Seven years ago tomorrow, Japan was devastated by a 9.1 "megathrust" tsunami that killed nearly 16,000 people. Now international scientists are taking a look at New Zealand's biggest geological threat: the Hikurangi Subduction Zone. Jamie Morton explains why we should care.
They knew it as the taniwha.
People travelling along the wild coast between Cape Campbell and the Wairau River would meet it as it suddenly rose up from the sea and swept them from the land.
Maori lore holds many such stories of this ocean monster causing death, destruction and peril for people living near the water.
It's likely they were told to explain the causes of natural hazards, to record loss of life, and to serve as warnings about the nature of particular places.
With the growing importance of matauranga Maori, or Maori knowledge, scientists have since looked to these tales as our first human evidence of tsunamis - and signposts to sites that may have been devastated many centuries ago.
It was even possible that the terrible Wairau River taniwha came from New Zealand's greatest geological threat.
Its name: the Hikurangi Subduction Zone.
Massive fault boundaries like it are scattered all around the vast and volatile Pacific Ring of Fire – and are known to have unleashed some of our planet's most catastrophic earthquakes and tsunamis.
One of them triggered the 9.1 eruption that, seven years ago tomorrow, sent a wall of water, up to 40m high in some places, tearing through Japanese ports and cities.
The surreal scenes of March 11, 2011 – stacks of logs swept down rivers like match-sticks, fishing vessels hurled into buildings like bath toys – were unforgettable.
Minami Sanriku was already equipped with a 5m seawall and a disaster readiness centre, but that wasn't enough to save most of the seaside fishing settlement.
Mayor Jin Sato had only that day chaired a meeting about how best to protect the community from such a disaster.
When I visited what was left of his town, he tearfully related to me the story of 24-year-old Miki Endo.
Through a loudspeaker mounted on the centre, she'd desperately urged residents to move to higher ground.
Sato was underwater for three minutes, clinging desperately to the building's radio antenna as his town vanished around him.
By the next morning, only a third of the 30 who had been on the roof with him were still alive; the courageous Endo was among the 42 people who perished in the three storeys below.
Her body wasn't found for a month.
"The town was swallowed in an instant by the tsunami," said Noriko Abe, whose bayside hotel lost its bottom floors.
"We were all devastated."
The Tohoku quake and tsunami cost some 15,883 lives and $286 billion, in a country already famously conscious of its natural threats.
Far greater was the death toll of 2004's Boxing Day Sumatra tsunami, which killed as many as 280,000 people in 14 countries.
That, too, was triggered by a 9.0-plus quake along an undersea subduction zone.
Both tragedies raised an unescapable question for those communities dotted along a wide expanse of our own eastern coastline: could the taniwha return?
"EXTREMELY DAMAGING AND DEADLY"
The two world wars were the bloodiest events in New Zealand's history, yet recent research has found a cataclysmic tsunami had the potential to be costlier for us than both put together.
Headline projections in an EQC-commissioned report estimated worst-case scenario impacts from a one-in-500 year event could include 33,000 fatalities, 27,000 injuries and $45b worth of property loss.
That death toll was also one or two orders of magnitude larger than our largest national disasters: the Erebus air crash, the Napier earthquake and Tangiwai.
It was odd reality to consider, given the fact only one person has ever been killed in the 10 tsunamis higher than 5m that New Zealand has experienced since 1840.
But it was because of such luck that the size and nature of our tsunami risk had not been widely understood, nor had the required responses been "widely appreciated", the authors noted.
A new simulation released this week suggested tsunami waves - up to 12m high in places - could inundate the coastline within an hour if a "megathrust" earthquake struck here.
Tsunami plans warn of three threats: distant, regional and local.
Distant tsunamis, generated in far-off locations like Chile, are held to pose the least risk as people would have ample time to flee over the 11 or so hours the wave would take to roll here.
A regional tsunami, perhaps triggered by an undersea volcanic eruption in the Kermadec Arc above New Zealand, would still offer us one to three hours' warning.
It's the threat of a local tsunami that worries Civil Defence officials the most.
One could hit within seconds to minutes, leaving no chance to order evacuations, and past records suggest local tsunamis may occur here every 40 to 50 years.
Tsunamis themselves came in different forms.
Some are precisely how we picture the beasts: turbulent, foaming walls of water filled with debris and sand that crash ashore and sweep inland.
Others can be simply described as rapidly rising or falling water levels, lasting minutes to an hour.
But both are dangerous, and both can travel over land at paces faster than any of us can run.
The Tohoku tsunami rushed across the Pacific at 800km/h, the speed of a jumbo jet.
How destructive a tsunami also depended on its run-up – and where it hit.
One of the most deadly scenarios for Wellington was a two-punch 8.9 earthquake and tsunami in the Hikurangi Subduction Zone, assuming a rupture extending across Cook Strait.
One GNS Science report found this catastrophic event could cause 40 deaths due to earthquake shaking - but a further 3200 in a tsunami that followed.
Although scientists were uncertain whether a rupture could extend from the subduction zone to the Cook Strait, such an eventuality could send tsunami waves 5m to 10m high barrelling toward the capital.
It would bring "extremely damaging and deadly inundation" in low-lying southern suburbs of Wellington, along with damaging and moderately deadly inundation in busy areas bordering Lambton harbour.
In Napier, a tsunami up to 8m high could plough through the low-lying city within minutes of an earthquake, reaching as far as 5km inland – and more waves could come over following hours.
Gisborne was also vulnerable to a quick-fire tsunami, with the city's beach front, and north of Muriwai at the end of Poverty Bay, most at risk.
Christchurch would be hit particularly hard, said University of California Santa Cruz geophysicist Steven Ward, who created the new simulation.
"Banks Peninsula, near Christchurch, acts like a tsunami magnet. The run-up is amplified there," he said.
"New Zealand hazard planners have their work cut out."
The entire danger zone bundles up most of the East Coast, thanks to an ancient and angry dividing line endlessly transforming and contorting our country.
LIVING ON THE BOUNDARY
New Zealand uneasily straddles the Australian and Pacific plates, two massive pieces of Earth's tectonic jigsaw puzzle that slide and collide.
Their boundary snakes past the North Island's East Coast, weaves into Marlborough, cuts across the South Island and then follows a 600km-long spine in the form of our postcard Southern Alps.
This geological scrum can be measured in just tens of millimetres of movement each year; under that speed, the hour-and-a-half journey from Auckland to Hamilton would take 3.6 million years.
Over millennia, however, the clash has shaped our mountainous landscape through uplift, stoked our volcanoes and forced "folding" where the rock deep beneath us crumples like a pushed tablecloth.
The resulting tension is relieved, of course, through some 15,000 earthquakes that GeoNet sensors pick up each year.
Understanding this intersection has allowed scientists to identify the two big bogeymen of New Zealand's natural hazard profile - one of them on land, the other beneath the sea.
The Alpine Fault has ruptured four times in past 900 years, each event causing a magnitude 8 earthquake.
A big event along the fault would devastate West Coast communities and perhaps re-draw the South Island's geography.
But the title of New Zealand's largest fault belongs to the Hikurangi Subduction Zone, which stretches from the northeastern South Island to waters well north of the East Cape.
Part of a bigger 3500km Hikurangi-Kermadec-Tonga system that fans up into the Pacific, this largely-offshore margin is where the Pacific Plate dives – or subducts – westward beneath the North Island.
If we drained the ocean, it would appear as an immense mountain range, rising up from the seabed, not far from the East Coast.
Massive subduction zone shakes can abruptly displace the seafloor and the ocean above it, causing a tsunami, and are sometimes called "megathrust" earthquakes.
The 9.1 quake that ruptured just east of Japan's Tohoku region started from a relatively shallow depth of 32km and continued all the way to the seabed.
It lasted about six minutes, producing the amount of energy that could power Los Angeles for a year.
Was Hikurangi capable of such might?
That remained largely unclear, as its location had made it tough to investigate.
But analysis of sediment cores taken from coastal areas – time capsules that can store a millennium of our geological history – tells us that our sleeping giant has woken before.
Buried around the Wairau Lagoon area, where the taniwha pounced, was evidence of two tsunamis: one that occurred between 800 and 900 years ago, and another that struck around 500 years ago.
"Beyond that, we haven't got a record of further quakes," GNS Science paleoecologist Dr Kate Clark said.
In Hawke's Bay – bang in the middle of Hikurangi's margin – there were signs of nine quakes in 7000 years that were big enough to have caused the land to subside by a metre or more.
"And further north, we don't really have any records of what we are certain are subduction earthquakes."
In any of the cases, it was difficult to be sure whether the subduction zone had been to blame.
But scientists could look to the tell-tale clue of buried soils that had dropped down and then been immediately overlaid by sediments, which signalled both an earthquake and tsunami.
"This is what we've seen around the southern part of the margin, around Wellington," Clark said.
"So we're trying to fill in the gaps and work hard on narrowing the age ranges, because the more places that we can identify past earthquakes, the more we can combine that information together with other evidence along the coast, and then infer what the magnitude may have been."
If the last event, 500 years ago, could be tied to sites both in Marlborough and Hawke's Bay, its scale would have been enormous.
"An earthquake that affected both of those places, several hundred kilometres apart, had to be greater than magnitude 8 – and possibly up to 9."
Clark and colleagues have also calculated the recurrence intervals of large events, which suggested one happening every 550 to 1400 years.
When it came to the whole margin going at once – think a 9.0 quake the size of Tohoku – the recurrence was probably far less frequent, although the evidence for such monster events remained sparse.
And Clark also emphasised that cataclysmic quakes didn't keep a tidy schedule.
"When we talk about recurrence intervals, people shouldn't assume we are talking about something that's fairly regular – and we know enough about action on the Hikurangi Margin that it does not appear be regular."
A SILENT THREAT
After Tohoku, scientists trawled through data from thousands of earthquakes in the region to pin-point one peculiar culprit behind the sudden megathrust.
Their research pointed to a deep-seated, almost imperceptible motion known as "slow slip" – a phenomenon common in our own subduction zone, where they'd actually been first observed using GPS in 2002..
Also known as silent earthquakes, slow-slip events can last from days to years, and can produce up to tens of centimetres of displacements along faults without seismologists even realising it.
When one of these slow-slip quakes are under way, scientists are now paying close attention, as they may provide clues into the occurrence of future earthquakes.
Other studies have tied these slow slip episodes to two big quakes in 2014: slow slip events were thought to precede the 8.1 Iquique earthquake in Chile, and a 7.2 shake off the coast of Mexico that hit just two months after a slow slip started.
"The discovery of slow slip events in the last 20 years has caused us to completely re-think how faults work," said Dr Laura Wallace, a geodesist at GNS Science.
"Everything we learn about them makes us realise how little we actually know."
When the Kaikoura Earthquake erupted in November 2016, with the equivalent energy release of 400 atomic bombs, shockwaves travelled hundreds of kilometres through the ground and triggered a series of faults as they spread.
It's believed they also set in motion a remarkable slow-slip event along the Hikurangi subduction zone more than 600km from Kaikoura spanning an area of more than 15,000sq km – comparable to the land area of New Caledonia.
Other signs of slow slip were recorded after the Kaikoura earthquake beneath Kapiti and Marlborough.
Silent quakes aren't always stress-builders; along some parts of the plate boundary, they are actually relieving tension.
"But in the shorter term," she added, "they can really affect the likelihood of earthquakes, if nearby faults are close enough to failure – so it's bit of a double-edged sword.
"Overall, most of the time they don't trigger those really big quakes – we've seen dozens of them in New Zealand so far and none of them have done this – but on very rare occasions, it's possible."
They could be a more of a worry if they kicked off in a fault region that had already become seized up and unable to move – as has been the case for the subduction zone, beneath the lower North Island, for many decades.
We might assume that when it finally became unstuck, a massive amount of pent-up energy would be rapidly unleashed.
"We know from GPS measurements that the subduction zone beneath the lower North Island is basically hung up right now, and one day it will slip," Wallace said.
"The thing is, we don't know when the earthquake will happen, how big it will be, or even how often these types of earthquakes happen – it's why we need to understand it all a lot better than we do."
INTO THE SUBDUCTION ZONE
Wallace is now helping lead a massive scientific effort that will give us a clearer picture of what's happening within Hikurangi.
Combined with a number of other international efforts to investigate the subduction zone over the next few years, it will help make Hikurangi one of the best-studied subduction zones on the planet.
Some $60m had been poured into the project by international organisations, such was its potential for far-reaching insights.
"Scientists are realising there are a many unique characteristics about the Hikurangi subduction zone that can teach us a lot of lessons about how the other subduction zones in the world work," Wallace said.
"So it's becoming a bit of a mecca; to my knowledge I don't know that there's been any other New Zealand-based problem that's had as much international investment in it."
Next week, the JOIDES Resolution, a specialist drilling ship operated by the International Ocean Discovery Programme (IODP) fresh back from probing the West Antarctic Ice Sheet, will set out for the slow-slip area off the East Coast.
They will drill boreholes 500m below the seabed, into which the scientists will install what Wallace called "sub-seafloor observatories".
These two listening posts, containing high-tech measuring and monitoring equipment inside steel casing, would stay under the seafloor for the next five to 10 years.
Their sensors would gather data on the how the rocks are strained during slow-slip events, as well as changes in temperature and flow of fluids through the fault zones.
As part of the project, jointly led by GNS and Pennsylvania State University, and funded by the IODP and the US National Science Foundation, scientists will retrieve sediment cores to give them a first-time look at the region's geological record and the types of rocks that host slow slip events.
There's been a recent flurry of scientific activity around the area.
A 2014 US-led project deployed two dozen sensors along the East Coast seafloor to pick up a slow-slip quake that lasted two weeks and caused 15-20cm of movement, equivalent to three to four years' worth of steady plate motion.
"It will take everybody working on the drilling project results a several years to fully wade through all of the data and get everything we can out of the drilling project," Wallace said.
"And if we can better understand what happens before, during, and after slow slip events, and their relationship to earthquakes in the area, then that might be able to give us an idea about what might is likely to happen in the future.
"So, while we'll never be able to actually predict earthquakes, the more we understand how these faults work and what their relationship to slow-slip events are, the better we might be able to better forecast them."
AN EARLY WARNING SYSTEM?
At 2.46pm on March 11, 2011, in downtown Tokyo, Japan's state broadcaster NHK was airing the dull debates of a Parliament night sitting when all hell broke loose.
To the sound of chimes, an urgent alert flashed upon the screen: "This is a tsunami warning."
Here was Japan's Earthquake Early Warning (EEW) system, a billion-dollar, state-of-the-art network based on hundreds of seismographs and models that could instantly model tsunamis from 100,000 different quakes in 10,000 different locations.
When a quake strikes off Japan, the system immediately estimates the location, magnitude and seismic intensity of the event by picking up P-waves that travel as fast as six kilometres a second, preceding the strong shaking caused by more damaging S-waves.
Within 20 seconds, scientists have an even more accurate estimate as several more seismographs fire back data.
In large quakes, warnings are instantly relayed live to several TV channels - popping up on screens with eerie bell chimes - while several of Japan's major mobile phone carriers alert users with texts, and Tokyo company Weathernews Inc alerts clients via computer.
Within five minutes on the Tohoku megathrust, seismometer data had incredibly calculated tsunami height and arrival time for each point of the coast.
Countless lives were likely saved as bullet trains came to a halt, factories secured their assembly lines and those at home had a quick opportunity to shut off their gas.
Tohoku's capital of Sendai got its first tsunami warning just three minutes after the alert, but it wasn't enough to save everyone.
Eight-hundred people would soon die when the ocean surged through the city's seaward suburbs and across its airport.
Minutes before 3pm, the tsunami warning was extended to the entire coastline.
After the Tohoku event, Japan has made further upgrades with its $400m S-Net system, relaying data from 150 sensors placed around the Japan Trench.
While New Zealand has a warning system for tsunamis caused by distant earthquakes, such as in South America, it doesn't have one for those caused by local events.
That was simply because tsunami generated by local earthquakes can potentially arrive at the nearest coast before scientists could calculate the location of the earthquake and issue a warning.
It was also possible that in some instances - such as a quake directly under a city, as happened in Christchurch – would make even an EEW system too slow.
New Zealand's current GeoNet capability includes hundreds of seismic instruments on land, a range of tsunami gauges that measure water level, and geodetic data fed in by more than 180 continuous GPS (CGPS) stations.
Our scientists have, however, investigated what a Kiwi EEW system might look like.
In 2013, a GNS Science report used a scenario similar to the March 1947 tsunami earthquake off the coast north of Gisborne to assess GeoNet's detection capabilities and potential required updates to the network.
After testing a range of detection and classification algorithms with the simulated data, the report authors concluded such an event could be detectable by the network in real time.
However, it found a large portion of the geodetic sensor network would need to be upgraded to stream the data and provide accurate information.
"The GeoNet continuous GPS network is presently far from being readily available for a tsunami early warning system," the report found.
At the time the report was written, only 37 of the CGPS sites provided data in real time, and a real time processing procedure wasn't available.
Creating a system would require a "substantial effort" from GeoNet staff, a "significant increase" in funding, along with the development of procedures and technology to process data in real time.
But a visiting Japanese seismologist said a New Zealand system wouldn't necessarily need to be as complex as that of Japan's.
The Hikurangi Subduction Zone lay comparatively much closer to the coastline, perhaps making communications logistics more manageable, and sensors might not have to be placed along the entire system.
"You might only need a smaller number of instruments along the Hikurangi," said Dr Kimi Mochizuki, of the University of Tokyo's Earthquake Research Institute.
"We may also find there could be sweet spots where we can focus on, meaning you don't have to cover the entire margin, but we also need to do some more research on this."
Meanwhile, GNS was running a two-year EQC-funded study to ask how – and if – Kiwis would use early warnings.
"One of the main things we want to find out is how people would use the time given by a warning system to make themselves and others safer," explained Dr Julia Becker, a social scientist at GNS.
"For instance in Japan, warnings for large earthquakes are automatically texted out and used immediately by train drivers to slow the trains down, for surgeons to make a patient safe during an operation and for the general public to take safety steps.
"Here we'll be looking at how it might be used for hospitals, rail and road transport, manufacturing and the general public.
"We'll also be looking at what the most effective channels for sending out warnings would be."
What ultimately remained clear was there was much that could be done to reduce our risk.
Researchers have suggested that spending just $50m extra on mitigation efforts each year had the potential to save thousands of lives.
LONG, STRONG, GET GONE
If authorities want four words to stick, it's these: long, strong, get, gone.
"If it's long and strong, don't fill your car with supplies or take any detours – just go," said Hawke's Bay's emergency management co-ordinator, Lisa Pearse.
Civil Defence was marking the anniversary of the Tohoku disaster by encouraging people to make a "tsunami hikoi" and work out the route we'll take to get there.
"The 2004 event and subsequently the Japanese event in 2011 heightened our awareness that we also have a risk here."
Research had shown awareness levels had lifted from 30 per cent a decade ago to 60 per cent today.
"But what worried me was a survey we carried out in Napier that asked what would you do in an event: most said they'd help their neighbour, or check on their children."
Just 17 per cent said they'd do the correct thing.
That finding was reflected by results of Civil Defence's latest disaster preparedness survey.
While the Kaikoura Earthquake had bumped up awareness – and more than eight in 10 Kiwis now had the necessary emergency items needed to get through – one in five still thought there would be adequate warning before disaster struck.
And to most Kiwis – perhaps thanks to Christchurch and Kaikoura – "disaster" effectively meant earthquakes; only 10 per cent of us considered the tsunami risk.
Discussing natural disasters could be unsettling, Pearse said, but it was a conversation New Zealanders should never stop having.
"It's a scary subject and people don't necessarily like to think about it or talk about, but it's like anything: if you're aware, you can prepare."