Kiwi scientist's shell research could help predict NZ's next big earthquakes

A New Zealand scientist has found a potential way to predict big earthquakes that might hit New Zealand in the future by looking back in time.

Dr Kate Clark and Dr Jocelyn Turnbull Source: Supplied: Margaret Low/GNS Science

The research, funded by the Earthquake Commission, aims to pinpoint the size of large past earthquakes and tsunamis with more accuracy on New Zealand’s east coast, in turn helping with future predictions along the Hikurangi subduction zone.

GNS Science researcher Dr Kate Clark has uncovered the method that can forecast future events by radiocarbon dating shells.

The Hikurangi subduction zone, according to EQC, is thought to have generated around 10 large or super-large earthquakes over the past 7000 years and runs offshore from Gisborne to the top of the South Island.

“Shellfish that died when parts of the coastline were suddenly lifted up, or dropped down, by previous earthquakes can give us information about the size and locations of past earthquakes,” explains Dr Clark.

“But we are limited in the precision of how we date the shells because we lack some critical local information about carbon in our marine waters,” she says.

Dr Clark says her new research aims to pinpoint each event to within about 50 years. Her key to unlocking the carbon dating puzzle is located in Te Papa’s shell collection.

“We are very lucky that Te Papa has a collection of over 300,000 lots of shells, comprising several million specimens, dating back into the 1800s with notes of exactly where and when they were collected and when the shellfish died,” says Dr Clark.

By radiocarbon dating the Te Papa shells in relation to the time when they were collected, Dr Clark can calculate the age of other shells that died in past earthquake and tsunamis along the east coast.

“Previously we used a New Zealand-wide average correction, but we suspected that was not suitable for the Hikurangi subduction zone and our earthquake ages could be wrong by up to 200 years at some locations,” she says.

Dr Clark says the team has processed the first samples and believes they are on track to significantly improve the precision of ages for past earthquakes.