Modeling convective transistions in the presence of sea-ice

A substantial part of the variability in the thermohaline circulation on decadal and longer timescales as found in ocean models is strongly related to the deep-water formation (DWF) process. Many of these studies, however, neglected the role of sea ice. Nevertheless, it is known that sea ice strongly influences the heat and salt budget of the polar ocean. In this study, DWF is studied in the presence of sea ice. This is done within the context of a simple geostrophic 6-level ocean model with a rectangular basin coupled to a thermodynamical sea ice model. The model is forced by mixed boundary conditions (MBCs). In general, the area of DWF changes rather discontinuously due to the self-sustaining character of convection under MBCS. For long times the DWF area remains basically unchanged, then, suddenly. the extension and/or location of the convective area dramatically changes in a few years time. Also, multiple equilibria are observed under MBCS. In this study, the main emphasis will be put on a case with rapid growth of the convective area. The authors start from a stationary state with a relatively small convective area and a large ice covered area. Then a rapid growth of the convective area is triggered by introducing a local salinity anomaly. The rapid growth is accompanied by a rapid increase in heat loss to the atmosphere. At the same time the ice cover retreats. The release of freshwater by melting is a negative feedback, which tends to stabilize the stratification and hence interrupt convection and prevent further ablation of the ice cover. This negative feedback is locally very effective; however, it can (but not necessarily) be compensated by the horizontal advection and diffusion of saline water beneath the ice. In the experiments the source of saline water is provided by convection south of the ice cover. The growth of the convective area, the retreat of the ice cover and, hence, the atmospheric heat flux strongly depend on the strength of this salt source. Also, a method is -presented to estimate the importance of convection for the simulated changes in the ocean temperature and salinity fields and atmospheric heat flux.