The morphodynamics of fluvial sand dunes in the River Rhine, near Mainz, Germany. II. Hydrodynamics and sediment transport

The dynamics of large isolated sand dunes moving across a gravel lag layer were studied in a supply‐limited reach of the River Rhine. During daylong surveys, suspended sediment concentration, bedload transport rate, water depth, flow velocity, turbulence intensity, near‐bed shear stress and water temperature were recorded over individual isolated dunes. This paper considers the hydrodynamic environment and sediment transport over the dunes. A companion paper details the sedimentology and morphology of the dunes. Flow over the flat gravel lag upstream of large dunes is more uniform than that over dunes, and gravels are rarely entrained by in‐bank discharges. Unsteady and non‐logarithmic velocity profiles are common within the boundary layer above the stoss side of large dunes, and the near‐bed flow demonstrates evidence of large‐scale, coherent low‐frequency flow structures; these may reflect stacked sequences of separated boundary layers generated by secondary dunes. However, the low‐amplitude morphology of large dunes does not affect the statistical properties of turbulence production over the stoss sides. Bed roughness and near‐bed shear stress commonly increase steadily over the stoss of dunes, but may decrease near the crest, especially where a crestal platform exists that is devoid of secondary bedforms. Bed roughness scales with the physical size of bed roughness elements. However, variability in roughness lengths is large, owing to the composite nature of bed roughness. Bedload transport over stoss slopes is spatially variable, but often shows an increase with increasing bed shear stress over the stoss. Well‐formed wakes only develop downstream of lee slopes, which are close to the angle of repose; otherwise, separation is weak, and suspension and settling of fine sediments is of little consequence to dune evolution. Wake flow is characterized by turbulence production one order of magnitude greater than over the stoss side, which may be related to vortex shedding from the dune. However, wake current speeds are extremely low and variable in direction; reverse flow is not sustained, and no retrogressive bedforms occur in the lee of large dunes. Near‐bed shear stress and bed roughness are usually low within the wake, reflecting low current speeds. Weak wake‐flow reattachment occurs at a variable distance downstream, up to several times the duneform height. A two‐dimensional numerical model for flow over dune topography, calibrated using average parameter values, provided a reasonable description of the flow upstream, and over the backs of large dunes as far as the crestal region. Wake flow could not be modelled. However, the temporal and spatial complexity of natural three‐dimensional flow over the dunes resulted in variance in parameter estimates; this variance precluded modelling of flow and bedload transport. Consequently, it was not possible to model dune evolution in a deterministic sense.