Predicting Disaster: A Risky Business

24 October 2012

Predicting Disaster: A Risky Business

By Gwynne Dyer

Six years in jail and an average fine of over a million dollars: that was the punishment given to six Italian scientists on 22 October for getting their earthquake advice wrong. So what will the expert geologists and vulcanologists in Italy say the next time they are asked about the likelihood of an earthquake? They will refuse to say anything, of course.

More than 5,000 scientists have signed a letter supporting their colleagues who found themselves standing trial for manslaughter in the medieval city of L’Aquila, where 309 people died in an earthquake in 2009. But the case is a bit more complex than it first appears.

People always look for a scapegoat when disaster strikes, and it’s understandable that the bereaved people of L’Aquila wanted someone to blame. Most of them were glad when the six Italian scientists were convicted: at least somebody was being punished for the crime. But it wasn’t exactly the crime that those 5,000 foreign scientists thought they had been accused of.

Even lawyers and judges know that you cannot predict an earthquake with any certainty. What the six were actually accused of was being too reassuring about how likely an earthquake was.

There were hundreds of small shocks around L’Aquila in the weeks before the big one struck, and the six scientists were sent to the city to assess the level of danger. They judged the risk as minor, and one, foolishly, said there was “no danger”.

On the basis of this scientific advice, it is claimed, thousands of citizens decided to sleep in their houses rather than outside – and 309 of them were crushed in their houses a week later when the magnitude 6.3 quake brought them down. So the scientists’ crime was not a failure to predict the quake, but a failure to state clearly that it COULD happen.

It’s still a stupid charge. Half of the really big earthquakes are preceded by a flurry of smaller shocks, true – but such clusters of small shocks are quite common, and only 5 percent of them are followed by a major quake. So the scientists were caught on the horns of a familiar dilemma.

Fail to issue a warning before a big quake, and you will be discredited (and maybe, if you are Italian, charged with manslaughter). But issue warnings every time there is a 5 percent risk, and you will cause 19 needless mass evacuations for every necessary one. You will be “crying wolf”, which is usually counter-productive.

The scientists’s conviction will probably be reversed on appeal, bringing this whole foolish episode to an end. For the rest of us, however, this just illustrates how hard it is for human beings to deal sensibly with big but incalculable risks.

The biggest incalculable risk of a purely natural order that we know about is the mega-tsunami that will be unleashed when the western flank of Cumbre Vieja volcano on the island of La Palma in the Canaries slides into the Atlantic Ocean. In an eruption in 1949, a chunk of rock about 500 cubic km. (120 cubic miles) in size, with a mass of 150 billion tonnes, became detached from the main ridge and slid two metres (7 ft.) down towards the sea.

This is bad news for people living around the Atlantic Ocean. In some future volcanic eruption (there have been six in the past 500 years), that whole mass may slide all the way into the ocean and generate a tsunami that would initially be about 600 metres (2,000 ft.) high.

It would travel outwards in an expanding circle at some 1,000 km. per hour (600 mph), destroying everything on the western coast of Africa in one hour. It would inundate England’s south coast in three, and reach the east coast of the United States, Canada iand Cuba in six. Brazilians would have to wait a little longer. The waves would reach up to 20 km. (13 mi.) inland in low-lying areas. Many tens of millions would die.

So let’s imagine that there’s another eruption on Cumbre Vieja, and a committee of global experts is convened to watch the western flank for signs of movement. Should they advise evacuation along all the vulnerable coasts? That’s several hundred million people. Who will give those people food and shelter? How long must they stay inland? And the economic damage would be huge.

The experts can’t wait until the last minute to give their advice: you can’t evacuate the entire US east coast in six hours. If they advise evacuation, and nothing bad happens, they will be the most unpopular people on the planet. If they don’t, and the worst does happen, they will be seen as guilty of mass manslaughter, just like the Italian scientists at L’Aquila.

Since it will always be much likelier that no catastrophe is going to happen this time, the experts will almost certainly issue reassuring statements intended to keep people in their homes. Just like the Italian scientists. And yet some day, next week or a thousand years from now, that mass of rock on Cumbre Vieja will really fall into the sea. Damned if you do, and damned if you don’t.


To shorten to 725 words, omit paragraphs 3, 4 and 9. (“People…was”, and “The scientists’…risk”)