Why do published models for fluvial and estuarine morphodynamics use unrealistic representations of the effects of transverse bed slopes?

Abstract

The sediment transport direction is affected by the bed slope. This effect is of crucial importance for two- and three-dimensional modelling of the interaction between the flow of water and the alluvial bed. It is not uncommon to find applications of numerical morphodynamic models in the literature that exaggerate the effects of transverse bed slopes on sediment transport compared to results from laboratory experiments. We investigate mathematically the consequences of such an approach, and we analyse through numerical simulations different explanations for the need to apply deviating values. The study reveals that the reason often lies in the setup of the numerical models, such as the choice of mesh resolution or the necessity to comply with specific aspects of the numerical scheme. The missing or inadequate implementation of physical processes in the model is another cause. All of these effects can be compensated by artificial diffusion added through the bed slope effect coefficients. Since increased diffusion strongly alters the physical processes of self-formed bed morphology, we recommend that modellers address the root causes of inflated erosion and deposition. Bed slope effect coefficients should be applied within the range found in the original publications.