An earthquake of magnitude 8 happens somewhere in the world on average once a year. A magnitude 9 quake, like the one in the Indian Ocean near Aceh on 26 December, happens once every 30 or 40 years. Had the Aceh quake been magnitude 8, the resulting tsunami would probably have spared Sri Lanka. But this quake was unusual not just because of its severity. Three weeks later, experts are still working out exactly what happened as the Earth’s crust shifted under the Indian Ocean.
Some basics we do know. The rupture started 200 kilometres west of Sumatra and travelled north to the Andaman Islands, over 1000 kilometres away. Pressure had been building for decades between the Indian tectonic plate and the Burma plate, under which it is slipping at a rate of 6 centimetres per year. “The two big plates are chugging along, but the surface is stuck by friction,” explains Rob McCaffrey of Rensselaer Polytechnic Institute in Troy, New York. Then came the rupture, when the surface of the Indian plate suddenly slipped by up to 20 metres.
Many big earthquakes, including the five largest since 1900, occur where there is a sudden long rupture along a boundary between plates that normally slide under each other at a shallow angle. By comparing the size and timing of signals recorded by some of the world’s 410 seismic stations, researchers determined the magnitude, time and location of the Aceh earthquake. The Indian plate slipped just before 8 am local time, and the rupture probably travelled 400 kilometres north-west in just 200 seconds.
Advertisement
For most earthquakes, 200 seconds of data would be sufficient to gain an understanding of the whole event. But not for this one, says Chen Ji of the California Institute of Technology, who is working on an extended model of the earthquake. Even with so many seismic monitors in place, he does not have enough readings he can rely on. Seismic monitors cannot record over the full range of the earthquake spectrum. They are set at a level that will record most earthquakes, but may have failed to capture the full severity of this one.
The first of more than 150 aftershocks started immediately after the main earthquake, and may continue for another year. These readjustments to the positions of the tectonic plates make it hard to establish exactly how they moved during the main event.
The first aftershocks at the north end of the earthquake were recorded as long as 83 minutes after the first shock, suggesting that after the first 200 seconds the rupture spread much more slowly than the 2 to 4 kilometres per second at the beginning. A possible explanation is that a second rupture emerged 400 kilometres from the first. “There is a case for two primary ruptures as there is a big bend to the north-west of the first rupture,” says Roger Bilham of the University of Colorado, Boulder. It is possible that this second rupture spread so slowly that the seismic monitors around the world did not pick it up. If that is what happened, researchers may have to recreate the rupture event by using computer models that work backwards from the arrival times of the tsunami waves around the Indian Ocean.
“There is a case for two primary ruptures, as there is a big bend to the north-west of the first”
GPS positioning data is also being used to determine how the tectonic plates shifted. But as GPS stations are on land, while the main movement happened beneath the sea, the value of this data is limited. A global monitoring network of seismometers known as the International Monitoring System, installed to help enforce the Comprehensive Test Ban Treaty, also yielded data that is now being examined. But it too has its limitations.
“The IMS is designed to measure relatively small events such as a nuclear test, not a magnitude 9 earthquake,” says Phil Cummins, a seismologist at Geoscience Australia, a government agency in Canberra. And under the treaty, national governments can keep any data they receive from the IMS confidential.
