Hillslope-Channel Coupling at the Edge of a Relict Landscape

Knickpoints on fluvial channels are often thought to separate a relict landscape upstream from an active, incising landscape downstream. However, immediately upstream of some knickpoints, we may find the channel has steepened due to its downstream boundary condition. Channel narrowing and an increase in the hillslope angle immediately uphill from the channel may also accompany this channel steepening. The edge of a relict landscape can therefore serve as a natural laboratory in which coupled channel and hillslope processes are strongly influenced by the transient process of knickpoint retreat. We use the topography of the Roan Plateau in western Colorado, revealed by Airborne Laser Swath Mapping, to guide the development of a numerical model in which channel and hillslope processes are coupled. We model knickpoint retreat along an upstream-dipping caprock that results in an effective base level fall for the upstream channel. This base level fall increases channel erosion within a reach hundreds of meters upstream from the waterfall; this erosion is further enhanced due to flow acceleration near the free overfall. We develop a physically-based model for self-formed bedrock channels to assess how much of the base level fall is accommodated by channel narrowing and how much by channel steepening. This steepening will in turn lower the basal boundary condition controlling adjacent hillslopes. By substituting space for time, we use the Roan Plateau to explore two time scales predicted by our model. The first is the channel response time, evaluated via the length scale over which oversteepening occurs. This will be influenced by the degree of channel narrowing, and requires that we capture adequately the dynamics of channel response to base level fall. The second time scale is reflected by the uphill extent and degree of hillslope steepening, which will depend on both the channel steepening (as a necessary trigger) and processes and rates of regolith production and transport. Evaluation of this time scale requires that we capture the hillslope dynamics, particularly if regolith transport is nonlinearly related to slope. Comparison of model results for different end member scenarios against topographic metrics from a real landscape should provide a rigorous test of the importance of different processes in modifying the edge of a relict landscape.