Figure 11

N. pachyderma (sin) percentage in Troll 3.1, plotted to age using the 14C dates in Figure 9, and compared with mean annual air temperature based on d18O of ice from core taken at Dye 3, Greenland. The age-to-depth relationship in Dye 3 is roughly linear prior to approximately 10,000 yr BP, as suggested by correlation to dated variations at Lake Gerzensee. The correlation between Troll 3.1 and Dye 3.1 after about 10,000 yr BP is based on annual layer counting and 14C-to-calendar year conversion, as discussed in Lehman and Keigwin (1992). As a result, the age model after 10,000 yr BP differs slightly from that used in Figure 1b of Hammer et al. (1986).


A series of abrupt changes in the poleward flow of warm surface waters has been documented in the northeastern North Atlantic during the last deglaciation (15,000 to 8,000 yr BP). They appear to correlate with evidence of atmospheric temperature changes over Greenland and downstream over Europe. Most important, they provide the first direct indication of the capacity of the conveyor circulation to shift abruptly in strength, as recorded by changes in temperature on the order of I C per decade across a range of more than 5°C at the surface of the northern North Atlantic. Numerical models of ocean circulation produce large changes in vertical and lateral ocean-heat fluxes in the North Atlantic as a result of changes in thermohaline circulation arising from fresh-water forcing at the ocean surface (Manabe and Stouffer, 1988; Maier-Reimer and Mikalojewicz, 1989; Stocker and Wright, 1991). Although the sensitivity of the different models to fresh-water forcing is debated and cannot yet be empirically evaluated, the rates and magnitude of the simulated changes now find support from high-resolution proxy records of ocean-circulation change. Owing to associated changes in heat transport, these non-linear aspects of circulation change must be considered a potent agent of climate variability.

Unlike the deglacial interval discussed in this paper, the last 8,000 years have been marked by relatively stable climate conditions in the North Atlantic region (Figure 11). But, as pointed out earlier by Broecker (1987), the conditions that conferred stability over this interval are not yet understood; thus, it cannot be assumed that such conditions will persist in the face of the gradual changes already forecast as a response to anthropogenic perturbation of the climate system.


I wish to thank Delia Oppo and Bill Curry for stimulating discussions and their reviews of an early draft of this manuscript. Julia Cole also provided useful criticism. Ernest Joynt III and Fritz Heidi assisted with graphics. This work was supported by NSF grants OCE-90819660 and OCE-91816259.

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