of the tropical portion of the Hadley circulation by 5 to 10 percent. Because reductions in Hadley-cell intensity are a possible cause of ENSO events, this result is qualitatively consistent with the Enfield and Cid (1991) correlations of strong El Niños and low solar activity. As will be discussed below, volcanic aerosol could also play this role, perhaps more efficiently.
An additional problem is the wavelength distribution of the radiation fluctuations. During the last solar cycle, 20 percent of the observed solar irradiance variations have been in the ultraviolet range, having wavelengths short of 300 nm, with magnitudes as presented in Figure 6 (Lean, 1987; Cebula et al., 1992). This radiation is all absorbed in the middle and upper atmosphere, above the tropopause. In a
GCM experiment with an increase in total solar radiation of 0.5 percent, but with all wavelengths restricted to less than 300 nm, we found that there was no increase in solar heating at levels below the tropopause. Thus, the ability of solar irradiance variations to replicate observed climate variability will depend on the frequency of the radiation affected, in addition to its magnitude. A change in UV radiation absorbed in the stratosphere might conceivably be able to influence climate if the impact on tropospheric dynamics described by van Loon and Labitzke (1988) proves to be real.
Previous studies (Lamb, 1970; Bray, 1974; Porter, 1981; Grove, 1988) have suggested that the trends in global climate observed during the years A.D. 1000 to 1800 could largely be explained by variations in volcanic aerosol forcing. Porter (1986) related acidity records in the Greenland ice cores to alpine glacial fluctuations in the Northern Hemisphere and found a reasonably good correlation, with increased acidity during periods of glacial advance. It is recognized that high-latitude ice cores may overemphasize mid- and high-latitude volcanic eruptions (Bradley and Jones, 1992b), which should be less effective in altering global climate than low-latitude eruptions. Furthermore, cold periods with low precipitation may have higher acidity concentrations because there has been less dilution from snowfall. Nevertheless, as the recent estimates of a 0.5°C cooling associated with Mt. Pinatubo (Hansen et al., 1992) suggest, volcanic aerosols (primarily sulfuric acid) have the capability to significantly alter the global climate (Rampino and Self, 1984).
What magnitude and frequency of volcanic eruptions are required to produce the cooling estimated for the Little Ice Age? We can use several recent GCM experiments with Mt. Pinatubo (Hansen et al., 1992; Rind et al., 1992; Pollack et al., 1993) to provide estimates. Aerosol radiative forcing depends on geographical distribution, optical depth, size distribution, composition, and altitude; however, as shown by Lacis et al. (1992), the forcing of tropospheric climate is primarily a function of aerosol-column optical depth, assuming a global distribution. In a study of transient volcanic aerosol, a volcanic eruption with a peak global optical thickness of 0.18 (on the order of what has been observed for Mt. Pinatubo) and an e-folding time of 12 months (so that in 3.5 years 97 percent of the aerosols will have been removed) produced a peak cooling of 0.5°C in year 2, and a three-year average cooling of 0.3°C (Hansen et al., 1992). In another experiment, with one-half the optical thickness and the same residence time, the cooling was reduced by a factor of 2. If this linearity holds for increased optical thickness values as well, it implies that to produce a global cooling of some 0.5°C to 1.0°C on a three-year average, via a transient volcanic eruption, would require a volcanic-