Investigations of grain size dependent sediment transport phenomena on multiple scales

Abstract

Sediment transport in coastal and fluvial environments resulting from short time-scale processes of disturbance such as urbanization, mining, agriculture and military operations have significant impact on channel network and shoreline morphology, downstream water quality and ecosystems, and the integrity of land use applications. The scale and spatial distribution of these effects are largely attributable to the size distribution of the sediment grains that become eligible for transport due to disturbance. An improved understanding of advective and diffusive grain size dependent sediment transport phenomena will lead to the development of more accurate predictive models and preventative measures. To this end, three studies were performed that investigate grain-size dependent sediment transport on three different scales. Discrete particle computer simulations of sheet flow bedload transport on the scale of 0.1-100 millimeters were performed on a heterogeneous population of grains of various grain sizes. The relative transport rates and diffusivities of grains under both oscillatory and uniform, steady flow conditions were quantified. These findings suggest that, due to preferential vertical sorting of the largest grains to the top of the bed, a representative grain size that is functionally dependent on the applied flow parameters should be employed when parameterizing bed roughness. On the scale of 1-10m, experiments were performed to quantify the hydrodynamics and sediment capture efficiency of various baffles installed in a sediment retention pond, a commonly used sedimentation control measure in watershed applications. Analysis indicates that optimum sediment capture effectiveness may be achieved based on baffle permeability, pond geometry, and/or flow rate. Finally, on the scale of 10-1,000m, simulations were performed using a path sampling bivariate watershed erosion / deposition model in which grain size dependent terrain modification and pattern formation were integrated. Results correspond well to field observations and suggest that, with further refinements, the presented model may prove a valuable tool for further scientific advancement and engineering applications. Although a unique set of governing equations applies to each scale, an improved physics-based understanding of small and medium scale behavior may yield more accurate parameterization of key variables used in large scale predictive models.