Feature

Giant Avalanche Alert!

Imagine an avalanche the size of Coromandel Peninsula falling off Mt Cook.

Keith Lewis and Jean-Yves Collot

About 170,000 years ago, a piece of seabed as big as the Coromandel Peninsula fell down a three-kilometre high submarine scarp off the East Coast. The giant avalanche then roared for 50 kilometres out across the flat, abyssal ocean floor. Did it produce a giant tsunami? Could it happen again?

In the Southern Alps, big landslides and rock avalanches can send hundreds of millions of cubic metres of mountain-side crashing into the valley below. Huge ones may even contain more than a billion cubic metres or one cubic kilometre of debris. An ancient one in Fiordland involved about 26 cubic kilometres and may be the world's largest onshore landslide.

In contrast, marine geological surveys over the last decade have shown that submarine "landslides" can be orders of magnitude bigger than any on land. Lubricated by water, they can involve the catastrophic collapse of thousands of cubic kilometres of seabed. New Zealand and French marine geologists recently discovered the debris of a truly awe-inspiring avalanche off the East Coast near Ruatoria. The discovery posed questions, not only of "when" and "how", but also of what are the risks to people living close to the east coast.

Huge Slip

Most people have seen slides on muddy hillsides after winter rains. The image that appeared on the French research vessel Atalante's state of the art "swath-mapping" monitor looked just the same, except for the scale. Instead of being a few hundred metres across, the "slide" off Ruatoria was 40 km wide and 100 km long, wider and longer than Coromandel Peninsula. The top of the slide-scar is off the mouth of the Waiapu River near East Cape. The jumbled mass of debris at the toe extends 50 km out across the flat floor of the Hikurangi Trough, which is about as deep as Mt Cook is high. Within the mass of broken rock and mud there are dozens of individual blocks larger than Mt Ngauruhoe. The largest, which is near the front of the mass of debris, is the size of Mahia Peninsula.

The huge landslide debris path can be seen spilling out across the Hikurangi Trough.

If we imagine, for a moment, the Coromandel Peninsula suddenly cascading from the top of Mt Cook with sufficient velocity to carry it 50 km out across a flat plain, we get some idea of enormity of the event. Even so, it was not the biggest on earth. An even bigger one off Hawaii travelled 160 km out across an abyssal plain, even moving slightly uphill. These were clearly catastrophic events. They were not creeping slides. They were huge avalanches. It is a frightening thought.

An even more frightening thought is what happens at the sea surface during one of these avalanches. On a miniature scale, fishermen working around Kaikoura Canyon have reported a flat, calm sea suddenly erupting into violent waves that threatened to capsize their boat. They think this occurs when small pieces of the groper-rich canyon-wall beneath them fall away.

Tsunami Generator

What would happen if a vast section collapsed? Initially, the overlying seawater would be sucked down with it, but it would bounce back, forming tsunami waves very much bigger than those the Kaikoura fisherman experienced. A relatively minor slope collapse after an earthquake off New Guinea in July 1998 produced a tsunami that appeared along the coast at dusk as a red-crested wave 10-15 m high -- as high as a three-storey building. For over 2,000 people, who had looked from their homes at the curious withdrawal of the sea that often precedes a tsunami, it was the last thing they ever saw.

Perhaps the most dramatic evidence of a tsunami generated by a submarine landslide is from Hawaii. Long before the first Polynesians arrived, the western flank of the 5 km high submarine volcano that protrudes from the sea as the island of Hawaii collapsed. It generated a wave that many people think washed white coral boulders to a height of nearly 200 m up the black lava slopes of neighbouring Lanai Island and removed the thick, red, tropical soils to a height of over 370 m -- higher than the Auckland Sky Tower. It has been suggested that the same wave devastated coastal areas in Australia, although this is debated.

The wave that ran up the side of Lanai was nowhere near Sky Tower height in the open ocean. As an example of this, in 1958, a landslide into a mercifully unpopulated Alaskan fjord generated a wave 60-80 m high (the height of 20-storey high-rise) but, as it travelled down the fjiord at 180 km/hour, it washed up the side of a projecting spur destroying pine trees to a height of 524 m. Incredibly, the crew of a fishing boat in the mouth of the fjord, who watched it coming, survived the wild ride over the bar and out to the open ocean.

In both the Hawaiian and Alaskan examples, the tsunami was enormous because the head of the avalanche was near or above sea-level. What of the Ruatoria Avalanche? Was there a tsunami? The search for evidence has begun onshore, but searchers will need to be very lucky because most of the evidence will have been eroded away. The soft muds of the East Coast eroded more rapidly than the basalt and coral of Hawaii, even before bush was destroyed. In addition, the evidence is older than on Hawaii.

Layers of volcanic ash in cores from the mud now blanketing the avalanche can be dated and show that the avalanche is about 170,000 years old, pre-dating the last ice age. If, as we suspect, there was a tsunami off Ruatoria, how big was it? Frankly, we don't know. Probably smaller than the Hawaiian one because the head of the main avalanche was not close to sea-level. The main collapse was probably over 500 m deep, so that the effect on the sea-surface was partly dissipated in the surrounding water.

Computer models have been developed to estimate the sizes of tsunamis generated by earthquakes, including those on fault lines that rupture the seabed. Models are also used to predict what happens to tsunami waves when they reach headlands and bays of the adjacent coast. There, the waves slow down in the shallows, but their height increases, causing them to run up the shore to many times their height in open water. The problem is that modelling tsunamis from submarine avalanches is far more complicated because the drop in the seabed moves down the slope behind a bulge of collapsing debris and pressurised water. This causes a whole series of waves that interfere with one another in complex ways.

Will it Happen Again?

Was this a "one-off", or are giant avalanches likely to happen again around New Zealand? If so, where? To answer that, it helps to understand why the Ruatoria Avalanche occurred where it did. Ruatoria lies near the boundary of two of the earth's vast crustal plates. To the east, the Pacific Plate creeps towards and dives under the edge of the North Island at about 45 mm per year -- roughly the rate at which toe-nails grow.

The Pacific plate is like a conveyor belt, but it is not smooth. It has long-extinct volcanic seamounts, many rising more than one kilometre above the abyssal seabed. When these reach the toe of the steep slope off Ruatoria, they are not scraped off like conveyor-belt luggage. Instead, they plough into and then under the steep submarine slope as though it were made of feta cheese. Each one penetrates like a slow motion bullet, crumbling the feta-like rocks, which collapse back into the tunnel in its wake. The result is a trough that marks the passage of the seamount into and under the continental slope. These impact-troughs occur widely around the Pacific, accompanied by small avalanches.

What makes the New Zealand margin different is that the abyssal volcanic cones hit the margin obliquely, so that the troughs leave an unstable triangle of rock between themselves and the steep slope of the Hikurangi Trough. It was the collapse of this badly fractured triangle that caused the huge avalanche off Ruatoria.

This was not the first time it had happened. There are ghostly scars of earlier seamount impacts, and even older avalanches, at several places along the East Coast, notably off Poverty Bay. Dating of these features is very uncertain, but we think that large impacts and their avalanches have occurred at intervals of several hundred thousand years. As confirmation, the scatter of Pacific plate seamounts approaching the margin from the east also suggests a frequency of many hundreds of thousand years.

Certainly, the big avalanches and their tsunamis may fascinate film-makers, but they are not something most of us will loose sleep over. This is not to suggest that we should be complacent about tsunamis. The coast from Bay of Plenty to Kaikoura is close to the plate boundary, where submarine volcanoes, deep earthquakes, fault ruptures, and small seabed avalanches may generate New Guinea-sized tsunamis often enough for one to be expected in a human life-time. Also, tsunamis that have travelled vast distances from places around the Pacific rim are a constant threat, although international monitoring may give sufficient warning for evacuation from these "far-field" events.

But what of homes built close to the shore? In New Guinea, a few fragments of foundation remained. There, legislation has been suggested to prevent building close to the sea in susceptible areas. Here, the catalyst for awareness and change has yet to happen.

Jean-Yves Collot is with the Observatoire Océanologique Villefranche sur Mer in France.
Keith Lewis is a marine geologist with the New Zealand Ocean-ographic Institute.