A pair of massive earthquakes which struck off the coast in Sumatra in April may have triggered quakes around the globe, according to new research.
The magnitude 8.6 and 8.2 earthquakes may also herald the formation of a new plate boundary between the Indian and Australian oceanic plates.
Three papers published in the online journal Nature have analysed the April 11 quakes, the largest of which was the tenth strongest in the past 100 years.
The quakes are said to have triggered small quakes during the three hours it took for seismic waves to travel through Earth's crust, however some faults were not rattled enough to fail immediately, instead breaking up to six days later.
Seismologists from UC Berkeley and the United States Geological Survey found that in the six days following the quakes, the rate of remote earthquakes - those more than 1,500 kilometres from the epicentre - with magnitudes of 5.5 or greater increased nearly fivefold.
While aftershocks have traditionally been defined as those smaller earthquakes that happen after and nearby the main fault rupture, the scientists now say this definition is wrong.
Instead, aftershocks are simply earthquakes of any size and location that would not have taken place had the main shock not struck.
"Until now, we seismologists have always said, 'don't worry about distant earthquakes triggering local quakes'," Roland Burgmann, professor of earth and planetary science at Berkeley and co-author of one of the studies, said. "This study now says that, while it is very rare - it may only happen every few decades - it is a real possibility if the right kind of earthquake happens.
"We found a lot of big events around the world (following the East Indian Ocean earthquakes), including a 7.0 quake in Baja California and quakes in Indonesia and Japan, that created significant local shaking," Burgmann said. "If those quakes had been in an urban area, it could potentially have been disastrous."
USGS seismologist Fred Pollitz suggested the delayed triggering of earthquakes may have occurred because of the strike-slip nature of the initial Sumatran earthquake. The 8.6 quake is the largest strike-slip event ever recorded.
"Most great earthquakes occur along subduction zones and involve large vertical motions. No other recorded earthquake triggered as many large earthquakes elsewhere around the world as this one," said Pollitz, "probably because strike-slip faults around the globe were more responsive to the seismic waves produced by a giant strike-slip temblor."
The low number of global earthquakes before the East Indian Ocean events may also have played a part in the increase in activity following the quakes.
"Imagine an apple tree, with apples typically ripening and then falling at some steady rate," co-author Ross Stein said. "If a week goes by without any falling, there will be more very ripe apples on the tree. Now shake the trunk, and many more than normal might drop."
Two other studies, also published in Nature, have suggested the earthquakes may herald the formation of a new plate boundary between the Indian and Australian oceanic plates.
Seismologists have suspected since the 1980s that the Indo-Australian plate may be breaking up, however Matthias Delescluse, a geophysicist at the Ecole Normale Superieure in Paris, said the April 11 earthquakes represent "the most spectacular example" of that process in action.
"It's the clearest example of newly formed plate boundaries," he said.
While most large earthquakes occur near tectonic plate boundaries, when one slides underneath the other, the two jolts were centred on the plate itself, 100 km and 200 km to the southwest of the major subduction zone that defines the plate boundary between the Indo-Australian and Sunda plates offshore Sumatra.
Delescluse and his team modelled stress changes before the April earthquakes, and suggested that two earlier earthquakes along the eastern plate boundary - the magnitude 9.1 tremor in 2004 that unleashed a devastating tsunami across the Indian Ocean, and another quake in 2005 - probably triggered the April quakes by adding to pent-up stresses in the plate's middle region.
In another paper, University of California seismologist Thorne Lay and colleagues found that unlike most quakes which rupture along a single fault, the initial quake had an extraordinarily complex four-fault rupture lasting around 160 seconds, which they said occurred as a result of the accumulated stresses across the plate.