| description abstract | The zonally averaged, latitude-depth ocean model, developed in Part I, is extended to a two-basin system representing the Atlantic and Pacific. Steady states are calculated under two different surface boundary conditions to study a possible global thermohaline circulation linking the Pacific and the Atlantic. If the surface temperature and salinity profiles are identical functions of latitude in both the Pacific and the Atlantic, the steady state under restoring boundary conditions consists, in spite of the different basin extensions, of a two-cell structure in each basin. Upon switching to mixed boundary conditions the state is unstable, and a one-cell circulation with downwelling at high northern latitudes develops in both basins. For a realistic surface salinity profile that is fresher in the Pacific, the steady state under restoring boundary conditions is completely different. It exhibits a global thermohaline circulation with strong interocean man exchange. Deep water is formed primarily in the North Atlantic, from which a deep flow spreads into the Pacific, where it upwells. This state is stable under mixed boundary conditions. The meridional heat flux is to the south in the Pacific and to the north in the Atlantic with maximum values of order 0.7 PW (1 PW = 1015 W). Both temperature and salinity structures of the steady state compare favorably with the observed zonal averages. The model also demonstrates that this ?conveyor belt? circulation is maintained by net evaporation in the Atlantic and net precipitation in the Pacific. Two deglaciation experiments simulating the termination of the last ice age are performed by applying a freshwater flux anomaly of 0.12 Sv and 0.06 Sv (1 Sv ≡ 106 m6 s?1) in the North Atlantic. The strong anomaly shuts off the interocean exchange, and deep water is formed subsequently only in the Southern Ocean. For the small anomaly the interocean circulation weakens but remains in operation. When the anomalous flux is switched off, the final equilibrium state for the first experiment is the mode with no interocean exchange, while in the second experiment the state returns to the original conveyor belt. The global thermohaline circulation thus exhibits more than one stable equilibrium under realistic surface forcing. | |
