About 50 volcanoes erupt each year, but the number that might erupt is more like 1000. Over time, volcanic eruptions are not a major natural hazard to humans compared with floods, landslides, earthquakes, or hurricanes, but they are locally destructive and in some circumstances can be powerful enough to change human history. The worst volcanic eruption in terms of human casualties was the Tambora eruption of 1815, which probably killed 92,000 people directly and something over 100,000 indirectly; in comparison, the great hurricane of 1970 killed close to 500,000 people in the Ganges Delta region of Bangladesh and India, and great earthquakes killed more than that in China, in 1556 and again in 1976.

Nevertheless, volcanoes are universally recognized as a threat to people, and they can serve as very good examples of the ways in which people respond to natural hazards. We need to define some terms as we discuss these issues. The hazard is the phenomenon that poses the threat‹not the volcano itself, but its eruption. Sometimes a hazard can be assigned a probability of happening in a given time frame. Value is what is at risk‹property, lives, or money; vulnerability is the what would be lost if a particular hazard were to occur. Risk is the probability of a loss, and is a function of value, vulnerability, and hazard. A crisis occurs when risk rises to a threatening level; and a disaster occurs when a hazard causes an extreme loss.

No volcanic hazard can be prevented: but the value, the vulnerability, and the risk can be reduced (mitigated) ahead of time by actions taken to reduce loss of life, property, or money. Those actions may be passive: that is, they involve decisions not to place valuable lives or structures or investments in areas of significant risk: they involve judicious planning. They may be proactive, that is, they involve preparations that anticipate a future crisis: installing equipment to forecast a disaster, or planning evacuation procedures.

Mitigation can be reactive, that is, rushing resources to an area after the event to reduce the after-effects. Proactive policies are more likely to save lives and property, and they may often be cheaper than reactive measures. Obviously there is a trade-off: proactive measures always cost money, whereas reactive measures may never be necessary, or may be needed so far into the future that they cost nothing now. Proactive policies may be more desirable than reactive policies in some areas, as long as they are judicious and well-focussed to be most effective. However, societies seem to rely more often on reactive mitigation, which is often more wasteful.

The first step toward mitigation is to assess the magnitude of the risk posed by a given volcano. This is easier for volcanoes with a long historical record, but many dangerous volcanoes can be identified, even when there are no conventional historical records. Volcanic hazards may occur within a time-frame that varies from seconds to millenia. The human response is interesting: any hazard that occurs on a time-scale less than 20 years‹about a generation‹is known and respected. Once a hazard skips a generation, it becomes much less of a perceived threat, and the respect and response can become dangerously casual. Once a hazard recurs on a time-scale of centuries, then it is basically out of the realm of reality, no matter how much the volcano may dominate the landscape.

THE HAZARDS: Vesuvius, AD 79.

Vesuvius towers over Naples, in Italy. It is only one of a series of volcanoes in an arc that includes Stromboli, Vulcano itself, and Mount Etna in Sicily, the largest volcano in Europe. Most of the eruptions of Vesuvius are explosive, and the volcano is a massive pile of fragments of ash and dust welded together by heat and pressure, symptomatic of a dangerous volcano.

Vesuvius erupts violently, with little warning, and seems to have done so for at least 17,000 years. Its eruption of 79 AD was a classic explosive event, and remains one of the most famous of all time. The eruption totally destroyed the cities of Pompeii and Herculaneum. Their fate and their locations had been recorded, but no physical evidence of them remained, and they disappeared from maps. When rediscovered in the 1700s, the riches of both cities were mined as if they were veins of ore. Spectacular marble and bronze statues were removed by miners tunneling into the volcanic deposits, to decorate the 18th-century palazzi of Naples.

Piecing together all the evidence, events of AD 79 look like this. Late on August 23, or early in the morning of August 24, Vesuvius began erupting with moderate explosions that generated a tall cloud over the mountain. Light falls of very fine ash began, but they totalled no more than 2 inches deep, with the winds taking them mainly eastward over villages and vineyards. The noise and the sight must have alerted everyone in the region, however. We can recognize from the ash that it was generated by explosions as the magma flashed ground water into steam, cracking, shattering, and clearing out the surface rocks in the crater.

The explosive eruption of magma followed quickly. Shortly after midday on August 24, tremendous vertical explosions formed an enormous mushroom cloud over the summit. This time the cloud must have been close to 30 km high, well into the stratosphere, and southerly winds carried it over the cities and ports to the south, including Pompeii. The cloud was maintained for hours, and in that time a lot of coarse ash fell round the volcano. By late afternoon roofs of villas that were not kept swept began to collapse as the thickness of ash reached 2 feet or so, and people began to flee. (People who felt safe under strong roofs left it too late, and were trapped by later events.) It must have been very dark.

The first cloud collapsed and nuées ardentes began shortly after midnight. The earliest surges and flows set fire to vegetation and destroyed villas and vineyards as they raced down the slopes of Vesuvius. Other surges later in the night came close to Pompeii, and should have warned people even more urgently that they should leave. But by this time 8 feet of ash had fallen, and only the upper stories of the buildings showed above the ash. It was the middle of the night: lanterns would have been useless in the choking ash and the darkness, and no-one can travel through eight feet of volcanic ash. Anyone left in the city‹perhaps 2000 people, 10% of the original population‹had no choice but to wait out the night.

The fourth and fifth surges of the sequence destroyed Pompeii about 7:30 the next morning, blasting across the city at hundreds of km/hour. Roofs and upper walls were sheared off by the surge, which would then have contained timber and tiles as well as ash as it shot across the city. Everyone who had stayed in the city for safety, rather than try to flee across country to the south, or west to the shore, was killed directly, or was burned or buried beneath roof debris, or asphyxiated by dust, ash, and poisonous gases. Further ash falls and surges only entombed the dead further under the debris.

Herculaneum had a different fate. It lay upwind of the volcano, and very little ash fell on it in the afternoon and evening of August 24. But it was only 4.5 miles from the crater of Vesuvius, directly down a relatively steep slope of the volcano. It was right in the path of the first of the base surges that cascaded down the western slopes toward the sea when the eruption cloud collapsed shortly after midnight of the 24th/25th. Within 2­5 minutes the surge reached Herculaneum, engulfing the entire city in hot gas and ash. People had had a good view of the afternoon's eruptions, and many of them seem to have gone down to the waterfront to find shelter in the stone structures built into the harbor wall.

Hundreds of skeletons have been discovered crowded together in the small section of the waterfront that has been excavated, and most of the city's population of several thousand is probably entombed there. The victims were overcome there by hot gas and ash. Most of the victims of a similar surge at Mt. St. Helens choked to death on the volcanic ash; others died from the heat of the surge or from injuries from the impact of the debris in the surge. In one way or another, the first base surge at Herculaneum would have been universally lethal. It was followed minutes later by the nuée ardente, which was much hotter, hot enough to turn the boats that it covered to charcoal. The major destruction to people was thus complete in a few minutes. Later surges and nuées ardentes ripped walls and roofs off the buildings, and then simply buried the city deeper and deeper, pushing the shoreline 400 m out to sea. The ancient port and beach were covered under 70 feet of volcanic debris.

Vesuvius typically erupts as violently as this only at intervals of several centuries. Surges killed at least 4000 people in 1631, for example. But it is simply not worth leaving the region uninhabited when the interval between disasters is so great: the daily risk of death and destruction by eruption is extremely low compared with the other hazards of normal everyday life.

Nevertheless, of all volcanic areas in the world, perhaps the Naples area is the most vulnerable to a truly catastrophic disaster. We know that Vesuvius can erupt with at most a few hours warning. More than three million people live within range of pyroclastic flows that might blast down the slopes of Vesuvius in an eruption. The road and rail system would be completely inadequate to evacuate the region in an emergency, which probably accounts for the fact that there is no evacuation plan, and apparently little general concern among the populace. The last major eruption was in 1944, now 50 years ago, and that occurred in the middle of a war and caused relatively little destruction to a nation that was being destroyed city by city at the time. Meanwhile, in 1996, the housing developments climb further and further up the slopes of the volcano.


The eruption of Nevado del Ruiz in 1985 shows a worst-case scenario in action. The volcanic hazards were predicted almost exactly, and there was a period of days of increasing activity well before the major eruption, giving enough time for appropriate action by local and national authorities. Yet the eruption, when it finally occurred, came as a disastrous surprise for many thousands of people, many of whom were killed. The disaster of Nevado del Ruiz was the worst volcanic catastrophe since the eruption of Mt. Pelée in 1902. It is worth some discussion, because it carries so many lessons for future action in areas of volcanic hazard.

Nevado del Ruiz is a volcano in the Andes in central Colombia. It is snow-capped, and drains both east and west. Its lower slopes are covered with the debris of frequent previous eruptions, as layers of volcanic debris that fell as ash, and as volcanic mud-flows or lahars. The lahars in particular have flowed down the steep-sided volcano in narrow canyons, to deposit as flattened sheets in lower districts. This means that local historical records are available (at least since the Spanish conquest) to supplement the geological surveys that defined areas of past destruction and future hazard.

Nevado del Ruiz erupted in 1595, and three times between 1828 and 1845. Each eruption was explosive, and deposited airfall ash. But the major damage was caused as ash fell on the snow-covered mountain slopes to cause massive lahars‹1000 people were killed by lahars in 1845. The town of Armero was built on the mud of an old lahar.

Small earthquakes began near Nevado del Ruiz in November 1984, and on 11 September 1985 a small explosion scatted ash up to 35 km west and northwest of the volcano: a small lahar ran down the northern slope of the volcano, in sparsely settled country. On the afternoon of 13 November, another explosion scattered ash to the east as far as the Rio Magdalena. A major explosion after dark began a major eruption about 9:30 pm, and although it dropped a thin layer of ash, it was again lahars that caused major destruction. Dropping 4000 m in elevation, but constricted into narrow canyons, lahars surged out along the main river valleys around the volcano with great speed and power.

To the east, the town of Armero lies below the junction of the two rivers that drain from the volcano slopes in steep narrow valleys. Lahars surged down both valleys, beginning as very watery floods generated from melting ice and snow. They stripped soft sediments from the upland canyons, and were carrying large amounts of new ash and old mud as they spread over the flatter landscape downstream. The lahars united to flood Armero about 11 pm. Most of the 25,000 inhabitants were killed, many of them buried by the mud. To the west, a lahar down the Rio Molinas hit Chinchina about 11 pm, killing about 1000 people.

The physical damage could not have been prevented, but the casualties could have been much lower. The destruction was limited in area because the lahars were directed by topography, even in the flatted areas of the lowland slopes. Timely warning could have allowed an evacuation in a very few hours, because safer higher ground was not far away, and it could have saved literally thousands of lives. The question, then, is why evacuation was not accomplished, in the face of a long warning period and a historical appreciation that the main danger would come from lahars?


The 1991-1994 eruption of the Japanese volcano Mount Unzen provides a good example of successful prediction and evacuation in a first-world setting, though the volcano cooperated by giving plenty of warning. Unzen had been dormant for 196 years before small earthquakes began near the volcano in late 1989. A small steam eruption began in the summit crater in November 1990, and small ash eruptions then and through the winter were followed by a seismic peak of more then 30 earthquakes a day in March and April 1991. When shallow earthquakes began to occur just beneath the crater on May 13, a lava eruption was predicted, and it began on a small scale on May 15. More than 1200 people were evacuated on May 19. A thick lava dome was observed in the crater on May 21, near a steep canyon on the east side of the volcano. Avalanches of hot debris were produced from the edge of the dome as pyroclastic flows that swept down the east side of the volcano, beginning on May 24. The government issued an evacuation order on May 26 for an anticipated "danger zone," and defined a "restricted zone."

The Tragedy of June 3, 1991.‹On June 3 an explosion led to a major pyroclastic flow that killed 43 people, observers largely made up of reporters, film crew, and scientists. Three volcanologists were killed: Maurice and Katia Krafft, who had produced spectacular films and books on volcanoes, often by risking their lives very close to eruptions; and Harry Glicken, an American expert on Mt. St. Helens, who would have been sensitively aware of the dangers of pyroclastic flows. Glicken had in fact the week before compared Mt. Unzen with Mt. Pelée, the dangerous volcano on Martinique. It is clear that most of the people killed understood the danger, and chose to accept it. They were in the evacuated "forbidden zone" when they were killed.

The Eruptive Pattern of Mount Unzen.‹The eruption continued to increase in intensity, with an even larger pyroclastic flow on June 8. 12,000 people had been evacuated by June 10, and remained so through September. The lava dome continued to swell in the crater, and activity reached a new peak in mid-September. Pyroclastic flows occured at a rate of 17 a day in September, with the largest of them all on 15 September. This flow reached Shimabara city and destroyed many of the evacuated homes. Activity remained high all winter: the seismographs were recording over 200 earthquakes a day in February 1992, and there were 500 pyroclastic flows down the mountain sides in March (16 per day). The winter rains caused mud slides on the unstable ash, cutting rail and road routes. The Unzen eruption eventually slowed down and had stopped by late 1995.


The Westmann Islands are rocky volcanic islands off the southern coast of Iceland: their livelihood depends on fishing and sheep farming. The largest island, Heimaey, has the largest town, Vestmannaeyjar, built on it: Vestmannaeyjar is the major fishing port in Iceland, with the best harbor on the entire southern coast of Iceland. The island is dominated by the peak of Helgafell, a dormant volcano that last erupted 6000 years ago.

A fissure about a mile long opened on the east side of Helgafell on January 23, 1973, only a kilometer from the center of Vestmannaeyjar. It erupted spectacular fire fountains along most of its length at first, but activity soon centered on a smaller area, and a cinder cone quickly built up 100 meters high. Easterly winds brought ashfall right over Vestmannaeyjar, burying some of the homes on the eastern edge of town.

The Icelanders had emergency evacuation plans ready. Within six hours after the eruption began, 5000 people had been evacuated to the mainland, most of them on ferries and local fishing boats. Crews of volunteers remained to deal with the eruption, which dropped several million tonnes of ash on the town. The ash and lava bombs buried some buildings, caved in the roofs of others, and set fire to wooden structures. Volunteers were able to save some structures by shovelling and hosing ashes off roofs.

After a couple of weeks, the ash eruption was gradually replaced by flowing aa lava, and lobes flowed generally northward, along the east edge of town, threatening to invade and perhaps ruin the harbor. The Icelanders decided to try to control the flow. Beginning February 7, they used the normal city water supply on the front of the lava flows, and were encouraged when the rate of flow seemed to slow down.

The eruption rates remained high, however, and by the end of February the cone was 200 m high and growing. For the next month the lava moved at 3­8 m/day, reaching the edge of the harbor and beginning to flow into the sea. There was every prospect that the lava would completely fill in and obliterate the harbor that was the entire rationale for the existence of Vestmannaeyjar.

The Icelanders brought in a pump ship that could blast seawater on to the harbor face of the flow. This was supplemented in March and April by high-capacity pumps brought in from the United States. Eventually water was delivered to the lava by 43 pumps through 19 miles of pipe.

At the end of March, lava was invading the eastern part of the town, and had completely covered several blocks, including the power generator and a fish processing plant. Seventy homes were lost in one day, and many more had been damaged or partly buried by ash up to 5 m deep. Carbon dioxide gas from the lava flowed invisibly through the town. Its density kept it low to the ground, but it killed cats and stalled cars. One man was suffocated going inside a buried building.

The Icelanders kept pumping. They found that water pumping began to slow down a section of lava after about a day, and that the best results were obtained by hosing down a band about 50 m deep. This would slow down enough to form a barrier for the freer flow behind it, causing the lava flow to thicken rather than flow further. Once the lava surface was cooled, roads could be bulldozed on the surface of the flow, and pipe laid across the surface to pump water deeper and deeper into the interior of the flow.

By early May the lava flow was 10­20 m high at the front, and 40 m thick, but the flow from the vent was slackening. Lava flow finally stopped in July. From February 7 to July 10, the Icelanders sprayed over 7 million cu m of water on to the flows.

After the eruption, the harbor at Vestmannaeyjar was deeper and more sheltered than before, though erosion of the softer ash at the edge of the flow posed a silting problem, and tephra mud was still being dredged out fifteen years later. 350 homes had been destroyed, but most of the town had been saved, and some of the damaged areas were being restored. The island was 20% larger, even if the new land was a barren waste of jagged lava. More than a million tonnes of ash was cleared from the town, some of it used to build a new runway at the airport. And geothermal energy generated from the lava flow had more than paid for the damage done by the eruption.


Mount Etna rises over 3300 m to dominate eastern Sicily. Its activity has been recorded for over 2500 years, making it one of the best documented volcanoes on Earth. It lies in a major earthquake zone, and volcanic eruptions are often accompanied or punctuated by earthquakes.

Tourists have come to Etna for millenia. Viewing an active volcano can be hazardous to your health. In 1843, 36 people were killed when advancing lava reached marshy ground as they stood watching it: the resulting steam explosion was the lethal agent. Nine tourists were killed at the summit in an unexpected eruption on 12 September 1979.

The eastern slopes of Etna are particularly prone to earthquake damage, directly and by tsunamis generated by earthquakes offshore to the east. Of all the cities in the world that have suffered repeated disasters without learning anything, it's possible that Catania receives first prize.

Catania is a port city on the east coast of Sicily. It is recorded as destroyed by lava in 693 BC, but was rebuilt on the same site. The north side of the city was destroyed on 425­424 BC, and rebuilt on the same site. Roofs were caved in by loads of ash in 122 BC: the Roman authorities encouraged rebuilding by forgiving the city its taxes for ten years. Catania was threatened by a lava flow in 252­253 AD: the inhabitants rushed to the lava front carrying the veil of St. Agatha, who had been martyred the year before. It was said that the lava flow halted immediately. Much of modern Catania is now built on that lava flow. In 1169, Catania was destroyed by an earthquake: 15,000 people were killed. The city was rebuilt on the same site. In the 1370s, a lava flow again reached the city and ran into the sea: modern Catania has extended over this flow too.

In 1669, earthquakes beginning on 25 February preceded a major eruption that began on 11 March and continued to be accompanied by earthquakes. Concerned that the lava flows were heading toward Catania, 50 citizens led by Diego Pappalardo tried to divert them by opening a breach in the side of the flow, using wet animal hides to protect them from the heat. Apparently this attempt was successful: lava broke out of the flow into a new path, heading for the town of Paternò. It wasn't long before 500 armed citizens of Paternò arrived on the scene, and re-directed the flow into its original course toward Catania.

The lava arrived at the city wall of Catania on 12 April 1669 after covering 12 km down the mountain. It finally overtopped the wall, demolishing 40 m of it. It filled many of the ancient streets and covered large sections of the town. The lava finally reached the sea on 23 April, diverted, it is said, away from the remnants of the city by the veil of St. Agatha. The Norman castle was the only substantial structure to survive, and the population was reduced from 20,000 to 3000 in the aftermath. Many villages in the region were devastated: one estimate reckons that 27,000 villages were made homeless. Nevertheless, the city was rebuilt on the same site.

The attempted diversion of 1669 eventually resulted in a Sicilian law that explicitly placed liability on the shoulders of anyone who diverted a lava flow: this law lasted until 1983.

Catania suffered a very large earthquake on January 11, 1693, which killed 18,000 people out of a population of 24,000.

In 1985, Zafferana Etnea was listed as the most vulnerable of thirteen towns and large villages on the slopes of the mountain.


A sensing station on Etna showed that the mountain was inflating from March to October 1991, suggesting that magma was moving upward. In the early morning of 14 December 1991, hundreds of small earthquakes accompanied the opening of two fissures high on the mountain, and Etna began to erupt after 2 years of quiet. Lava fountains up to 300 m high were seen before dawn. Lava emerged at about 2400 m altitude. By 24 December, lava began to accumulate on the floor of a basin called Val Calanna. It advanced slowly, and did not threaten lives. But by the end of the month it had destroyed the water supply system for Zafferana Etnea, a town of about 7000 people on the flanks of the volcano. The lava flows expanded further in Val Calanna, moving eastward to within 2 km of Zafferana Etnea itself on January 1.

Barriers Don't Work.‹At that point the Italian Ministry for Civil Defense ordered the building of an earth barrier to protect the town. The barrier was built at a natural choke point at the east end of Val Calanna, where it narrows into a deeply eroded canyon. The barrier was meant to prevent or delay the flow advance, not to divert it, by creating an obstacle that would encourage lava ponding in the large Val Calanna basin, giving it time to cool before flowing downhill again.

Army and Fire Brigade personnel built the barrier in 10 days by diking the valley bottom in front of the advancing lava and accumulating loose material (lava, scoria, and earth fragments) on a small natural scarp. The final barrier was 250 m long and about 20 m high.

As it turned out, the lava flow slackened. On 7 January, lava stopped just before reaching the barrier, and Italian volcanologists declared, "there is therefore little chance of further advance of the front, as the flow seems to have reached its natural maximum length." They were wrong.

Unspecified plans were prepared in case of renewed lava advance. This was just as well, because the lava finally reached the base of the barrier on 14 March. By 7 April, it was close to the top of the 20-m barrier. That afternoon, it flowed around the south side of the barrier, and the next afternoon it spilled over the central part and began to move down the gorge of the Portella Calanna valley. Now it advanced rapidly, aided by the steep slope, covering 1 km in 5 days.

Zafferana Etnea was now clearly threatened directly, and more barriers were hastily built across the projected flow direction. The uppermost three, each 3­5 m tall, were built on April 10 and 11 and were buried and overtopped one after the other within a few days. The last barrier, about 10 m tall, was built on April 12 and overtopped on April 13. Each served as a "speed bump" in slowing the flow, which then resumed its advance after overtopping the barrier.

Explosives and Blockages Probably Don't Work.‹A second technique was used to try to stop the flow. The vents were bombarded with large blocks of cement in an attempt to plug up the main channel and divert the lava into other flows. The main flow did overtop the channel near the vent, thereby making a new, adjacent flow, which stopped near the outskirts of town.

On 14 April, lava was overflowing the last barrier, 1.5 km from the inhabited center of Zafferana Etnea and 7.5 km from the main vent. Lava destroyed 2 isolated houses above Zafferana that day and covered nearby orchards.

There was no point in trying to diverting the lava flow low on the mountain. Zafferana Etnea is scattered along the road at one altitude: any lava diverted from one populated area will flow into an adjacent populated area. This is why the Italians had attempted to stop the flow rather than divert it. After a brief pause, new lava approached Zafferana over earlier flows, and was again within a kilometer of the town by 20 April.

Success‹Divert, Then Block the Main Flow.‹Preliminary efforts had also been made to slow or halt the advancing lava by disrupting the feeder tube system high on the mountain. Experiments with directed explosives, designed to blast holes in the lava field and encourage lava breakouts, began on 13 April in the upper Valle del Bove and Val Calanna, and they continued at skylights in the main lava tube on 17, 21, and 29 April, and 4 May. The idea was to cause lava overflows and thus reduce the amount of lava moving toward inhabited areas. In early May, it was difficult to evaluate whether these experiments had affected the course of the eruption. On 4 May an overflow began from a skylight in the main lava tube at around 2100 m altitude, where blocks of cement had been dropped and explosives detonated on previous days.

Finally, the first unquestioned diversion was achieved from the main lava tube only 500 m downslope from the primary vent. An artificial channel was excavated leading away from the right-hand side of the lava tube, and then, on 27 May, engineers blew open the side of the flow with 7 tonnes of explosive. Two-thirds of the tube's lava flowed into the diversion channel. Finally, on May 29, bulldozers obstructed the main channel by pushing several hundred tonnes of large lava blocks into it. By nightfall on May 29, all of the lava output was flowing into the artificial diversion channel, and the former flow froze in position, 850 m short of Zafferana Etnae.

In March 1993 the eruption ended. It had been Etna's longest eruption in duration, and the largest for 300 years. The volcano erupted again in 2001-2002, is still active, and some people have predicted that Etna is in the early stages of a prolonged active period of eruption.

The Worst Nightmare.‹However, so far the worst nightmare has not occurred. When Etna first emerged as a volcano several million years ago, its lava and ash fell and began to pile up on the surrounding rock, which was soft clay and silt. Now there is a mountain nearly two miles high piled on that soft basement. The underlying rocks tilt slightly to the east, toward the coast, and there is some evidence that some or all of the volcano is beginning to slip to the east, sliding on that soft base of silt and clay. The worst imaginable catastrophe would occur if some or all of the volcano were actually to collapse in a giant landslide into the sea, not only killing hundreds of thousands of people directly, but causing a tsunami that would affect much of the Mediterranean coast in a much more horrific disaster.

Such landslides have occurred on the other side of the world, generated from the volcanoes of Hawaii. The debris of giant landslides lie on the deep sea floor all round the present islands, probably dating from tens or hundreds of thousands of years ago. One study even suggests that a tsunami from one of these catastrophes affected the coast of Australia. Volcanoes do collapse catastrophically, though so rarely that there has not been such a disaster in recorded history (remember that recorded history is less than 100 years in some parts of the world!).

Obviously there is nothing one could do in the face of a threat like this. One would simply hope that there would be enough warning time to evacuate some of the people threatened.