Alternate bar response to sediment supply termination

Sediment supply is widely held to be one of the primary controls on bar topography in alluvial channels, yet quantitative linkages between sediment supply and bar topography are not well developed. We explore the conditions under which alternate bars form and how they respond to the elimination of sediment supply in two linked laboratory experiments. The first set of experiments was conducted in a 28 m long, 0.86 m wide flume channel using a unimodal sand-gravel mix. The second set of experiments was conducted at field scale in a 55 m long, 2.74 m wide channel using a unimodal gravel mixture. In both experiments, alternate bars and patchy surface grain-size distributions developed under steady flow and sediment supply conditions. The cessation of the sediment supply induced a reduction in the surface grain-size heterogeneity and the bars were eliminated. In both flumes, mean boundary shear stress had declined, but were capable of moving sediments after the bars disappeared, albeit at relatively small rates compared to when the bars were present. In the smaller flume, the previously stationary bars migrated out of the flume and were not replaced with new bars. A nearly featureless bed formed with limited surface grain-size heterogeneity, a slightly coarsened surface and a slightly reduced slope. In the larger flume, the formation of alternate bars was induced by an imposed upstream flow constriction and as such, the bars did not migrate. Termination of sediment supply led to progressive erosion of bed topography and loss of the bars, coarsening of the bed surface, loss of bed texture patchiness and significant slope reduction. The original alternate bar topography redeveloped when the sediment supply was restored once sufficient deposition had occurred to reconstruct the original channel slope. This shows that the bar loss was reversible by establishing the previous conditions and highlights the importance of sediment supply for bar formation. The role of sediment supply in bar formation and stability is not often recognized in stream restoration. Our results suggest that the loss of sediment supply can significantly affect alternate bar topography and that considerable volumes of sediment may be needed restore channel bars.