Dr. Tim Newfield, Princeton University.
The sun, dimmed by fog. Source: 0xefbeadde.wordpress.com
The June 1991 Pinatubo eruption in the Philippines was one of the largest volcanic eruptions of the twentieth century. It is well documented. There are living witnesses, newspaper articles, detailed surveys of the mountain before and after it blew its top, and satellite maps of the ejecta. The eruption was photographed from the ground and the air, and today you can even YouTube it.
Pinatubo released up to 20 megatons of sulphur dioxide as many as 35 kilometers into the sky. It turned into fine sulphuric acid aerosol, and, within weeks, enveloped much of the Earth. The aerosols were suspended in the atmosphere for around two years. While there, they "veiled" the sun by absorbing or "backscattering" solar radiation. That heated the stratosphere but cooled Earth's surface. The volcano caused a sudden (but non-uniform) fall in average global temperatures of at least .5 degrees Celsius that was still in effect as late as late 1992. In the Northern Hemisphere, temperatures in summer 1992 fell by about 2 degrees Celsius.
Pinatubo on 12 June 1991, a few days before the big eruption. The mountain shrunk 300 meters in the 15 June explosion. Lonely Planet’s Philippines describes a hike up the mountain as ‘most accessible’, but warns not to attempt it in ‘dodgy weather’. Source: USGS.
Earlier and much larger volcanic eruptions in the late Holocene are more obscure. Take Tambora, which erupted in April 1815 and pumped 60 to 110 megatons of sulphur dioxide into the air, leading to one of the most infamous ‘Years Without a Summer’. No volcanologist or climate scientist doubts it dwarfed Pinatubo, but far less is known about the earlier eruption. There are fewer firsthand accounts, and no films or photos (though some argue Joseph Turner and other artists captured its far-reaching atmospheric effects). While the available instrumental data are useful, they are limited and local. Nevertheless, scientists have determined that it was one of the biggest volcanic episodes of the last several thousand years. A recent study estimated that it resulted, in some regions, in 2 to 4 degrees Celsius of cooling from June to August, 1816.
There were other large events, deeper in the historical past. Yet these episodes are far more mysterious. Often the culpable volcano (or volcanoes) is not known, and firsthand accounts (if any) are more than vague: they are cryptic.
For example, something traumatic appears to have affected the world in around the year 536 CE. The five reports that survive for this "536 event" say nothing of an eruption. They merely describe in vague terms a sort of unusual sun dimming or atmospheric veiling. The Roman statesman Cassiodorus, for example, describes a dim moon, and a sun that lost its "wonted light" and appeared "bluish," as if in "transitory eclipse throughout the whole year."
These reports leave room to doubt that the phenomenon they describe was really volcanic in origin. Their mysteriousness, however, has spurred intense interest from scholars and enthusiasts since the phenomenon first appeared in the pages of the Journal of Geophysical Research, in 1983. NASA geoscientists Richard Stothers and Michael Rampino discovered a stratosphere-clouding volcanic episode tucked away in four (but, by 1988, five) late antique texts. They also found it in sulphate in Greenlandic ice, and they discovered pumice-lodged wood they date to 540 ±90 CE (meaning give or take 90 years), on Rabaul, a volcano in Papua New Guinea.
Since 1983, much has changed. Rabaul is long gone. Even before it seemed the dust veil witnessed (inconsistently) over the Mediterranean was not a volcanic dust veil, but instead some sort of "damp fog," the mountain was considered an unlikely source. In the 1980s, assessments of Antarctic ice did not turn up major mid sixth-century volcanism, but rather a signal from about 505 CE. That exonerated all Southern Hemispheric volcanoes from causing the 536 event. Rabual’s eruption chronology was re-dated with greater precision at least twice within eleven years, and it was determined that the 540±90 date was, in fact, an uncalibrated mix-up of the ages originally returned for the pumiceous wood. Rabaul actually exploded sometime in the interval of 633-670 CE, or (as of 2015) 667-699 CE.
Other volcanoes got their share of attention too. Before Rabaul, the Greenlandic sulphates were associated with the great ‘White River Ash’ eruption of Alaska’s Mount Churchill, which was dated roughly in 1975 to 700 ±100 CE, but in 2014 to 833-850 CE. They were also loosely associated with Iceland’s Eldgjá, which is well-known for erupting in the 930s. After, they were tied to the Chiapanecan El Chichón, Indonesia’s infamous Krakatoa, the now-dormant stratovolcano Haruna, and the El Savadorian Ilopango. The latter received considerable press in 2010, when palaeoecologist Robert Dull asserted its ‘paroxysmal’ Tierra Blanca Joven event, considered the largest Central American eruption of the last 84,000 years, and previously given third- and fifth- century dates, actually caused global cooling in 536 CE.
El Salvador’s largest and deepest (crater) lake, Lago de Ilopango. A survey of archaeological excavations suggested a 100-kilometre radius around the site was little- or un-inhabitable for a century after the eruption (whether that happened in 535/536 CE or not). A recent edition of Frommer’s Central America describes a visit to the (always) warm-water caldera as "overrated." Lago de Coatepeque, another volcanic lake 50 kilometres to west, is preferable. Source: NASA Earth Observatory
Yet for a while after 1983, scientists could find no eruptions in 536 CE. The original ice dates of 540 ±10 and c. 535 CE that Stothers and Rampino used to explain the abnormal Byzantine veiling were adjusted in 1984, at around the same time that Stother’s second, more influential article on a volcanic 536 event appeared in Science. This does not now seem surprising. The dates that scientists have given for most first-millennium eruptions have shifted back or forward in time at some point or another. Analyses of the remnants of eruptions in eruption-site sediments often produce ages that disagree by a half century or more. Studies of sulphate layers in ice cores also vary: a couple years in some cases, a decade or five in others.
For more than a decade after 1983, it seemed that the 536 event had other causes. Explanations were diverse. Some held that the clouding Procopius and his peers had witnessed was tropospheric and regional, not a stratospheric phenomenon of hemispheric or global proportions. Volcanism that was local and remarkable, but globally inconsequential, was the cause of some kind of low-hanging ‘damp fog’.
Others held firm: volcano or no volcano, the event was global. Oceanic outgassing, an interstellar cloud, and an asteroid or comet impact event were proposed. The latter, advanced in the early ‘90s, was not immediately popular. Some scholars considered an impactor a "much less likely" explanation for the 536 event than a major volcanic eruption, despite the then-complete lack of evidence for such an eruption. Yet the impact theory eventually gained some credibility. Different types of rocks and impacts were envisioned. A comet might have "air-bursted"’ in the upper atmosphere and ignited one or more vast forest fires, or alternatively a "medium-sized asteroid" struck an ocean and threw marine aerosols into the stratosphere. The impact of a comet less than one kilometer in diameter could have loaded the sky with enough debris to generate multiple successive years of cooling. Even after volcanic eruptions could again be convincingly tied to 536 CE cooling, some scientists argued that an asteroid 640 metres in diameter crashed into Australia, compounding the chilling effect of volcanic eruptions and carving out the Gulf of Carpentaria.
The impactor theory failed to convince many for long. Michael Baillie, a tree ring expert (or "Dendrochronologist") who first advocated the theory in a 1994 article, sided with volcanic explanations after glaciologist Lars Larsen and his team found evidence for a major eruption in multiple ice cores at both poles. This big, low-latitude, Tropical event was affixed a date of 533/534 ±2 CE. It seemed to explain why the "sun’s rays," according to John of Ephesus, "were visible for only two or three hours a day" in 536/37 CE. Larsen also drew attention to "an even larger" Northern Hemisphere deposit, which he dated to 529 ±2 CE. This may not have seemed important at the time, since there are no written sources that suggest anything strange about 529 CE. Yet, only months later, Baillie drew on a growing quantity of tree ring data to suggest that both newly discovered eruptions be moved forward by six or seven years. This adjustment offered an explanation for the unusual tree-ring signals he had highlighted in the early 1990s.
Tree ring data significantly altered scientific understandings of what happened in the sixth century. Independently of texts and ice, tree rings suggest a major disturbance in 536 CE. Tree ring data, unknown to Stothers and Rampino in the 1980s, give perhaps the best record of the sixth-century event. They give annual information with an objectivity that sixth-century historians cannot match. Together, they have a temporal and spatial "awareness" no written source can rival.
Mediterranean texts describe the 536 event as 12 or perhaps 18 months long, but Baillie surveyed trees from Ireland, Germany, Scandinavia and the U.S.A. that clearly show that the event lasted for roughly a decade. Tree rings also demonstrate that the 536 event was not a Byzantine oddity. Rather, it was vast: hemispheric or even global. Trees also reveal not one steady stretch of poor growth but a marked departure from normal growing conditions, with acute troughs and peaks. Some scholars therefore believed that a cluster of stratosphere-clouding phenomena were to blame, not a single cataclysm. The first nadir was in 536-537 CE, while the second, and more pronounced, was in 540-541 CE. More recent tree ring studies have highlighted a third low in 546-547 CE. This one, and another in the early 550s, were already visible in Baillie’s original work, but they were not much discussed.
Over the last twenty years, tree ring studies have confirmed that the 536 event was hemispheric, and at a point global, and that it lasted for more than a decade. Multiple tree ring temperature reconstructions have found several of the coldest growing seasons (typically June-August) of the last two (or, in some cases, seven-and-a-half) thousand years fall within the sixth-century downturn.
A few examples: a 1993 paper identified 536, 535, and 541 CE as the second, third, and fourth-coldest growing seasons in a 2,000-year-long chronology from Sierra Nevada. A 2001 paper used a Mongolian tree ring series that was nearly as long, and found unusually chilly temperatures from 536 to 545 CE, with low points in 536 and 543 CE. A 2015 study used a composite northern hemisphere chronology stretching back to 500 BCE, and established the successive decades of 536-545 and 546-555 as the coldest and tenth-coldest decades in the series. According to the same series, six of the thirteen coldest years between 500 BCE to 1250 CE happened during the sixth-century climatic downturn.
The "Baillie bump," the forward-pushing of Larsen's eruptions (and now most first millennium eruptions detected in ice), placed major volcanism at each of the cooling episodes identified in tree ring data. Michael Sigl and a team of scientists recently included these results within an important synthesis of glacial volcanic eruption chronologies. It is still not clear which volcanoes erupted in 535/536 and 539/540 CE, but a cluster of volcanoes seem to have caused the downturn.
Still, there may be room to doubt whether Cassidorus and company took in a hemispheric event in 536 CE. They may well have witnessed a local disturbance. Procopius has Vesuvius bubbling, but not erupting, in 536 CE. Whether this ‘extinguisher or all things green’ erupted around then - or perhaps another nearby mountain - we do not know. Minor, nearby volcanism may have coincided with a much larger, distant eruption. One would have veiled Mediterranean skies, while the other marked the world’s trees. Tree rings from Constantinople’s hinterland may support this theory, since they have failed to reflect a major change in growth from 536 to 550 CE.
Of course, it may still be that an impactor near-simultaneously fell to Earth from space. Dallas Abbott and her team have recently found iron oxide, silicate spherules, and other ejecta indicators in the melt-water of a portion of a sixth-century Greenlandic ice core. They interpreted a high concentration of calcium as calcium carbonate, a main component in seashells, and detected tropical aquatic microfossils: a first for Greenlandic ice. It is evidence for an impact at sea, which then sent marine aerosols into the stratosphere.
For years, the 536 event or 536-550 CE downturn figured as a particularly cold stretch (in fact the coldest) in a long cool phase that set in more than a century before 536 CE and has many names: The "Vandal Minimum," the "Early Medieval Cold Period," or the "Migration Period Pessimum." Very recently, a multidisciplinary study concluded that that the 536-550 event triggered a longer cold period within this Minimum. They call it the "Late Antique Little Ice Age," and argue that it was possibly even chillier and more unstable than the better-known early modern Little Ice Age.
Did this cooling have profound consequences for sixth-century societies? Maybe, yet historians came to the 536 event rather late. In 2005, historian Antti Arjava wrote an interdisciplinary appraisal of the evidence for a sixth-century cooling event. Aside from Arjava, the few historians who have wrestled with the clouding have not attempted a complete or current synthesis of the written and scientific evidence. Arjava's paper has therefore served as the main conduit for historians and archaeologists for the science surrounding the 536 event. However, Arjava wrote his paper in the years when scientists could not match the event with a volcanic eruption. The paper plays up the cloud’s mysteriousness, and diminishes its extent and impact. A reading of John the Lydian’s account, one fuller and closer than that offered by Stothers, led to the conclusion the event was Mediterranean specific, more of a fog than a veil, and damp, not dry. That and the lack of consistent evidence for poor harvests and food shortage in the 530s suggested the cloud had little effect on contemporary societies.
Much has changed since 2005. It is more difficult now to diminish the downturn or doubt that it triggered a marked, though temporary, demographic contraction in many regions of the world through its effects on plants. However, minimalist readings remain popular. They are still, if mostly through Arjava, a reaction to a pair of catastrophist books on 536 published in 1999 by Keys (Catastrophe) and Baillie (Catastrophic Encounters with Comets). The books argued for far-reaching and at times unfathomable historical consequences from mystery clouding, from Teotihaucan’s fall to China’s reunification, from Islam’s emergence and Charlemagne’s birth to England’s colonization of North America and Japan’s modern nation state. A reluctance to engage with the palaeoclimate sciences and a willingness to write nature out of history have allowed historians to dismiss the significance of the 536 event for contemporary peoples.
Recently, more scientifically-minded historians, such as Michael McCormick, have offered more appropriate (if maximalist-leaning) narratives, in which cooling had moderate implications for sixth-century peoples. A vast, near-unparalleled environmental event need not have cataclysmic consequences to warrant study. Histories of resilience and adaptation to sudden and dramatic climate change should be as important and intriguing as histories of failure and collapse. This is clear in new work on the effects of the downturn, from the Yucatán to Fennoscandia, which emphasizes coping strategies and a certain hardiness in those that lived beneath the veils.
Mayan Calakmul’s largest structure, Structure II. With roughly 6,200 constructions spread out over about 30 square kilometres in the late Classic period, the city of Calakmul (now in Mexico’s Campeche State) experienced rapid growth during the sixth-century "hiatus." This interval of debated tumultuousness between the early and late Classic phases saw a leveling off (or decline) in stelae and monumental building at several Mayan locales as well as (perhaps dramatic) population contraction. Richardson Gill (in his The Great Mayan Droughts) argued the downturn caused this break in activity and drawing on palaeoclimatology (from other world regions) assigned the hiatus a firm start date of 536 CE. Built atop a preclassical stucco-decorated plaza, Structure II was an important building throughout the Classic period. Fodor’s Cancún, Cozumel, Yucatán Peninsula recommends a stop at the ‘vast’ and ‘lovely’ but little-visited Calakmul, which in its "heyday" (between 542 and 695 CE) numbered at least 50,000 people. A climb up the pictured pyramid allows for a "soaring vista."
Although not everyone would have come out from under the dust worse off, it is important to not let the pendulum swing back too far. After all, there are indications from across Eurasia of subsistence crises. Read together, these reports suggest a rather uneven occurrence of downturn-triggered crop failure and genuine famine. That clouding density and duration undoubtedly varied, and people were not everywhere equally vulnerable, might account for this patchiness. So too the concurrence of other natural and cultural pressures in some areas.
It should be emphasized that large eruptions do not simply chill the world. The effects on weather and climate are non-uniform. They are regional and can differ markedly, as Pinatubo and Tambora have shown. Tropical eruptions, such as the 539/540 event, also exercise a different force on climate than high latitude Northern Hemispheric ones, like 535/536. For instance, major near-equatorial volcanism is known to cause winter warming in North America, Europe, and Russia, but winter cooling in Western and Eastern Asia. Extratropical Northern Hemispheric volcanism cools hot and cold seasons alike. Seasonality matters too. That high latitude eruptions seem to be more impactful if they occur in summer could indicate that the 535/536 eruption happened in that season.
A few contemporary reports of despair and devastation seem hyperbolic. Did Italian mothers really eat their daughters? Did three quarters of the population north of the Yellow river really die off? Yet neither they, nor less-sensational descriptions, should be written off as lacking any grounding in the immediate post-eruption reality. Most sixth-century societies were able to absorb one bad year, but very few were able to absorb two or three. Back-to-back(-to-back) years of poor growing conditions, caused by a sharp cooling of average temperatures, were certain to take a toll.
An eruption cluster - multiple Tambora-like events within a few years of each other - caused the mid sixth-century downturn. Whether an impactor was roughly coincident is uncertain. The variability of the effects of both eruptions on climate and the extent and regionality of the loss of life are uncertain as well. Sixth-century cooling may well have helped cause the outbreak of the "Plague of Justinian" - the so-called "First Bubonic Plague Pandemic" - with profound demographic consequences. This link, and other enduring mysteries of the sixth-century downturn, will be the subject of a future article on this site.
Dr. Newfield will write a synthesis of the scholarship on sixth-century cooling, like this but more complete, in the The Palgrave Handbook of Climate History, edited by Franz Mauelshagen, Christian Pfister and Sam White.
D. Abbott et al, ‘What Caused Terrestrial Dust Loading and Climate Downturns between A.D. 533 and 540?’ Geological Society of America Special Papers 505 (2014).
A. Arjava, ‘The Mystery Cloud of 536 CE in the Mediterranean Sources’ Dumbarton Oaks Papers 59 (2005).
M. Baillie, ‘Dendrochronolgy Raises Questions About the Nature of the AD 536 Dust Veil Event’ The Holocene 4 (1994).
M. Baillie, ‘Proposed Re-Dating of the European Ice Core Chronology by Seven Years Prior to the 7th Century AD” Geophysical Research Letters 35 (2008).
U. Büntgen et al, ‘2500 Years of European Climate Variability and Human Susceptibility’ Science 331 (2011).
U. Büntgen et al, ‘Cooling and Societal Change during the Late Antique Little Ice Age from 536 to around 660AD’ Nature Geoscience 9 (2016).
B. Dahlin and A. Chase, ‘A Tale of Three Cities: Effects of the AD 536 Event in the Lowland Maya Heartland’ in G. Iannone ed., The Great Maya Droughts in Cultural Context: Case Studies in Resilience and Vulnerability (University of Colorado Press, 2014).
R. Dull et al., ‘Did the Ilopango TBJ Eruption Cause the 536 Event?’ American Geophysical Union Fall Meeting 2010 Abstract V13C-2370.
C. Hammer et al, ‘Greenland Ice Sheet Evidence of Post-Glacial Volcanism and its Climatic Impact’ Nature 288 (1980).
C. McKee et al, ‘A Revised Age of AD 667-699 for the Latest Major Eruption at Rabaul’ Bulletin of Volcanology 77 (2015).
L. Larsen et al, ‘New Ice Core Evidence for a Volcanic Cause of the A.D. 536 Dust Veil Event’ Geophysical Research Letters 35 (2008).
J. Luterbacher and C. Pfister, ‘The Year Without a Summer’ Nature Geoscience 8 (2015).
M. McCormick et al, ‘Climate Change During and After the Roman Empire: Reconstructing the Past from Scientific and Historical Evidence’ Journal of Interdisciplinary History 43 (2012).
E. Rigby et al, ‘A Comet Impact in AD 536?’ Astronomy and Geophysics 45 (2004).
A. Robock, 'Perspectives, Pinatubo Eruption: The Climatic Aftermath' Science 295 (2002).
A. Robock, ‘Volcanic Eruptions and Climate’ Reviews of Geophysics 38 (2000).
M. Sigl et al, ‘Timing and Climate Forcing of Volcanic Eruptions for the Past 2,500 Years’ Nature 523 (2015).
R.B. Stothers and M.R. Rampino, ‘Volcanic Eruptions in the Mediterranean before A.D. 630 from Written and Archaeological Sources’ Journal of Geophysical Research 88 (1983).
R.B. Stothers, ‘Mystery Cloud of AD 536’ Nature 307 (1984).