WARMER TROPICS DURING THE MID PLIOCENE? COMPARING TROPICAL AND SUBTROPICAL PROXY DATA TO CLIMATE MODEL OUTPUTS

Fossil assemblage-based reconstructions of sea-surface temperatures (SSTs) produced by the Pliocene Research Interpretations and Synoptic Mapping (PRISM) Group indicate that mid-Pliocene surface ocean temperatures were unchanged or slightly cooler than modern at the tropics and low latitudes and significantly warmer at higher latitudes, particularly in the North Atlantic. This change in the latitudinal pattern of SSTs has been attributed to enhanced meridional ocean heat transport generated by more vigorous surface ocean gyres and/or thermohaline circulation. A recent study (Haywood et al. 2005), which reconstructed SST in the subtropical and tropical Pacific Ocean using alkenone paleothermometry, combined with previously published alkenone SST estimates from the Atlantic Ocean, has provided the first evidence to suggest that tropical ocean temperatures during the mid Pliocene were, in contrast with PRISM estimates, several degrees warmer than present-day. This result has been supported by coupled ocean-atmosphere GCM studies that also predict significant warming at the tropics. If these estimates are accurate, they have profound implications for our understanding of the Pliocene world and what made it warm. In particular warming at all latitudes may indicate that higher concentrations of atmospheric CO2, rather than enhanced meridional ocean heat transport, were responsible for mid-Pliocene warmth.

A potential criticism of the warmer tropical SST scenario is that GCM simulations that have utilised PRISM SSTs, which display no warming at the tropics, have been found to be in good agreement with terrestrial climate proxies in tropical and subtropical regions. Here we present a new comparison between Pliocene proxy climate data and two experiments using the Hadley Centre GCM, one running with prescribed PRISM SSTs and the other with warmer than present SSTs in the tropics and subtropics. The comparison indicates that both SST scenarios produce climates that are in broad agreement with Pliocene terrestrial proxy data. The result underlines the limited ability of terrestrial proxy data to distinguish between two different SST scenarios and demonstrates that, in this particular case, they cannot be used to support or undermine either a scenario of unchanged tropical SST or warmer than present SSTs.