Off the North Island’s east coast lies New Zealand’s most active and fastest-moving fault. The Hikurangi subduction zone, where the Australian tectonic plate is clawing its way over the top of the Pacific plate, has the potential to rupture in a mega-thrust earthquake of magnitude 8 or more and set off a massive tsunami.
It stretches from the East Cape to Marlborough, and regularly rumbles in earthquakes. But it also produces silent, long-running quivers known as slow-slip events, which release the equivalent amount of energy of a big quake but over weeks or months.
Thanks to a fleet of ocean-going monitoring instruments and two permanent observatories embedded hundreds of metres deep in the sea floor off the coast of Gisborne, scientists are beginning to decipher the “crosstalk between slow-slip events and earthquakes”, says Laura Wallace, who studies how the Earth’s crust deforms during seismic events.
In March this year, Wallace set off on a voyage to retrieve data from the observatories that were entombed in boreholes in the bottom of the ocean five years ago, continuously measuring pressure fluctuations, fluid flows and temperature changes.
Even at a first glance and without detailed analysis, she says it’s clear the observatories have captured two recent slow-slip events, in 2021 and 2022. “With the straining of the crust, we can determine exactly where the slow-slip events were happening, how much movement along the plate boundary was involved, and exactly when they started and finished,” she says. “In both cases, it looks like they initiated just off Gisborne, but then, over a couple of months, propagated right out to the trench where the Hikurangi subduction zone emerges at the seabed. This is something we could never see in any other way.”
The pressure and temperature recordings also provide insights into how a sequence of three large earthquakes off the East Cape and in the Kermadecs in March 2021 influenced the plate boundary closer to New Zealand. “We can see in the pressure data that this sequence triggered a very large slow-slip event. It’s clear that some of these large, regional earthquakes do have an impact on what’s happening on the plate boundary closer to home.”
The 2016 Kaikōura earthquake also tripped off a sequence of slow-slip events along most of the Hikurangi subduction-plate boundary. “The Kermadecs earthquakes are really interesting, because they spurred on a slow-slip event despite being so far away,” says Wallace.
But there are examples of the opposite – a period of slow slip that precedes a large earthquake. Subduction zone plate boundaries, where one plate dives beneath another, ring the Pacific Ocean and have produced the world’s largest earthquakes and tsunami, including the magnitude 9.2 Indian Ocean earthquake on Boxing Day, 2004 and the magnitude 9 Tohoku-Oki megathrust in 2011 off northern Japan. In the latter, a slow-slip event may have helped to set it off. “That could happen due to stress changes. A slow-slip event is relieving a lot of built-up tectonic stress, but it’s also redistributing stress to other faults surrounding that plate boundary and can push those towards failure.”
There have been hundreds of slow-slip events around the world that aren’t linked to earthquakes, Wallace says. “The real trick is trying to understand the characteristics of a slow-slip event that does eventually lead into a bigger earthquake. That’s the big reason we need to be studying these things.”
Wallace observed her first slow-slip event not long after arriving in New Zealand from the US in 2001 to help set up the GeoNet network to monitor seismic and volcanic activity. Checking the GPS time series at a site near Gisborne, she realised it had moved eastward by three or four centimetres over the course of a few weeks. For many years, the GeoNet instruments remained the main tool to detect slow-slip movements on land as well as offshore, but with limited resolution. “The ocean-bottom observatories give us our first true understanding of what’s happening and reveal the interconnectedness of the entire system.” l
