Clues scratched into rock may help reveal where big quakes in the future will strike hardest.
The new findings, published this week, could prove useful in helping better prepare the country for big shakes from some of our most hazardous faults – including the massive Alpine Fault.
"Large faults are like a heavyweight boxer - you know they are going to pack a wallop," said Russ Van Dissen, a GNS Science earthquake geologist and co-author of the paper.
"So if you want to prepare yourself, it's helpful to know which direction the energy is going to coming from."
The study, led by Victoria University PhD student Jesse Kearse, suggested the warning signs could be out there waiting to be found, in the form of scratch lines left in rock faces.
These markings, or striations, have been observed in fault rock faces for decades – but the new study proved the first time that scientists had been able to link the shape of them, and the direction in which a fault propagated, or "un-zipped", during a quake.
This was crucial as the direction of this un-zipping had much to do with where seismic energy was focused.
The findings stemmed from an analysis of the Kekerengu Fault, which runs from North Canterbury through Marlborough and out to sea, and tore open the countryside when it violently ruptured during the 7.8 Kaikoura Earthquake.
The phenomenon of "directivity" was clearly evident during the November 2016 event, with Wellington experiencing much stronger ground shaking than Christchurch even though the earthquake epicentre was closer to Christchurch.
This was a result of the south-to-north direction of the rupture, which sent a large amount of seismic energy to the north.
Van Dissen saw a good reason to go hunting for more striations.
"If scientists can confidently deduce past directions of fault rupture propagation, that will improve the characterisation of damaging earthquake ground shaking."
If it could be shown that the South Island's Alpine Fault in the South Island had ruptured a number of times in a certain direction – either north-to-south, or south-to-north – then that information could be used to better plan for future earthquakes, and in designing more resilient buildings and infrastructure.
Presently, there's no way to predict where, when, or with what force a large earthquake will be unleashed in the future.
As quakes begin on faults often buried kilometres below the surface, it's extremely difficult for scientists to know how close these faults are to failing in an earthquake.
The best technology that agencies had to work with today were earthquake early warning systems – used overseas but not yet here – that could give a few minutes' warning.
Beyond that, scientists could give "earthquake forecasts", which gave a rough probability of something happening over a certain period of time.
The Alpine Fault, for example, has a 30 per cent probability of rupturing in the next 50 years.
While the new findings represented a hypothesis that needed to be further tested, they could also prove a potentially useful tool.
"It enables us to bring a more nuanced approach to the way we analyse past fault ruptures and better estimate future ground shaking," van Dissen said.
"Maybe instead of digging up softer sites along faults and looking for material to radiocarbon date, scientists will go to bedrock sites specifically to look for these striations. That wouldn't have happened in the past as there was no reason to do it."
The other authors of the study, published in the scientific journal Geology, were Yoshihiro Kaneko of GNS Science and Professor Tim Little of Victoria University.