Scaling of sediment dynamics in a laboratory model of a sand-bed stream

•A strategy is explained to scale mixed load sand-bed streams in the laboratory.•The Shields number, the particle Reynolds number and the size distribution of sediments are the most important scaling parameters.•Similarity in the Shields number produces bedform dimensions that are close to the geometric scale.•Lightweight sediments decrease the time needed to achieve equilibrium.

A movable bed model was designed in a laboratory flume to simulate a mixed load sand-bed stream. The modelling objectives were to reproduce bedload and suspended sediment transport as well as downstream and transverse sediment fluxes in ratios similar to the field site. To meet these objectives the model contained an exact geometric scale and graded lightweight sediments to simulate migrating dunes and suspended load transport. The experiments are somewhat novel in that most mobile bed models have vertical exaggeration, whereas in these experiments exact geometric similitude of channel dimensions was maintained. The goal of this paper is to review the scaling strategy and the level of similarity among dimensionless parameters between model and field. Similarity in dimensionless bed shear stress and the particle Reynolds number enabled the experiments to replicate the dominant sediment dynamics present in the stream during a bankfull flow. There was a conflict in the strategy, in that grain roughness was exaggerated with respect to nature. However, the paper shows that geometric similarity of bedforms and the resulting drag is much closer to what is predicted for nature. In addition, measurements of sediment transport are compared to values computed from well-supported formulations, which is shown to reinforce the validity of the scaling strategy. Lastly, criteria for movable bed equilibrium are defined and it is shown that lightweight sediments contributed to the rapid development of near-equilibrium conditions. Overall, the paper shows a methodology that can be used to model mixed load streams at an exact geometric scale.