Variation of surface air temperatures in relation to El Niño and cataclysmic volcanic eruptions, 1796–1882

https://doi.org/10.1016/S1364-6826(99)00055-3Get rights and content

Abstract

During the contemporaneous interval of 1796–1882 a number of significant decreases in temperature are found in the records of Central England and Northern Ireland. These decreases appear to be related to the occurrences of El Niño and/or cataclysmic volcanic eruptions. For example, a composite of residual Central England temperatures, centering temperatures on the yearly onsets of 20 El Niño events of moderate to stronger strengths, shows that, on average, the change in temperature varied by about ±0.3°C from normal, being warmer during the boreal fall–winter leading up to the El Niño year and cooler during the spring–summer of the El Niño year. Also, the influence of El Niño on Central England temperatures appears to have lasted about 1–2 years. Similarly, a composite of residual Central England temperatures, centering temperatures on the month of eruption for 26 cataclysmic volcanic eruptions, shows that, on average, the temperature decreased by about 0.1–0.2°C, typically, 1–2 years after the eruption; although for specific events, like Tambora, the decrease was considerably greater. Additionally, tropical eruptions appear to have produced greater cooling than extratropical eruptions, and eruptions occurring in boreal spring–summer appear to have produced greater cooling than those occurring in fall–winter.

Introduction

More than 200 years ago, Benjamin Franklin (1784) noted a “coolness” during the summer months of 1783 which occurred in conjunction with a “constant fog” over Europe and North America. He conjectured that the decreased surface air temperature was caused by a reduction in incident sunlight, owing to the presence of the fog. Of several possible causes of the fog, Franklin speculated that volcanic activity in Iceland (he was unaware of the devastating eruption of Asama in Japan) might be responsible for the observed meteorological effects (Damon, 1968, Bryson and Goodman, 1980, Sigurdsson, 1982).

Later, Abbot (1913) noted that shortly after the Katmai eruption in 1912 changes were observed in the solar radiation received at observatories of the Smithsonian Institution. He speculated that the two phenomena might be associated (cf. Humphreys, 1913, Landsberg and Albert, 1974). It should be noted that a similar reduction in insolation, one lasting about 3 years in length, had previously been observed by researchers at the Montpelier Observatory in the South of France following the eruption of Krakatau in 1883 (Wexler, 1952).

More recently, changes in surface air temperature have been recorded following the eruptions of El Chichón in 1982 and Pinatubo in 1991 (e.g., Bluth et al., 1992, Dutton and Christy, 1992, Trepte et al., 1993, McCormick et al., 1995, Hansen et al., 1996). In particular, the June 1991 eruption of Pinatubo produced the largest sulfur dioxide (SO2) cloud detected by the Total Ozone Mapping Spectrometer aboard the Nimbus-7 satellite during its operational life. The SO2 cloud (and other byproducts of the explosion) was observed to encircle the Earth within about 3 weeks, straddling the equator ±20–30° in latitude, and, over time, to migrate poleward (toward both poles), thereby essentially blanketing the entire Earth. The aerosol loading from this event, being about 30 Tg, was somewhat smaller than those estimated for Krakatau in 1883 (50 Tg) and Tambora in 1815 (100 Tg), and appears to have been globally effective for several years.

On the basis of the 15 largest stratospheric-aerosol-producing volcanic eruptions since 1866, Robock and Mao (1995) have concluded that the volcanic climatic timescale is probably about 2 years in length and that associated with these eruptions is a 0.1–0.2°C cooling — the amount of cooling being dependent upon the location (latitude) and time of year (season) of the eruption. In addition, they found that the cooling was evident in the temperature record whether it was corrected for ENSO (i.e., El Niño–Southern Oscillation) or not. Global cooling has also been suggested to follow the great Tambora blast of 1815 (Stommel and Stommel, 1979, Stothers, 1984).

Typically, a stratospheric-aerosol-producing volcanic eruption is one having a volcanic explosivity index (VEI)≥3 (Newhall and Self, 1982, Rampino and Self, 1984, Simkin and Siebert, 1994), although, strictly speaking, the most important aspect as to whether or not an eruption will have a climatic effect is the size of the sulfur content of the emissions that reach into the Earth’s stratosphere, rather than its estimated size of explosivity (Sato et al., 1993, Carroll, 1997, Hansen et al., 1997, Rowntree, 1998). Signatures of stratospheric-aerosol-producing eruptions are often found in the record of ice cores from Greenland and/or Antarctica (Hammer, 1977, Hammer et al., 1980, Legrand and Delmas, 1987, Moore et al., 1991, Delmas et al., 1992, Langway et al., 1994, Langway et al., 1995, Zielinski et al., 1994, Robock and Free, 1995, Stuiver et al., 1995, Zielinski, 1995, Clausen et al., 1997, Cole-Dai et al., 1997), as well as in the record of tree rings (Briffa et al., 1988, Briffa et al., 1994, Briffa et al., 1998, Jones et al., 1995).

Recently, Wilson (1998c) found evidence for a connection between the surface air temperatures of the Armagh Observatory (Northern Ireland) for 1818–1858, i.e., a subset of the series I dataset (Butler and Johnston, 1996), volcanism, and a paucity of sunspot observing days. Specifically, he found that the cataclysmic eruptions (i.e., those having VEI≥4) of Tambora (1815), Galunggung (1822), and Cosguina (1835) appeared to be closely associated with decreases in temperature as recorded at the Armagh Observatory and with decreases in the number of sunspot observing days as recorded by Samuel Heinrich Schwabe, the discoverer of the sunspot cycle, during the interval of 1826–1858 at Dessau, Germany, and as reconstructed by Rudolf Wolf during the interval of 1818–1848 (Wilson, 1998a). Presumably, the number of sunspot observing days was sometimes greatly reduced due to persistent overcast meteorological conditions that may have prevailed during these times of colder clime following these specific volcanic eruptions.

In this investigation, a comparison is made between the published series I listing of annual temperatures from the Armagh Observatory, 1796–1882 (Butler and Johnston, 1996) and the published listing of annual Central England temperatures (Manley, 1974). Also, the behavior of 4-month moving averages of residuals (i.e., deviations from normal) of the Central England temperatures is examined in relation to composites of El Niño and cataclysmic volcanic eruptions (i.e., using the method of epoch analysis). One finds that the two temperature datasets are indeed linearly correlated and that both El Niño and cataclysmic volcanic eruptions seem to have a measurable effect on Central England temperatures.

Section snippets

Comparison of Central England and Armagh Observatory temperatures

Fig. 1(bottom panel) shows the variation of Manley’s (1974) Central England temperatures, denoted T(CE), during the interval of 1796–1882, the interval limits being dictated by the beginning and ending dates of the Armagh Observatory series I dataset (shown in Fig. 1(middel panel)). During this interval T(CE) had a mean temperature of about 9.1°C and a standard deviation of about 0.6°C. The highest individual annual average occurred in 1868 (10.4°C), while the lowest occurred in 1879 (7.4°C).

Monthly adjusted residuals

In order to examine more closely the individual roles and effects of El Niño and cataclysmic volcanic eruptions in short-term climatic change (especially, as they may be recorded in the surface air temperature records), it is imperative that monthly averages be used (rather than annual averages). Because the published listing of Central England temperatures by Manley (1974) also includes monthly averages — in contrast to the published listing of Armagh Observatory temperatures by Butler and

The composite residuals in relation to El Niño years

Fig. 7(bottom panel) plots the average residual of Central England temperatures in the vicinity of an El Niño year for the events occurring during 1796–1882 (counting the two moderate events of 1866 and 1867–1868 as one event). The composite signature, based on 20 events, suggests that just before the start of an El Niño year (i.e., during the boreal fall–winter leading up to the El Niño year) the average residual transitions from near normal conditions to typically warmer values (about +0.3°C

Summary and concluding remarks

This investigation has examined Central England and Northern Ireland (Armagh Observatory) temperatures during the interval of 1796–1882. It is apparent that the two datasets are correlated with each other, although there is also evidence that the temperatures of Northern Ireland may have been somewhat cooler in comparison to that of the Central England temperatures, particularly prior to 1825. More importantly, this investigation has shown that both El Niño and cataclysmic volcanic eruptions,

References (53)

  • H.B. Clausen et al.

    A comparison of the volcanic records over the past 4000 years from the Greenland Ice Core Project and Dye 3

    Journal of Geophysical Research

    (1997)
  • J. Cole-Dai et al.

    Annually resolved southern hemisphere volcanic history from two Antarctic ice cores

    Journal of Geophysical Research

    (1997)
  • P.E. Damon

    The relationship between terrestrial factors and climate

  • R.J. Delmas et al.

    1000 years of explosive volcanism recorded at the South Pole

    Tellus

    (1992)
  • C. Deser et al.

    El Niño events and their relation to the Southern Oscillation: 1925–1986

    Journal of Geophysical Research

    (1987)
  • E.G. Dutton et al.

    Solar radiative forcing at selected locations and evidence for global lower tropospheric cooling following the eruptions of El Chichón and Pinatubo

    Geophysical Research Letters

    (1992)
  • J.A. Eddy

    The historical record of solar activity

  • Franklin, B., 1784. Meteorological imaginations and conjectures. Memorial Library Philosophical Society, Manchester 2,...
  • C.U. Hammer

    Past volcanism revealed by Greenland ice sheet impurities

    Nature

    (1977)
  • C.U. Hammer et al.

    Greenland ice sheet evidence of post-glacial volcanism and its climatic impact

    Nature

    (1980)
  • J. Hansen et al.

    Global surface air temperature in 1995: return to pre-Pinatubo level

    Geophysical Research Letters

    (1996)
  • J. Hansen et al.

    The missing climate forcing

    Philosophical Transactions Royal Society of London Series B

    (1997)
  • D.V. Hoyt et al.

    A revised listing of the number of sunspot groups made by Pastorff, 1819–1833

    Solar Physics

    (1995)
  • P.D. Jones et al.

    Tree-ring evidence of the wide-spread effects of explosive volcanic eruptions

    Geophysical Research Letters

    (1995)
  • H.E. Landsberg et al.

    The summer of 1816 and volcanism

    Weatherwise

    (1974)
  • C.C. Langway et al.

    A 10-century comparison of prominent bipolar volcanic events in ice cores

    Journal of Geophysical Research

    (1995)
  • Cited by (8)

    View all citing articles on Scopus
    View full text