Characterization and Prediction of Hyporheic Flow Induced by Bars in Gravel Streams
Hyporheic flows induced by bars in gravel streams are vital to the overall health of a river but can also be the primary mechanism for contamination to fish habitats. It is therefore of primordial importance to characterize such hyporheic flows under varying hydraulic and morphological conditions, and develop methods and tools for their prediction. This thesis is intended as a contribution to these ends. The work involves an extensive series of laboratory experiments, guided by a dimensional analysis of the phenomenon under consideration, and revealing a possible dependency of vhyp /k on Δ/Λ and Δ/hav. Hyporheic flows were visualized in a gravel bar and underlying substrate constructed in a 21 m long flume under nine different test cases (combination of varying free surface flow depths (18, 15 and 12 cm) and bar lengths (1.0, 1.6 and 2.5 m)). Dye injections were recorded and processed to reveal hyporheic flow paths and velocities. Hyporheic flows upstream and downstream of the bar top traveled longitudinally, converged at the flow divide line and upwelled into the stream flow at the flow divide line. For any given bar, smaller free surface flow depths produced hyporheic flows that penetrated deeper into the gravel and emerged into the stream flow further downstream. Hyporheic flow was always turbulent with the 1.0 and 1.6 m long bars producing similar hyporheic flow velocities that were approximately 30% higher than those produced by the 2.5 m long bar. The present analysis shows that for any given value of Δ/Λ, vhyp /k varies with Δ/hav by first increasing from 0 to a maximum at Δ/hav ≈ 0.45, and then decreasing. Although Darcy’s equation is only valid for laminar flows, a simple model based on this equation was developed for determining hyporheic flows through a gravel bar with the goal of investigating if Darcy’s equation can provide any reasonable approximation of hyporheic flow. The computational grid represents the gravel bar and underlying substrate of the laboratory experiments. The model produced hyporheic flow paths which were in good agreement with the laboratory experiments. However, the hyporheic flow velocities were greatly underestimated. In spite of this, the model offers a quick and computationally simple method for preliminary assessments in practical applications.