Testing for reach-scale adjustments of hydraulic variables to soluble and insoluble strata: Buckeye Creek and Greenbrier River, West Virginia

Abstract

Listen

Translate article

An open question exists as to whether channel geometries and hydraulics are adjusted in bedrock streams with stable, concave profiles in a manner analogous to alluvial rivers. As a test of this problem, a comparison was undertaken of channel geometries and hydraulics among reaches with substrates that are of high mechanical resistance, but of variable chemical resistance. Reaches were selected from Buckeye Creek and Greenbrier River, West Virginia, USA because these streams flow over sandstones, limestones, and shales. The limestones have Selby rock resistance scores similar to those of the sandstones. A total of 13 reaches consisting of between 6 and 26 cross sections were surveyed in the streams. HEC-RAS was used to estimate unit stream power ( ω) and shear stress ( τ) for each reach. The reaches were selected to evaluate the null hypothesis that that ω and τ are equal atop soluble versus insoluble bedrock. Hypothesis tests consisted of paired t-tests and simultaneous, multiple comparisons. Geomorphic setting was included for Greenbrier River because previous studies have suggested that bedrock streams are intimately coupled with hillslopes. Holding geomorphic setting constant, three separate comparisons of ω and τ reveal that these variables are lowest atop soluble substrates in Greenbrier River (significance ≤0.05) and that changes in ω and τ are mediated by changes in channel geometry. Similarly, headwater reaches of Buckeye Creek developed atop shale and sandstone boulders are statistically distinguishable from downstream reaches wherein corrosion of limestone is the primary means of incision. However, comparisons in each stream reveal that channel geometries, ω and τ, are not strictly controlled by bed solubility. For constant substrate solubility along the Greenbrier River, ω and τ are consistently higher where a bedrock cutbank is present or coarse, insoluble sediment enters the channel. The latter is also associated with locally high values of ω and τ in Buckeye Creek. Assuming that incision by corrosion requires lower values of ω and τ because the channel need not be adjusted for block detachment and tool acceleration, we posit that the statistically lower values of ω and τ are tentative evidence in favor of differential geometric and hydraulic adjustments to substrate resistance. We observe that these adjustments are not made independent of geomorphic setting.