Modeling forced pool–riffle hydraulics in a boulder-bed stream, southern California

Abstract

The mechanisms which control the formation and maintenance of pool–riffles are fundamental aspects of channel form and process. Most of the previous investigations on pool–riffle sequences have focused on alluvial rivers, and relatively few exist on the maintenance of these bedforms in boulder-bed channels. Here, we use a high-resolution two-dimensional flow model to investigate the interactions among large roughness elements, channel hydraulics, and the maintenance of a forced pool–riffle sequence in a boulder-bed stream. Model output indicates that at low discharge, a peak zone of shear stress and velocity occurs over the riffle. At or near bankfull discharge, the peak in velocity and shear stress is found at the pool head because of strong flow convergence created by large roughness elements. The strength of flow convergence is enhanced during model simulations of bankfull flow, resulting in a narrow, high velocity core that is translated through the pool head and pool center. The jet is strengthened by a backwater effect upstream of the constriction and the development of an eddy zone on the lee side of the boulder. The extent of flow convergence and divergence is quantified by identifying the effective width, defined here as the width which conveys 90% of the highest modeled velocities. At low flow, the ratio of effective width between the pool and riffle is roughly 1:1, indicating little flow convergence or divergence. At bankfull discharge, the ratio of effective width is approximately 1:3 between the pool and downstream riffle, illustrating the strong flow convergence at the pool head. The effective width tends to equalize again with a ratio of 1:1 between the pool and riffle during a modeled discharge of a five-year flood, as the large roughness elements above the pool become drowned out. Results suggest that forced pool–riffle sequences in boulder-bed streams are maintained by flows at or near bankfull discharge because of stage-dependent variability in depth-averaged velocity and tractive force.