INVITED: THE INFLUENCE OF TROPICAL OCEAN CONDITIONS ON PLIOCENE WARMTH AND THE ONSET OF NORTHERN HEMISPHERE GLACIATION

During the Cenozoic, a gradual cooling culminated in the onset Northern Hemisphere glaciation approximately 3 million years ago. Superimposed on this general cooling trend were periods of relative warmth, such as the early and mid Pliocene (5 to 3 million yeas ago). The cause of these relative warm periods remains largely unknown. For example, why was the early and mid Pliocene characterised by relatively warm conditions when continental positions were approximately the same as now and when the evidence for higher concentrations of atmospheric greenhouse gasses is equivocal? Philander and Federov (2003) suggested that the warm Pliocene might have exhibited significant differences in the thermal structure of the oceans. In particular the prevailing warm climate may have deepened the thermocline leading to perennial El Nino conditions. As climate cooled the thermal structure changed (approximately three million years ago) leading to the first appearance of cold surface waters in regions that today have intense oceanic upwelling (the eastern equatorial Pacific and the coastal zones of southwestern Africa and California). This is important as it may have brought into play feedbacks involving ocean-atmosphere interactions of the type associated with El Nino and also mechanisms by which high-latitude surface conditions can influence the depth of the tropical thermocline and vice versa.

We present results from a series of experiments using the Hadley Centre GCM that examine the role of oceans and ocean structure on Pliocene warmth and the onset of Northern Hemisphere glaciation. Firstly, we present results from a fully coupled ocean-atmosphere GCM simulation for the mid Pliocene that examines the thermal structure of tropical oceans and longitudinal temperature gradients in the Pacific Ocean, with the objective of identifying if the mid Pliocene was indeed a period characterised by perennial El Nino conditions. Secondly, we present results from GCM sensitivity tests in which the gradient of SSTs in the tropical Pacific are altered so that the influence of tropical ocean conditions on high latitude precipitation and temperatures estimates, and therefore indirectly Northern Hemisphere glaciation, can be quantified.