He and his team, who predicted the March 28 quake about 2 weeks before it occurred, said the area under the Mentawai islands west of Sumatra is most at risk of an earthquake with a magnitude of 8-8.5 or stronger.

"The potential for something bigger is there," he added.

Unlike the December 26 rupture that triggered the Indian Ocean tsunami in which 300,000 people perished, the March earthquake did not did create a giant wave because there was no rapid vertical movement of the earth's floor.

But McCloskey believes the stress changes and historical evidence could mean the next one could be similar to the earthquake of 1833 which set off a tsunami in the region that killed many people.

STRESSFUL EVENTS

The scientists, who reported their findings in the journal Nature, studied stress changes caused by the March earthquake. It is the same technique they used after the December tsunami when an estimated 1,200 km (750 miles) of fault line slipped up to about 20 metres (65 feet).

Both earthquakes were caused by the Australian tectonic plate grinding under the Indonesian plate. The displacement changes the stress values everywhere in the region.

"The area where we are most concerned about the earthquake rupturing is probably under Siberut island. That part hasn't ruptured since 1797," said McCloskey.

About 100 km (60 miles) south of Indonesia's Siberut is the area of the 1833 earthquake which hasn't had any major movement for 150-170 years.

"Slip on the southern portion could be as great as in 1833: that is, up to 10 metres," the scientists said in the journal.

Seismologists can determine which areas are vulnerable to an earthquake but they can not predict when one will occur, which is why early warning systems are so important.

"Our work is strongly indicating that the foot should not be taken off of the accelerator in going ahead full steam with high-tech warning systems," said McCloskey.

But he added that low-tech preparedness training is also highly effective.