Fears that a volcanic eruption in the Canary Islands could send a 25-metre-high tsunami crashing into the US coast have been gaining momentum. But rival scientists dismiss the prediction as a hugely unlikely worst-case scenario. Who do we believe?

By Trevor Butterworth

Published in the Financial Times
October 9, 2004

The single most riveting scene of destruction from Hollywood's summer of catastrophe surely goes to the giant Atlantic storm surge that rolled through Manhattan in The Day After Tomorrow. It was not merely a sublime special effect - the sea as Jaws - it was the one disaster in a movie vilified for outlandish science that will, at least according to some scientists, eventually come to pass.

The unnerving promise of life imitating Hollywood first came to public attention in October 2000, with a BBC documentary, Mega- tsunami: Wave of Destruction. Based on research carried out by the Benfield Hazard Research Centre at University College London, the programme warned that the next eruption of the Cumbre Vieja volcano in the Canary Islands could send a massive chunk of La Palma island hurtling into the sea. The splash created by trillions of tons of rock would produce a wave of gnarly proportions that would speed across the Atlantic and lash the Caribbean and Americas.

Still, God's wrath is in the details, and the story didn't really pick up speed until August 2001, when Steven Ward, a geophysicist at the University of Santa Cruz in California, and Simon Day, a geologist at the Benfield Centre, released a model describing the potential collapse and tsunami. As with many headline-grabbing scientific claims, the study was marbled with hypothesis and inference. But the scientists' model gave birth to a monster: the worst of all plausible collapses, in which a mountainous block of the La Palma island, 15 miles long, 10 miles wide and 1 mile thick, accelerates into the sea at speeds reaching 100 metres per second.

According to Ward and Day's study, the block would remain solid for the first 10 miles, driving a wave 900 metres high in front of it. Then, over the course of the next 23 miles, the block would crumble and expand to cover 3,600sq km of ocean. La Palma would block most of the energy of the tsunami heading north toward England, whose south coast would be hit by waves five to seven metres high; but no such break would deflect a series of waves from racing across the Atlantic and depositing waves up to 25 metres high on Florida.

"Wave Goodbye to the USA", as the Daily Express put it, before bidding sayonara to Bristol and Liverpool too. Adding to the carnage, The Sun warned that Brighton and Portsmouth would be pummelled by waves, which were now twice as high as those described in the study. In a similar vein, The Independent forecast 50-metre waves for Florida, along with the Telegraph, The Times and the Mirror. Going to the seaside would never be the same again. No matter where it was reported, the La Palma tsunami roared unchecked, unquestioned and exaggerated, even though Ward and Day's paper stressed that if the parameters of their model were less severe - say, a slower slide speed or a smaller collapse - the resulting waves would be considerably less fraught with doom.

Oddly enough (or not, if one considers what actually crashed into the United States in September 2001), the La Palma tsunami failed to make much of a splash in the American media - until, that is, this August, when the Spanish government provided a news peg by ending its funding of a programme to monitor the fault bisecting Cumbre Vieja. Rather than vilify Madrid, which had paid for the Benfield Hazard Centre's research on La Palma, Benfield director Dr Bill McGuire spotted an opportunity.

"The US government must be aware of the La Palma threat," he told The Hartford Courant newspaper. "They should certainly be worried, and so should the island states in the Caribbean that will really bear the brunt of a collapse." Cable news channel MSNBC warned of a "highly deconstructive" (sic) wave of "Biblical proportions", even though the Bible, big as it is on tempests and floods, is pretty much free of deconstructive tsunamis. New York's Daily News reported "mountainous waves up to 75 feet high crashing into New York and other East Coast cities". The tsunami, explained McGuire, would sock every 300 feet of shoreline with the energy of the World Trade Center collapse.

These repeated warnings of an inevitable, if not quite imminent, catastrophe have left some American tsunami experts stirred rather than shaken. "Grossly overstating the tsunami hazard risk from a hypothetical event - as the Benfield scientists have done - really undermines the credibility of legitimate warning systems," says George Pararas-Carayannis, who served as director of the Hawaii- based International Tsunami Information Center for more than 20 years. "I get calls all the time from people worried about whether they should sell their homes," he says, "especially from Florida. Even the governor - Jeb Bush's office - called. It's ridiculous. This wave is not going to happen."

Pararas-Carayannis, who has just completed a study on volcanic flank collapses for America's National Science Foundation, believes the assumptions that went into Ward and Day's model are wrong, beginning with the idea that an eruption on Cumbre Vieja will result in a vast monolithic block of rock careering into the sea. "All kinds of friction and turbulence would slow the slide," he says, pointing, in particular, to the topography of the sea bed around La Palma, which is composed of a series of dykes from previous volcanic eruptions.

Furthermore, despite their destructive power, tsunamis generated by collapsing volcanoes or landslides are short period waves, which attenuate rapidly in the ocean. They just don't have the oomph to journey across thousands of miles - unlike the tsunamis unleashed by earthquakes on the ocean floor that may draw energy from ruptures hundreds and even thousands of miles long.

Charles Mader, editor of the journal the Science of Tsunami Hazards and an expert on wave modelling, concurs: Benfield has "constructed what we call a 'cold fusion tsunami' - it puts out more energy than it takes in from the collapse". He believes the fundamental physics of the wave predicted by Benfield are wrong, and based on his modelling of Benfield's worst-case scenario, the tsunami reaching out to North America would end up being no more than 1 metre high.

History would appear to be on the side of Mader and Pararas- Carayannis. When a powerful earthquake along the Fairweather fault in Alaska created a massive landslide in Lituya Bay in 1958, it produced a wave 516 metres high - the largest in recorded history - which shot up the spur of a facing mountain, stripping it of trees, soil and even barnacles. But by the time the rest of the wave reached the mouth of the bay, six miles away, the crest had dropped to 15 metres; and once in the Pacific, it simply dissipated.

Similarly, the tsunamis generated by Krakatoa in 1883 and Santorin in 1490BC - two of the most destructive events in recorded history - rapidly diminished in height as they travelled through the sea. Jakarta, for instance, recorded waves of about 2.4 metres from Krakatoa. And geological evidence shows that the wave from Santorin, which exploded with a force approximate to 60,000 Hiroshimas, was only 7 metres high by the time it reached Jaffa, in modern-day Israel.

"These critics are grasping at straws, or worse," says Steven Ward, who claims his numerical modelling accounts for dispersion and frequency and other factors that Pararas-Carayannis and others argue are missing or wrong. But Benfield's McGuire is a little more conciliatory. "It's a case of our tsunami modeller against theirs," he says, "and in this debate the jury is still out. But the modelling Ward has carried out on landslides leaves me convinced that he is right."

One might say that it is hardly an academic matter as to which group of tsunami-modellers is right about the potential aftershock of a collapse at La Palma; after all, some 100m people and Dollars 3,000bn of property reside on the eastern seaboard of the US. One just has to look at the only hurricane to pass directly over New York City to gauge the scale of destruction created by even a modest wave: in just one hour, the sea rose 4 metres, causing the East and Hudson Rivers to converge over Lower Manhattan as far up as Canal Street.

So it comes as a bit of a surprise to discover the Benfield Group, one of the world's largest reinsurance intermediaries (they arrange catastrophic insurance for insurance companies, many of them American) shrugging off the Benfield Hazard Centre's predictions. "Such an event would be so immense as to be uninsurable," says David Bogg, head of media relations for the group, which sponsors the Benfield Hazard Centre.

And even if the La Palma tsunami were not to be quite so immense, it would still be classified as a flood, explains Bob Hartwig, chief economist of the Insurance Information Institute in New York. And floods are, generally, not covered under most property insurance. As far as catastrophic events go, insurance companies are more worried by what lurks in the heavens. "Meteors," says Hartwig, "they're covered by insurance, but not factored into premiums. If a meteor similar to the one that hit Siberia in 1908 were to hit western Europe or the north-east United States, we could go bankrupt."