Scientists say new research is an important step in being able to forecast the likelihood of damaging earthquakes and tsunamis on the Hikurangi subduction zone.
But they admit they are still a long way away.
The study, led by GNS Science, has provided a better understanding of the way plate tectonic stresses build up and are dissipated along the zone, on the east coast.
It involved the detailed analysis of several hundred earthquakes between Hawke's Bay and East Cape to produce the first direct physical evidence of the way stresses change before, during and after slow-slip events.
The Hikurangi plate boundary is where the Pacific tectonic plate subducts the Australian tectonic plate and is what scientists call a subduction zone.
Subduction zones are a type of fault and are responsible for the most powerful earthquakes and tsunamis in the world.
This latest research paper titled Episodic stress and fluid pressure cycling in subducting oceanic crust during slow slip was published in the journal Nature Geoscience this week.
It clarified the mechanisms at play deep inside the fault by focusing on a zone between 15km and 35km below the surface where the Australian and Pacific plates meet.
Lead author and seismologist at GNS Science Emily Warren-Smith said the study had shown accurate monitoring of earthquakes made it possible to keep track of stress changes where the plates met.
It included data recorded by the national seismic and geodetic network operated by GeoNet alongside 20 Ocean Bottom Seismometers deployed off the East Coast of the North Island in 2014-2015 as part of collaborative research with organisations in Japan and the United States.
Prior to this research, scientists had hypothesised that a build-up of stress and pressure from fluids on the deep parts of faults may trigger them to slip.
The study found that a certain type of small earthquake started occurring in the subducting plate in the lead up to slow-slip events.
"This type of small earthquake can only happen when there is a lot of fluid around at very high pressure," Warren-Smith said.
"They started disappearing once the slow-slip event finished, suggesting the pressure had dropped again."
It concluded that highly pressurised fluid released from the subducting plate travels upward and lubricates the interface between the two plates.
This initiates a slow-slip event where a large patch of the fault moves slowly and benignly for weeks or months and then stops.
Deformation during the slow-slip event then acts to temporarily relieve the pressure before the process starts over again.
Repeated cycles of this pressure build up and release seem to control how often slow-slip earthquakes occur on particular parts of the subduction zone.
Warren-Smith said the findings had international significance and would open up new avenues of research at other plate boundary zones around the world.
"Now we know what signals to look for, we can start to examine other subduction faults to see if they behave the same way, not just in New Zealand, but worldwide."