«RICHARD B. STOTHERS Institute for Space Studies, Goddard Space Flight Center, NASA, 2880 Broadway, New York, NY 10025, U.S.A. Abstract. Somewhere in ...»
CLIMATIC AND DEMOGRAPHIC CONSEQUENCES OF THE
MASSIVE VOLCANIC ERUPTION OF 1258
RICHARD B. STOTHERS
Institute for Space Studies, Goddard Space Flight Center, NASA, 2880 Broadway, New York,
NY 10025, U.S.A.
Abstract. Somewhere in the tropics, a volcano exploded violently during the year 1258, producing
a massive stratospheric aerosol veil that eventually blanketed the globe. Arctic and Antarctic ice cores suggest that this was the world’s largest volcanic eruption of the past millennium. According to contemporary chronicles, the stratospheric dry fog possibly manifested itself in Europe as a persistently cloudy aspect of the sky and also through an apparently total darkening of the eclipsed Moon. Based on a sudden temperature drop for several months in England, the eruption’s initiation date can be inferred to have been probably January 1258. The frequent cold and rain that year led to severe crop damage and famine throughout much of Europe. Pestilence repeatedly broke out in 1258 and 1259; it occurred also in the Middle East, reportedly there as plague. Another very cold winter followed in 1260–1261. The troubled period’s wars, famines, pestilences, and earthquakes appear to have contributed in part to the rise of the European ﬂagellant movement of 1260, one of the most bizarre social phenomena of the Middle Ages. Analogies can be drawn with the climatic aftereffects and European social unrest following another great tropical eruption, Tambora in 1815.
Some generalizations about the climatic impacts of tropical eruptions are made from these and other data.
1. Introduction In the year 1259, a rain of sulfuric acid aerosols and tiny glass shards originating from one of the greatest volcanic eruptions of the past two millennia fell out of the stratosphere onto the North and South polar ice caps. Samples of the debris have been retrieved in modern times from dated ice cores in Greenland (Hammer et al., 1980; Langway et al. 1988; Johnsen et al., 1992; Palais et al., 1992; Zielinski et al., 1994; Zielinski, 1995), the Canadian Arctic (Fisher and Koerner, 1988; Zheng et al., 1998), and Antarctica (Zanolini et al., 1985; Langway et al., 1988; Palais et al., 1990; Delmas et al., 1992; Hammer et al., 1997). On the plausible assumption that this event had to have been a tropical eruption, the total production of sulfuric acid has been estimated as ∼300 megatons (Hammer et al., 1980; Langway et al., 1988;
Zielinski, 1995; Hammer et al., 1997) and possibly as much as ∼600 megatons (Delmas et al., 1992). Thus, this eruption exceeded in magnitude all of the greatest known historic eruptions of post-medieval times, such as Tambora, Indonesia, in 1815 and Laki, Iceland, in 1783.
To the present day, though, the source volcano for the 13th century aerial fallout remains unknown. A Madinah, Saudi Arabia, eruption in 1256 (Camp et al.,
1987) seems to have been too early in date and much too small in magnitude to be relevant. A huge eruption of El Chichón, Mexico, occurred at roughly this time (Tilling et al., 1984) and the composition of glass shards from ﬁeld samples appears to match what is found in Greenland and Antarctic ice cores (Palais et al., 1990, 1992). Yet the match is not perfect in detail and, in any case, glass chemistry is not a highly discriminatory diagnostic for volcanic eruptions.
The year of the volcanic eruption, however, is accurately known. Although the peak fallout of aerosols over the poles occurred in 1259, the rise to maximum started in 1258 (Hammer et al., 1980). Therefore, if modern tropical eruptions provide an applicable analogy, stratospheric injection must have occurred in 1258.
Such a large eruption would probably have had easily observable meteorological consequences. Tropical eruptions in modern times generate globe-girdling stratospheric aerosol veils (dry fogs) that persist for several years, slowly settling out. The aerosols block some of the incoming sunlight and alter atmospheric circulation patterns, and by these means cool much of the Earth’s surface. This temporary disturbance of the world’s climate, often involving increased precipitation, can adversely affect agriculture. Consequences may be a greater human susceptibility to famine and disease, leading ultimately to social and political unrest. In earlier times, with their more primitive technological conditions, these consequences often occurred in exaggerated form. On the other hand, human beings are ﬂexible as well as vulnerable in the face of adverse climatic events, large or small. There is obvious risk, therefore, in linking natural and social phenomena unless cause and effect are clearly discerned. In the present instance, the largest known volcanic eruption of the past millenium is the subject of inquiry. With such an unambiguous premise, the climatic and social consequences can be readily sought, although the scales that are manifested will need some interpretation in the light of other contingencies.
The present paper traces the aftermath of the eruption of 1258 in Europe and the Middle East, as determined mostly from contemporary historical sources. Although there were no instrumental measurements made at that early time, the anecdotal remarks by careful observers reporting on their own times and localities can serve as limited proxies for such information. Even if these human recorders (usually monks) were not scientiﬁcally trained or widely informed, they had at least a formal education. It is always necessary, however, to beware of methodological pitfalls when compiling climatic and related information from the medieval chronicles (Bell and Ogilvie, 1978; Ingram et al., 1978; Alexandre, 1987; Pﬁster et al., 1996). It is less well understood that this information needs also to be placed in the contemporary historical context, in order to interpret the climatic events correctly.
CONSEQUENCES OF THE MASSIVE VOLCANIC ERUPTION OF 1258
2. Dry Fog in 1258
Two pieces of evidence suggest the presence of a widespread dry fog in 1258. First, there is the continued cloudy appearance of the sky in the summer of that year across France. Was this appearance due at all times to the abundant rain clouds?
According to the reliable, but loquacious Richer (1267) of Sens:
What, then, shall I say about the fruits of the earth that year, when the weather was so remarkably unseasonable that the warmth of the Sun was hardly able, even a little, to reach the earth, and the fruits of that year could barely attain maturity, if at all? For so great a thickness of clouds covered the sky throughout that whole summer that hardly anyone could tell whether it was summer or autumn. The hay, drenched incessantly by strong rains that year, was unable to dry out, because it could not collect the warmth of the Sun on account of the thickness of the clouds.
In other parts of France, western Germany, and northern Italy, the summer and autumn of 1258 were also very rainy and chilly (Notae Constantienses, c. 1260;
Chronicon Savigniacensis, c. 1300; Annales Spirenses, c. 1259; Girard de Fracheto, 1271; Alexandre, 1987). Although in England a brief hot summer did ripen the crops, these were undone by heavy autumn rains, beginning in August (Matthew Paris, 1259; John de Taxter, 1265; Titow, 1960).
Although we are not explicitly informed about the sky conditions that summer in England, the cloudiness in France seems to have persisted between rainstorms.
This suggests, but does not prove, that a constant dry fog was present in the atmosphere. Certainly, the known volcanic eruption would be expected to have produced a global stratospheric dry fog. Polar ice cores indicate that the volcanic output of sulfur in 1258 was twice the output from Tambora, implying a worldwide peak aerosol optical depth of ∼2, which is comparable to the aerosol optical depth over Europe during the visibly thick dry fog created by the Laki ﬁssure eruption (Stothers, 1996a). Though very large, the drop in direct solar radiation by itself would not have interfered much with surface heating of the earth and with crop growth, as we know from the reported course of the Laki dry fog over England during the summer months of 1783 (Grattan and Charman, 1994).
More conclusive evidence of a large aerosol veil in 1258 comes from the contrasting accounts of two lunar eclipses during this year and a later year. The English chronicler John de Taxter (1265) states that ‘the Moon underwent a complete (totalis) eclipse’ on 18 May 1258, and that ‘there was a typical (generalis) eclipse of the Moon, of a bloody color’ on 24 December 1265. The Englishman Matthew Paris (1259) also uses the term generalis for a typical lunar eclipse.
The normal color of the eclipsed Moon is in fact red. When, however, the Earth’s stratosphere contains an abundance of volcanic aerosols, the incident sunlight cannot be refracted and scattered into the shadow cone, and therefore the Moon appears completely dark (Link, 1963). For the Moon to effectively vanish, an aerosol optical depth of 0.1 or more is needed. This appears to have been the 364 RICHARD B. STOTHERS case on 18 May 1258, conﬁrming that both hemispheres of the Earth were obscured by aerosols. But by 31 August 1262 the stratosphere seems to have cleared, because the eclipsed Moon on that date showed ‘a bloody color’ (Chronica Minor Auctore Minorita Erphordiensi, 1265).
A last point to be noted about the weather in 1258, whose relevance will be explained below, is that a long cold spell occurred in England between February and June that year (Matthew Paris, 1259). The same winter is also reported to have been a severe one at Prague in Bohemia (Continuator of Cosmas, 1283) and the springtime was noted as harsh in northern Iceland (Ogilvie, 1990).
3. Weather Patterns after 1258
After the very rainy autumn of 1258, the following winter in England was unexceptional. Matthew Paris (1259), who regularly reports in detail on the weather near London, indicates nothing unusual for that winter. The Chronicle of Novgorod (1471) mentions an odd frosty day in Russia during April 1259. The summertime afterward was hot and dry in Austria and Germany (Continuatio Lambacensis, 1283; Annales Wormatienses, c. 1300) and hot and stormy in France (Notae Constantienses, c. 1260), while it rained a lot in England (Flores Historiarum, 1265).
Less is known about the weather in the following year, 1260. After a very mild winter, central France experienced severe cold and snow during April (Chronicon Savigniacensis, c. 1300). But the summer weather was alternately dry and stormy, with a lot of hail, near Prague (Continuator of Cosmas, 1283) and likewise near London (Flores Historiarum, 1265).
It was not until later that year that Europe suffered another very cold winter.
The winter of 1260–1261 struck Iceland so severely that people were forced to slaughter many of their livestock (Thórdarson, 1284) and ice formed in the sea all around the island (Storm, 1888). Very harsh winter conditions are also reported for England (Flores Historiarum, 1265; Continuator of William of Newburgh, 1298) and for northwestern Italy (Ventura, c. 1325). In Alsace, the Ill River froze (Annales Colmarienses Minores, c. 1300), but it is not clear whether this happened in the winter of 1260–1261 or of 1261–1262, or in both winters.
It is apparent that very cold European winters occurred in 1257–1258 and 1260–
1261. Does this pattern of cold winters match the behavior of weather anomalies after modern large explosive eruptions in the tropics? Within a month of such a modern tropical eruption, northern land masses experience sudden and prolonged declines of surface air temperature, which last 3–6 months (Kelly and Sear, 1984;
Sear et al., 1987; Bradley, 1988; Kelly et al., 1996). If the eruption of 1258 occurred in January, the suddenly cold weather in England between February and June might thus be explained. At least we suspect from that year’s dark lunar eclipse on 18 May that the eruption must have taken place before mid-May. After a modern eruption, a
CONSEQUENCES OF THE MASSIVE VOLCANIC ERUPTION OF 1258second period of cooling ensues anywhere from 1 to 5 years after the eruption, the most typical elapsed time being 2 or 3 years (Rampino et al., 1988; Angell, 1988;
Bradley, 1988; Mass and Portman, 1989; Groisman, 1992; Robock and Mao, 1995;