Workshop Abstracts

Previous Abstract | Back to Abstracts Page | Next Abstract

Internal Tide propagation sensitivity to model vertical coordinate and resolution

Le Boyer Arnaud, Cyril Lathuilière, Annick Pichon, Flavien Gouillon
SHOM
(Abstract received 04/27/2015 for session X)
ABSTRACT

The discretization of the physical space to apply numerical methods can lead to a misrepresentation of physical processes. This is particularly the case at Internal Tides (IT) scales where mixing is involved and Oceanic General Circulation Models (OGCM) do not have a sufficient spatial resolution to represent these processes. This work focuses on the role of the model vertical resolution on the IT propagation and on the modal representation of IT using idealized experiments. We show how the model vertical resolution and the choice of isopycnic or eulerian coordinate impact the IT propagation in an idealized context. The IW are generated over a continental slope and forced by an S2 barotropic tide in a two dimensions domain. Three different stratifications are used to allow for IT propagation: a two-layer ocean, a uniformly stratified ocean and a mixed layer above a uniformly stratified ocean. Step stratifications can be challenging for OGCM and any model vertical coordinate inducing abrupt changes in the water masses properties. These changes induced by mixing would imply a modification of the IT characteristics. Despite a diapycnal mixing associated with the fixed vertical coordinates, we show that model solutions converge for high model vertical resolution and for all the stratifications. Coarser model vertical resolutions show that there is a wave phase lag and/or amplitude modulation between the isopycnic and eulerian vertical coordinates. The actual model vertical resolutions are likely to induce biases in the internal wave field and their propagation thus inducing unrealistic vertical mixing rates in global/regional simulation. With this study, we aim to define a vertical model resolution threshold for OGCM below which, these biases could vanish or be negligible.

Previous Abstract | Back to Abstracts Page | Next Abstract


2015 LOM Workshop, Copenhagen, Denmark June 2nd - 4th, 2015