China devastation: measuring stress build-up can help predict aftershocks. Credit: Nir Elias/REUTERS

An analysis of the seismic stress changes in China’s Sichuan basin has been published just 8 weeks after a magnitude 7.9 earthquake there killed more than 60,000 people and left millions homeless. The rapid calculation and publication1 brings accurate forecasting of aftershocks a step closer, experts say.

Tom Parsons of the US Geological Survey in Menlo Park, California, and his colleagues describe several large faults in the basin that faced increases in stress as a result of the earthquake’s main shock on 12 May. These fault sections are more likely than others to produce large aftershocks in the near future.

It is still not possible to accurately predict whether and when massive after-quakes might occur, such as the 6.0 aftershock northeast of the original epicentre that killed eight people and destroyed thousands of buildings. But experts say the study shows that it is now feasible to rapidly calculate the propagation of stress in the crust in the immediate aftermath of a major earthquake, even in areas as geologically complex as the Sichuan basin.

Such data could be used to produce large-scale maps of high-risk zones, significantly improving efforts to protect civilians in disaster regions. “We can’t predict earthquakes — many scientists think we may never be able to — but we can do a lot better than nothing,” says John McCloskey, a geophysicist at the University of Ulster in Colerain, UK.

Current global earthquake hazard maps are too vague to help local emergency planners mitigate risks effectively. So a main goal of earthquake physics is to identify zones of particularly high seismic risk.

McCloskey and his colleagues analysed seismic stress perturbations in the surrounding crust after the devastating 2004 Sumatra earthquake and tsunami2. Using essentially the same methods that Parsons and his team applied in estimating the stress redistribution after the Sichuan rupture, McCloskey’s team successfully forecast the large quake in Sumatra in March 2005 that was triggered by the earlier rupture.

The geology of the Sichuan region is poorly understood and much more complex than the Sumatran zone, making it harder to forecast aftershocks. The seismic history of the region is also unclear. Large earthquakes are much less frequent there — every 2,000 to 10,000 years, according to an analysis published this month3 — than in Sumatra.

“It is easy to retrospectively explain why an earthquake has happened at a particular place,” McCloskey says. “But prospectively, what Parson and his colleagues have done is the best science we have at the moment.”

McCloskey is setting up a group to systematically analyse changes in seismic stress in the immediate aftermath of future large earthquakes. The goal is to forecast within 12 hours the likelihood of big aftershocks, and map the high-risk ‘red zones’.

“We have the methods to say strong things about what might happen days, months or years after the main shock,” he says. “Aftershock forecasts will never be perfect. But in the absence of knowing something definite we’re still able to advise people on the balance of probabilities.”

McCloskey hopes that a scientific journal, such as Nature, will provide a forum for rapid dissemination of information such as daily hazard maps after a major quake.

But Philip Campbell, editor-in-chief of Nature, thinks such information would be better placed elsewhere because the peer-review process can slow things down. “Rushing the publication of such studies would not be good, as they usually improve quite a bit during peer review,” he says.*