FEEDBACKS BETWEEN SEDIMENT SUPPLY AND CHANNEL EVOLUTION FOLLOWING KNICKPOINT RETREAT IN CANYONS ON THE ROAN PLATEAU, WESTERN COLORADO

Canyons downstream of multiple knickpoints on the Roan Plateau provide a unique opportunity to explore feedbacks between sediment supply and channel evolution in a transient landscape. We have previously explored the pattern of knickpoint retreat, initiated by late Cenozoic incision of the Colorado River, with a kinematic model in which the retreat history of each knickpoint is a power-law function of its declining upstream drainage area. We now make use of the strong similarity in shape between canyon planforms and channel longitudinal profiles downstream of knickpoints, well documented using LiDAR 1-m and USGS 30-m digital elevation data, to develop and test a model for channel evolution. Morphologic self-similarity requires that retreat of the canyon walls be tied to lowering of the channel profile. We ask how the bedrock channel profile and alluvial thickness distribution are maintained in the face of the sediment supply this linkage demands. We employ a simple numerical model founded upon conservation of sediment: the rate of change of alluvial thickness at a point is the sum of the sediment supplied from retreat of canyon walls, the loss of sediment (to solution or suspension) by weathering, and the divergence of sediment (bedload) transport. Consistent canyon width-to-depth ratios (typically ~2.5) allow us to scale canyon wall retreat rates to local channel incision rates. We assume higher weathering rates closer to each knickpoint, where waterfall misting is greatest and any loss of water discharge to infiltration is minimal. Bedload sediment transport scales with excess shear stress. Changes in sediment thickness modify both the channel slope and the protection of the bedrock channel floor, providing a feedback mechanism that either enhances or reduces the channel incision rate (and hence canyon wall retreat rate), depending upon the channel incision rule employed. Model iterations are evaluated by their ability to predict observed canyon planforms and longitudinal channel profiles, estimates of sediment thickness, and knickpoint retreat histories consistent with our previous modeling of multiple canyons. Our approach therefore marries detailed topographic data with numerical models, allowing quantitative testing of the effects of sediment supply on transient longitudinal profile evolution.