Geologists hope a massive trove of shellfish kept at Te Papa will help forecast when the next big earthquake might strike.
Last year geologists said the Hikurangi subduction zone was capable of magnitude 9 'megathrust' quakes, which could potentially unleash thousands of times the energy of the quake which rattled Christchurch in 2011, and create tsunami waves 30m high.
The little-understood faultline hasn't ruptured in recorded history.
"My previous research has shown that area from East Cape right down to Marlborough, it's had 10 large subduction earthquakes over the last 7000 years," GNS Science earthquake geologist Kate Clark told The AM Show on Monday.
"This is telling us that even though we haven't seen a large subduction earthquake in historic times, it is capable of producing large earthquakes - so we should expect one sometime in the future."
But when? Working out how frequently it ruptures will give us a clue. Dr Clark's contribution is looking at shellfish and figuring out when they died.
"When there's a large earthquake on the Hikurangi subduction zone, it causes parts of the coastline to rise up - like we saw in the Kaikoura 2016 earthquake. Big areas of coastline were raised up suddenly, and that kills all the shells there, because they're out of their living depth.
"Subduction earthquakes can also cause coastlines to subside, so lagoons will suddenly drop down and same thing - shells will die off because they've suddenly changed living depth.
"Large tsunamis can also pick up lots of shells from the beach and scatter them inland - they die as well."
Since the 1800s, collectors and scientists have picked up shells and dated them. This helps Dr Clark and others compare the dates their radio carbon dating gives them with the real thing, allowing them to calibrate the dates for shells as old as 40,000 years.
"We compare what our radiocarbon dating tells us the age is, with what the actual age is... This helps us correct all of our other shell ages from along the East Coast that we've used to date earthquakes and tsunamis."
While they can't predict exactly when the next quake will strike, the more they know about the risks the better we can prepare for the inevitable.
"We're using the timing of those earthquakes to understand the frequency of these earthquakes. Say, within the next 1000 years are we likely to have one? Yes, yes we are."