I mainly work on Holocene and late glacial palaeoclimate, occasionally delving back as far as the last glacial maximum, but the tools I use are of value for reconstructing climate over a longer period, albeit with greater uncertainty. The Pliocene, a generally warm period before the advent of large northern hemisphere ice sheets, is becoming an important target for palaeoclimate research, as its high atmospheric CO2 concentration make it an analogue for future climate.
High latitudes were warmer during the Pliocene — the ice sheets were much smaller, consequently sea-level was ~20m higher than modern, and coniferous forests grew on Greenland. In contrast, reconstructions of Pliocene tropical climate have suggested little warming.
The apparent stability of tropical climates is part of the evidence that led Prof Richard Lindzen (a man who “never met a negative feedback he didn’t like“) to propose that negative feedbacks cause low climate sensitivity in the tropics, a sort of thermostat that regulates temperature change. His suggested mechanism — known as the iris — that high-cloud cover contracts when tropical sea surface temperatures (SST) are high causing an increase in outgoing long-wave radiation, has not fared well in the literature. Now the evidence for tropical stability is looking shaky.
The best estimates of Pliocene SST came from the ratio of magnesium to calcium in the shells (tests) of planktonic foraminifera which are constrained by their algal symbionts to live in the photic zone near the sea surface. The Mg/Ca ratio is temperature dependent, but also depends on other factors, in particular the Mg/Ca ratio of sea-water. If the Mg/Ca ratio of sea-water has changed since the Pliocene, Mg/Ca based reconstructions will be biased.
This is what O’Brian et al (2014) argue in a recent paper in Nature Geosciences (see also the associated editorial). O’Brian et al compare reconstructions of tropical SST derived from Mg/Ca ratios of planktonic foraminifera from two biomarkers, alkenones, and TEX86. Both the biomarkers indicate tropical SST ~2 °C warmer than today, suggesting that Mg/Ca reconstructions are biased low. This has several implications such as the global estimate for Pliocene temperature anomaly is too low and that there is no need to invoke a thermostat to regulate topical temperatures.
When I read this paper, my first thought was that coral reefs would struggle to survive under the reconstructed temperatures. A quick check shows that the Caribbean was dominated by free-living corals rather than reef-building corals.
O’Brien, C.L. et al 2014. High sea surface temperatures in tropical warm pools during the Pliocene. Nature Geoscience 7, 606–611 (2014) doi:10.1038/ngeo2194