How Mega Tsunamis Are Formed

RECENT EVENTS IN Japan and southeast Asia have ensured we are all too aware of the tsunami-triggering potential of enormous, submarine earthquakes.

Less well known, however, is the fact that volcanoes are also very effective tsunami generators, with ‘flank collapse’ – when a sizeable chunk of a volcano collapses into the sea – spawning tsunamis that have taken close to 20,000 lives in the last 400 years.

In 1979 a tsunami caused by the collapse of Indonesia’s Iliwerung volcano took several hundred lives. And in 1792, the failure of part of Japan’s Unzen volcano launched a tsunami that battered coastal villages, resulting in 14,000 deaths. Now scientist say a collapse of the Cumbre Vieja volcano in the Canary Islands could create a tsunami that would devastate the East Coast of the USA and hatter the UK’s western shores. It’s the subject of a new BBC dramatisation called Mega-tsunami, airing this month on BBC Two.

More often than not, volcanoes are not solid, unmoving bastions of strength, but wobbly piles of ash and lava rubble just looking for an excuse to collapse.

Mega TsunamisThe evidence for this is all around us, with many hundreds of massive landslides now identified at volcanoes right across the planet. Typically, these leave behind enormous collapse scars, such as the great rocky amphitheatre torn from the east flank of Mount Etna and, most recently, the 3km-wide bite taken out of the north flank of Mount St Helens by the landslide that triggered its 1980 eruption.


Once a volcano’s flank has become unstable, it can be shaken off by an earthquake, pushed off by an injection of new magma, or sometimes just fall off as the flank becomes too steep. It doesn’t even need an eruption to start things moving. As the extraordinary footage of the collapse of Mount St Helens north flank reveals, once a volcanic landslide gets going, there is no stopping it. The mass of moving rock hurtles downslope at velocities matching those of a Formula 1 racing car, typically travelling many kilometres before coming to rest and obliterating anything and everything in its path.

With a volume of a couple of cubic kilometres, the Mount St Helens landslide was just a tiddler. Compare this with the 45km3 prehistoric flank collapse at neighbouring Mount Shasta, or the staggering 5,000km3 volume of the Nuuanu landslide, which took an enormous bite out of Hawaii’s Ko’olau volcano a few million years ago.

Nuuanu landslideBut the Nuuanu landslide is just one of around 70 mammoth collapses whose debris is scattered about the sea floor surrounding the islands of Hawaii. Another, known as the Alika 2 slide, formed about 100,000 years ago when a monumental chunk of the Big Island’s Mauna Loa volcano fell into the Pacific, sending a towering tsunami surging throughout the archipelago. Marine shell deposits now stranded up to 60m above sea level on the flanks of neighbouring Kohala volcano testify to colossal waves, but this is only half the picture. Over the last hundred millennia, the Kohala volcano has actually been subsiding, so that the true height of the tsunami looks as if it was nearer 400m. That’s a quarter as high again as London’s Shard – Western Europe’s tallest building.

It’s difficult to grasp the impact such an event happening today would have on our world’s crowded coastlines. But with major collapses of ocean island volcanoes taking place somewhere on the planet as often as every 10,000 years or so, it may not be too long before we find out.

One prime candidate for the next big collapse is the Cumbre Vieja volcano on the Canary Island of La Palma, which has been behaving in a rather disturbing manner since an eruption in 1949. Then, the eruption was accompanied by some particularly strong earthquakes beneath the volcano’s western flank, together with the opening of a 3km-long line of fractures along the crest of the volcano, down which part of the west flank dropped seawards by a few metres. This might have been no big deal but for the results of a survey of the volcano, undertaken by my research team between 1994 and 1997. It hinted that the west flank of the Cumbre Vieja might still be on the move, albeit extremely slowly.

Cumbre Vieja volcanoMore than a decade on and we are convinced that something a bit special is going on. Comparisons between GPS readings undertaken in 1997 and 2007, designed to monitor relative displacements, reveal some astonishing results. Not only is the entire west flank of the volcano deforming independently of the rest of the edifice, but over the intervening decade it moved westwards and upwards by more than 10cm. This may not sound like much, but it means that Cumbre Vieja’s west flank qualifies as the most recently activated giant landslide, albeit moving at an incredibly slow speed… for now.


While the current rate of seaward movement is tiny, all the evidence points to the likelihood that at some time in the future, the west flank of Cumbre Vieja will fail, plunging into the North Atlantic. No-one has ever observed the formation of a megatsunami as a consequence of such an event, but we can build a realistic picture of what it might look like.

For a terrifying worst case, which envisages 500km3 of rock sliding into the sea at 100m/s, a computer model built by Steve Ward at the University of California, Santa Cruz predicts an initial bulge of water close to a kilometre high, subsiding into waves merely hundreds of metres high. This may sound like something from a science fiction film, but tsunami deposits identified more than 180m above sea level on the neighbouring island of Gran Canaria show that something similar has happened before. A wave of this size would have been unleashed after one of at least 14 collapses of volcanic flanks that occurred in the archipelago during prehistoric times.

Of course, it’s perfectly possible that the future collapse, when it comes, will be smaller than the worst case, as a consequence of which the resulting tsunami will be reduced in height.

MegatsunamiCertainly, marine geologist Dr Russell Wynn and colleagues at the UK’s National Oceanographic Centre in Southampton have provided evidence for at least one ancient collapse in the Canaries occurring in a piecemeal manner. If the Cumbre Vieja were to fail in the same way, with the rock making up the flank sliding into the sea bit by bit over a period of hours, then although there would be more tsunamis, each would be smaller and less destructive.

However, everything we know about past volcano collapses points to the fact that when a volcano sheds a part of its flank, it usually does so very quickly, with most of the material sliding off in one go. While we can’t be certain, the weight of evidence suggests that when the west flank of the Cumbre Vieja eventually plunges into the North Atlantic, the resulting tsunami will be prodigious and unprecedented in the historical record.


This news is not good for the Canary Islands, but what about further afield?

One of the controversies about tsunamis is just how well they conserve their energy as they travel further and further from the source, and how high they will be when they reach distant shorelines.

The original tsunami model for a future collapse of the Cumbre Vieja, published in 2001 by the aforementioned Steve Ward and University College London’s Dr Simon Day, supports the idea that sufficient energy is conserved as the tsunami spreads out across the North Atlantic so that the waves remain big enough to cause major destruction as far away as the UK, West Africa and even the east coast of North America. Other experts disagree, suggesting that a future megatsunami would lose energy more rapidly as it travelled, resulting in waves along the east coast of North America that were just a few metres high.

Ward and Day, however, stick to their guns. As Steve Ward observes: “The 2011 Japan tsunami struck a few hundred kilometres of coast. Imagine the same level of damage spanning shores 10,000km long – from Nova Scotia to Brazil, from Casablanca to Keflavik.”

Japan tsunamiWard and Day do have some independent support for their predictions. Far out in the middle of the Atlantic, the island of Bermuda sits bang in the path of any tsunami heading west. In addition to its eponymous shorts, the island is also known – among geologists, at least – for some enigmatic deposits exposed along its coastline. These take the form of shell and coral debris resting 20m above sea level. Gary McMurtry of the University of Hawaii and Dave Tappin of the British Geological Survey are of the opinion that this material was dumped by a passing tsunami hundreds of thousands of years ago.

The obvious source is an ancient collapse of one of the Canary Island volcanoes. Should this prove to be the case, then it will go a long way towards supporting the idea that a future collapse of the Cumbre Vieja could present a serious threat to the entire Atlantic Basin.


But when is this all going to happen? Unfortunately, we don’t know. The average frequency of flank collapses in the Canary Islands is about every 100,000 years, but given that the Cumbre Vieja’s west flank is already on the move, it is likely to meet its watery grave much sooner. We might well have to wait thousands of years, but the collapse could happen at any time. As Dr Simon Day points out, this is most likely to happen during an eruption, “when the volcano is subjected to the additional forces imposed upon it”.

In theory, there should be warning signs in the form of an acceleration in the rate of sliding. Provided the volcano is being monitored, this should allow an alert to be raised, ensuring the evacuation of threatened coastlines. In the meantime, don’t be dissuaded from visiting the beautiful island of La Palma. Soak up the Sun and visit the volcano.

You’d be very unlucky to get caught up in a megatsunami-forming landslide.