Modeling fluxes and form in landslide-prone terrain
Abstract
Landslides dramatically alter the Earth's surface over short timescales. The mass transfer associated with a limited number of slope failures can dominate the sediment budget of a region for decades or longer. The initiation, failure geometry, and runout of individual landslides depend on a range of factors and cannot be predicted from current models. Given these realities of landslide behavior over human timescales, it is challenging to reasonably represent these processes in landscape evolution models. Here, we evaluate the ability of two landslide models, both of which are formulated to apply at geomorphic timescales, to generate topographic patterns and sediment flux rates observed in natural landscapes. Episodic debris flow activity is ubiquitous in steep, low-order mountainous catchments and generates valley networks with low concavity. A physically-based model for debris flow incision (Stock and Dietrich, GSA Bull, 2006) proposes that incision rates depend on the frequency, volume, and velocity of debris flows as well as the density of trigger sites and the state of bedrock weathering in low-order valleys. Valley slope angles are predicted to decline with drainage area according to how these properties vary spatially. We calibrated the model for a well-studied small catchment in the Oregon Coast Range using cosmogenic radionuclide erosion rates and then analyzed the slope-area signature of low-order valleys across much of the Central Oregon Coast Range to explore spatial variations in baselevel lowering. This endeavor shows that baselevel lowering rates vary significantly due to patches of resistant bedrock, drainage reorganization, and tectonic forcing. In regions with weak sedimentary bedrock, earthflows can reduce hillslope gradients, promote gullying, and dominate sediment yield through their downslope translation. A one-dimensional, physically-based model for earthflow-prone hillslope evolution (Booth and Roering, JGR, in press) incorporates earthflow, gully, and soil creep transport and predicts that average hillslope gradient increases nonlinearly with erosion rate. We applied the model to interpret catchment adjustment in the wake of a well-characterized knickpoint in the Waipaoa River, New Zealand. The model predicts the spatial pattern of erosion rate in the catchment based on average gradients and can be further tested by characterizing the morphology of individual hillslopes. Landslide models that integrate across multiples events or episodes of landslide activity should be crafted with testable (and thus rejectable) morphologic and sediment flux constraints. Airborne lidar and erosion rates from cosmogenic radionuclides have greatly facilitated this endeavor although additional constraints are needed.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFMEP43D..01R
- Keywords:
-
- 1815 HYDROLOGY / Erosion;
- 1826 HYDROLOGY / Geomorphology: hillslope