trends bear a strong resemblance to the Northern Hemisphere’s leading pattern of variability, the Arctic Oscillation/North Atlantic Oscillation (AO/ NAO) (Thompson and Wallace, 1998; Wallace, 2000). That is, the observed trends in the wintertime temperatures and precipitation appear to be largely regional expressions of the positive trend of the AO/NAO. There are indications of analogous trends in the Southern Hemisphere, in association with its annular mode (Thompson et al., 2000).
Several ideas have been put forth concerning the cause of those trends. Shindell et al. (1999, 2001) have proposed that they might be occurring in response to increasing concentrations of greenhouse gases, which warm the lower atmosphere but cool the stratosphere. Indeed, similar trends have been simulated in an atmospheric model driven by increasing concentrations of greenhouse gases. It has also been suggested that stratospheric ozone depletion (Volodin and Galin, 1999) and trends in sea-surface temperatures (Rodwell et al., 1999; Hoerling et al., 2001) might be capable of inducing similar trends in the AO/NAO. Because favored patterns of variability tend to be highly sensitive to external forcing, it is indeed quite plausible that they can be changed in a number of ways. Fyfe et al. (1999) give a contrasting view, that the AO/NAO did not appear to be altered substantially by greenhouse forcing, yet its Southern Hemisphere counterpart, the annular “AAO” mode, was indeed strongly amplified.
Whatever its cause, the observed trend in the AO/NAO is influencing other components of the earth system. The trend has altered the distribution of deep convection over the North Atlantic (Lilly et al., 1999, Dickson et al., 1996), caused glaciers in Norway to advance (Hurrell, 1995), and contributed to the thinning of the stratospheric ozone layer over middle and high latitudes of the Northern Hemisphere during winter and early spring (Thompson et al., 2000). Interactions with the oceanic thermohaline circulation (THC) have the potential to exert positive feedback on the AO/NAO itself, increasing its variability on decadal and longer scales. Owing to such feedback, it is conceivable that shifts in the AO/NAO contributed to some of the large climatic swings in the paleoclimate record.
If trends observed in the AO/NAO during the last 30 years were to continue into the twenty-first century, they could have important societal impacts. For example, the recovery of the stratospheric ozone layer from human-induced depletion during the twentieth century could be delayed by several decades (Shindell et al., 1999).