Gwynne Dyer: Atlantic needs wave warning device too

"We may have severely underestimated the level of the tsunami hazard along the margins of the Atlantic Ocean," says an unnamed researcher at the Benfield Hazard Research Centre at University College London (quoted in the Financial Times).

Scientists do jump on the bandwagon at times like these, but you can hardly blame them - it's the only time they can really get our attention.

By the end of this year, the countries of South Asia will almost certainly agree to install a system to give advance warning of tsunamis such as the one that killed an estimated 150,000 people on December 26.

The countries around the Pacific Ocean have operated a similar system since 1968, and one could be installed around the Indian Ocean for about $14.25 million, although individual countries would then have to spend further amounts to get timely warnings to their people who live in low-lying coastal areas.

Such a system wouldn't have saved many people in northern Sumatra, where the tsunami struck within half an hour of the mighty earthquake (at 9 on the Richter scale, the biggest in 40 years) that occurred just off the west coast.

But it would have given people in Thailand and Malaysia more than an hour to get to higher ground, and more than two hours in Sri Lanka and India.

The property damage would have been the same, and five or six million people would still be homeless, but as many as 50,000 lives could have been saved by a proper warning system.

Another time, when the undersea quake didn't happen so close to a densely populated coast, the great majority of the lives at risk might be saved. So why is no system in place? Basically, because big earthquakes are rare in the Indian Ocean.

The Pacific tsunami warning system has been in business for almost 40 years because the Pacific rim - the "ring of fire" - has more than half the active volcanoes and earthquakes on the planet, and an average of 10 tsunamis a year.

No quake has caused a big tsunami in the Indian Ocean since Krakatoa exploded in 1883, and human beings don't readily respond to merely theoretical dangers.
Which brings us to the Atlantic, an area of even less seismic activity. The last subsea earthquake big enough to cause a tsunami in the Atlantic was the great Lisbon quake of 1755, whose 60,000 victims included tens of thousands killed by the tsunami that struck the city only minutes after the original shockwaves. But there is another problem in the Atlantic - La Palma, in the Canary Islands.

It was Dr Simon Day, of the University College of London's Benfield Hazard Research Centre, in co-operation with the Swiss Federal Institute of Technology in Zurich, who modelled what will happen when one day, probably during a volcanic eruption, Cumbre Vieja, the western half of La Palma, falls into the sea.

Five hundred cubic kilometres of rock weigh about 500 billion tonnes, and when they plunge into the ocean, the model predicts, they will cause a mega-tsunami. It will be 650m high in the Canaries themselves, and still "several tens of metres high" when it strikes the east coasts of the United States, Canada, Cuba and Brazil eight to 10 hours later.

"There's no 'if' about it," says geophysicist Bill McGuire, also of the Benfield centre. "It will happen; it's just a question of time." But although La Palma is the most volcanically active in the entire Atlantic region, it could still be quite a lot of time before Cumbre Vieja falls off.

The volcano has erupted in 1470, 1585, 1646, 1677, 1712, 1949 - that was when the whole western side of La Palma slid down 4m, alerting the scientists - and most recently in 1971. It has probably been erupting every century since long before people started recording it.

Yet Dr Day estimates there is only a 5 per cent chance that Cumbre Vieja will collapse in any given century. So what do you do if the volcano erupts again?

You certainly don't evacuate the whole east coast of North America, most of the Caribbean and northeastern South America, plus all the coastal areas from Guinea to Morocco in West Africa and from Spain to Ireland in Europe.

You can't make hundreds of millions of people refugees for months or years and cripple entire national economies just because a volcano is erupting in La Palma and there's a 5 per cent chance of disaster.

But a warning system that gave people a decent chance of survival if the worst happened would certainly help.

For many coastal areas, a couple of hours' warning of a tsunami is all that would be needed for people to make it to higher ground. For some large coastal cities, where traffic bottlenecks prevent rapid evacuation, refuges in the upper stories of tall building that can withstand a tsunami might be needed, but nothing too expensive or elaborate.

We are talking about human nature here, and against a low-probability catastrophe, people will pay only for low-cost insurance.