Sustained Embayment of Shoulder Type Escarpments through a Feedback between Knickpoint and Divide Migration

Topographic escarpments are characterized by an extreme slope change across the escarpment edge, promoting escarpment retreat through time. Escarpments are typically divided into arch- and shoulder-type. In arch-type escarpments, the drainage divide is located inland from the escarpment edge and knickpoints that are formed where channels flow across the escarpment can retreat and embay the escarpment. In shoulder-type escarpments, the drainage divide aligns with the escarpment edge, which is expected to cause a slow and uniform escarpment retreat. However, shoulder-type escarpments are sometimes associated with deep embayments that may influence their long-term evolution. Yet, the processes that sustain embayments in shoulder-type escarpments remain largely unexplored. In exploring this process, we note that shoulder-type escarpments are typically embayed along antecedent channels that used to flow away from the escarpment but are now flowing across it (hereafter reversed channels). The reversed channel typically develops over the sedimentary fill of the antecedent valleys, that is often underlaid by more resilient rocks. We hypothesize that a feedback between knickpoint retreat and divide migration along such reversed channels sustains the escarpment embayment. In this feedback, divide migration increases discharge at knickpoints and facilitates knickpoint retreat. Knickpoint retreat, in turn, changes the local base level for the divide and drives further divide migration. To investigate this mechanism, we analyzed reversed channels in West Virginia and used numerical simulations to show that the coupled dynamics of divide migration and knickpoint retreat can reach a steady state, such that embayment extend at a constant rate while the distance between divide and knickpoint remains fixed. We show that the steady-state distance between divide and knickpoint depends on lithologic differences (erodibility) between bedrock and channel-fill sediment. Overall, our findings suggest that embayment of shoulder-type escarpments can be sustained through a feedback between the divide and knickpoint migration along reversed channels, and that the rate and pattern of retreat is influenced by lithologic variations. This feedback could explain the global diversity of escarpment morphologies.