Abstract
Examined is the effect on surface temperature of the volcanic eruptions of Asama and Laki in 1783, Tambora in 1815, Coseguina in 1835, Krakatoa in 1883, Santa Maria, Soufrière and Pelée in 1902, and Agung in 1963, using temperature records extending back to 1781. These records include New Haven, Connecticut, in North America; Edinburgh, De Bilt, Copenhagen, Berlin and Vilnius in Northern Europe; Geneva, Basel, Hohen-peissenberg, Vienna and Budapest in Central Europe; the “Central England” data of Manley; and the merged Northern Hemisphere data of Groveman and Landsberg and Jones et al. At New Haven and in Europe there is more evidence of a cooling following eruptions in subtropical and temperate latitudes than in equatorial latitudes (despite the similarity in mean dust-veil index), with a cooling most evident following the Asama and Laki eruptions in Japan and Iceland, and next most evident following the Coseguina eruption in Nicaragua. Following the tremendous Tambora eruption, the eruption with the largest dust-veil index, there is obvious cooling at New Haven, but not in Europe and perhaps not for the hemisphere as a whole. Hemispheric cooling is indicated to have been most pronounced following the Agung eruption—of the six eruption episodes the one with the smallest dust veil index but the best temperature data. Based on an application of Student's t-test to station, regional and hemispheric data, on 27 occasions (out of a possible 96) the average temperature for the 5-year period after the eruption is significantly (at the 5% level) lower than the average temperature for the 5- year period before the eruption, but in no case is the average temperature after the eruption significantly higher. It is proposed that cooling is not more apparent following some eruptions because of the tropospheric warming associated with strong and persistent El Niño episodes occurring shortly after the eruptions.
