École Nationale Supérieure de
Techniques Avancées

Abstract

Studying the ocean mechanism turns out to be primordial for better climate forecasts given its major role in climatic regulation. Its energetic budget seems to verify the mechanic energy conservation at each cycle thanks to what we call the mixing. The latter is essential for completing the ocean model and researchers are trying to parameterize it with dimensionless numbers. In this study we use in-situ measures taken in the Caribbean Sea to assess which of the parameterizations proposed in the literature of this coefficient is the most reliable to the data series. We realize some data processing to operate the raw data and to assess the measured mixing coefficient. The results that we have progressively obtained make us question a numerous times our processing method, as for example the one for determining the mean flow, the spectrum integration for the energy with the integration bounds fixed by us and in particular for distinguish signal from noise. We analyze the parameterizations of Shih and al. using the buoyancy Renolds number $Re_b$ and the one of Maffioli, Brethouwer et Lindborg with the perpendicular Froude number $Fr_\perp$, that we confront to processed data. These parameterizations have a two orders of magnitude difference, in the greater ones for Shin and al. and in the smaller ones for Maffioli, Brethouwer et Lindborg, which seems to better follow the evolution of measured $\Gamma$. A study of the dependence of measured $\Gamma$ against dimensionless numbers shows that the law power of Shih and al. is well-overlaid with measurements. The gap observed before is certainly due to the front coefficient. The processing method for the Froude number might be deficient and has to be reviewed.