Earthquakes that unfold slowly and silently deep beneath the North Island could be key to forecasting future shakes and tsunamis unleashed by our largest fault.
A million-dollar, three-year project will boost scientists' growing understanding of
"slow-slip" earthquakes along the Hikurangi Subduction Zone.
Scientists believe the subduction zone, which runs up alongside the North Island's east coast, can generate "megathrust" quakes larger than magnitude 8, such as those which created tsunamis that devastated Indonesia in 2004 and Japan in 2011.
Worst-case scenarios of a major Hikurangi event could include thousands of fatalities and injuries, and billions of dollars worth of property loss.
But slow slip quakes - where plate boundary faults release pent-up strain slowly over days to months, rather than seconds in a typical quake - may help us better gauge the threat.
Their discovery 20 years ago has revolutionised seismology and our understanding of fault mechanics.
While they happen off the east coast every few years, no one feels them when they're happening - and their driving force remains unclear.
The new project, led by GNS Science, is designed to detect subtle physical changes inside the fault before slow-slip quakes happen, to reveal the mechanisms that regulate their timing.
"This will clarify if there are observable physical changes within the fault that could enable the development of more accurate forecasts of when the fault might fail, in both slow and, possibly, fast slip quakes," project leader Dr Laura Wallace said.
Tantalising evidence has emerged in recent years that a build-up of water pressure near the fault exerts a major control on timing of slow-slip quakes in New Zealand.
GNS seismologist Dr Emily Warren-Smith said if this build-up influenced slip timing, then monitoring the accumulation of water in the fault might enable better forecasting of slow, and possibly, fast earthquakes in the future.
But it was possible that fluid pressure changes inside the fault might be a symptom of slow-slip earthquakes rather than a direct cause, Wallace said.
Alternatively, there might be other processes such as the steady build-up of stress from tectonic plate motion that controlled the tempo of slow-slip quakes.
The project was aiming to resolve this dichotomy by mounting a large-scale deployment of undersea and land-based monitoring instruments in southern Hawke's Bay and Wairarapa.
These would monitor changes before, during and after a regularly recurring slow-slip event expected offshore in the region sometime in the next two years.
Wallace said the project would forge new ground in the field of seafloor geodesy and help put New Zealand at the forefront of global efforts to monitor offshore faults that can generate large quakes and tsunamis.
The team was heading out this weekend on Niwa's research ship Tangaroa to do the first set of seafloor sensor deployments.
"The project will lead to new evidence-based information that will help significantly in planning and preparedness and make New Zealand safer and better able to recover after a major earthquake."
A separate voyage out to the Hikurangi subduction zone - where the Pacific Plate dives down, or "subducts" beneath the east coast of the North Island - has just finished.
US scientists have just dropped their own specialised equipment onto the seafloor to visualise the structure of the sub-surface, and investigate how fluids are distributed within sediments.
Programme leader Dr Jess Hillman, of GNS Science, said this would enable scientists to better understand how the movement of fluids was related to activity on our largest offshore faults and the occurrence of subseafloor gas.
Voyage specialist Dr Peter Kannberg, of Scripps Institution of Oceanography in the US, said earthquakes, seafloor slope stability, and seafloor gas release are all governed in some part by the presence of fluids.
"Our instrumentation can detect where these fluids are in the Earth, allowing us to better understand the role of fluids in modulating these natural hazards."
The new three-year project is supported by a $960,000 grant from the Marsden Fund.