Was Antarctica kept warm by subtropical waters in the Eocene? Part 2: Climate Model results

Opening of the Tasmanian Gateway (TG) during the Eocene/Oligocene transition has long been invoked as the causal mechanism for the global climate shift from the "Hothouse" world of the early Cenozoic to the "Icehouse" world of the past 35 million years. ODP Leg 189 was designed to test the hypothesis that Antarctic cryospheric evolution resulted from the thermal isolation of Antarctica, caused by the opening of the TG. The proposed mechanism specifically being investigated was the cessation of poleward penetration of the heat-transporting, warm East Australian Current as the cause of the climate cooling. Five sites (1168-1172) were drilled to document paleoceanographic and paleoclimatic changes associated with the opening of the TG as Australia moved northward from Antarctica during the early Cenozoic. Demonstrating that this climatic transformation occurred synchronously with the rapid opening of TG is one of the major results of ODP Leg 189. To test the TG hypothesis we compare model fully coupled (ocean-atmosphere-sea ice-land) climate model results for Late Eocene conditions with proxy data and isotopic climate reconstructions. In this part, we demonstrate that (1) the East Australia Current (EAC) never extended far poleward, bending eastward around the northern edge of New Zealand instead, (2) even if this current had extended to Antarctica it is unlikely that turning the current off would have initiated glaciation, (3) that the proxy data agree with the paleocurrent predictions of the model, and (4) that the geologic record is not consistent with changes in the EAC as being the control on Antarctic glaciation. One alternative hypothesis --that the climate changes were driven by changes in greenhouse gas concentrations--is explored and appears consistent with the proxy data record. The model-predicted climate sensitivity and the proxy record of climate change allow us to estimate the most likely concentrations of carbon dioxide (or its radiative equivalent) through the Paleogene. The model is most consistent with values of pCO2 greater than 1120ppm in the early Eocene, values ~1120ppm in the middle Eocene, values ~560ppm in the late Eocene, and near modern values in the Oligocene.