Evidence is growing the South Island's Alpine Fault, capable of devastating earthquakes of magnitude 8 and above, ruptures more frequently than previously believed.
And the towns of Hokitika and Greymouth are sitting right in the firing line.
Last year, GNS Science earthquake scientist Rob Langridge and Victoria University paleoseismologist Jamie Howarth uncovered evidence of a previously unknown quake that took place sometime between 1740 and 1840. This is more recent than the 1717 quake, which has been estimated at magnitude 8.1 - nearly three times stronger than the 2016 Kaikoura earthquake - and shifted one side 8m relative to the other.
Research has found a big one happens every 300 years or so, and GNS Science estimates there's about a 30 percent chance of another happening in the next 50 years. Some studies looking 100 years out have concluded it's as much as 85 percent likely.
But Dr Langridge and colleagues have now confirmed smaller quakes can happen on the fault in between the big ones. Using radiocarbon dating, they've narrowed down the likely date of the most recent quake, measuring magnitude 6 or 7 (so potentially as strong as the Christchurch quakes) to between 1813 and 1848.
"One of the real challenges with the Alpine Fault - because it is so bush-covered - is actually finding sites that have been cleared and therefore can be studied," said Dr Langridge.
Scientists trenched a site near the Toaroha River, south of Hokitika, and found evidence of the more recent quake in the northern part of the 600km-long fault's central section.
"Once we started working there, the story really grew in large part because of the richness of dateable organic material in the trenches."
In addition to the newly discovered quake, they found more evidence for three big quakes - one between 1084 and 1276, a second between 1250 and 1580, and the most recent big one, which they dated to between 1673 and 1792. Analysis of tree ring patterns in 2018 suggested it happened in 1717, prior to European settlement when there were few people in the South Island.
The 19th century quake was perhaps only a partial-section slip, Dr Langridge said, proving activity on the Alpine Fault is more complex than previously thought - particularly in its north, where it meets up with other fault zones, such as the Marlborough Fault System.
Alternatively, the quake might have been centred further north, the rupture spreading along the fault to the southwest where Dr Langridge and his team did their digging; or possibly triggered by a separate quake in the Marlborough Fault System.
There isn't enough evidence yet to favour one scenario over the others, Dr Langridge said.
"One of the outcomes of this study is that you should expect a shorter recurrence interval of strong shaking at fault section ends. Because of the recurrence times of earthquakes though, you obviously have to wait a long time to see the effects of such fault behavior."
If the Alpine Fault does trigger another magnitude 8-plus quake in the near future, GNS says it will "have a major impact on the lives of many people", and likely be "one of the biggest earthquakes since European settlement of New Zealand".
But the findings prove there could be strong quakes on the fault in between the big ones.
"An important outcome is that sites or towns near fault intersections and section ends may experience strong ground motions more frequently due to locally shorter rupture recurrence intervals," the study, published in the Bulletin of the Seismological Society of America this week, says.