Laboratoire des Écoulements Géophysiques et Industriels




Nos tutelles

CNRS

Nos partenaires

Rechercher


Accueil > Actualités > Soutenances de doctorat > Doctorats 2014

Jeudi 20 février 2014, soutenance de thèse de Cyrille Quam AKUETEVI - 11h30, Amphithéâtre K118, site Bergès ENSE3

Dynamique des courants de bords ouest turbulents de faibles latitudes, étude numérique / Dynamics of turbulent western boundary currents at low latitude, a numerical study

Encadrant

- M Bernard Barnier.
Directeur de recherche CNRS, Co-directeur de thèse
- M Achim Wirth.
Chargé de Recherche CNRS, directeur de thèse

Résumé

Strong western boundary currents are a dominant feature of the worlds oceans, also at low latitudes. It is called the North Brazil Current in the Atlantic and the Somali Current in the Indian ocean. They exhibit a turbulent dynamics and their region is the source of strong kinetic energy production and internal
variability of the worlds oceans. Several places exists where the western boundary currents retroflect (i.e separation from the coast) and generate coherent structures as anticyclonic, eddies and dipoles.

The dynamics of oceanic western boundary current has so far not been extensively study in the viewpoint of turbulent boundary-layer theory.

This doctoral thesis is directed towards the study of the turbulent boundary-layer throughout the determination of the turbulent structures, fluxes, balances in the low latitudes turbulent western boundary currents.

The approach is to use a fine resolution reduced-gravity shallow water model to understand the turbulent boundary-layer processes in the first part of this thesis and then apply these results to the Ocean General Circulation Model NEMO in the DRAKKAR configuration in the second part of this thesis. The case of Somali Current is considered for this application.

The dynamics of low latitude turbulent western currents, subject to two different types of idealized wind forcing, Monsoon Wind and Trade Wind, is considered using numerical results from integrations of a dedicated fine resolution (2.5km) reduced gravity shallow-water model. For viscosity values
of 1000m2 s−1 and above, the boundary layer dynamics compares well to the analytical solutions of the Munk-layer and the inertial-layer, derived from quasi-geostrophic theory. Modifications due to variations in the layer thickness (vortex stretching) are only important a few kilometers from the boundary. When the viscosity is reduced the boundary layer becomes turbulent and coherent structures in form of anti-cyclonic eddies, bursts (violent detachments of the viscous sub-layer) and dipoles appear. Three distinct
boundary layers emerge, the viscous sub-layer, the advective boundary-layer and the extended boundary-layer. The first is characterized by a dominant vorticity balance between the viscous transport and the advective transport of vorticity. The second by a balance between the advection of planetary vorticity and the advective transport of relative vorticity. The extended boundary layer is the area to which turbulent motion from the boundary extends.

The scaling of the three boundary layer thicknesses with viscosity is evaluated. This scaling is used to revisit the validity of the laminar Munk-layer theory for the high Reynolds number turbulent western boundary currents. It is shown that small westward velocities have a stabilizing effect (inertial effect)
on the boundary current and alter the vorticity balance near the boundary. This inertial effect must be taken account in the parameterization of a turbulent western boundary current. The consequences for the grid resolution, refinement and the complexity of numerically modelling western boundary currents are discussed.

A pragmatic approach to determine the eddy viscosity for coarse resolution numerical models base on the Munk’s formula, the scale of the anticyclones and Prandtl’s formula is proposed.