Geographical distribution of the first empirical orthogonal function (EOF) of the surface air temperature changes from Figure 1. The percentage of variance represented by this EOF is given for the different levels. (From Barnett et al., 1992; reprinted with permission of the American Geophysical Union.)

of a degree Celsius, although this has not been proven satisfactorily. At the very least, the model results imply that oscillations of this magnitude may be expected to occur without changes in external forcing.


The El Niño/Southern Oscillation (ENSO) phenomenon represents a well-studied example of atmosphere-ocean oscillations. The ability of this type of localized forcing to alter global patterns of temperature and precipitation indicates that it could have an effect on globally averaged climate (Ropelewski and Halpert, 1987), and the El Niño occurrence itself is generally associated with a global mean temperature change of about 0.1 to 0.2°C. The prevalence of El Niños during the past several decades and the lack of pronounced La Niñas undoubtedly have influenced global temperatures (Trenberth, 1990). A key question is whether the frequency of El Niño occurrences fluctuates with time. The results of analysis are equivocal; for example, there are some indications that fewer events took place during the period from the 1920s to the 1950s (Cooper et al., 1989), while others claim that 1925 to 1932 was a period of unusual activity (Quinn et al., 1987). Cole (1992) has investigated the variance of ENSO activity as reflected in coral records for the past century and found that the dominant frequencies have changed significantly. This observation serves to emphasize that ocean-atmosphere oscillations may themselves be the result of external forcing, a subject we will return to below.

Changes in the ocean heat transport, associated with changes in the wind-driven circulation or the large-scale thermohaline circulation, represent an obvious method of inducing climate variability (see, e.g., Weyl, 1968; Watts, 1985; Rind and Chandler, 1991). The "Great Salinity Anomaly" in the North Atlantic was apparently associated with colder temperatures during the late 1960s and 1970s and reduced deep-water formation (Brewer et al., 1983; Lazier, 1980). The Younger Dryas cooling event (from ca. 11,000 to 10,000 years B.P.) is a much-cited example of North Atlantic Deep Water (NADW) reduction that led to cooling, at least in the region of the North Atlantic (e.g., Broecker et al., 1985). The same type of process may have played

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement