Aftershock data revised to assess risk

Scientists should have a better picture by the end of the month of when Canterbury's shakes will settle into a "background" level, GNS Science says.

Dr Kelvin Berryman, of GNS Science, said about 100 scientists from around the world were helping build a better model for predicting earthquakes in Canterbury, taking into account the thousands of shakes in the past 15 months.

Results from the new model were expected by the end of the month and would include the revised likelihood of another big one and where it might strike.

Although quakes will continue in the region for decades, the results should also give weary Cantabrians a clearer picture of when it would drop below a disruptive level, more like the occasional shaking experienced in Wellington or Napier.

"It's about when it's at a background level that isn't likely to be producing further damage," he said. "The main part of that story is just where are they going to be? You can still have a magnitude-6.0 earthquake, but if it goes another 10 kilometres offshore then it's not really an issue."

While shakes triggered by the September 2010 quake could continue for 30 years, they would fade in strength and frequency.

"Compared with before September, earthquakes are going to become more part of the fabric of Canterbury," he said. "But in 30 years' time, there are not going to be damaging earthquakes, one hopes."

Most of the stress directly beneath Christchurch appeared to have been released, meaning even an improbable larger quake was likely to occur elsewhere in the region, causing less damage.

After the September 2010 quake, scientists thought Canterbury would experience aftershocks for one to two years before settling. Damaging quakes on February 22, June 13 and December 23 had forced a rethink.

Berryman said it was clear Canterbury was following a more unusual, unpredictable "cluster" pattern, with the September 2010 quake triggering sequences of shakes along different faults.

"In September, we thought it was going to be like other earthquakes in New Zealand where you have the main shock and you have the aftershocks tailing away and you are done. But if you don't get rid of the stress all at once, you end up with a sequence – much less common and much more difficult to characterise."

With information from the past 15 months, including the December 23 quakes, the new model would more accurately reflect this cluster, he said.

While the December 23 quakes would increase the chances of another damaging earthquake, other factors might lessen the risk, he said.

"We are in that living and learning situation."

- © Fairfax NZ News