Sediment transport and bed topography for realistic unsteady flow hydrographs of varying length in a laboratory flume

Authors: Wren, Daniel; Kuhnle, Roger; Langendoen, Eddy; MCALPIN, TATE - Us Army Corp Of Engineers (USACE)

Submitted to: Journal of Hydraulic Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/2/2024
Publication Date: 4/25/2024
Citation: Wren, D.G., Kuhnle, R.A., Langendoen, E.J., Mcalpin, T.O. 2024. Sediment transport and bed topography for realistic unsteady flow hydrographs of varying length in a laboratory flume. Journal of Hydraulic Engineering. 150(4):18pp. DOI: 10.1061/JHEND8.HYENG-13769.

Interpretive Summary: Measurement and prediction of sediment transport are needed for management decisions that concern rivers and streams. Since it is difficult and expensive to measure sediment transport, numerical models are often used to estimate sediment loads. The models typically assume that stream conditions do not change in time and that the bed of the stream contains bed forms that are appropriate for the existing flow conditions. This is often not the case, since runoff from large rain storms can generate high flows that create larger bed forms that persist after the flow rate returns to normal, resulting in larger errors in estimating sediment transport rates. In this paper, we used a laboratory flume for experiments with flows that changed in a similar manner to a small stream, with a rapid rise in flow rate followed by a more gradual decline. We measured sediment transport, bed topography, and the slope of the water surface as they changed in response to the changing flow. We found that maximum bedform heights increased as the hydrographs got longer, but bedform lengths did not get longer for hydrographs greater than two hours long. When the measured water surface slope measurements were used, a common equation for sediment transport was able to provide reasonable estimates of the transport rate through the hydrographs. This information will be used to improve the prediction of sand transport and the timing of bed adjustment to water flow, which will help improve sediment transport modeling. The work can also aid in flood forecasting for streams, since the bed configuration also affects water depth. The results will be used by river managers, model developers, and other researchers studying river engineering and sediment transport.

Technical Abstract: Complex interactions between flowing water and bedforms in rivers with sand beds during changing flow rates result in non-equilibrium conditions where bedforms, sediment flux, and flow resistance are often different on the rising and falling limbs of flow hydrographs. In order to help bridge knowledge gaps and improve the ability to model transport rates and bedform dimensions for unsteady flows, a series of laboratory experiments in a recirculating flume at the USDA-ARS-National Sedimentation Laboratory was initiated. Six hydrographs with durations of 1-6 hours in 1-hour increments were created by scaling a real hydrograph from Goodwin Creek Experimental Watershed, near Batesville, MS, USA. Bedform conditions and water-surface slope were measured over a 15-meter section of the flume throughout the experiments. Sediment transport rate was also monitored during all experiments. Counter-clockwise hysteresis between discharge and sediment transport rate as well as bedform dimensions was present for all hydrograph lengths. Clockwise hysteresis was found for the relationship between discharge and bedform migration rates. The amount of hysteresis in transport rate, bedform amplitude, and bedform length varied according to the length of the hydrograph, although the relationship was not monotonic. The Engelund-Hansen method was able to approximately predicted sediment transport rates based on the real-time water surface slope measurements that were available during the experiments. Maximum bedform amplitudes increased with hydrograph length, while bedform lengths did not continue to increase for hydrographs of 2-hour and longer duration.