The megaquake connection: Are huge earthquakes linked?
March 18, 2011 The recent cluster of huge quakes around the Pacific Ocean has fuelled speculation that they are seismically linked. New Scientist examines the evidence
AT 2.46 pm local time on Friday last week, Japan shook like never before. The magnitude 9.0 earthquake wrenched the main island of Honshu 2.5 metres closer to the US and nudged the tilt of Earth's axis by 16 centimetres. At the epicentre, 130 kilometres offshore, the Pacific tectonic plate lurched westwards, and a 10-metre-high tsunami sped towards the coastal city of Sendai and the surrounding region.
The devastation caused by the events is difficult to exaggerate - estimates suggest the number of fatalities could top 10,000. One of the few consolations is that quakes of magnitude 8.5 and above are rare: the Sendai earthquake is in the top 10 of the highest-magnitude quakes of the last 100 years.
Yet three of these - the December 2004 Sumatra quake, the February 2010 Chile quake, and now Sendai - have struck in just over six years. This presents a horrifying possibility: that there is a link between these megaquakes and that, as a result, more could strike.
Most geologists say that the number of megaquakes is too small to be able to make a statistically convincing case for a link. "You will get a lot of different answers from different people, but inevitably the ability of any one of those to convince everyone else that they're right is going to depend on the statistics of small numbers," says Ross Stein of the US Geological Survey (USGS) in Menlo Park, California, "and we're never going to get anywhere."
A handful, however, feel there must be a link between recent events. "What is clear is that for the 6.2 years since 2004, there have been more great earthquakes around the world than in any 6.2-year period throughout the 110-year history of seismic recordings," says Thorne Lay at the University of California, Santa Cruz. His colleague Emily Brodsky goes further: "The recent spurt of magnitude-8-plus earthquakes may be an extended aftershock sequence of the 2004 Sumatra earthquake."
While demonstrating a domino effect is a challenge, Lay does have a geological mechanism that can link some large quakes that occur several months apart. He cites doublets - pairs of comparably large quakes that happen on the same or neighbouring faults within months of each other. In November 2006, an 8.3-magnitude quake shook the Kuril Islands north of Japan as the Pacific plate pushed beneath them. Two months later, in January 2007, the islands felt the force of a second large quake, this time an 8.1-magnitude event. When the Pacific plate lunged beneath the islands in the first of those quakes, it left the oceanic crust under tension. The January quake was the result of a new rupture that allowed the plate to stretch and thin to compensate (Nature, vol 451, p 561; and see diagram).
Doublets like this show that large earthquakes months apart and on separate faults can be linked by a sound geological mechanism. It does not prove that this is also true of events separated by longer periods and greater distances. For this, Lay says you would want to show, for instance, that a region which has experienced a large quake recorded unusual seismic activity and perhaps even some small tremors during a previous large event elsewhere on the planet. This might suggest that the earlier event unsettled a fault, effectively priming it.
For an example, Lay cites the 7.9-magnitude event that hit Sichuan in China in May 2008. "The 2004 Sumatra earthquake increased seismicity in that area as the [shock] waves passed by," he says. "Was the 2008 earthquake a delayed, large aftershock, or a totally independent event?"
The trouble is that large earthquakes generate tectonic waves that ripple around the world's surface and routinely trigger smaller quakes on distant faults, so increased activity in China is hardly a surprise. "If you have a quake of, let's say, 6.2 or larger, every sand grain on the planet is moving to the music of that event," Stein says.
Perhaps the biggest stumbling block to proving that large earthquakes trigger others on a global scale is statistics. To find out if quakes really do cluster in a non-random way, Tom Parsons, also at the USGS in Menlo Park, and Aaron Velasco at the University of Texas at El Paso, crunched some numbers. They compensated for the rarity of megaquakes by widening their focus to events of magnitude 7 and higher. The pair then searched for statistical links between these quakes and others of magnitude 5 and above around the world.
After reviewing a 30-year catalogue of events, they found no significant evidence that large quakes regularly trigger tectonic activity 1000 kilometres or more away.
Discussions over the clustering of megaquakes will rumble on. Unfortunately for the people of Japan, what is beyond doubt is that large quakes can cluster locally: aftershocks are common in the wake of a large quake, and occasionally they can be as large as the primary shock.
"We are in a cluster right now and it's centred on Japan," says Stein. "Japan could be in for other big quakes in the wake of this one. That part we should agree on."