Equatorial Glacial Sea Surface Temperatures
Determining the past equatorial sea surface temperature for the last glacial maximum (LGM), about 21 ka ago, is critical to understand global climate sensitivity to greenhouse gases changes. Since CLIMAP in the early 1980's, the tropical estimates for the glacial sea surface temperature (SST) have been reestimated toward a larger LGM to present amplitude. CLIMAP estimates were based on microfossil transfer functions in some low sedimentation rate, carbonate dissolved and somewhat ill-dated cores (at least in the equatorial Pacific ocean).
In collaboration with Luc Beaufort, Edouard Bard (CEREGE) and Alan Mix (OSU), we investigated past changes in Sea Surface Temperatures of the Western Pacific Warm Pool during the last 180,000 years, by applying two independent methods: alkenones and transfer functions based on planktonic foraminifera in core MD97-2138, located north of Papua New Guinea. For the latter, to circumvent dissolution and biases linked to hydrological structure we developed a new regional transfer function (TROP-2), which we calibrated with intertropical core-tops excluding the Eastern Pacific area where anomalous high abundances of Neogloboquadrina dutertrei may result more from the High Nitrate Low Chlorophyll ecosystem than from temperature or thermocline depth gradients. Using this new transfer function, the core-top-LGM SST difference is about 0.1-1.3oC, consistent with CLIMAP estimates. However, the total Late Glacial-to-Holocene SST range here, based on both proxies, is substantially greater, about 2.2°C. Within the past ~30 ka, the coldest conditions occurred near 18 cal. ka BP, younger from the L.G.M. s.s, and the warmest interval occurred between 5-9 ka, in the early Holocene. The stage 6 - stage 5 difference is about 2°C using TROP-2 and only 0.9°C using the alkenones method, and overall the time series from the two different proxies are different. Nevertheless, we conclude that in the WPWP, a strategy of calibrating regional, ecosystem-relevant transfer functions yields better estimates of SST change, and based on this proxy, peak warm and cold conditions here were not coincident with modern and LGM time intervals. The total range of SST variability in the region is 2-3oC in late Pleistocene time.
In Rutgers, I generated the Mg/Ca SST record from this very same core, allowing to compare transfer functions/alkenones and Mg/Ca estimates in the core of the WPWP.
