Bedforms, macroturbulence, and sediment transport at the fluid-bed interface

Abstract

Small scale morphodynamic processes at the fluid-bed interface in coastal environments are still not fully understood. Deformation of the seabed by hydrodynamic forces results in the occurrence of complex bedforms on a variety of scales. Owing to their interaction with the water column and turbulence production, large bedforms, such as dunes, significantly influence sediment transport and hydraulic roughness. Turbulence production above bedforms is predominately associated with large-scale coherent flow structures (macroturbulence). It has been suggested that macroturbulence is the principle mechanism behind the entrainment and transport of sediment in suspension. The objective of this thesis was to investigate the formation and occurrence of macroturbulent structures and their coupling to suspended sediment transport forming downstream of natural bedforms. Specifically, the effect of flow unsteadiness and bedform geometry was examined. The research is based on multiple research approaches including two field campaigns into the German Elbe Estuary and Danish tidal inlet Knudedyb. Ship-based observations and lander-based moorings deploying acoustic Doppler current profilers and optical measurement probes were conducted. The field research was complemented by laboratory experiments and a 2DV numerical model study. Results show that flow unsteadiness has significant implications on the production of macroturbulence. In both, the Elbe Estuary and the tidal inlet Knudedyb, very large compound bedforms occur in water depths of < 20 m. The primary bedforms remain ebb-oriented during a tidal cycle while smaller superimposed bedforms reverse direction with each tidal phase. Water-depth scale macroturbulence develops when flow direction and primary bedform orientation are aligned (ebb phase), once the accelerating ebb flow overcomes a velocity threshold. The frequency of macroturbulent structures falls into the Strouhal range of macroturbulence associated with flow separation zones downstream of bedforms. The flow structures originate in the region of high velocity gradients in the bedform lee and are traceable over the downstream bedform stoss-side. The magnitude of turbulence production decreases with the slope of the bedform lee-side. Our laboratory and model results show that the velocity gradient downstream of bedforms decreases with lower lee-slopes. Low turbulence intensities above lower lee-slopes are attributed to a smaller scale of macroturbulent structures though this has yet to be verified. In the field, water-depth scale macroturbulence is absent when bedform orientation and flow direction are opposed (flood phase). The former gentle-sloping ebb stoss-side now serves temporally as the hydraulic lee-side. High velocity and pressure gradients are absent preventing the generation of macroturbulence of this scale. Macroturbulent structures are responsible for the rapid upward transport and mixing of suspended sediment. In the Elbe Estuary, sediment within flow structures forms distinct clouds of sediment with cohesive properties. Confined clouds can be observed for up to one hour after their first occurrence, even after they merge to form larger structures. In the Knudedyb, such cohesive properties are absent. Sediment disperses more rapidly under the high flow velocities of the upper water column. During the flood tide, due to the absence of water-depth scale macroturbulence, suspended sediment transport is smaller in magnitude and believed to be associated with turbulence generated at the secondary bedforms. Transport distances are limited by the steep flow-facing flank of the primary bedforms. The complexity of these processes currently lacks adequate description and might result in under- or overestimation of regional sedimentary budgets. The variability of macroturbulence production due to flow unsteadiness causes a change in hydraulic roughness between tidal phases. Hydraulic roughness is also shown to vary for bedforms with equal height and length, based on the lee-slope. Our results demonstrate that there is a need for current generations of morphodynamic models to take into account the variability of hydraulic roughness induced by flow unsteadiness, the relative orientation of flow to bedforms, and lee-slope characteristics.