'Witness mountains' on scarplands are expected outcomes of drainage capture dynamics

Scarplands worldwide, underlain by tilted or undeformed sedimentary successions, display a distinct morphology with relief asymmetry between steep escarpments and gentle plateaus. This asymmetry drives escarpment retreat towards the plateau, gradually eroding the old plateau. Residual flat-topping landforms, commonly known as 'residual hills', 'mesas', 'buttes', or 'witness mountains,' prominently rise above the surrounding areas. Traditionally seen as remnants of an ancient plateau, these landforms are believed to endure escarpment retreat due to their capping by resistant rocks (e.g., sandstones, basalts, or duricrusts). This widely accepted conceptual explanation has yet to undergo testing through numerical modelling.

Here, we propose a distinct mechanism for the rapid formation of 'witness mountains' — drainage capture dynamics during escarpment retreat. Using a 2-D numerical landscape evolution model incorporating stream-power-driven river incision and linear hillslope diffusion (representing mass wasting processes), we simulate escarpment evolution in tilted and undeformed multi-layered stratigraphy. As rivers draining down the escarpment capture low-sloping plateau rivers, a knickzone rapidly propagates upstream through the captured river system, followed by steepening hillslopes behind the escarpment front. This process results in flat-topped drainage divides bounded by marginal escarpments in both the front and back of the initial escarpment, forming 'witness mountains'. With equally steep slopes on both sides, these 'witness mountains' do not retreat towards the plateau. Instead, they gradually decay in elevation in situ until their distinctive morphology is no longer discernible. Similar results emerge in simulations assuming uniform lithology, suggesting drainage capture can sculpt 'witness mountains' beyond differential erosion resistance.

Our simulations suggest that 'witness mountains' can be rapid outcomes of local drainage capture events affecting the scarpland foreland. This insight, supported by empirical observations in the Southwestern German Scarplands, challenges conventional interpretations of landscape evolution, offering valuable perspectives for understanding scarpland transience.