Ocean dynamics is the expression of physical processes in continuous interaction, which cover spatial scales ranging from kms to planetary scale, and temporal scales ranging from seconds to century. In this context, we study the mechanisms of multiscale oceanic variability, with a particular interest in fine-scale structures (intense eddies, fronts, filaments), their predictability, their mutual interactions, the associated energy transfers and their role in climate variability.
Our recent research focuses on the explicit resolution of fine scales in numerical models based on the community code NEMO. This is achieved by carrying out realistic simulations at very high resolution (1/12°-1/60°) including in some cases the coupling with the atmosphere and its boundary layer, the ice pack, the ice caps and marine biogeochemistry (see axis 2), and by comparing the results with observations and process studies.
These simulations produce simulated data sets used for the preparation of the SWOT mission (to be launched in 2023 to observe for the first time ocean dynamics at scales resolved down to ~ 10 km) and the exploitation of altimetry data, once the satellite is flying. Our objectives are to analyze the observed signals and to study energy exchanges between scales (spectra, structure functions) as well as wave-current interactions. These data allow us to develop deterministic or stochastic parameterizations of sub-grid scale effects (kinetic energy dissipation, role of fine scales in convection, etc.) in intermediate resolution models, in particular via machine learning methods (sparse regression, deep learning) fed by simulated data. The predictability of mesoscale and sub-mesoscale flows is studied by means of very high resolution regional ensemble simulations, with a view to contributing to the design of future space missions (see axis 3). This activity contributes to the long-term development of the NEMO code used as the dynamic core of operational forecast suites of the marine Copernicus CMEMS service, and in the Earth System Models used for climate projections.