The Big One
The Alpine Fault ruptures—on average—every 330 years with a magnitude 8 earthquake. Geologists and authorities are racing to quantify what might happen, and how they might respond in the event of the next one, likely to occur some time in the next 50 years.
At 3 AM on May 29, 2013, the South Island’s technological unconscious roars suddenly from its thousand severed throats, then subsides.
The technological infrastructure every civilisation relies on—the hydro-electric stations, electricity transmission lines, sewage pipes, water pipes, telephone lines and internet, the roads, railways, airports, even the ethereal mobile network—all of it is either broken, crippled, overloaded or, for safety reasons, shut down.
The South Island shakes for three minutes. The technology fails. Taps disgorge their last rope of water, then croak.
Welcome to the Alpine Fault earthquake. Its epicentre is Fox, but its rupture extends from Milford Sound, 400 kilometres north to Springs Junction. The shaking extends out in a 200-kilometre aureole around the rupture, triggering landslides, liquefaction, building damage, infrastructure damage and casualties from Nelson in the north, to Christchurch in the east and Dunedin in the south.
But as the day dawns, the epicentre of this magnitude-8 quake is revealed to be not Fox at all, but a 10-metre-by-6-metre white marquee at 41 Chester St, Christchurch, on hire from Party Warehouse, with a gas heater keeping its four occupants warm.
Tom Robinson, an Imperial-Collegetrained PhD candidate at Canterbury University, emerges from the tent and gestures back to it with the quiet satisfaction of a malevolent god.
“In there,” he says, “is the power of nature.”
Six months earlier, Canterbury’s Civil Defence Emergency Management (CDEM) recruited Robinson to produce the realistic aftermath of a magnitude 8 earthquake. He has parcelled up a litany of catastrophic effects, and today his tent team will supply them as successive ‘injects’ to the Canterbury team, who are gathered in the CDEM headquarters adjacent, and to other South Island civil defence hubs. None of them know what’s coming.
Exercise Te Ripahapa—loosely translated as Boundary Fault—would be the most comprehensively scripted Civil Defence exercise ever undertaken in New Zealand. It begins in Christchurch at 9 am with a situation report on the hours since 3 am: Power outage across the entire South Island was almost instant. The hydro-electric dams held, but as the shaking increased, the generators shut down automatically to protect their turbines. At 8 am, after inspection, Manapouri began generating again, and its electricity kick-started the eight Mackenzie Basin stations. Electricity is leapfrogging back up the island, but damage to the trans-alpine pylons and lines means the outage on the West Coast and Queenstown is permanent. State Highway One is blocked by a rockfall at Kaikoura. Christchurch Airport, initially shut for inspection, is functioning again, but Greymouth Airport has suffered severe liquefaction, and the condition of other airports is unknown.
Robinson has the rigour of a scientist. And though the sitrep is hypothetical, it’s entirely realistic. Over previous months, he’s interviewed the power companies, the telcos, the New Zealand Transport Agency, investigated runway substrates to see which might liquefy. He’s pulled up research, such as the paper published in 2012 that documented four large rockfalls into Milford Sound rockfalls triggered, so the paper supposed, by previous Alpine Fault earthquakes, and large enough to create a significant wave across the confined waters of the sound.
The only degree of freedom Robinson and Canterbury CDEM’s Te Ripahapa organiser, James Thompson, have allowed themselves involves reporting times. Most damage in a major quake is done and dusted within minutes. But reporting of that damage to the CDEM hubs might, given the breakdown of communication implicit in a big quake, take many hours.
Inside the Canterbury CDEM building, the duty manager, James Kirner, fields his first inject. A fisherman who works out of Milford Sound has reported Milford village destroyed by a local tsunami.
Kirner sits at his computer desk and dispenses injects. The CDEM team is divided into specialist sections, each with its own job, and the room becomes a haze of crosscutting activity. In response to information cascading through the duty manager’s desk, Te Ripahapa players clamp magnetised buttons onto large-scale Canterbury wall map: damage to the Otira viaduct, Lewis Pass bridges down, the litany goes on, the buttons go up—orange for confirmed road blockage, blue for suspected river blockage, red for rockfalls, black for casualty locations.
In rooms adjoining the CDEM team, the fire, police and St John Ambulance emergency services run their own show, but feed back into the main room. Dave Berry, the Fire Service area commander for Canterbury, has a team flying by helicopter to Methven for a hotel collapse, another en route to Hanmer. Warrant Officer Simon Cole, representing the Army, has requested a P3 Orion for aerial reconnaissance. Eight helicopters are heading to Nelson from the North Island, including the Air Force’s 18-man-capacity NH90s. Cole can’t just commandeer these aircraft, but his request has gone up the chain of command and it’s being assessed. Meantime, he is working with Kevin Moran, who heads CDEM’s operations team, organising local helicopters to undertake urgent reconnaissance. They’ve organised a priority reconnoitre of the Waimakariri River. At first light, Environment Canterbury engineers identified low flows on the river, and the chopper pilot has moseyed 70 kilometres upstream and detected a 40-metre landslip in a gorge. The threat to all points downstream—not least the city of Christchurch if the dammed river backs up too high—is too obvious to need comment.
The Milford Sound and Waimakariri emergencies clarify an important distinction. The Canterbury CDEM has been pelted with sufficient disturbing reports that a cooler structure must prevail. The CDEM regional controller, Neville Reilly, oversees that structure. He’s gathered reports from all of Canterbury’s territorial local authorities (TLAs) and fired the information up to the National Crisis Management Centre (NCMC) in Wellington. It’s the job of staff there to co-ordinate the national response. Now he turns to Canterbury’s own regional action plan. His brief, as is the brief for every regional Civil Defence hub in the South Island, is to look after his own backyard first, then see what’s available for emergencies outside the region. He can disregard Milford Sound. It’s Southland’s territory, and beyond that, almost certainly the NCMC’s job to get a Navy vessel or an NH90 into the devastated village. But evacuation downstream of the Waimakariri landslip is an urgent regional priority.
Only after each region has its backyard in order will surplus resources spill to other areas—specifically, the West Coast. For the purposes of today’s exercise, the West Coast is something close to a black hole from which few signals escape. It’s suffering in silence.
There’s an alarming echo between today’s hypothetical silence and the future, for the time will come.
All of the coast’s population centres are built near the fault. At Franz Josef, the 24,000-litre storage tank for the local gas station is embedded in the fault’s crush zone.
Like so much of old New Zealand, the Coast was built without thought of seismic hazard.
New Zealand’s geology was once understood either through the search by the old Department of Scientific and Industrial Research (DSIR) for coal and minerals or, at its more exalted end, geomorphology. In the 1940s, Victoria University geology professor Charles Cotton had held the country in thrall to his Geomorphology: An Introduction to the Study of Landforms, then under revision for its third edition. Geomorphology was mostly topography,the action of wind and water and ice on landscape formation, and stratigraphy, the study of rock strata, which underlay and shaped the surface. There was little specialist attention to tectonics, which was simply stratigraphy on the move.
Old geology acknowledged an alpine uplift, but the prevailing explanation for coastal mountains was an offshore trough, the geosyncline, that sagged under an increasing burden of sediment, put pressure on the mantle, and the mantle—by the principle of isostasy—responded to its load and rebounded like a trampoline. There was no reason to look for a fault at the base of the Southern Alps, and if there was no-one looking, the fault was easy to miss, the strike of it softened by rapid erosion and concealed by dense forest. Recognition of the sudden and more brutal mechanism of New Zealand’s alpine tectonics awaited a more primitive consciousness than could be supplied by the prevailing academia. In 1941, the right man was on his way.
Photos of Harold Wellman in any group usually catch him glancing sideways, above and beyond any grinning human assembly. Smiling at the camera simply wasn’t in his nature. He was smart, practical, argumentative and intuitive. He’d trained as a surveyor, but as the 1930s depression took hold, he travelled south and went black-sand gold mining on the coast. Tiring of that, he began odd-jobbing, including intermittent work for the DSIR. That contact morphed into a full-time job, and a suggestion that he fledge as a geologist by taking a degree. In between fieldwork, he’d completed a masters degree, but when the bus dropped him and Max Willett, a fellow freshman geologist, at Hokitika in August 1941, Wellman was still a raw recruit.
The 19th-century surveyor Charlie Douglas had reported a mica lode near Paringa, and the two geologists were under instruction to assess its mining potential. They studied the Douglas maps at the Hokitika Lands Office and then, hitching south from the town and gazing left, Wellman saw continual slight offsets on the moraines, and notches in the hills. His training was minimal, but his eyes were fresh. He saw a continuing fault. [Continues below…]
Suddenly curious, the two pushed into the West Coast’s tumbled riverbeds and found the rounded cobbles of two dominant rocks—granite and greywacke. The rivers were excavating across a major unconformity. The men picked up a lift on the back of a Model-A truck and as they rolled along, the passing landscape fell suddenly into an enlightening pattern. The solid low humps of granite lay on the western side of the fault, and the brittle ridges of greywacke climbed to the east. The fault was keeping pace with them as surely as the moon follows a child on a night-time street, and they mapped it at a steady 20 miles per hour.
They assessed the mica at Paringa, then walked on to fulfil a larger self-appointed mission, past Haast to the Jackson River, then followed the coast around to the Hollyford River mouth and up to Lake McKerrow. They were then 100 kilometres due south of Haast, but still the fault scarp ran on, into the northern side of the lake and out the other side headed for Milford Sound.
It was the longest straight line on Earth, and it bespoke pure tectonic power. Wellman named it the Alpine Fault and in his biographical notes described his approach to it, near Paringa: “The broken schist looked like the result of explosions, or perhaps we were seeing the heart of old earthquakes.”
Back at the CDEM headquarters, Neville Reilly has finished teleconferencing with his immediate bosses. At regional level, Civil Defence is under the control of the local authorities, and a top committee of mayors takes advice but retains control. The chair of that committee, Janie Annear of Timaru, has just declared a civil emergency. Reilly now has powers of control and compulsion. If he wants to commandeer a helicopter from—say—the media, he can do it.
By the afternoon, Te Ripahapa has settled down from the seething ants-nest of the morning to some semblance of control. The action plan for Canterbury is not only being implemented, including such detail as explosives to blow the Waimakariri dam, the managers are now organising the next action plan.
Te Ripahapa is also starting to exhibit a few of the farcical elements that are inevitable in a long role-play. Kevin Moran from the operations team moves across to resources to check the action on two trampers badly injured and needing evacuation from the head of the Boyle River and is informed they’re still looking for a spare helicopter.
“I sent that email two hours ago!” says Moran. “Forget it, they’ll have bled out by now.”
In a separate room altogether, another of the independents, Paul Schoolderman from the public health division of the Canterbury District Health Board, is head down and earnest. He’s preparing a checklist for the West Coast District Health Board in Greymouth. He’s listed water sampling, sanitation and infectious diseases as priority concerns. He’s compiling instructions for efficient pit latrines. And he’s listed “temporary morgue”. He turns with a bright smile. “We’re going to have to look at disposal of the dead.”
While there are moments of good humour, there’s also a shared seriousness around Te Ripahapa. After Christchurch’s two big quakes, all are aware that New Zealand’s earthquake hazard is real, and the Alpine Fault is now recognised as one of the major hazards. That recognition didn’t come easily.
Christchurch 1949: twenty geologists pose for a group picture. The American seismologists Beno Gutenberg and Charles Richter stand at the front left, two obvious stars. As usual, Harold Wellman is all but divorced from the group, back row, off to the right. Everyone smiles for the camera, except Wellman. He’s not looking. He’s gazing right, and westward to an event horizon beyond the Alps.
The occasion is the Seismology Symposium section of that year’s prestigious Pacific Science Congress. Richter and Gutenberg had already developed their magnitude scale, a standard seismometer for measuring that magnitude, and backed by a Defence Department budget they’d salted those seismographs around the world. Richter’s paper to the symposium listed global earthquake patterns, and he noted a “vigorous deformation” on a “circum-Pacific active zone”. That zone showed the San Andreas Fault across Southern California, and also marked in the Alpine Fault, but gave it no specific mention.
Wellman wasn’t scheduled to speak, but he’d hinted big things and his impromptu lecture was well attended. As it began, he stood beside a large handmade geological map one metre by two metres, mounted on an easel. The map showed the South Island’s distinctive rock bands in colour, including the ultramafic rocks that make up Dun Mountain and the Red Hills in the north, and which reappear as the Red Hills Range in Westland. Partway through his lecture, Wellman suddenly slid Westland 480 kilometres along the Alpine Fault to match up the strata. Worldwide, no geologist had ever proposed such an astonishing offset.
Simon Nathan’s book Harold Wellman: A Man Who Moved New Zealand recorded eyewitness accounts of a vivid moment. “One professor clapped his hands over his head, taking it in at once and accepting it.” Not everyone did. A former geology professor from Otago University, Patrick Marshall, rose to dispute it and Wellman tossed him the chalk and suggested he provide his own interpretation. Offended, the professor walked out.
The American reaction was not recorded, but in 1958, when Richter published Elementary Seismology, his standard textbook on earthquakes, he referred specifically to his experience in New Zealand and listed the Alpine Fault as a world-ranked master fault. Under the heading ‘Large Strike Slip?’ he acknowledged Wellman’s “bold suggestion” of an accumulated 480-kilometre shift “since the Mesozoic”. It was a cautious assessment. The Americans had evidence, most lately from the 1906 San Francisco earthquake, that the San Andreas Fault could slip sideways up to five metres at a time, but the conventional wisdom was that the fault’s accumulated shift might total 40 kilometres. [Continues below…]
But by 1958, Wellman was already far ahead of the scholarship, rampant with his own vision. He rejected the diffuse age range implicit in the Mesozoic’s 250-to 65-million-year-old time span. He’d done more fieldwork along the fault and he’d flown over the Haast-Milford section. In a 1959 paper, he noted a “1000-foot horizontal shift of glaciated surface on the south bank of the Martyr River”. The paper ended with a simple logical sequence: if the last glacial advance was 20,000 years ago, then the annual rate of movement to create the Martyr River offset was around half an inch a year, and at that speed, “the postulated 300-mile shift could have been accomplished since the Oligocene”.
In geological terms he’d lopped off over 100 million years, with huge implications for earthquake probability—radical stuff that was immediately challenged. In an authoritative 1963 paper, Pat Suggate, later to become director of the New Zealand Geological Survey and one of New Zealand’s most respected geologists, acknowledged Wellman’s work, then summed up all the research on the Alpine Fault and reiterated the horizontal shift had been largely completed between 150 and 140 million years ago.
Wellman’s response, in a 1964 paper, was to up the rate of offset to an average of more than 25 millimetres a year and a total time for the offset of 19 million years—an astonishing rate of travel, and accepted at that time by almost no-one.
By the 1980’s with plate tectonics firmly in place, analysis of the fault would show its horizontal movement averaged 23 millimetres a year—very close to Wellman’s estimate—and that it was taking up an estimated 61 per cent of the oblique pressures exerted by the Pacific and Australian plates.
As to earthquake frequency, the fault was still mute, but there were ways of making it speak. Big earthquakes cause landslides and the rivers issuing down onto the West Coast would cut through the detritus, leaving terraces either side of the downcutting that might contain dateable organic debris.
First off the blocks was a young geologist, John Adams. He prised fallen timber out of three river terraces and carbon-dated them by layer. His 1980 paper was the first to suggest a regular ticking of the tectonic clock. The fault ruptured on average, he claimed, once every 500 years, at around magnitude 8.
The range of the research was slim, but good enough to put the geological community and engineers on alert. Yet South Island seismometers recorded few shakes on the fault, so maybe it was aseismic, maybe it simply crept, the compression of the plates taken up by elastic strain across the Alps, or absorbed in ductile zones.
In 1998, the Earthquake Commission commissioned Geotech consultant Mark Yetton to do a major study, but even during his initial deskwork, Yetton knew the aseismic school was wrong. Overseas studies suggested that the lack of earthquakes on a major fault did not signal quiescence. Rather, many of the world’s big slip-strike faults were seismically evolved, their fault plane polished over time so that small and medium earthquakes dropped away and the big faults waited century by century for their single magisterial rupture.
Yetton extended Adams’ river terrace investigation from three to 19 sites, and his analysis went beyond the river terraces. He hand-dug two trenches across the fault, and machine-dug three others with a 15-tonne Mitsubishi excavator. He found unmistakeable earthquake signatures, including liquefaction. Radiocarbon dating of both the river terrace and the trench material gave a broad band age for two large quakes, one within the range 1480–1645, the other 1700–1750.
Partway through the three-year study, Yetton met Andrew Wells, just then pioneering the new field of dendrology. Wells had a Scandinavian corer that extracted a four-millimetre-radius ‘straw’ out of tree trunks. It was hand-operated, and tough work coring the trunks of hardy old mataior rimu, but the two persisted and the final results were astonishingly precise.
In 1620 and 1717, the tree growth rings consistently diminished, even plunging in many cases from a regular average width of one or two millimetres, to zero. Yetton backed up these dates with evidence for rupture length, using trench data and forest disturbance patterns. The 1620 event ruptured along 250 kilometres, he concluded. The 1717 event ruptured from Milford Sound to Springs Junction, more than 400 kilometres. The energy release implicit in those ruptures bespoke earthquakes around magnitude 8.
Studies that went on into the 2000s gradually pushed the earthquake record back to 1000 AD, but in 2008, in deep bush, GNS scientists Kelvin Berryman and Ursula Cochran found the geological equivalent of the Rosetta Stone. Not far from where Wellman had first confirmed the fault’s passage down to Lake McKerrow, they uncovered an 18-metre bank, downcut by the Hokuri Creek. It was heavily vegetated, but if you cleared it with a spade, the bank was a virtual barcode of earthquake sequence—18 vertical metres of alternating dark swampy and grey sedimentary layers. The bank held one of the longest continuous quake records of any on-plate fault in the world.
The two scientists radiocarbon-dated the black organic layers, and their 2012 paper reported 24 ruptures going back 8000 years. The maths was simple—an average of one Alpine Fault earthquake every 330 years, the last one in 1717.
They noted “a fairly regularly repeating earthquake cycle” and insofar as straight maths can provide probabilities, they did their duty of predicting seismic hazard: a 30 per cent chance of a large earthquake on the Alpine Fault within the next 50 years.
If you wanted a headline from that research, it was that the fault was ticking like a grandfather clock, and chiming on the hour. New Zealand was within the last five minutes of the next chime. The big black hands were moving towards the next awful hour and everyone should beware, for being very old and with worn cogs and suspect springs, the old granddaddy sometimes skipped forward whole minutes, and chimed early anyway.
As to te Ripahapa, James Thompson, its originator and organiser, gave it a tick in some areas, and a could-do-better in others.
On the debit side, some territorial local authorities breached protocol by going straight to the National Crisis Management Centre, bypassing their regional Civil Defence Emergency Management hubs. The internal server that connected the CDEM hubs with the NCMC in Wellington slowed as more people joined, and its software had unfamiliar functions that caused some confusion. On the plus side, the high-level communications worked smoothly between Canterbury CDEM and its civilian bosses, and between Canterbury CDEM and its sister hubs throughout the South Island. The co-ordination between the Canterbury team and the Fire Service, police, St John Ambulance and Army also worked well. Above and beyond all that, Te Ripahapa had raised awareness throughout the South Island’s Civil Defence structure of a major seismic hazard. There were graphic lessons, and mistakes that wouldn’t be repeated next time.
True, a number of territorial local authorities did pack up and go home at 5 pm, when the exercise still had two hours to run, but no-one blamed them too much for that. There was only so much earthquake you could take before the home comforts beckoned, the evening meal awaited, the nightly telly offered its happy blandishments, and so to bed.