CHANGES IN THE VERTICAL STRUCTURE OF THE NORTH-ATLANTIC OCEAN BETWEEN GLACIAL AND MODERN TIMES

The distribution of planktic foraminiferal deltaO-18 over the North Atlantic Ocean has been interpreted by Duplessy et al. (1991, 1992) in terms of the distribution of surface water salinity during the last glacial maximum (LGM) and the Younger Dryas (YD). We present here the implications of this surface salinity distribution for changes in deep convection during these periods. Temperature/salinity/density diagrams of the water column have been constructed using the planktic and benthic foraminifera deltaO-18 values together with sea surface temperature estimates obtained by using micropaleontological transfer functions. The precision of these reconstructions is limited by the remaining uncertainties in the two basic assumptions upon which the methods are based: - the sea surface temperature reconstruction by CLIMAP are valid within the statistical uncertainties; - the water salinity/deltaO-18 relationship is well constrained for surface and intermediate ocean waters. A significant difference between the modem and the reconstructed glacial water structure is observed. During the LGM, the modern relatively warm and salty Norwegian Sea surface water and the North Atlantic Deep Water (NADW) were replaced by low salinity surface water in the northern Atlantic Ocean and the Norwegian Sea and cold deep water (0-1-degrees-C). However, surface water salinity in the central North Atlantic Ocean (52-54-degrees-N., 25-40-degrees-W.) was sufficiently high to permit deep water convection as a source of North Atlantic intermediate and deep waters. Open ocean convection at the northern limit of the subtropical surface waters (around 40-degrees-N.) may have contributed to the ventilation of the intermediate waters. The waters deeper than 2000 m were the result of mixing between poorly-oxygenated southern waters and well-ventilated North Atlantic deep waters. A similar reconstruction for the cold YD period indicates that the modern type of circulation, with warm and ventilated NADW, was largely in operation and that the hydrologic pattern was different from that prevailing during the LGM.