麻豆传媒

ON THE evening of 31 October 1755, the Portuguese capital of Lisbon was basking in its reputation as one of the world鈥檚 great cities. It was the thriving centre of a growing empire, a bustling port and home to a wealth of art and culture. Twenty-four hours later it lay in ruins, effectively wiped from the face of the Earth by one of the most destructive earthquakes ever recorded (see 鈥淩esurrecting Lisbon鈥�).

Seismographs had not yet been invented, but contemporary descriptions of the shaking and the level of destruction point to a massive earthquake. The US Geological Survey estimates a magnitude of 8.7, which makes it the largest earthquake ever recorded in Europe, and almost 30 times as powerful as the 1906 quake that destroyed San Francisco. Reports of shaking came from as far afield as Switzerland and northern Italy, and in Finland lake waters oscillated in response to the shock.

The quake occurred not below Lisbon itself but beneath the Atlantic Ocean a few hundred kilometres from the Portuguese coast (see Map), where it jolted a huge area of seabed and triggered a giant avalanche of sediment. This generated a series of tsunamis that rapidly advanced on the Portuguese capital. Within an hour, three great waves over 13 metres high surged up the Tagus estuary, crashed through the sea wall and penetrated far into the heart of the city. Already-tottering buildings were flattened and thousands of stunned and injured survivors drowned. In Lisbon alone, the quake and tsunamis combined probably killed at least 30,000 people, and possibly as many as 75,000.

The devastation and loss of life due to the huge waves, however, stretched far beyond the city limits. Like the Indian Ocean tsunami of 26 December 2004, the waves travelled far from their source, battering the Atlantic coasts of Portugal and southern Spain, where they caused severe damage and led to many deaths at C谩diz. Morocco, Madeira and the Azores were also hit hard, and several hours after the quake the tsunamis reached the Caribbean, where 7-metre waves were recorded in the Lesser Antilles. There are no contemporary accounts from the US, but models suggest that the east coast must have been struck by waves at least 2 metres high. Tsunamis were also reported in Newfoundland in Canada, France, Belgium, the Netherlands, Ireland, south Wales and the south coast of England.

With the 250th anniversary of the Lisbon earthquake approaching and the Indian Ocean tsunami fresh in people鈥檚 minds, there is considerable interest in the future threat from tsunamis in the Atlantic 鈥� in particular to the densely populated North Atlantic rim. Even before the scale of death and destruction in the Indian Ocean was established, the US National Oceanic and Atmospheric Administration (NOAA) announced its intention to establish an Atlantic tsunami warning system similar to the one it set up in the Pacific Ocean in 2003. The system would consist of five DART (deep-ocean assessment and reporting of tsunamis) buoys, which use pressure detectors to measure changes in water depth overhead, four off the east coast and a fifth in mid-ocean.

Meanwhile, the Canadian government has decided to buy a supercomputer capable of modelling the tsunami threat to the country鈥檚 shores. In Nigeria, scientists have called for the establishment of a tsunami early warning system, and in June the UK Department of Environment, Food and Rural Affairs (DEFRA) published a comprehensive report that evaluated the threat posed by tsunamis to the UK.

But just how serious is the threat? Although the ocean floor of the Atlantic is much less prone to tsunami-causing quakes than those of the Pacific or Indian oceans, and only 2 per cent of recorded tsunamis have occurred in the Atlantic basin, the threat is serious enough for NOAA to decide that it wants the warning system in place within 18 months. And as the Lisbon disaster shows, the potential for lethal and destructive waves is real.

There are other examples, too. On 18 November 1929, a magnitude 7.2 earthquake 300 kilometres south of Newfoundland triggered a submarine landslide with a volume of around 200 cubic kilometres. The water displaced by the slide created tsunamis up to 13 metres high, which killed 28 people in Newfoundland and Nova Scotia and were recorded in Portugal and the Azores.

There is also evidence of a major inundation striking Britain in 1607. According to Simon Haslett of Bath Spa University, UK, and Ted Bryant of the University of Wollongong in New South Wales, Australia, on 30 January that year a wave up to 7.5 metres high devastated the Bristol Channel and parts of south Wales, taking at least 2000 lives. The reality of the event is unquestioned 鈥� it is recorded in contemporary documents 鈥� but a definitive cause remains to be determined.

The damage described in the documents could equally well have been caused by a severe storm surge, such as the one produced by hurricane Katrina. What is more, tsunamis and storm surges leave a similar legacy 鈥� large boulders above the normal high-water mark, a layer of sand dumped inland and coastal erosion. However, the reports talk of the 1607 inundation occurring during fine weather, and a possible source for the tsunami has been highlighted by Roger Musson of the British Geological Survey, who points to a large, ancient submarine fault off south-west Ireland.

鈥淎bout 7900 years ago, an enormous mass of marine sediment collapsed off the coast of Norway, sending waves up to 30 metres high towards the Shetlands鈥�

A quarter of a century earlier, in 1580, the British and French coasts appear to have taken another battering. Following a magnitude 5.9 earthquake somewhere in the North Sea, giant waves sank 20 ships and killed hundreds in the towns of Dover, Calais and Boulogne. Again, though, it may well be that the waves were due to a severe storm that occurred shortly after the quake. Certainly, the earthquake would seem too small to generate a significant tsunami even in the immediate vicinity.

We need to go back to prehistoric times to find the first indisputable evidence for a major tsunami in the North Atlantic. Stein Bodevik of the University of Troms酶 in Norway and colleagues identified two deposits in the Shetland Islands that date back 5500 and 1500 years. The deposits extend too far inland to have been caused by a storm surge, and so are tsunami deposits. The source of the tsunami, though, is not known.

The best-understood and most closely studied Atlantic tsunami was triggered not by an earthquake but by an underwater landslide. About 7900 years ago an enormous mass of marine sediment with a volume far in excess of 1000 cubic kilometres collapsed from the continental shelf off the west coast of Norway. Known as the Storegga slide, it was one of the world鈥檚 biggest landslides. A tsunami deposit linked to the collapse was discovered in eastern Scotland in the 1980s by David Smith and Alastair Dawson, then at Coventry Polytechnic in the UK, and Dave Long of the British Geological Survey (麻豆传媒, 4 August 1990, p 46). Similar deposits have since been identified all down Scotland鈥檚 east coast, as well as in Norway, the Faroe and Shetland Islands, Iceland and eastern Greenland. The waves that dropped the deposits were on the order of 3 to 6 metres high in Scotland, up to 12 metres along the Norwegian coast and a whopping 27 to 30 metres in the Shetlands.

Clearly then, tsunamis do occur in the Atlantic, some large enough to be massively destructive. So what are the prospects for another Storegga-sized event, or a rerun of the Lisbon tsunami?

First of all, it looks as if there鈥檚 no need to worry too much about the Norwegian continental shelf. Studies undertaken on behalf of oil companies suggest that the Storegga collapse and similar landslides have largely exhausted a huge pile of unstable sediment that accumulated during the last ice age. Little remains to trigger another large tsunami. But one could occur elsewhere around the Atlantic rim, and now there is another factor to consider: climate change.

鈥淎 tsunami triggered by an earthquake off the coast of Portugal remains a real risk. With the coast now densely populated, that could be devastating鈥�

In the sediment around the margins of the Atlantic are large deposits of solid methane known as gas hydrates. As sea temperatures rise, these might start to break down, causing instability and ultimately failure of the surrounding sediment. Similarly, rapid melting of the Greenland ice sheet has the potential to trigger earthquakes as the newly exposed crust 鈥渞ebounds鈥� after the weight of the ice has gone.

A tsunami triggered by an earthquake off the coast of Portugal remains a real risk. A particular area to watch is the Gorringe Bank, a ridge astride the Azores-Gibraltar fault. Radiocarbon-dated sediments in the C谩diz region of Spain suggest that events similar to the 1755 quake also occurred in 216 BC and perhaps also 60 BC. With the coast between Lisbon and Gibraltar now densely populated, a repeat performance could be devastating. Warning times would be short, too.

Inevitable collapse

The biggest threat, though, is the inevitable collapse of the unstable western flank of Cumbre Vieja, a large volcano on La Palma, one of the Canary Islands. At some point 鈥� it is impossible to predict when 鈥� this 500-cubic-kilometre chunk of rock is going to break off and plunge into the sea. And the consequences could be devastating.

Steve Ward and Simon Day of the University of California, Santa Cruz, predict an initial bulge of water an astounding 900 metres high if the entire west flank collapsed in one go, subsiding to 20-metre waves along the east coast of North America and perhaps 7 to 10 metres along the western European coast. These wave heights are comparable to those that brought devastation to the margins of the Indian Ocean last December. The worst affected areas would be the Canary Islands themselves, where waves are predicted to be well in excess of 100 metres high, and the coasts of north-west Africa and northern Brazil, which will be battered by waves several tens of metres high.

The recent report on UK tsunami risk, compiled by the British Geological Survey for DEFRA, concludes that waves created by a Cumbre Vieja collapse would have a maximum height of 2 metres when they reach the shores of the UK. This is enough to be damaging, but not enough to cause major destruction and loss of life. But worryingly, this prediction is based on the idea of gradual collapse 鈥� the 鈥渟low failure鈥� model put forward by Russell Wynn and colleagues at the National Oceanography Centre in Southampton. In this scenario the unstable rock breaks up and enters the sea bit by bit over a period of time, never rapidly enough to trigger a large tsunami. It is, however, very much a best-case scenario that flies in the face of considerable evidence from elsewhere for much more catastrophic behaviour.

In 2004, Gary McMurtry of the University of Hawaii and colleagues announced an astonishing discovery. They showed that a giant underwater landslide from Hawaii鈥檚 Mauna Loa volcano around 120,000 years ago generated a tsunami that reached more than 400 metres up the flanks of the neighbouring Kohala volcano and penetrated over 6 kilometres inland. No break-up and gradual entry into the sea here, but a wholesale and devastating failure of a massive chunk of volcano generating an equally huge tsunami.

Evidence for a similar catastrophe can also be found in the Canary Islands themselves. On La Palma鈥檚 neighbour El Hierro there is a huge slab of rock that at some point in prehistory slipped 300 metres before grinding to a halt. Critically, it did not break up, and it moved so fast that it melted the rock along the slip surface. More recently, Juan Acosta of the Spanish Institute of Oceanography in Madrid and colleagues identified an enormous block of rock sitting on the seabed off the west coast of the volcanic island of Fuerteventura. At 22 kilometres long and 11 kilometres wide, it is comparable in size to the unstable west flank of Cumbre Vieja and clearly arrived at its present position largely intact.

It seems, then, that giant volcanic landslides collapse pretty much in one piece, move very rapidly and make a very big splash. When this will happen at Cumbre Vieja is anyone鈥檚 guess. But when it does, we should be prepared for the worst, not hoping for the best.