope erosion in steep, actively incising landscapes around the world, yet few models of bedrock river incision and landscape evolution explore feedbacks involved with particularly large individual landslides or an extremely high number of landslides associated with a large earthquake. Here, we highlight the influence of landslides on landscape evolution along the eastern margin of the Tibetan Plateau. Large landslides within the Dadu, Yalong and Min river gorges draw attention to the interaction between landslide dams, channel morphology, longitudinal river profiles, and river incision - emphasizing that over time-scales relevant to landscape evolution (>10,000 yr), large landslides can inhibit incision and prevent the complete adjustment of rivers to regional tectonic, climatic and lithologic forcing. The extremely high number of landslides (and landslide dams) associated with the Wenchuan earthquake, meanwhile, highlights the high magnitude and episodic nature of hillslope erosion along the Longmen Shan front. Furthermore, the pulse of erosion brought on by the earthquake, coupled with long recurrence intervals for such events (>2000-4000 yr), suggests that measured rates of short-term erosion in the region may have missed the input from previous earthquake-related landslide events, and thus underscores the need for caution when comparing short-term and long-term average erosion rates. These examples highlight the fundamental role that landslides play in the evolution of a steep, active eroding landscape such as the eastern margin of the Tibetan Plateau. However, they also point out an important lesson in landscape analysis: given the snapshot of on-going landscape evolution observable at any moment of time, the suite of landforms and measurable process rates may reflect the direct legacy of past events (e.g., landslide dams), or, in the absence of clear evidence of a big event (e.g., Longmen Shan, prior to the earthquake), may omit evidence of extreme but rare events that nonetheless exert significant influence on long-term landscape evolution. Finally, these examples serve to emphasize that the sediment delivered to river channels during large individual landslides and earthquakes presents an interesting problem for landscape evolution models, particularly when it comes to the feedback between hillslope erosion and channel incision.