Info

Note: A, reconstructed temperature series; B, 2- to 49.9-year composite wave mode; C, 10- to 49.9-year wave mode; and D, 2- to 9.9-year wave mode. The 5-year interval measurements of the sulfate series were linearly interpolated to annual values. The boldface coefficients indicate significant values if the series were totally independent. No correction has been made for the degrees of freedom lost due to autocorrelation. Therefore, the significance should be interpreted as indicative only.

Note: A, reconstructed temperature series; B, 2- to 49.9-year composite wave mode; C, 10- to 49.9-year wave mode; and D, 2- to 9.9-year wave mode. The 5-year interval measurements of the sulfate series were linearly interpolated to annual values. The boldface coefficients indicate significant values if the series were totally independent. No correction has been made for the degrees of freedom lost due to autocorrelation. Therefore, the significance should be interpreted as indicative only.

a cluster clearly separate from the other variables. The third component explains 11.6% of the total variance.

Running correlation coefficients were calculated by using a window of 50 years between the sulfate series against the reconstructed temperature series. The results are presented in Fig. 3. They show a pattern of negative correlation coefficients spatially distributed toward the northern latitudes in North America. No similar pattern appears in the South America reconstructions. These results also indicate a more consistent and strong negative relation of the sulfate contents with the Northern Hemisphere composite reconstructions. Mann et al. (1998) ran a similar analysis on their series using another index of volcanic activity. They concluded that explosive volcanism exhibits the expected marginally significant negative correlation with temperature during much of 1610-1995 period, most pronounced in the 200-year window centered near 1830, which in

FIGURE 3 Running correlation coefficients (50-year window) between the volcanic sulfate loads and 2- to 50-year wave modes of the reconstructed temperature series. The darkest spots correspond to coefficients lower than —0.36, the limit of significance (p <0.05).

cludes the largest number of known explosive volcanic events.

In Fig. 4, we present a plot of the standardized 5-year running mean of each reconstructed temperature series and a plot of the sulfate loads. Note that high peaks of sulfate content correspond in general, but not always, with cool periods in the series.

Our data suggest a volcanic influence on climate that is not particularly strong, but is discernible in time and space. Several papers have expressed doubts about the existence of a volcanic influence on surface climate (Landsberg and Albert, 1974; Ellsaesser, 1977). After examining the temperature evolution accompanying the largest eruptions since the late eighteenth century, Angell and Korshover (1985) concluded that while volcanic eruptions certainly do not cause warming, the evidence that they cause cooling is not overly impressive. It has become clear in the last decade (Robock and Mao, 1995) that the effect of a volcano on climate is most directly related to the sulfur dioxide content of emissions that reach into the stratosphere and not to the explo-sivity of the eruption, although the two are highly correlated.

Rampino and Self (1982) signaled that the largest eruptions in terms of total ejecta and eruption column height did not necessarily produce the greatest climatic effects (Self et al., 1981). It seems that significantly smaller, but sulfur dioxide-rich explosive eruptions can produce similar temperature deviations. Past increases in stratospheric optical depth may have been controlled to a large extent by sulfur dioxide-rich explosive eruptions, which, by the nature of their magma bodies, tend to be relative small in volume (VEI 3 or 4). If these eruptions occur in clusters, this might provide an explanation for certain prolonged periods of high stratospheric optical depth (1883-93 and 1902-06). However, a complete historical or volcanological record of these events does not exist for years prior to ca. 1950 (Newhall and Self, 1982). Assigning specific atmospheric opacity variations to known historic eruptions is often not possible. Nevertheless, volcanic aerosols are potentially significant climate forcing mechanisms on timescales of years, decades, and perhaps longer periods (Zielinski, 1995; Overpeck et al., 1997), and the correlation of temperature reductions with volcanic eruptions can be observed even in the instrumental record for the Northern Hemisphere (Bradley, 1988).

Some authors indicated that because the climatic response to ENSO is of the same amplitude and timescale as volcanic responses, it is necessary to separate them to examine the volcano signal (Jones, 1988; Robock and Mao, 1995). Handler (1986) has supposed that tropical explosive volcanoes could cause a warming of the SST, and he assumes a connection as triggering El Niño events. In our series, we have not filtered the ENSO signal, which could mask the volcanic influence. Howev

FIGURE 4 (a) Five-year moving average of reconstructed temperature series, standardized (0,1). (b) Standardized volcanic sulfate series.

er, experiments conducted by Robock and Mao (1995) have shown that spatial patterns are clear with or without removing ENSO.

0 0

Post a comment