(2) Armor Development from Decapitated Flash Flood Bores in Supply-Limited Flume Experiments
Kealie Goodwin1, Rebecca Rhodes2, Joel P Johnson3
1. University of Texas at Austin -email@example.com
2. The College at Brockport, SUNY -firstname.lastname@example.org
3. University of Texas at Austin -email@example.com
In rivers assumed to have quasi-normal flow, three main processes have been used to explain bed surface armoring: i) selective entrainment and transport of smaller grains, ii) limited supply of smaller grain sizes, and iii) equal mobility of grains of different sizes, which develops through natural feedbacks such that larger, less inherently mobile grains are enriched on the surface relative to smaller grains. Flash flood-dominated river channels in arid environments often completely lack surface armoring, yet it is unclear whether increased sediment supply or transport of all grain sizes prevents armor development. In order to examine armor development in an end-member case of non-normal flow, we conducted a series of laboratory experiments using flash flood bores. The flume is 33.5 m long, 0.5 m wide, 0.8 m tall, and capable of creating reproducible flood bores by raising a high-speed computerized lift gate and releasing impounded water. For each experiment, the gate was quickly lowered as soon as the flood bore traveled the length of the flume, “decapitating” the bore from subsequent flow, to better isolate the effects of the bore alone on entrainment and transport. Sediment was not fed into the upstream end of the flume and only sourced from the gravel bed (2 mm to 40 mm), resulting in supply-limited experimental conditions. In response to repeated flood bores, the surface grain size distribution rapidly coarsened. We interpret that kinetic sieving was the dominant cause of surface armoring in these experiments. Digital gravelometry from photographs taken after each bore show increased armoring, while sediment transported out the downstream end of the flume tended to be as coarse or coarser than the bed surface. Travel distances of three sizes of RFID-tagged tracer clasts show that the largest particles were transported farthest, while the smallest particles were preferentially buried under the surface. In these experiments, the bores disrupt the bed surface and entrain grains of every size class so that the smallest size fraction is able to fall in between the larger grains, coarsening the surface and preventing the smallest grains from being transported. While the combination of decapitated bores and supply-limited gravel sizes do not directly mimic natural channel conditions, our experimental design uniquely isolates the effects of bores on transport and grain sorting. They also suggest that transport of all grain sizes is not the primary control on armoring in natural flash flood-dominated channels. Future experiments will investigate the role of sand supply on armoring and transport in flash flood bores.