| description abstract | This numerical experiment aims to investigate the dynamical effects of the vertical convection of waters, occurring in regions where the cooling of the upper ocean is due to periodical, local, rough winter atmospheric conditions. The simulation is performed with a 2-layer quasi-geostrophic model, which configuration yields a small internal radius of deformation (10 km), and a high spatial resolution (5 km). The sinking of waters is parametrized by a conversion, in a limited area, of some of the upper-layer fluid into lower-layer fluid. This forcing is continuously active for the first 3 months of the year and is zero otherwise. This cycle is repeated for several years. The model response results in an elevation of the interface (a dome), which generates a geostrophic circulation first limited to the region where the forcing is active. The dome grows and reaches a critical amplitude where it becomes baroclinically unstable, and breaks into eddies of smaller size, in a way very comparable to what happens in experiments done in a rotating tank. After 15 forcing cycles, a general circulation is established to the whole basin, with mean and eddy currents in statistical equilibrium. Eddy-driven mean flows are then superposed on, and interact with, the geostrophic circulation due to the mean signature of the dome. The eddy currents are mostly generated by the baroclinic instability of the dome, but also by the barotropic (horizontal shear) instability of the mean currents. The oceanic relevance of the experiment is discussed in the context of the Western Mediterranean Sea, with some success for the upper-layer circulation. | |
