LONDON - Scientists were still grappling yesterday with the mystery of why a large earthquake under the Indian Ocean did not produce a sizeable tsunami.

One thing is certain, however, the earth did not move in yesterday's quake in the way it did on Boxing Day last year when 10-metre high waves killed up to 300,000 people around the Indian Ocean.

Brian Baptie of the British Geological Survey in Edinburgh said that it would take some days before seismologists can work out precisely what happened on Monday when northern Sumatra was rocked by an earthquake that measured 8.7 on the Richter scale.

"It is very difficult at the moment to understand why there was not a tsunami following an earthquake this size. One possibility is that its epicentre may have been deeper," Dr Baptie said.


The Boxing Day earthquake, which measured 9.3, resulted from a massive jolt that took place some 10 kilometres below the seabed, whereas preliminary estimates suggest that the latest quake was focussed on a fault zone three times deeper.

This may have dissipated some of the energy before it reached the seabed, or at least caused the ground to move in such a way that it failed to displace the water column above to generate a large tsunami.

"An 8.7 magnitude earthquake occurring 150 miles off the coast is certainly big enough to generate a large tsunami and it was exceptionally lucky that we didn't have one," Dr Baptie said.

By far the most destructive tsunamis are generated from large, shallow earthquakes with an epicentre or fault line near to or on the ocean floor. They usually occur in areas where one of the Earth's tectonic plates is slipping, or subducting under another plate.

This is the case off the west coast of Indonesia's Sumatra -- one of the most geologically active regions of the world -- where the Indian-Australian plate is crashing into and subducting under the Burma plate at a rate of about 7cm a year.

As the plates slip past one another, huge stresses build up which can be suddenly released with a jolt to generate an earthquake. When sudden vertical displacements of the seabed occur, massive volumes of water are lifted up to produce the potential energy that drives a tsunami.

These waves can travel great distances at rapid speeds in deep water, where they can pass by barely noticed, but when a tsunami enters shallower water the wave slows down and increases in height and destructiveness.

Elements of the Indian Ocean tsunami warning system, which is not due to be fully operational until mid-2006, alerted governments in the region to the latest earthquake and helped to assess whether a tsunami had occurred.

Indian Ocean nations agreed in Paris earlier this month to establish a full warning system by next year, but the first stages of it were already in place.

Tuesday was the deadline for governments to provide contact details to Japanese and American warning centres, and many had already provided the necessary information.

The contacts made it possible for news of the 8.7 magnitude earthquake off Sumatra to be relayed to authorities in the region. Thailand, Indonesia and Sri Lanka then ordered a protective evacuation.

Two sea-level gauges in Sri Lanka and off Australia more than 1000 miles from the epicentre of the quake enabled experts to call off the tsunami warning. The gauges recorded a rise of less than 30cm in sea level three hours after the quake, which meant that the danger of a large tsunami had passed.

It is generally accepted that an earthquake smaller than 7.5 is unlikely to cause a tsunami. However, one that is 8.7 is more than capable of producing one, said Greg Romano of the US National Oceanic and Atmospheric Administration in Washington DC.

"A tsunami is generally generated when you have a large displacement of water in a vertical direction and it is likely that we did not have nearly as large a displacement as with the December 26 event," Mr Romano said.

Scientists predicted earlier this month that a second major earthquake was imminent because of a build-up of underground stress following the Boxing Day earthquake.

Seismologists at the University of Ulster in Coleraine had calculated that the December earthquake had increased the stress on the nearby Sumatra fault and on the associated Sunda trench, making another earthquake more likely.

The Sunda trench, which separates Indonesia from the Australian tectonic plate, runs southeast to the site of December's earthquake, which shifted some 250,000 square kilometres of subduction zone. This rupture has increased the chances of further movement along the Sunda trench and the Sumatra fault, potentially triggering an earthquake of an estimated 7.5 magnitude, Professor John McCloskey said two week ago.

"Our results indicate unambiguously that there is a real danger of another quake in the region," Professor McCloskey said at the time.

"We have all heard that lightening does not strike twice in the same place, but earthquakes do. One of the strongest observations in seismology is that earthquakes cluster in space and time. Where you have one earthquake you are likely to have others," Professor McCloskey added.

His calculations suggested that stress was building up at two points, one of them at a point somewhere further south than the epicentre of the Boxing Day earthquake, which was indeed where Monday's quake struck.