EP51A-0888
Local Three-dimensional Flow and Sediment Patterns as an Evidence of Pool-riffle Self-maintenance

Friday, 18 December 2015
Poster Hall (Moscone South)
Elham Vahidi1, Esmaeel Bayat2, Jose Fernando Rodriguez2, Patricia M Saco2 and Gustavo A. M. de Almeida3, (1)University of Newcastle, Civil and Environmental Engineering, Callaghan, Australia, (2)University of Newcastle, Callaghan, Australia, (3)University of Southampton, Southampton, United Kingdom
Abstract:
For decades pool-riffle morphodynamics has been described based on the cross- sectional averaged flow characteristics, using episodic shifts in higher shear stress or velocities from the riffles to the pools (i.e. reversal conditions) as an indication of the long term self-maintenance of the structures. Recently more attention has been paid to three-dimensional flow features and sediment transport characteristics, but this has been done in a compartmentalised way, with studies either focusing on one or the other aspect. In this study, we present for the first time a three-dimensional analysis of sediment transport processes as an indication of self-maintenance mechanisms due to erosion and aggradation in pools and riffles. We do that by first reconstructing from experimental data the 3-D flow patterns in a typical pool-riffle sequence and computing bed shear stress distribution and near-bed streamlines for low and high flow conditions. We then use these 3-D features to complement a one-year dataset from an unsteady fractional sediment transport and bed evolution model applied to an existing stream. Local instantaneous bedload transport is obtained using the bed shear stress distribution corresponding to the flow condition (low or high) and assumed to follow the streamlines for that flow condition (low or high). Streamlines laterally contract and expand on pools and riffles, respectively, but in a different way during low and high flow condition. We apply the streamtube concept to compute instantaneous local rates of sediment transport in pools and contiguous downstream riffles and compute reversal conditions and identify self-maintenance episodes. Comparison with reversal episodes identified using cross- sectional average values reveal that when considering the 3-D effects stronger reversal episodes were detected, but the actual number of episodes increased only slightly (8%). Most reversal episodes occurred in the streamtubes that originated near the centre of the pool. Results support the idea that local flow can produce near-bed self-maintenance conditions even in conditions which average flow could not, but only to a certain extent.