surface processes research group


Ph.D. (Geology) student, Jack Zunka, and I are investigating meandering where rivers are incising into bedrock. Valleys with rivers eroding bedrock both laterally and vertically evolve asymmetric cross sections wherein the slopes adjacent to the eroding banks at the outsides of bends are over-steepened, sometimes to the point of forming overhanging cliffs, and slopes adjacent to the passive banks at the insides of bends are more gradual, so-called slip-off slopes. This cross-section asymmetry is the hallmark of a river actively meandering in bedrock, although there have been periods of vigorous debate regarding whether bedrock meanders might instead be forms inherited from past conditions.

Fueling the debate are observations that bedrock meanders are bigger than alluvial meanders, i.e., where rivers are eroding unconsolidated sediment. There exists a power-law relationship (i.e., a straight line in log-log space) between alluvial meander wavelength and bankfull channel width (or bankfull discharge) over many orders of magnitude of both variables, where meander wavelength is twice the straight-line distance between the straight bits on either end of a meander bend. There is a similar relationship for bedrock meanders, but for the same discharge, the wavelength of bedrock meander bends is generally larger than for alluvial meanders.

We hypothesize that meander wavelengths reflect discharges of frequencies and magnitudes that are most effective for bank erosion. For alluvial meanders, effective discharges for bank erosion are similar to bankfull events, which typically occur with an annual flood recurrence interval of approximately 1.5 years. We reason that the flows necessary to both mobilize coarse sediment (e.g., boulders) mantling the bases of over-steepened bedrock slopes and erode the underlying bedrock are probably larger and less frequent than the flows necessary to erode alluvial banks. Meander wavelength is a consequence of the effective flow hydraulics, because loci of erosion of alternate banks must be separated by the distance over which the high-velocity flow moves from one bank to the other. This distance is greater for greater discharges with wider, faster flows capable of eroding bedrock banks.

Complementary to a closer look at the flows eroding bedrock is an investigation of the relationship between valley cross-section morphology and the relevant process rates. For example, what are the conditions that promote formation of cliffs rather than soil-mantled slopes? This and other questions led us to develop a two-dimensional particle-based valley cross-section model.

Bedrock meandering

In the field, the focus of our investigations is the Buffalo National River in Arkansas, where we are collaborating with another research group investigating the incision history of the Buffalo. Broadly, this group comprises Amanda Keen-Zebert (Desert Research Institute), Stephanie Shepherd (Bloomsburg University of Pennsylvania), Matt Covington (University of Arkansas), Mark Hudson (USGS). We are also collaborating with Darryl Granger (PRIME Lab), James Syvitski (University of Colorado at Boulder), Greg Tucker (University of Colorado at Boulder), Robb Jacobson (USGS), Chuck Bitting (Buffalo National River), and Bob Cross (Ozark Society).

Animation showing soil/sediment (light gray), bedrock (dark gray), water (blue), and cosmogenic nuclide (TCN) concentration (red in soil/sediment, yellow in bedrock) in 2D simulation of weathering, slope transport, rockfall, left-side-biased fluvial erosion/evacuation, and TCN production. “River” cuts down and to the left, but relief is regularly re-centered vertically. Boundary conditions on left and right are periodic, and two periods are shown.