Mixing in Rhine outflow

Sustainable engineering of coastal systems in Regions of Freshwater Influence.

Aim: To obtain a better understanding of turbulent mixing in stratified coastal flows.

A Region of Freshwater Influence (ROFI) is region where fresh water from a river enters the coastal sea. The ROFI is dominated by a competition between freshwater runoff which tries to build up a stable stratification, and mixing of water column by tides and waves.

Location of the Rhine Mouth (source: Thesis Gerben J. de Boer, 2009)

The Rhine river, the second largest river in Europe, has such a ROFI. The huge input of freshwater creates a salinity deficit and as a result, the region surrounding the river-mouth can be stratified even when the rest of the shelf sea is well-mixed. However, the tide itself tries to mix the stratified water-column. Depending on the relative strength of both effects both well-mixed and stratified conditions may occur, usually in an oscillatory manner known as tidal straining. In the current research we try to understand turbulent mixing processes under those complex conditions. Final objective is to develop simplified process descriptions (parameterisation) that can be used in practical modelling of estuarine, coastal waters.

The Rhine ROFI during well-mixed conditions (top) and during stratified conditions (bottom), (source: Simpson et al. 1990). During well mixed conditions, the iso-density planes are parallel to the coast (homogenous in the vertical direction). During stratified conditions, vertical density gradients are observed and the iso-density planes are tilted.

 The previous oscillations are of crucial importance for issues such as water quality, coastal management and understanding the consequences of engineering works in the coastal zone. They are however quite difficult to model. The turbulence closures that are usually used in classic coastal models are often based on the k-ε model. These closure schemes give more satisfying results for mixed conditions than for strongly stratified conditions because they have difficulties to distinguish sharp density transitions.

In order to deal with this problem, a computer code 'LES-Coast' will be used to understand tidal straining processes. This code is based on the Large Eddy Simulation approach in which the large scales of motion of the spatially filtered Navier-Stokes equations are solved. Also Direct Numerical Simulation (in which all the scales of motion are solved) will be used for Lagrangian particle tracking.

These new approaches for coastal modeling will help to achieve the overarching goal of the research project which is a better understanding of turbulence in stratified flows and the development of improved turbulence closures to account for anisotropy. At the end, we hope to develop improved turbulence closures in order to bridge the gap between idealised high resolution modeling that resolves the small scales of turbulence and coastal modelling that uses traditional turbulence closures.

Group members

Steven Kaptein, Matias Duran Matute, Herman Clercx