OS23B-1199:
Turbulent Fluxes of Suspended Sediment from Coupled Acoustic Doppler Current Profilers

Tuesday, 16 December 2014
Maximiliano Sassi1, Ton Hoitink2 and Bart Vermeulen2, (1)Royal Netherlands Institute for Sea Research, Den Burg, 1790, Netherlands, (2)Wageningen University, Wageningen, Netherlands
Abstract:
Turbulent diffusion is a cornerstone in geophysical fluid mechanics, controlling the exchange of momentum, heat and mass in surface flows occurring in the atmosphere, in rivers and in the ocean. In fluvial and coastal systems, modeling turbulent diffusion of momentum and suspended sediment requires knowledge about turbulent diffusivities, which is generally derived from parameterizations based on laboratory experiments. Field determinations of momentum and sediment diffusivities are cumbersome, requiring an instrumental array to simultaneously sample turbulence and mean flow quantities in time and in space. Recently, a new technique to analyze geophysical surface flow turbulence was introduced, appropriate for large scale systems, based on coupling of acoustic Doppler current profilers (ADCPs). Here, we extend this approach to obtain collocated profiles of both the Reynolds stress tensor and eddy covariance fluxes, to derive vertical profiles of turbulent momentum and sediment diffusivity in a tidal river. Shear and normal stresses are obtained by combining the variances in radial velocities measured by the ADCP beams. The covariances between radial velocities and calibrated acoustic backscatter allow to determine the three Cartesian components of the turbulent flux of suspended sediment. The main advantage of this new approach is that flow velocity and sediment concentration measurements are exactly collocated, and that it allows to profile over longer ranges, in comparison to existing techniques. Results show that vertical profiles of the inverse turbulent Prandtl-Schmidt number is coherent with corresponding profiles of the sediment diffusivity, rather than with profiles of the eddy viscosity. This implies modelling suspended sediment dynamics requires knowledge about the sediment diffusivity, as the Prandtl-Schmidt number cannot be estimated from the eddy viscosity alone.