Last summer, earthquake geologists Russ Van Dissen and Tim Little dug a 15-metre-long trench across a paddock on North Canterbury’s Kekerengu Fault, hoping to answer a question that had long eluded scientists: how often did it rupture, and how far did it slide when it did? They suspected it was one of the fastest-moving faults in the country, and new dating techniques gave them a chance to find out for sure.

Their team spent several weeks in the field, tracing the fault’s story into the past. They cut down through the earth in a series of steps, taking soil samples from the walls, until they were four metres below the surface in a muddy hole at the heart of the fault.

Though Van Dissen works for GNS Science and Little for Victoria University, they’ve been best friends since they met while making a similar trench on the Awatere Fault two decades ago, sharing a taste for beer and music around the campfire at the end of a day’s work (Van Dissen on guitar, Little on harmonica).

They have both spent their professional lives trying to understand a handful of connected faults that score Marlborough and North Canterbury’s farms and river valleys; to infer from faint echoes left in the landscape what happens when they detonate in sudden movement. New Zealand’s historical record is short—we’ve had only a handful of really large earthquakes since records began in the 1800s—so figuring out their patterns relies on this kind of ‘palaeo-seismic’ research.

In February, the geologists sent the samples taken from the trench for dating. They use both radiocarbon and ‘optically stimulated luminescence’ techniques, which can detect how long ago a mineral particle last saw sunlight. The results confirmed their suspicions—the Kekerengu Fault had a very fast slip rate, and had ruptured three times in the past 1200 years… Make that four.

Less than a year after Van Dissen and Little dug across where they thought the fault would be, the Kekerengu came to life—one of at least nine faults ignited in the magnitude 7.8 Kaikōura Earthquake on November 14.

“Our study found it was a very active fault by New Zealand standards—but we couldn’t have told you it was going to move in a few months’ time!” Van Dissen says.

Though they had filled in their survey trench, a faint rectangular outline remained in the grassy paddock. Now, though, it is split clean down the middle by a scar of tortured soil, the far side ripped sideways along the hill, the two halves nine metres apart.

[Chapter Break]

“What unknown affinity / Lies between mountain and sea / In country crumpled like an unmade bed,” asked Denis Glover in his 1953 poem Arawata Bill.

The poet was thinking of gold, but the answer could equally be plate tectonics, the great unifying theory of geological science that rose to prominence a decade after Glover published his poem sequence about the old prospector.

This country is crumpled by the shoving, straining might of two great sections of the Earth’s outer shell: these islands are the result of their meeting.

The Australian Plate dives beneath the Pacific Plate off Fiordland, collides with and crunches past it through the Alpine Fault, then rides over the Pacific Plate off the North Island’s east coast, forcing it down into the molten mantle. Marlborough is scrunched in the middle—and its complex network of faults carries the plate-boundary movement from the Alps through to Cook Strait at around three centimetres per year.

It tends to be less of a slow creep, more a series of sudden leaps—periodic, violent releases of pressure in the form of earthquakes.

At two minutes past midnight on November 14, part of the ‘Humps’ fault zone ruptured 15 kilometres below Culverden. Over the following hundred seconds, the earth unzipped.

The quake ruptured the Conway and Leader faults, jumped to the Hundalee fault, flew along the Uwerau and Fidget faults, up the Jordan Thrust and down the Papatea Fault, set off the Kekerengu, and ran out to sea along the Needles fault under the light of the full moon. New Zealand’s geography changed in a matter of minutes.

Along a 120-kilometre stretch of the Kaikōura coast, 80 kilometres was lifted clear out of the water, permanently altering the shoreline and leaving sea creatures gasping and exposed. Landslides dammed rivers and destroyed roads and railways. Kaikōura was cut off, homes collapsed, and two people died.

Many people didn’t sleep that night: quake-zone residents fled shaking houses and stumbled up the nearest hill as tsunami warnings were issued; civil defence and emergency services readied the response; and scientists stirred, electrified, wondering what would be revealed at first light.

GNS geologist Nicola Litchfield woke as the quake rocked Wellington. “I didn’t actually get out of bed and drop-cover-hold or anything, I just lay there thinking, ‘I wonder what fault that was?’”

When the shaking stopped, she checked the GeoNet website and confirmed it was a big one. Tsunami concerns had closed the road around the harbour for a few hours, but at 4.30AM she made it to the GNS office in Lower Hutt. “It was packed; there was a lot of adrenalin pumping.”

By 9AM, Litchfield and a colleague were strapping themselves into a helicopter to assess the damage from the air—a little apprehensive, not knowing what turmoil they might find.

They flew down the coast and over to the epicentre, the multiple rake-lines of ruptured faults immediately obvious—and unexpected. “It broke so many faults over a very large area. We knew about most of them, we just didn’t know that so many could rupture together,” Litchfield says. “If you’d asked me two weeks ago if an earthquake down near Culverden could trigger the Needles Fault off Cape Campbell, I’d have said, no way.

“That’s what we’re now trying to figure out—how the earthquake moved from fault to fault.”

Within days, teams of scientists from universities and research institutes around New Zealand mobilised and fanned out across the earthquake zone—landslide experts, tsunami specialists, coastal-uplift teams, marine biologists. Litchfield co-ordinated GNS’s largest squad—the fault team.

They needed to get there fast, before engineers straightened roads and farmers mended warped fences, before rainstorms blurred the detail of slips, before the high-tide bull kelp rotted away on rocks raised from the sea.

For the geologists who had spent decades detecting the subtle traces of earthquakes hundreds or thousands of years old, here was a chance to see, raw and up close, the effects of one on landscapes they knew so well—the scars fresh, the wounds still open, the dark soil still spilling from beneath the Earth’s skin.

[Chapter Break]

Ten days after the earthquake, Russ Van Dissen introduces me to the Kekerengu Fault. A metre-high clay-coloured cliff sticks straight up out of a paddock, crowned with a mohawk of grass. This part of the fault has been lifted both up and along: “We’re seeing this face as it gets on a conveyor belt and goes that way—it’s gone nine metres sideways.”

The cicatrices run across the paddock in jagged ribbons until they’re drowned by a newly formed lake. They cross the splintered road, tear through buckled railway lines and dive into the ocean, where they join up with the Needles Fault and rupture the sea floor.

This earthquake was full of surprises, but at least here in the northern part of the fault zone it’s unfolded pretty much as Van Dissen imagined it might.

“You’re never sure you’re going to be right, but in retrospect it was what I was envisioning. A large amount of slip, travelling along the Kekerengu and onto the Needles Fault.”

The Kekerengu Fault ripped the surface for 36 kilometres. In some places it caused displacements of 12 metres, which by international standards is a huge amount of slip for a fault this size.

Van Dissen has been probing and prodding this part of New Zealand for nearly 30 years, since he arrived in Kaikōura in the late 80s as a young American masters student on a Fulbright Scholarship.

He had a 1961 geology map with the Kekerengu Fault marked on it, strung high along the Seaward Kaikōuras. He set off to try to find it before the snows came, tramping up a different streambed each day, and each day finding evidence of a big thrust fault low down in the range, far below where he’d expected the Kekerengu to be. “I was frustrated for a while until I realised I’d actually discovered a new fault.”

He named it the Jordan Thrust—it links the Hope to the Kekerengu faults and is responsible for the uplift of the Seaward Kaikōuras—and the study of the area’s tectonic movements became his life’s work. (That year, he also met his future wife, and decided to stay in New Zealand—“it’s been a rather significant year in my life!”)

Now, Van Dissen has returned to the Kekerengu with a team of GNS geologists, two drone pilots and several international experts from a group called GEER (Geotechnical Extreme Events Reconnaissance)—“they’re kind of like Thunderbirds”, says Van Dissen. They mobilise in the wake of large earthquakes and bring manpower and technology to help with the scientific response, while learning lessons they can apply at home.

The team is trying to record as much information as possible about the pattern of faulting in the landscape. Brand-new drone and 3D mapping technology is enabling them to capture much more detail than would have been possible even five years ago.

Damp fog has obscured the mountains all morning, and Greek drone technician John Manousakis has to pick his moment for sending his contraption up over the rugged, torn terrain. The fault has turned a corner here, and that has forced the rupture over a much wider area—rather than one sharp line, an area hundreds of metres across has cracked, crevassed and slipped. A farm track is dramatically displaced, the two halves no longer touching, each now leading to a vertical cliff.

Manousakis uses a tablet to control the drone and its attached camera. With a locust-swarm buzz the machine’s four rotors whirr into life and it levitates into the grey sky. At 50 metres up, it begins to trace a pre-programmed grid pattern, taking photos that can be combined with GPS measurements to create an accurate 3D model of the whole area.

“The technology is really changing things,” says Van Dissen. “Now we can start asking questions about how the topography interacted with the fault rupture, and also monitor how the landscape heals after an event like this. It’s a lot more descriptive than just having a flat map.”

[Chapter Break]

A hundred kilometres of scarred land away to the south, other teams of geologists from GNS and the universities of Otago and Canterbury are working on the faults closest to the epicentre. There, in the Humps fault zone near Waiau, the ground hasn’t always ripped open in a clean line—instead, the breakage is distributed across an area several kilometres wide. That’s unexpected, says Andy Nicol from the University of Canterbury—and it also makes the faults harder to trace and measure.

“You end up crawling around through the paddocks trying to find the cracks as they weave their way across the countryside. There’s a bush telegraph out there, though, so as soon as you meet one farmer, they tell you where the next lot of ruptures are.”

So far, they’ve turned up four faults in the southern zone that definitely ruptured to the surface, though Nicol thinks they could find up to four more. And they’re looking closely at how each joins up with the next.

“This earthquake is going to give us a lot of information about how, why and where faults interact with each other—we haven’t got to the bottom of that yet. That will help us with forecasting, to understand what situations we might expect to trigger events like this, so we can produce better hazard models.”

Because that is the ultimate aim of much of this research—to understand how and where this complex network of faults may cause problems in the future.

As new information about a fault’s history is discovered, geologists feed that into New Zealand’s National Seismic Hazard Model (NSHM). It allows for a generalised level of background seismicity, and incorporates palaeo-seismic data (of the kind Van Dissen and Little have collected) about the slip rates and estimated recurrence of 530 active faults. The model is then used to develop building codes and disaster response plans, and functions as our best estimate of earthquake risk in each area of the country.

“I do believe we’re doing New Zealand a good service by studying these fault lines, getting them into the building code, developing guidelines for land use planners and engineers,” Van Dissen says.

He tells a story about the Trans-Alaska Pipeline, which was built in the 1970s to transport oil across North America and had to cross right over the Denali Fault. Geologists and seismologists worked with engineers to analyse the risk and design giant sliders for the pipeline to sit on in case of ground shaking. In 2002, the Denali Fault did rupture in a massive magnitude-7.9 earthquake.

“The pipeline slid the way it was supposed to, and not a drop of oil was spilt. There was no ecological disaster,” Van Dissen said. “By understanding displacement, we can help engineers design infrastructure for when they have to cross faultlines, because that’s inevitable here in New Zealand.”

[sidebar-1]

But Andy Nicol says the Kaikōura Earthquake has thrown into relief ways in which the model needs to be improved.

“We’ve known for a while that we’re not seeing in the geological record all the faults that could be active,” he says, and that means they’re not making it into the model. The Papatea and Leader faults that broke the surface on November 14 were both thought to be inactive because there was no record in the rocks of recent ruptures—that is, in the past 125,000 years.

In addition, those widely distributed fault traces near the epicentre demonstrate just how hard it is to accurately capture the whole picture of an earthquake that occurred hundreds of years ago: the chances of finding and measuring all of the displacement that occurred are very slim.

“We’re underestimating the number of faults, and the amount of displacement,” says Nicol. “So both of those mean we’re likely to have more earthquakes with bigger magnitudes than we’ve allowed for.”

However, there’s a competing suggestion that might mean the model is over-estimating the hazard. In general, the model has a ‘one fault = one earthquake’ policy—but Nicol says recent research has shown multiple fault ruptures are actually the norm for big earthquakes in New Zealand. That means the model could be allowing for five earthquakes when there was only one.

“The big earthquakes that we’ve got good historical data for—Napier, Edgecumbe, Darfield—all of those ruptured multiple faults. And this earthquake has really brought that home.”

This is an area of contention among scientists—and there is still so much that is unknown—but all agree that the 2016 Kaikōura event will contribute hugely to increasing our understanding.

“Earthquakes like this are a terrible tragedy for a lot of people, but for science they’re a real bonanza,” Nicol says. “We learned a lot from the Darfield quakes, but I expect this one will be an order of magnitude better in terms of what we can learn from it and how it helps us improve our understanding of how the plate boundaries work.”

[Chapter Break]

Earthquakes are tragic. The scientists are acutely aware of the tension between their professional excitement and residents’ distress and displacement. Often they’re the first outsiders on the scene, and, though they’re not trained counsellors, end up doing a lot of listening.

Heading south along the battered highway, Van Dissen, photographer Rob Suisted and I stumble across a gathering at the Kekerengu Store. A hulking black helicopter roars on the beach, and Prime Minister John Key is taking selfies with the locals.

Van Dissen knows a lot of people, including Bridget Jessep, who runs Clarence River Rafting. She’s here partly to catch up with neighbours—and with her GP, so she can get a prescription renewed—but the PM’s visit is meaningful, too. She and her neighbours have felt abandoned since the quake.

“We’ve seen more of GNS than the emergency services. It’s just interesting being on the sharp end of the stick in an event like this,” she says.

Like most others, Jessep is preparing for a difficult summer, but hardship doesn’t mean people aren’t fascinated by the transformations around them—and hungry for more information. Van Dissen produces a map and geology pamphlet from his pocket, underlines faults and answers questions.

Over by the tea table we meet John Murray. His family has lived in the Clarence River valley for three generations. The morning after the quake, “one of the local rafters came back and said, ‘Something’s happened to your river flat down there,’” he says. “Something’s happened, all right.”

Murray farms a 70-hectare stretch along the northern bank of the Clarence, a flat area at the foot of the hills running towards the river. Overnight, the unspooling Papatea Fault had jacked up the far side by eight to 10 metres, leaving a vertical cliff in the middle of his formerly flat field and diverting the river in front of it. “I was absolutely gobsmacked,” says Murray.

A GNS team has already been down to check it out. “The woman in charge said, ‘This is Christmas for me, this is what we train for.’” He didn’t mind, though. “It pisses me off that I’ve lost some land, but you’ve got to look at the big picture. We’ve seen something phenomenal, that my father never saw, that my grandfather never saw.”

[Chapter Break]

As the sun rose on November 14, the earthquake’s dramatic effects on the landscape were clear—the fault traces, the uplift, some 100,000 landslides. For scientists who focus on the marine environment, finding out what had happened to the ocean floor was a lot more complex, though luck was on their side.

Marine geologist Philip Barnes from NIWA was at sea when the earthquake struck, on board the research vessel Tangaroa off the east coast of the lower North Island.

He was leading a team of 14 people studying the impacts of past earthquakes on the floor of the Hikurangi Trough, the vast deep basin that follows the plate boundary north from Kaikōura. They were taking cores of what are called ‘turbidites’: layers of debris from underwater landslides—‘turbidity currents’—that roar along the seabed after large earthquakes.

Because the heavier particles fall out first and the lighter ones settle over days and weeks, the cores look stripy, so it’s easy to see where one turbidite ends and another begins. They can also be dated. “We can build up high-resolution chronologies of when particular earthquakes affected that part of the sea floor—where, how big, what’s the recurrence between events, and, with a bit of luck, how much time since the last one,” says Barnes.

Now they had a chance to measure the effects of a large earthquake on the undersea environment, just days after the event. Conveniently, they had all the gear and expertise on board already, so they diverted the vessel to the Kaikōura coast.

In 2015, Barnes and colleagues had mapped parts of the underwater Needles Fault in detail, and they wanted to see if anything had changed. It had. The fault had slipped vertically by 1.4 metres and broken along 34 kilometres of seabed.

The Tangaroa team also drew up sediment cores, and soon found evidence that a brand-new turbidite had formed over a wide area of the Hikurangi Trough, 100–200 millimetres thick and extending 300 kilometres from Kaikōura. That means the earthquake must have caused massive submarine landslides, says Barnes—but until the offshore area is comprehensively surveyed, we won’t know where they occurred.

Joshu Mountjoy, also from NIWA, has spent years trying to understand the behaviour of these kinds of landslides—in particular, to establish the risk of one occurring at the head of the Kaikōura Canyon. Scoured out by sediment from Canterbury’s rivers, the canyon starts very close to shore in shallow water, quickly diving to 1200 metres deep: a big landslide here would cause a devastating tsunami.

That didn’t happen on November 14, despite the intense shaking and the Hundalee Fault forging offshore virtually into the canyon. That might suggest the risk isn’t as great as was once thought, Mountjoy says, but he hopes to take one of NIWA’s smaller vessels out this summer to find out just what has happened down there.

“This is probably the most dynamic canyon in New Zealand and the opportunity to see how it responds to an event like this is unprecedented in our lifetime,” he says. “There has been a lot of speculation about the influence of earthquakes and tectonics on submarine canyon development, but here’s an example where it’s just happened, and we can see what it does.”

Mountjoy’s team will also try to resolve the last puzzle pieces of this earthquake’s complex fault story. What happened when the Hundalee Fault ran offshore, what undersea fault was triggered? And what lifted the Kaikōura Peninsula?

These are the detailed questions that underlie the big puzzles of geology—and plenty of those, too, remain unanswered. How do earthquakes leap from one fault to another? How exactly did the Kaikōura Earthquake influence other seismic activity, like Rotorua’s reawakened geyser and the Hawke’s Bay slow-slip earthquake? And is ‘The Big One’—the anticipated magnitude-8 rupture on the Alpine Fault—more or less likely now? Did it ease stresses on the fault, or make them worse? (See sidebar.)

If anything, the mysteries of the Kaikōura Earthquake have underscored just how complex the Earth’s processes are. The detail gathered by the marine, uplift, fault, seismology and other science teams over the coming weeks and months, each a tiny fragment of a million-piece puzzle, will add incrementally to our understanding of how this plate boundary works.

Predicting with confidence where and when an earthquake will happen is the holy grail of this field of geology, and it’s still a long way off.

[Chapter Break]

In the late afternoon we reach Waipapa Bay. It’s the end of the road, the highway comprehensively buried in a series of huge slips. It’s also the site of the most striking coastal uplift. In most places, the seabed was elevated two to three metres. Here, where twin traces of the Papatea Fault split to the shore, they forced a 700-metre stretch of coastline five and a half metres out of the sea.

That sudden thrusting extrusion is hard to imagine—what must it have looked like in the moonlight, as tonnes of solid rock leapt from the seabed, sloughing off the ocean? Ten days on, limp tresses of dying kelp lie stinking on the rocks, and seabirds fight over the banquet raised up for them. There is still a sense something dramatic happened here, but it’s fading. Soon, it will seem that it has always been this way.

This is the great illusion of our landscape—it feels so solid, so permanent, but it is made in violence and constant change.

At the beach we find Jane and Derrick Millton, the Clarence valley farmers whose landslide-stranded cows made international news. An entire hillside has collapsed onto their property, a slip 300 metres high and a kilometre across, Derrick says. They’ve come down to show a friend what’s happened to their local bay.

Jane Millton was born here, spent most of her life here. “The second day after the earthquake I was so sad… But then I had this funny feeling that it’s quite exciting—what a privilege in my lifetime to see something I know so well, now so different.” But the novelty is wearing off. “The first few days, you’re buzzing on adrenalin, and then it hits you—the emotion and the tiredness and the reality of the massive job ahead of you to tidy it all up.”

She looks out at the unfamiliar beach. “It’s a pretty special place. It will be again. It’s just a bit broken.”