ERODIBILITY IN BEDROCK RIVERS: A TRICKY MIX OF MATERIAL PROPERTIES AND INTERNAL LANDSCAPE DYNAMICS (Invited Presentation)

Bedrock rivers evolve in response to climatic, tectonic, sediment supply, and lithologic controls. Unlike other key influences, the role of lithology, or substrate erodibility, has resisted encapsulation into a testable, mathematical theory. Here we present two end-member case studies, one at the channel cross-section scale and one at the landscape scale, whose results are best explained through the lense of bedrock erodibility as a state variable that depends both on inherited rock properties and the internal dynamics of eroding landscapes.

At the cross-section scale, we find that bedrock erodibility is far from steady and uniform even within a single rock unit. Erodibility, as measured with surficial proxies like uniaxial compressive strength, surface roughness, and density of cracks, varies as a function of elevation above the thalweg as well as the local stream erosion rate. This finding causes channel geometry to deviate from theoretical predictions, and implies that channel geometry emerges from a dynamic feedback between erosion and weathering. The strength of the feedback is modulated by rock properties. Cross-section-scale variability in rock properties within a single geologic map unit, which can be as great or greater than variability between units, adds further complexity.

At the landscape scale, erodibility is set by a combination of in situ bedrock properties and the supply of any boulders too large for the river to move. Significant correlations between river channel steepness and proximity to boulder-delivering hillslope failures in a well-constrained natural experiment suggest that boulder delivery drives resistance to river erosion and that the importance of this resistance varies by lithology. This control may be especially strong as boulder supply processes preferentially concentrate rock sourced from the most erosion-resistant exposures into channels. Boulder sizes and delivery rates depend on river incision history, such that boulder-driven resistance to erosion is a (potentially nonlocal) function of both rock properties and landscape evolution dynamics.

Results suggest that understanding lithologic controls on bedrock river evolution requires accounting for the influence of both inherited rock properties and the dynamic interplay between river erosion and rock erodibility.