Figure 3

Difference in sea surface and surface air temperatures between the "conveyor on" and "conveyor off" states in the GFDL coupled atmosphere-ocean model (after Manabe and Stouffer, 1988; reprinted with permission of the American Meteorological Society). Although the sensitivity of the conveyor circulation in this model (and others) is debated, and cannot yet be empirically evaluated, there is little doubt that a weakening or elimination of the overturning conveyor circulation will lead to cooler sea and air temperatures in and downwind of the northern North Atlantic region.

dicted) at the cold, wintertime values reconstructed for the last glacial maximum (the LGM, which occurred about 18,000 yr BP) by CLIMAP (1981). As can be seen in Figure 4, a similar pattern of cooling (in this case, wintertime) was achieved. Although the presence of large ice sheets contributed to oceanic cooling during the LGM (Manabe and Broccoli, 1985), in both this and the coupled atmosphere/ocean experiment ice sheets were not introduced, and all other climate boundary conditions were those of today. Figures 3 and 4 provide an impression of the influence of a cold ("conveyor off") ocean alone.

The main purpose of the present paper is to assess the capacity of the ocean to undergo circulation changes of the magnitude and rapidity suggested by the climate proxy

Figure 4

Difference in wintertime surface air temperatures between the control run and one with sea surface temperatures specified at glacial (LGM) values reconstructed by CLIMAP (1981). The results are similar to those seen for air temperatures in the GFDL coupled model (Figure 3, lower panel), although these are for wintertime and the area of maximum temperature difference is located somewhat further south, in accordance with the difference between the location of maximum ocean cooling as reconstructed for the LGM and simulated for the "conveyor off" state in the upper panel of Figure 3.

record around the Atlantic region, through a review of the evidence for such changes in the ocean sediments themselves. Following up on the discussion brought forward in this introduction, I will begin by assessing the magnitude of heat transport into the circum-Atlantic region by the ocean's conveyor circulation system. I will then outline the approach used in reconstructing past changes in conveyor circulation and present results from studies of new Atlantic sediment records with sub-century-scale resolution and climate sensitivity sufficient to capture the amplitude and rates of climate change implied by ice-core studies and numerical models (Lehman and Keigwin, 1992; Koc-Karpuz and Jansen, 1992). These provide a test of the proposed conveyor model of circum-Atlantic climate change, as well as an assessment of the frequency and abruptness with which the conveyor circulation may vary.

HEAT TRANSPORT AND THE ATLANTIC CONVEYOR

The net transports of heat and salt in the world's oceans are characterized by a southward, trans-equatorial flow of relatively cold, salty North Atlantic Deep Water (NADW) formed by cooling and sinking in the northern North Atlantic, and an apparently compensatory, northward, trans-equatorial flow of warm, salty water in the Atlantic thermocline (Gordon, 1986; Broecker, 1991 a). Cooling and sinking occurs predominantly in the Labrador and the Nordic seas (Figure 5). The dense products of surface cooling in the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement