Two hundred years ago the most powerful eruption in modern history made itself felt around the world. It could happen again at almost any time

Apr 11th 2015 | From the print edition

20150411_FBP001If aliens had been watching the Earth during 1815 the chances are they would not have noticed the cannon fire of Waterloo, let alone the final decisions of the Congress of Vienna or the birth of Otto von Bismarck. Such things loom larger in history books than they do in astronomical observations. What they might have noticed instead was that, as the year went on, the planet in their telescopes began to reflect a little more sunlight. And if their eyes or instruments had been sensitive to the infrared, as well as to visible light, the curious aliens would have noticed that as the planet brightened, its surface cooled. 

Mount Tambora (pictured), a volcano on the Indonesian island of Sumbawa, was once similar in stature to Mont Blanc or Mount Rainier. But in April 1815 it blew its top off in spectacular fashion. On the 10th and 11th it sent molten rock more than 40 kilometres into the sky in the most powerful eruption of the past 500 years. The umbrella of ash spread out over a million square kilometres; in its shadow day was as night. Billions of tonnes of dust, gas, rock and ash scoured the mountain’s flanks in pyroclastic flows, hitting the surrounding sea hard enough to set off deadly tsunamis; the wave that hit eastern Java, 500km away, two hours later was still two metres high when it did so. The dying mountain’s roar was heard 2,000km away. Ships saw floating islands of pumice in the surrounding seas for years.

In his book “Eruptions that Shook the World”, Clive Oppenheimer, a volcanologist at Cambridge University, puts the number killed by the ash flows, the tsunamis and the starvation that followed them in Indonesia at 60,000-120,000. That alone would make Tambora’s eruption the deadliest on record. But the eruption did not restrict its impact to the areas pummelled by waves and smothered by ash.

When the sulphur hits the stratosphere

The year after the eruption clothes froze to washing lines in the New England summer and glaciers surged down Alpine valleys at an alarming rate. Countless thousands starved in China’s Yunnan province and typhus spread across Europe. Grain was in such short supply in Britain that the Corn Laws were suspended and a poetic coterie succumbing to cabin fever on the shores of Lake Geneva dreamed up nightmares that would haunt the imagination for centuries to come. And no one knew that the common cause of all these things was a ruined mountain in a far-off sea.

While lesser eruptions since then have had measurable effects on the climate across the planet, none has been large enough to disrupt lives to anything like the same worldwide extent. It may be that no eruption ever does so again. But if that turns out to be the case, it will be because the human world has changed, not because volcanoes have. The future will undoubtedly see eruptions as large as Tambora, and a good bit larger still.

Mixed in with the 30 cubic kilometres or more of rock spewed out from Tambora’s crater were more than 50m tonnes of sulphur dioxide, a large fraction of which rose up with the ash cloud into the stratosphere. While most of the ash fell back quite quickly, the sulphur dioxide stayed up and spread both around the equator and towards the poles. Over the following months it oxidised to form sulphate ions, which developed into tiny particles that reflected away some of the light coming from the sun. Because less sunlight was reaching the surface, the Earth began to cool down.

The sulphate particles were small enough to stay aloft for many months, so the cooling continued into the following year. By the summer of 1816 the world was on average about 1ºC cooler than it had been the year before—an average which hides much larger regional effects. Because the continents are quicker to cool than the heat-storing seas are, land temperatures dropped almost twice as much as the global average.

This cooling dried the planet out. A cooler surface meant less evaporation, which meant less water vapour in the lower atmosphere and thus less rain. Rainfall over the planet as a whole was down by between 3.6% and 4% in 1816.

If such numbers seem suspiciously accurate, considering that most of the world of 1816 was devoid of thermometers and rain gauges, it is because they come from recent computer modelling of the climate that seeks to mimic the conditions Tambora created. Like all modelling results, such numbers need caveats. These results, though, and similar ones from other models, can be accorded the credence that comes from having been proved right in similar situations.

The 1991 eruption of Mount Pinatubo in the Philippines was about a sixth as large as Tambora’s in terms of the volume of lava, rock and ash, and about a third as large in terms of sulphur emissions. Satellites showed that in the summer of 1992 the sulphur it had spewed into the atmosphere was reducing the amount of sunlight getting to the Earth’s surface by well over three watts per square metre; for comparison, the warming effect of the 40% increase in the atmosphere’s carbon-dioxide level since the age of Tambora is just two watts per square metre.

With the energy absorbed by the Earth reduced, temperatures fell by around half a degree in the year after Pinatubo; rainfall dropped off significantly, too. Computer models run after the eruption but before these effects became visible captured the effects reasonably accurately (though they had a tendency to overestimate the cooling). This is one of the best reasons for thinking that such models capture the workings of the climate quite well.

The historical record largely bears out what the models suggest Tambora did. Across Europe the summer of 1816 was cold and wet, and the harvest terrible. The effects were most notable around the Alps; in Saint Gallen, in Switzerland, the price of grain more than quadrupled between 1815 and 1817. Starving migrants took to the roads in their hundreds of thousands; mortality rates climbed due to starvation and disease. Death also stalked Yunnan, where Tambora’s cooling shut down the monsoon and cold days in summer killed the rice harvest for three years running.

Monsoons, which are driven by the difference in temperature between hot land and cooler sea, are particularly vulnerable to the excessive cooling of the land that volcanoes bring. Their weakening can have effects on more than crops. In his excellent account of the global impacts of the 1815 eruption, “Tambora”, Gillen D’Arcy Wood of the University of Illinois draws on the writings of James Jameson, a doctor in Calcutta, who held the lack of fresh water which followed the failure of the 1816 monsoon responsible for the cholera epidemic that swept through Bengal the following year.

Was this all down to one volcano? Not entirely; nothing in the climate has a single cause. The global climate shifts in various ways on a number of timescales, and its particular disposition at the time a volcano strikes will influence the way the volcano’s effects play out. The fact that an El Niño event—a swing in the global climate driven by the slopping of warm water east across the Pacific towards South America—was getting under way at the time of the Pinatubo eruption in 1991 undoubtedly modulated its climatic effects.


Alan Robock, an expert on links between volcanoes and climate at Rutgers University, notes a particularly intriguing initial condition that could have influenced the world’s response to Tambora. There had been another large eruption—larger than Pinatubo—just six years before. No one knows where this 1809 eruption was, but its signature can clearly be seen in the Greenland and Antarctic ice sheets. The sulphur put into the stratosphere by volcanoes shows up quite clearly in the year-by-year records of what was going on in the atmosphere that climate scientists extract from polar ice cores. These records make it possible to give dates to large eruptions in the past even if no one recorded the event at the time (see chart).

Cooling Mr Knightley

The ice cores show that the 1809 eruption was easily large enough to have had effects on the climate, and there is some evidence of cooling in subsequent years. In Jane Austen’s “Emma”, which according to Euan Nisbet, a geologist at Royal Holloway, London, seems to follow the weather of 1814, spring is remarkably late, with apple trees blossoming in the middle of June. Pre-cooling along these lines might have made some of the subsequent effects of Tambora more marked, while possibly lessening others. Some researchers believe that a number of eruptions close together might be able to trigger a climate downturn that lasts considerably longer than the few years models normally predict; a set of eruptions in the late 13th century, this idea suggests, may have been part of the reason for the subsequent global cooling known as the “little ice age”.

If the prior state of the climate system constrains an eruption’s effects, so does that of the human world. The damage done to Europe by the preceding quarter-century of revolutionary and Napoleonic war could have left it particularly vulnerable to 1816’s “year without a summer”. The situation in Yunnan would hardly have been as dire had the population not been hugely expanded by the Qing dynasty’s encouragement of new settlers.

Similarly uncaptured in models, but even more fascinating to speculate about, are the after-effects of the Tambora downturn. In America, the spike in grain prices caused by Europe’s hunger drove a wave of farmers across the Appalachians to where the Ohio Valley was enjoying far more clement weather, with barges taking exports for Europe down the Mississippi in ever larger amounts. The collapse in the grain price when Europe’s harvest recovered contributed to the American economy’s first major depression.

The historian John Post, in a study of Tambora’s effects published in 1977, “The Last Great Subsistence Crisis in the Western World”, held that the volcano reshaped European politics. The disorder that sprang up in the bad weather from 1816 to 1818, and its subsequent repression, created a climate for authoritarian rule that held sway until the middle of the century. Mr D’Arcy Wood points out that it was in the aftermath of the Tambora famines that farmers in Yunnan started to plant opium poppies, the value of which as a cash crop offered some insurance against future failures of the grain harvest.

On top of such structural shifts, there are the personal stories. If Shelley, Byron and their romantic entourage had not been cooped up in a Swiss villa by incessant rain, would they have amused themselves by writing horror stories for each other—including John Polidori’s “The Vampyre”, the first novel to deal with seductive bloodsucking aristocrats, and Mary Shelley’s “Frankenstein”, which has shaped fears of scientific innovation from that day to this? If the summer frosts of “Eighteen-hundred-and-froze-to-death” had not driven Joseph Smith, a farmer, from Norwich, Vermont to Palmyra, New York, a place of vigorous religious enthusiasms, would his son Joseph junior still have been able to find the golden tablets to which the angel Moroni led him a few years later, or would the history of Mormonism have been very different?

Reappraising the risks

And what if this happened again? In general, volcanoes are not something people around the planet worry about very much. In lists of the 40 most expensive and most lethal natural disasters since 1970 recently produced by Swiss Re, a reinsurer, no eruptions feature at all. Models of the economic losses that large eruptions could cause are nothing like as well developed as those that the insurance industry uses for storms, floods or earthquakes, because such losses have mattered little. Some reinsurers, though, are beginning to put that right.

One worry is that even quite a small eruption could cost a lot if it hit a built-up part of a developed country. A study by Willis Re suggests that an eruption of Italy’s Vesuvius like the one which took place in 1631 (a much smaller event than that which destroyed Pompeii) could lead to an economic loss of well over €20 billion ($22 billion). Most of the property damage would be down to buildings collapsing under the weight of the ash that falls on them. The 1707 eruption of Mount Fuji produced only 2% as much ash as Tambora did, but Christina Magill of Macquarie University has calculated that if both eruptions were rerun today the urban area affected by heavy ashfall would be greater in the case of the Fuji eruption, since a great deal of that ash fell on what is now Tokyo.

The other reason for thinking more seriously about the damage done by volcanoes than recent history might seem to merit is that geology shows that they need to be assessed on much longer timescales. Today’s earthquakes, storms and floods—which make up the bulk of the natural disasters that insurers worry about—are doing more damage than yesterday’s did, but that is because they hit a world in which there is more valuable property that is likely to be insured, not because the disasters themselves are getting worse. The world’s worst storm or earthquake over a millennium is not all that much worse than the worst of a century. With volcanoes things get worse and worse the deeper in time you look.


In terms of direct effects, this is still not particularly worrying for most of the world’s population. Seven out of eight people on the planet live more than 100km from any potential eruptions. The “Global assessment report” (GAR) prepared for the UN summit on disaster-risk reduction held in Sendai, Japan, in March found that 95% of those at risk live in just seven countries. Five—Indonesia, the Philippines, Japan, Mexico and Guatemala—are on the circum-Pacific “ring of fire”, where clashing tectonic plates promote volcanism as well as earthquakes; the other two are Ethiopia and Italy. Two-thirds of the exposed population is in Indonesia.

The good news for the people who are at risk is that volcanoes—unlike earthquakes—provide a fair amount of warning before doing their thing. Scientists are increasingly good at looking out for such warnings, and most volcanoes that are close to lots of people are now pretty carefully monitored, though there are exceptions—the GAR points to the Michoacan-Guanajuato cinder-cone field in Mexico as a worrying one. Satellites and seismology are likely to pick up some signs of imminent eruptions from almost all the others. When the warnings seem to merit it, action can be taken. During the 2010 eruptions of Mount Merapi in Indonesia, the largest so far this century, 350,000 people were evacuated; as a result the death toll was only a few hundred. Evacuations kept the casualties at Pinatubo similarly small.

Unfortunately, predicting really large eruptions may be harder than predicting smaller ones like Merapi’s. Before a very large eruption you can expect a volcano to have been dormant for centuries; it takes time for the infernal forces to build up. But that does not mean that the first eruption of any long-dormant volcano will be catastrophic. It might have decades of throat-clearing to go through before it really lets rip. It might go back to sleep.

It was with this in mind that geologists embarked on a project to try to understand long-dormant Pinatubo’s history soon after it started to show signs of life in 1990. They found that the volcano seemed not to be the throat-clearing type, specialising instead in dramatic eruptions. Stephen Sparks of Bristol University says that understanding did a lot to make people feel justified in calling for a big evacuation.

Wherever the next big eruption happens, though, and whether predicted or not, it will, like Tambora, have global effects—and this time there will be a greater range of them. The climate is not the only global system now open to interruption.

All disasters now reverberate more than they would once have done. Disrupted supply chains transmitted the losses from the Japanese earthquake and tsunami in 2011 far and wide; tourism meant many more Swedes died in the Indian Ocean tsunami of 2003 than in any recent disaster on their home soil. Volcanoes, though, have the added ability to interfere with one of the ways in which such connections between far-off places are supported. As Eyjafjallajokull in Iceland showed five years ago, a quite small eruption’s ash cloud can have a big impact on air traffic if it is in an inconvenient place.

A really big eruption would shut down large swathes of airspace for a couple of weeks. If the airspace in question were hard to reroute around, that would have both direct impacts on the aviation industry—Eyjafjallajokull cost it about $1.7 billion—and indirect impacts on its users—valued at about twice the direct effects in that case. The losses would not be evenly spread or easily predictable. The Kenyan women who provide most of the labour for the country’s cut-flower industry suffered disproportionately when Eyjafjallajokull kept their blooms from market.

Another problem not seen when Tambora erupted would be damage to the ozone layer. The reactions by which chlorine destroys ozone are encouraged by the sulphate particles produced by volcanoes. In the 19th century that didn’t matter; there wasn’t any chlorine in the stratosphere. Now, thanks to human intervention, there is. Pinatubo saw global reductions in stratospheric ozone levels and a marked deepening of the “ozone hole” over Antarctica. If a Tambora-scale eruption were to happen in the near future it would have even stronger effects.

Warmer house on the prairie

And then there is the climate. If, like Tambora and Pinatubo, the volcano in question is close to the equator, Mr Robock says models predict an average cooling of perhaps 2ºC in the summer of the next year over much of North America, Europe, Asia and Africa, and decreased precipitation over the Amazon, southern Africa, India, South-East Asia and China. The models also make predictions about the weather in the intervening winter: the particles that cool the surface warm the stratosphere, which sets up a strong Arctic jet stream in a particular configuration. Expect a peculiarly warm winter in America’s prairies, western Europe and Central Asia, and a very cold one in eastern Canada, the Middle East and southern China.

20150131_irm000_290What these shifts would mean for agriculture is hard to say. The experience of Tambora suggests gloom, but this is not that world. For one thing, there is more agricultural land in more places. That gives more scope for bad harvests in some regions being offset by better ones elsewhere. Both models and studies of the years after Pinatubo suggest that, for various reasons, the world’s plant life as a whole gets more productive in the cooler, drier years that follow eruptions. It is also possible that some parts of a world stressed by global warming might experience sudden cooling as less of a problem than it was after Tambora—though the dryness might exacerbate their challenges.

Another reason for tempered optimism is that the world would know what was coming. Mr Robock and his colleagues would be spreading the word before the eruption was over. Futures markets would doubtless pay attention. So, one would hope, would governments.

The Red Cross/Red Crescent Climate Centre is dedicated both to providing warnings about the human impacts of climate shifts and extreme weather and to acting as an advocate for the people who suffer from them most. It spends a lot of time looking at how to get timely warnings of the likely regional effects of El Niño events to the countries and people they are most likely to harm, along with advice on how to limit the damage. Its head, Maarten van Alst, says he thinks that the climate impacts of a contemporary Tambora might be comparable to those of the big El Niño of 1997-98, which have been estimated at $36 billion, with 130m lives affected and 21,000 lives lost. And as with El Niños, forewarned would be forearmed. Mr van Alst and his colleague, Pablo Suarez, are trying to get a programme started that would study what actions should be given priority in that lull between the eruption and the cooling that would follow.

Such vigilance could come into its own well before there is another Tambora, since there is a way for considerably smaller eruptions to have climatic effects. Eruptions that take place well away from the equator cool only their own hemisphere, and these lopsided coolings have an impact on the intertropical convergence zone (ITCZ), a belt of rain around the equator. When the northern hemisphere cools the ITCZ shifts south, and that causes droughts in Africa’s Sahel. Of the Sahel’s four worst years of drought during the 20th century, three took place after northern-hemisphere eruptions: in the year after the Katmai eruption in Alaska, (1913) and the years of and after the El Chichón eruption in Mexico (1982 and 1983).

A repeat of the Tambora-sized blast at Taupo in New Zealand that took place 1,800 years ago, on the other hand, would push the ITCZ to the north and bring plentiful rain to the Sahel. The Amazon, though, which depends on the ITCZ staying put, would have a dry few years.

For a smallish volcano at high latitudes the effects on the ITCZ would probably swamp the local and regional effects. The direct damage a full-on Tambora would wreak in a populated region would be far greater, and its hard-to-foresee effects further afield, like those Eyjafjallajokull had on Kenya, might conceivably reinforce each other in calamitous ways, multiplying the economic damage. Still, in most cases it seems likely that here, too, the climate effects would trump the rest.

20150411_FBP002But that does not mean their impacts would be as dire as those felt two centuries ago. As well as having a wider agricultural base and more foresight, the world today is more developed and better governed. A lot of the damage done in famines such as those of the 1810s comes from agricultural workers losing income at a time of price rises and governments doing nothing about it. Today the proportion of the population working the land is in most places much lower than it was then, and most governments both perceive a need to act during famines and have the capabilities to do so. There might well be a need for humanitarian interventions in the weird-climate years that followed; but such interventions do now happen.

That said, there is no reason to limit concern to Tambora-sized eruptions. There are much larger ones on offer. Some 26,500 years ago the Taupo volcano in New Zealand erupted with well over ten times the power it mustered 1,800 years ago. The odds of a really big eruption in any given year are tiny. Over a century, though, they mount up to maybe a few percent. So, though few of those alive today would perish in a rerun of Tambora, the chances of something much worse over their lifetimes cannot be ruled out. And though forewarning would help, there is no way of forestalling. Humans have huge powers over the planet. But they cannot stop a volcano whose time has come.|a