(PhysOrg.com) -- Current earthquake prediction methods may need to be revised in the wake of the recent Haiti and Asian earthquakes and tsunami.
University of Queensland researcher Dr Huilin Xing said current prediction methods were so far mostly based on recorded historical earthquakes and so it was difficult to make reliable predictions.
“That may need a different way of thinking, such as to combine modern computational and observational technologies with conventional prediction methods,” said Dr Xing, who is Deputy Director of UQ's Earth Systems Science Computational Centre.
Dr Xing said a number of recent large destructive earthquakes had occurred after an earthquake silence of different periods.
It was 600 years between the 2004 Boxing Day Sumatra earthquake/tsunami and a previous destructive earthquake in the region; about 5000 years to the 2008 China Wenchuan earthquake; and about 200 years to the current Haiti earthquake.
“However, the recorded earthquake history is relatively shorter for some cases,” he said.
Dr Xing said the Caribbean tectonic plate was moving eastward with respect to the North American plate at around 20mm per year, while Haiti was located close to the northern edge of the Caribbean plate.
“The January 12, 2010, Haiti earthquake occurred as a left-lateral strike slip faulting on the Enriquillo-Plantain Garden fault system (EPGZ).
“This fault system accommodates about 8 mm per year, and it is the likely source of a series of historical large earthquakes in the region such as in 1751, 1761, 1770 and 1860,” he said.
“However, it had been quite silent (locked) along the EPGZ afterwards until the Magnitude 7 Haiti earthquake occurred on January 12, 2010.
“This earthquake and aftershocks are releasing the accumulated stress along it during the long time interval since the previous quake.”
The ESSCC is a Centre within UQ's School of Earth Sciences. It conducts research on the mechanics and physics of solid Earth processes.
Explore further: NASA balloons begin flying in Antarctica for 2014 campaign