OCEAN GENERAL CIRCULATION MODEL SENSITIVITY EXPERIMENT WITH AN OPEN CENTRAL AMERICAN ISTHMUS

We have tested the sensitivity of an ocean general circulation model to changes in solid boundary conditions by opening the central American isthmus in order to simulate possible circulation patterns existing prior to about 3.0-4.0 Ma. The model was driven by present winds, atmospheric temperatures, and moisture fluxes, with the only difference between control and perturbed runs being the open isthmus. Significant changes result for the perturbed case, almost all of which can be traced to changes in surface salinity distribution. With an open isthmus, lower-salinity waters from the Pacific dilute North Atlantic surface waters by >1.0 parts per thousand. T he lower surface salinities cause the thermohaline cell in the North Atlantic to collapse, with North Atlantic Deep Water (NADW) production reduced to near zero. This response greatly weakens poleward ocean heat transport in the North Atlantic. The "no-NADW" result appears to be insensitive to initial conditions and represents a circulation mode different than a "low-NADW" mode found in the same ocean model using present geographic forcing. There is some agreement between model predictions and geologic data, which indicate a significantly different deep-water circulation pattern prior to similar to 2.4 Ma. However, observations suggests that the time of major transition in circulation regimes occurs earlier than the final closure of the isthmus. This discrepancy may reflect the fact that near-closure of the isthmus by 7-10 Ma could have been sufficient to alter the circulation. As ocean heat transport is a key source of warmth to the high latitudes of the North Atlantic, some compensating factor (CO2?) may have been required to restore warmth to that region prior to similar to 3.0- 4.0 Ma.