North Atlantic sea surface conditions during the Younger Dryas cold event J. C. DUPLESSY, L. D. LABEYRIE & M. PATERNE Centre des Faibles Radioactivités, Laboratoire mixte CNRS-CEA, 91198 Gif sur Yvette cedex, France Abstract: A map has been drawn of the oxygen isotopic composition (ó18O) ol planktonic fo¡aminilera which deposited their shells in isotopic equilibrium with ambient surface waters during the Younger Dryas cold event. These ó!80 values were derived from the ó18O values of Neogloboquadrina pachyderma (left coiling) and Globigerina bulloides, which lived in their respective optimum temperature ranges. This map reflects the main trends of the sea surlace temperature field during the Younger Dryas, but does not exhibit any ár8O anomaly associated with continental ice-sheet melting. This result rules out all the theories relating the Younger Dryas cooling to injection of continental ice-sheet meltwater with very low ór80 value into the North Atlantic Ocean. This reconstruction suggests that drifting of Arctic sea ice transported by the East Greenland Current may be a possible cause lor the strong cooling which occurred in the middle of the last deglaciation. The most impressive of the deglacial climatic excursions is the Bolling/Allerod Younger Dryas oscillation and its termination leading to warm Holocene conditions. The Younger Dryas took the deglaciated areas of Scandinavia and north- \ryestern Europe back into almost glacial con- ditions after these areas had enjoyed warm temperatures for about two millennia. At the Allerod-Younger Dryas transition, l l 000 radio- carbon years ago, temperatures dropped by 6-8'C in the northwestern Atlantic (Ruddiman & Mclntyre 1973; 1981; Duplessy et al. l98l; 1992) and even more on the nearby European continent (Van der Hammen 1957; Coope 1977; Mangerud 1987). The cooling which marked the end of the warm Allerod period developed within a few decades (Taylor et al. 1993) and the Younger Dryas has been often considered as a typical example of abrupt climatic change. The origin of this cold event is still largely unknown, despite the large number of scenarios for it which have been proposed. Mercer (1969), and later Ruddi- man & Mclntyre (1981), suggested that its onset marks a major influx of tabular icebergs from a disintegrating Arctic ice shelf. Boyle & Keigwin (1987) called on shifts in the pattern of orographic winds in response to the retreat of the ice sheets as a possible cause. Johnson & McClure (1976) remarked that the beginning of the Younger Dryas event coincides with the diversion of Laurentide meltwater from the Mississippi to the St Lawrence River. Later Rooth (1982) and Broecker e/ ø/. (1988) sug- gested that the impact of this diversion was to decrease the salinity of the northern North Atlantic, eventually to the point of turning off deep water production in the North Atlantic and stopping the Nordic Heat Pump - that is, the advection of warm surface waters to the high latitudes. Other potential points of freshwater injection into the North Atlantic have been proposed, noticeably iceberg discharge from Hudson Bay, the discharge of the Baltic Ice Lake into the Norwegian Sea or the waning of the Barents Sea ice sheet into the Arctic Ocean (Berger 1990). Because the study ofthe inception of the Younger Dryas event offers an opportu- nity to analyse the positive feedbacks which am- plify a minor perturbation applied to a climate system, it is important to determine the causes of this event if it is not a simple manifestation of the stochastic behaviour of the climate system. In this paper, we reconstruct the mean condi- tions prevailing at the surface of the North Atlantic Ocean during the Younger Dryas by mapping the oxygen isotopic composition of planktonic foraminifera which lived in their respective optimum temperature ranges. We then use this reconstruction to describe the circulation of the North Atlantic Ocean and test the various mechanisms which have been pro- posed as causes of this abrupt cooling. Strategy Ideally, a reconstruction of conditions prevailing at the sea surface should rest on both a sea surface temperature (SST) and a sea surface salinity (SSS) map, as has been shown for the last glacial maximum (CLIMAP l98l; Duplessy et al. l99l). The SST estimates are derived from Froz Andrews, J. T., Austin, W. E. N., Bergsten, H. & Jennings, A. E. (eds), 1996, Late Quaternary Palaeoceanography of the North Atlantic Margins, Geological Society Special Publication No. I I 1, pp.167 175. llgL