Rivers meandering in bedrock: Lithologic, climatic, and process controls on form and evolution

Abstract:

Whereas meander wavelengths of alluvial rivers characteristically scale with bankfull discharge, bedrock meander wavelengths are typically 5–10 times greater than the scaling relationship for alluvial rivers would suggest, a fact that has led some to conclude that bedrock meanders are “underfit.” Others, however, have reasoned that larger dominant discharges should be expected for bedrock meanders to erode bank toes and mobilize sediment in bank-shielding scree piles, which often accompany steep, even vertical, outside banks capable of supplying coarse debris via landslide, debris flow, and rockfall. We attempt to test this hypothesis by finding dominant discharges for the Buffalo National River, Arkansas, by several methods. First, assuming that, as with alluvial meanders, bedrock meander wavelengths are 7–15 times hydraulic width at dominant discharge, we use cross-sections extracted from LiDAR-derived DEMs to find discharges corresponding to objectively-determined meander wavelengths. Second, assuming that dominant discharge must mobilize scree mantling outer-bank toe slopes, we use measured grain size distributions and cross sections to determine this discharge. For each of these calculated discharges, we use flow-duration curves to find corresponding recurrence intervals. Third, assuming that mean residence times of scree are similar to dominant discharge recurrence intervals, we use measured scree volumes and flux rates inferred from lateral migration rates to find those residence times. Preliminary results for the site with the longest gauge record yield a mean recurrence interval of 26 yrs corresponding to a meander wavelength-to-width scaling ratio of 11 (9–139 yrs for scaling ratios of 7–15). Recurrence intervals found by the other methods await field and cosmogenic isotope concentration measurements.