The role of deep-seated landslides in landscape evolution: Quantitative modeling and high-resolution topographic analysis

In many mountainous settings, landslides are the primary geomorphic agent that sets fundamental landscape characteristics, such as topographic relief and catchment-averaged erosion rate. The coupled system of landslides and other geomorphic processes responds to changes in climatic or tectonic forcing, but few studies have addressed these responses quantitatively, especially in terrain prone to deep-seated landslides (those thicker than the upper layer of soil). This study quantifies the topographic expressions and mechanics of deep-seated landslides using a combination of high-resolution topographic data and mathematical modeling. I first demonstrate that deep-seated landslides distinguish themselves from surrounding terrain by generating meter spatial scale surface roughness associated with gradients in strain rate of the deforming material. These methods are capable of mapping landslides with more than 80% accuracy in three study sites throughout the Pacific Northwest, United States. At longer, kilometer scale spatial wavelengths analysis of slope and drainage area data shows that landslides lengthen hillslopes and reduce ridge top elevations to leave their signature on the topography. I then develop and implement a mathematical landscape evolution model including a novel treatment of deep-seated landslide flux to simulate landslides at these longer spatial scales. The model generates topographic profiles for two different bedrock types in agreement with those observed in a study area in the Eel River catchment, California, United States. The sediment fluxes required to produce these profiles are in agreement with independently estimated modern rates. Two-dimensional simulations constrain two essential geomorphic conditions at which landslides occur. First, there must be pre-existing pockets of deep weathering, which allow landslides to erode large volumes of material at rates that episodically exceed the long term average erosion rate. Second, the characteristic time scale for landslide processes must be shorter than the time scales associated with both soil creep and river incision. As the landslide time scale shortens, landslides systematically reduce hillslope relief and increase valley spacing to reduce the mean topographic gradient. This dissertation therefore improves the objectivity of analyzing landslide-prone terrain and provides a framework for rigorously interpreting landscape response to changing climatic and tectonic forcing.;This dissertation includes both previously published and co-authored material.