Beyond the Angle of Repose: A Review and Synthesis of Landslide Processes in Response to Rapid Uplift, Eel River, Northern Eel River, Northern California

摘要

In mountainous settings, increases in rock uplift are often followed by a commensurate uptick in denudation as rivers incise and steepen hillslopes, making them increasingly prone to landsliding as slope angles approach a limiting value. For decades, the threshold slope model has been invoked to account for landslide-driven increases in sediment flux that limit topographic relief, but the manner by which slope failures organize themselves spatially and temporally in order for erosion to keep pace with rock uplift has not been well documented. Here, we review past work and present new findings from remote sensing, cosmogenic adionuclides, suspended sediment records, and airborne lidar data, to decipher patterns of landslide activity and geomorphic processes related to rapid uplift along the northward-migrating Mendocino Triple Junction in Northern California. From historical air photos and airborne lidar, we estimated the velocity and sediment flux associated with active, slow-moving landslides (or earthflows) in the mélange- and argillite-dominated Eel River watershed using the downslope displacement of surface markers such as trees and shrubs. Although active landslides that directly convey sediment into the channel network account for only 7% of the landscape surface, their sediment flux amounts to more than 50% of the suspended load recorded at downstream sediment gauging stations. These active slides tend to exhibit seasonal variations in velocity as satellite-based interferometry has demonstrated that rapid acceleration commences within 1 to 2 months of the onset of autumn rainfall events before slower deceleration ensues in the spring and summer months. Curiously, this seasonal velocity pattern does not appear to vary with landslide size, suggesting that complex hydrologic-mechanical feedbacks (rather than 1-D pore pressure diffusion) may govern slide dynamics. A new analysis of 14 years of discharge and sediment concentration data for the Eel River indicates that the characteristic mid-winter timing of earthflow acceleration corresponds with increased suspended concentration values, suggesting that the seasonal onset of landslide motion each year may be reflected in the export of sediments to the continental margin. The vast majority of active slides exhibit gullied surfaces and the gully networks, which are also seasonally active, may facilitate sediment export although the proportion of material produced by this pathway is poorly known.Along Kekawaka Creek, a prominent tributary to the Eel River, new analyses of catchment-averaged erosion rates derived from cosmogenic radionuclides reveal rapid erosion (0.76 mm/yr) below a prominent knickpoint and slower erosion (0.29 mm/yr) upstream. Such knickpoints are frequently observed in Eel tributaries and are usually comprised of massive (u3e10m) interlocking resistant boulders that likely persist in the landscape for long time periods (u3e105yr). Upstream of these knickpoints, active landslides tend to be less frequent and average slope angles are slightly gentler than in downstream areas, which indicates that landslide density and average slope angle appear to increase with erosion rate. Lastly, we synthesize evidence for the role of large, catastrophic landslides in regulating sediment flux and landscape form. The emergence of resistant blocks within the mélange bedrock has promoted large catastrophic slides that have dammed the Eel River and perhaps generated outburst events in the past. The frequency and impact of these landslide dams likely depend on the spatial and size distributions of resistant blocks relative to the width and drainage area of adjacent valley networks. Overall, our findings demonstrate that landslides within the Eel River catchment do not occur randomly, but instead exhibit spatial and temporal patterns related to baselevel lowering, climate forcing, and lithologic variations. Combined with recent landscape evolution models that incorporate landslides, these results provide predictive capability for estimating erosion rates and managing hazards in mountainous regions.