SEARCH

SEARCH BY CITATION

This paper reports the results from a series of mechanistic studies on the global ocean circulation by assimilating a set of observed surface thermohaline data into a low-resolution Geophysical Fluid Dynamics Laboratory (GFDL) global ocean general circulation model (OGCM). In most of the experiments the surface wind stress is set to zero. The surface thermohaline forcing features a strong relaxation of surface temperature and salinity towards observed values. The ocean circulation resulting from the assimilation of surface thermohaline fields depends heavily upon the various parameterizations of vertical mixing processes. Consistent with previous studies, a greater vertical diffusivity results in a greater meridional overturning stream function and a greater meridional heat transport. The vertical structure of stratification is particularly well reproduced if the vertical diffusivity is set to be buoyancy dependent. Even in the absence of explicit wind forcing, the major circulation features are reproduced. In particular, water mass properties are realistic, and the Antarctic Circumpolar Current (ACC) and the Gulf Stream have significant barotropic components (especially the ACC). The feature of realistic water masses can be attributed to the reproduction of convection-dominated heat and freshwater fluxes in the polar and subpolar regions, where these water masses are formed. In the absence of explicit wind, the barotropic transports are driven by bottom pressure torque, which is associated with variable bottom topography and upslope/downslope bottom flows. Convective processes play an important role in redistributing the water mass, generating the barotropic transport, and maintaining the thermohaline structure. In the absence of convective adjustment, the global circulation is characterized by a thermocline reversal in polar and subpolar regions. Deep convection also leads to larger bottom flows, and hence indirectly play a role in generating the barotropic transports. In the low-resolution model, the barotropic ACC is broad; this seals the Indian Ocean from the Atlantic Ocean, and prevents both the barotropic and baroclinic interbasin exchange between the Indian and the Atlantic oceans. In this situation, the North Atlantic Deep Water (NADW) outflow is compensated by the cool and fresh South Pacific water from the east of Drake Passage (cool water route). When convection is used as a device to suppress ACC, and to ensure that the NADW is the only deep water source, the NADW formation is seen to cause a large Indonesian throughflow. The teleconnection between the throughflow and the NADW is via the Agulhas leakage water. In the absence of ACC, the NADW outflow is compensated by the warm and salty Indian water through the leakage (warm water route). The results show that the realization of warm water route depends on the structure of the modeled ACC.