When the 2011 earthquake and tsunami struck Tohoku, Japan, Chris Goldfinger was 320 km away, in the city of Kashiwa, at an international meeting on seismology. As the shaking started, everyone in the room began to laugh. Earthquakes are common in Japan – that one was the third of the week – and the participants were, after all, at a seismology conference. Then everyone in the room checked the time.
Seismologists know that how long an earthquake lasts is a decent proxy for its magnitude. The 1989 earthquake in Loma Prieta, California, which killed 63 people and caused six billion dollars’ worth of damage, lasted about 15 seconds and had a magnitude of 6.9. A 30-second earthquake generally has a magnitude in the mid-sevens. A minute-long quake is in the high sevens, a two-minute quake has entered the eights, and a three-minute quake is in the high eights. By four minutes, an earthquake has hit magnitude 9.0.
The earthquake was not particularly strong. Then it ticked past the 60-second mark, making it longer than the others that week. The shaking intensified.
At a minute and a half, everyone in the room got up and went outside.
It was March. There was a chill in the air and snow flurries but no snow on the ground. The earth snapped and popped and rippled. It was, Goldfinger thought, like driving through rocky terrain in a vehicle with no shocks, if both the vehicle and the terrain were also on a raft in high seas.
The quake passed the two-minute mark. The flagpole atop the building that he and his colleagues had just vacated was whipping through an arc of 40 degrees. The building itself was base-isolated, a seismic-safety technology in which the body of a structure rests on moveable bearings rather than directly on its foundation. Goldfinger lurched over to take a look. The base was lurching, too, back and forth 30 cm at a time, digging a trench in the yard. His watch swept past the three-minute mark and kept going.
For decades, seismologists had believed that Japan could not experience an earthquake stronger than magnitude 8.4. In 2005, however, at a conference in Hokudan, a Japanese geologist named Yasutaka Ikeda had argued that the nation should expect a magnitude 9.0 in the near future – with catastrophic consequences because Japan’s famous earthquake and tsunami preparedness was based on incorrect science. Now, Goldfinger realised as the shaking hit the four-minute mark, the planet was proving the Japanese Cassandra right.
For a moment, that was pretty cool: a real-time revolution in earthquake science. Almost immediately, though, it became extremely uncool because Goldfinger and every other seismologist in Kashiwa knew what was coming.
One of them pulled out a mobile phone and started streaming videos from the Japanese broadcasting station NHK, shot by helicopters that had flown out to sea soon after the shaking started. Thirty minutes after Goldfinger first stepped outside, he watched the tsunami roll in, in real time, on a tiny screen.
In the end, the magnitude 9.0 Tohoku earthquake and subsequent tsunami killed more than 18,000 people, devastated northeast Japan, triggered the meltdown at the Fukushima power plant, and cost an estimated $220 billion. The shaking earlier in the week turned out to be the foreshocks of the largest earthquake in the nation’s recorded history. But for Chris Goldfinger, a paleoseismologist at Oregon State University and one of the world’s leading experts on a little-known fault line, the main quake was itself a kind of foreshock: a preview of another earthquake still to come.
The Coming Quake
Most people in the United States know just one fault line by name: the San Andreas, which runs nearly the length of California and is perpetually rumoured to be on the verge of unleashing ‘the big one’. That rumour is misleading, no matter what the San Andreas ever does. Every fault line has an upper limit to its potency, determined by its length and width and by how far it can slip. For the San Andreas, that upper limit is roughly an 8.2 – a powerful earthquake but, because the Richter scale is logarithmic, only 6% as strong as the 2011 event in Japan.
Just north of the San Andreas, however, lies another fault line. Known as the Cascadia subduction zone, it runs for 1130 km off the coast of the Pacific Northwest, beginning near Cape Mendocino, California, continuing along Oregon and Washington, and terminating around Vancouver Island, Canada. The Cascadia part of its name comes from the Cascade Range, a chain of volcanic mountains that follows the same course 150 km or so inland. The subduction zone part refers to a region of the planet where one tectonic plate is sliding underneath (subducting) another. Tectonic plates are those slabs of mantle and crust that, in their epochs-long drift, rearrange the Earth’s continents and oceans.
Take your hands and hold them palms down, middle fingertips touching. Your right hand represents the North American tectonic plate, which bears on its back the entire continent. Your left hand represents an oceanic plate called Juan de Fuca, some 23,000 km2 in size. The place where they meet is the Cascadia subduction zone. Now slide your left hand under your right one. That is what the Juan de Fuca plate is doing: slipping steadily beneath North America. When you try it, your right hand will slide up your left arm, as if you were pushing up your sleeve. That is what North America is not doing. It is stuck, wedged tight against the surface of the other plate.
Curl your right knuckles up so that they point towards the ceiling. Under pressure from Juan de Fuca, the stuck edge of North America is bulging upward and compressing eastward, at the rate of, respectively, 3-4 mm and 30-40 mm a year. It can do so for quite some time. But it cannot do so indefinitely. There is a backstop – the craton, that ancient unbudgeable mass at the centre of the continent – and, sooner or later, North America will rebound like a spring. If only the southern part of the Cascadia subduction zone gives way – your first two fingers, say – the magnitude of the resulting quake will be somewhere from 8.0 to 8.6. That’s the big one. If the entire zone gives way at once, an event that seismologists call a full-margin rupture, the magnitude will be somewhere from 8.7 to 9.2. That’s the very big one.
Flick your right fingers outwards, forcefully, so that your hand flattens back down again. When the next very big earthquake hits, the northwest edge of the continent, from California to Canada and the continental shelf to the Cascades, will drop by as much as 1.8 m and rebound 10-30 m to the west. Some of that shift will take place beneath the ocean, displacing a colossal quantity of sea water. The water will surge upwards into a huge hill, then promptly collapse. One side will rush west, towards Japan. The other side will rush east, in a 1100 km liquid wall that will reach the Northwest coast, on average, 15 minutes after the earthquake begins. By the time the shaking has ceased and the tsunami has receded, the region will be unrecognisable. Kenneth Murphy, who directs the Federal Emergency Management Agency (FEMA) Region X, the division responsible for Oregon, Washington, Idaho, and Alaska, says, “Our operating assumption is that everything west of Interstate 5 will be toast.”
In the Pacific Northwest, the area of impact will cover some 360,000 km2, including Seattle, Tacoma, Portland, Eugene, Salem, Olympia, and some seven million people. When the next full-margin rupture happens, that region will suffer the worst natural disaster in the history of North America. Roughly 3000 people died in San Francisco’s 1906 earthquake. Almost 2000 died in Hurricane Katrina. FEMA projects that nearly 13,000 people will die in the Cascadia earthquake and tsunami. Another 27,000 will be injured, and the agency expects it will need to provide shelter for a million displaced people and food and water for another 2.5 million. “This is one time that I’m hoping all the science is wrong, and it won’t happen for another thousand years,” Murphy says.
In fact, the science is robust, and one of the chief scientists behind it is Chris Goldfinger. Thanks to work done by him and his colleagues, we now know that the odds of the big Cascadia earthquake happening in the next 50 years are roughly one in three. The odds of the very big one are roughly one in ten. Even those numbers do not fully reflect the danger – or, more to the point, how unprepared the Pacific Northwest is to face it. Forty-five years ago, no-one even knew the Cascadia subduction zone existed, and its discovery stands as one of the greatest scientific detective stories of our time.
Reading the Trees
Almost all the world’s most powerful earthquakes occur in the Ring of Fire, the volcanically and seismically volatile swathe of the Pacific that runs from New Zealand up through Indonesia and Japan, across the ocean to Alaska, and down the west coast of the Americas to Chile. The Ring of Fire, it turns out, is really a ring of subduction zones. Nearly all the earthquakes in the region are caused by continental plates getting stuck on oceanic plates – as North America is stuck on Juan de Fuca – and then getting abruptly unstuck. And nearly all the volcanoes are caused by the oceanic plates sliding deep beneath the continental ones.
The Pacific Northwest sits squarely within the Ring of Fire. Off its coast, an oceanic plate is slipping beneath a continental one. Inland, the Cascade volcanoes mark the line where, far below, the Juan de Fuca plate is heating up and melting everything above it. In other words, the Cascadia subduction zone has, as Goldfinger put it, “all the right anatomical parts”. Yet not once in recorded history has it caused a major earthquake – or any quake to speak of. By contrast, other subduction zones produce major earthquakes occasionally and minor ones all the time. The question facing geologists was whether the Cascadia subduction zone had ever broken its eerie silence.
In the late 1980s, Brian Atwater, a geologist with the US Geological Survey, and postgraduate student David Yamaguchi found the answer – and a major clue in the Cascadia puzzle. Their discovery is best illustrated in a place called the ghost forest, a grove of western red cedars on the banks of the Copalis River, near the Washington coast. The cedars are spread out across a salt marsh on a northern bend in the river, long dead but still standing. Leafless, branchless, barkless, they are reduced to their trunks and worn to a smooth silver grey.
It had long been assumed that they died slowly, as the sea level around them gradually rose. But by 1987, Atwater, who had found in soil layers evidence of sudden land subsidence along the Washington coast, suspected that that was backwards – that the trees had died quickly when the ground beneath them plummeted. To find out, he teamed up with Yamaguchi, a specialist in dendrochronology, the study of growth-ring patterns in trees. Yamaguchi took samples of the cedars and found that they had died simultaneously: in tree after tree, the final rings dated to the summer of 1699. Since trees do not grow in the winter, he and Atwater concluded that sometime from August 1699 to May 1700, an earthquake had caused the land to drop and killed the cedars. That time frame predated the written history of the Pacific Northwest – and so, by rights, the detective story should have ended there.
But it did not. If you travel 8000 km due west from the ghost forest, you reach the northeast coast of Japan. That coast is vulnerable to tsunamis, and the Japanese have kept track of them since at least 599CE. One incident has long stood out for its strangeness. On the eighth day of the 12th month of the 12th year of the Genroku era, a 950 km-long wave struck the coast, levelling homes, breaching a castle moat, and causing an accident at sea. The Japanese understood that tsunamis were the result of earthquakes, yet no-one felt the ground shake before the Genroku event. When scientists began studying it, they called it an orphan tsunami.
Finally, in a 1996 article in Nature, a seismologist named Kenji Satake and three colleagues, drawing on the work of Atwater and Yamaguchi, matched that orphan to its parent. At approximately nine at night on January 26, 1700, a magnitude 9.0 earthquake struck the Pacific Northwest, causing sudden land subsidence and, out in the ocean, lifting up a wave half the length of the continent. It took roughly 15 minutes for the eastern half of that wave to strike the Northwest coast. It took ten hours for the other half to cross the ocean. It reached Japan on January 27, 1700: by the local calendar, the eighth day of the 12th month of the 12th year of Genroku. Once scientists had reconstructed the 1700 earthquake, certain previously overlooked accounts also came to seem like clues.
The reconstruction of the Cascadia earthquake of 1700 is one of those rare natural puzzles whose pieces fit together as tectonic plates do not: perfectly. It is wonderful science. It was wonderful for science. And it was terrible news for the millions of inhabitants of the Pacific Northwest. As Goldfinger put it, “In the late ’80s and early ’90s, the paradigm shifted to ‘uh-oh.’”
When the Dogs Bark
Goldfinger told me this in his lab at Oregon State. Inside the lab is a walk-in freezer. Inside the freezer are floor-to-ceiling racks filled with cryptically labelled tubes, 10 cm in diameter and 1.5 m long. Each tube contains a core sample of the sea floor. During subduction-zone earthquakes, torrents of land rush off the continental slope, leaving a permanent deposit on the bottom of the ocean. By counting the number and the size of deposits in each sample, then comparing their extent and consistency along the length of the Cascadia subduction zone, Goldfinger and his colleagues were able to determine how much of the zone has ruptured, how often and how drastically.
Thanks to that work, we now know that the Pacific Northwest has experienced 41 subduction-zone earthquakes in the past 10,000 years. If you divide 10,000 by 41, you get about 243, which is Cascadia’s recurrence interval: the average amount of time that elapses between earthquakes. That time span is dangerous both because it is too long – long enough for us to unwittingly build an entire civilisation on top of our continent’s worst fault line – and because it is not long enough. Counting from the earthquake of 1700, we are now 315 years into a 243-year cycle.
It is possible to quibble with that number. Recurrence intervals are averages, and averages are tricky: ten is the average of nine and 11 but also of 18 and two. It is not possible, however, to dispute the scale of the problem. The devastation in Japan in 2011 was the result of a discrepancy between what the best science predicted and what the region was prepared to withstand. The same will hold true in the Pacific Northwest – but here the discrepancy is enormous.
The first sign that an epic Cascadia earthquake has begun will be a compressional wave, radiating outwards from the fault line. Compressional waves are fast-moving, high-frequency waves, audible to dogs and certain other animals but experienced by humans only as a sudden jolt. They are not very harmful, but they are potentially very useful, since they travel fast enough to be detected by sensors 30-90 seconds ahead of other seismic waves. That is enough time for earthquake early-warning systems, such as those in use throughout Japan, to automatically perform a variety of lifesaving functions: shutting down railways and power plants and triggering alarms so that the public can take cover. The Pacific Northwest has no early-warning system. When the Cascadia earthquake begins, there will be, instead, a cacophony of barking dogs and a long, suspended, what-was-that? moment before the surface waves arrive. Surface waves are slower, lower-frequency waves that move the ground both up and down and side to side: the shaking, starting in earnest.
Soon after that shaking begins, the electrical grid will fail, probably everywhere west of the Cascades and possibly well beyond. In theory, those who are at home should be safest; it is easy and relatively inexpensive to seismically safeguard a private dwelling. But most people in the Pacific Northwest have not done so. Anything indoors and unsecured will lurch across the floor or come crashing down. Refrigerators will walk out of kitchens, unplugging themselves and toppling over. Water heaters will fall and smash interior gas lines. Houses that are not bolted to their foundations will slide off – or, rather, they will stay put, obeying inertia, while the foundations jolt westwards.
Other, larger structures will also start to fail. Ian Madin, who directs the Oregon Department of Geology and Mineral Industries (DOGAMI), estimates that 75% of all structures in the state are not designed to withstand a major Cascadia quake. FEMA calculates that, across the region, something in the order of a million buildings – more than 3000 of them schools – will collapse or be compromised in the earthquake. So will half of all highway bridges, 15 of the 17 bridges spanning Portland’s two rivers, and two-thirds of railways and airports; also, one-third of all fire stations, half of all police stations, and two-thirds of all hospitals.
The shaking from the Cascadia quake will set off landslides throughout the region – up to 30,000 of them in Seattle alone, the city’s emergency-management office estimates. It will also induce a process called liquefaction, whereby seemingly solid ground starts behaving like a liquid, to the detriment of anything on top of it. Fifteen per cent of Seattle is built on liquefiable land, including 17 day-care centres and the homes of some 34,500 people. So is Oregon’s critical energy-infrastructure hub, a 10 km stretch of Portland through which flows 90% of the state’s liquid fuel and that houses everything from electrical substations to natural gas terminals. The sloshing, sliding and shaking will trigger fires, flooding, pipe failures, dam breaches and hazardous-material spills. Four to six minutes after the dogs start barking, the shaking will subside. For another few minutes, the region, upended, will continue to fall apart on its own. Then the wave will arrive, and the real destruction will begin.
Among natural disasters, tsunamis may be the closest to being unsurvivable. The only likely way to outlive one is not to be there when it happens: to steer clear of the vulnerable area or get yourself to high ground as fast as possible. For the 71,000 people who live in Cascadia’s inundation zone, that will mean evacuating in the narrow window after one disaster ends and before another begins. They will be notified to do so only by the earthquake itself – “a vibrate-alert system,” Kevin Cupples, the city planner for the town of Seaside, Oregon, jokes. Depending on location, they will have ten to 30 minutes to get out. That timeline does not allow for finding a torch, hesitating amid the ruins of a home, searching for loved ones. “When that tsunami is coming, you run,” says Jay Wilson, chair of the Oregon Seismic Safety Policy Advisory Commission (OSSPAC). “You protect yourself, you don’t turn around, you don’t go back to save anybody. You run for your life.”
What Can We Do?
The time to save people from a tsunami is before it happens, but the region has not yet taken serious steps towards doing so. Hotels and businesses are not required to post evacuation routes or to provide employees with evacuation training. These lax safety policies guarantee that many people inside the inundation zone will not get out. Twenty-two per cent of Oregon’s coastal population is 65 or older. Twenty-nine per cent of the state’s population is disabled, and that figure rises in many coastal counties. “We can’t save them,” Kevin Cupples says. “I’m not going to sugar-coat it and say, ‘Oh, yeah, we’ll go around and check on the elderly.’ No. We won’t.”
Those who cannot get out of the inundation zone under their own power will quickly be overtaken by a greater one. A grown man is knocked over by ankle-deep water moving at 10.7 km/h. The tsunami will be moving more than twice that fast. Its height will vary with the contours of the coast, from 6 m to more than 30 m. It will look like the whole ocean, elevated, overtaking land. Nor will it be made only of water – not once it reaches the shore. It will be a five-storey deluge of pickup trucks and door frames and cinder blocks and fishing boats and utility poles and everything else that once constituted the coastal towns of the Pacific Northwest.
The inundation zone in a full-margin rupture will be scoured of structures from California to Canada. The earthquake will have wreaked its worst havoc west of the Cascades but caused damage as far away as Sacramento, California. FEMA expects to coordinate search-and-rescue operations across 160,000 km and in the waters off 700 km of coastline.
OSSPAC estimates that in the Interstate 5 corridor, it will take one to three months after the earthquake to restore electricity, a month to a year to restore drinking water and sewerage service, six months to a year to restore major highways, and 18 months to restore health-care facilities. On the coast, those numbers go up. Whoever stays there will spend three to six months without electricity, one to three years without drinking water and sewerage systems, and three or more years without hospitals. How much all of this will cost is anyone’s guess. But whatever the ultimate figure, the economy of the Pacific Northwest will collapse.
On the face of it, earthquakes seem to present us with problems of space: the way we live along fault lines, in brick buildings, in homes made valuable by their proximity to the sea. But, covertly, they also present us with problems of time. The Earth is 4.5 billion years old, but we are a young species, with an average individual allotment of threescore years and ten. The brevity of our lives breeds a kind of ignorance of or an indifference to those planetary gears, which turn more slowly than our own.
This problem is bidirectional. The Cascadia subduction zone remained hidden from us for so long because we could not see deeply enough into the past. It poses a danger to us today because we have not thought deeply enough about the future. Where we stumble is in conjuring up grim futures in a way that helps to avert them.”
The School in the Zone
The last person I met with in the Pacific Northwest was Doug Dougherty, the superintendent of schools for Seaside, which lies almost entirely within the tsunami inundation zone. Of the four schools that he oversees, with a total student population of 1600, one is relatively safe. The others sit 1.5-5 m above sea level. When the tsunami comes, they will be as much as 14 m below it.
In 2009, Dougherty told me, he found land for sale outside the inundation zone and proposed building a new K-12 campus there. Four years later, to foot the $128 million bill, the district put up a bond measure. The measure failed by 62%. Dougherty tried seeking help from Oregon’s congressional delegation but came up empty. The state makes money available for seismic upgrades, but buildings within the inundation zone cannot apply. At present, all Dougherty can do is make sure that his students know how to evacuate.
Some of them, however, will not be able to do so. At an elementary school in the community of Gearhart, the children will be trapped. “They can’t make it out from that school,” Dougherty said. “They have no place to go.” On one side lies the ocean; on the other, a wide, roadless bog. When the tsunami comes, the only place to go in Gearhart is a small ridge just behind the school. At its tallest, it is 14 m high – lower than the expected wave in a full-margin earthquake. For now, the route to the ridge is marked by signs that say Temporary Tsunami Assembly Area. I asked Dougherty about the state’s long-range plan. “There is no long-range plan,” he said.
Dougherty’s office is deep inside the inundation zone, a few blocks from the beach. About 130 km further out, 3000 m below the surface of the sea, the hand of a geological clock is somewhere in its slow sweep. All across the region, seismologists are looking at their watches, wondering how long we have, and what we will do, before geological time catches up to our own.
Courtesy of the author. Originally published in The New Yorker (June 20, 2015), newyorker.com.