Study of the energetics of the climate system using a coupled model: atmosphere, surface ocean, sea ice.

Summary The aim of this thesis is the exploitation of a coupled model of atmosphere, surface ocean, sea ice, for the study of the interactions between the surface ocean and the atmosphere. The coupled model, which has never been tested before, is described in a first part (chapter 2). A particular interest is given to the term representing the heat transport by the oceans: its impact on the thermodynamics of the ocean and on the general representation of is analyzed (chapter 3). Chapter 4 is devoted to the analysis of the model results at the seasonal scale. The behavior of the model is satisfactory and stable in its representation of the annual cycle. But there are systematic anomalies. Thanks to various sensitivity studies, the origin of these anomalies is better identified. But the interest of these experiments is wider since they allow to reveal some preponderant factors of the coupling (ocean-radiation interaction, surface wind friction) and to better understand the interactions between the surface ocean (or the sea ice) and the atmosphere. Finally, the last chapter (chapter 5) analyzes the climate drift experienced by our model. We obtain original results, in the sense that we highlight a slow atmospheric component that destabilizes the climate system during the first five years of simulation. In particular, we address the problem of the representation of the radiative properties of high clouds (cirrus) in the equatorial region: these cloud-radiation interactions are crucial phenomena but nevertheless little known, for the sensitivity of climate models to disturbances.