But in the late 1990s, the townsfolk were eager to show the visiting Americans the castle logbooks and family journals where the original reports were recorded. Documents older than the Declaration of Independence were stored in government filing cabinets and in private homes. “You go to these little town museums and they just put them in your hands,” Yamaguchi said. His favorite was from Miho, a pine-covered peninsula south of Tokyo. A family of innkeepers there preserved the writings of an ancestor, a peasant who has served as the village headman. The floodwaters struck Miho early in the morning, and his was one of the few eyewitness accounts. He described strange currents that flowed back and forth with the force of a river. Like Viola Riebe’s finger-wagging uncle, he cautioned future generations to remember that dangerous waves could hit with no ground shaking. “The whole village was puzzled,” he wrote.
Unlike the elegant calligraphy of the samurai scribes, the headman’s brushwork and language were crude. “I had to laugh,” Yamaguchi recalled. “This guy’s writing was so bad it looks like things I write in Japanese.”
The scientists quizzed residents on geography and any changes to the landscape over the past three centuries. Using historic maps they paced off the run-up described in the old documents. At the spot where water surged up a castle moat in 1700, they found a busy intersection. But on the margins of a wide bay, they could pinpoint the location of a fisherman’s shack and salt kilns flooded by the tsunami.
“A lot of what we saw was eerily similar to what the samurai described,” Yamaguchi said.
The field studies all pointed to a big wave in Japan—up to sixteen feet high in places—and a full-rip 9 from Cascadia. But the slam-dunk paternity test linking parent quake and tsunami offspring came from the Copalis ghost forest on the Washington coast.
When he first considered the problem, Yamaguchi didn’t see any way to wring more precise dates out of the old cedars. The trunks were too battered by time. None had the bark and outer rings he would need to establish an exact year of death.
It was Atwater who stumbled on the answer. Paddling up the Copalis on one of his many excursions there, he noticed that a cedar had succumbed to age and gravity and had toppled into the river. He steered his canoe in for a close look. Protruding from the riverbank were the broken ends of the tree’s roots. Grabbing his trenching tool, Atwater grubbed out the mud from around a root stub and saw that the airless muck had preserved its bark in pristine condition.
The ghost forest of the Copalis. (image credits 3.2)
It was a “D’oh” moment worthy of Homer Simpson.
“In retrospect we should have gone for the cedar roots first,” Yamaguchi said. “We just didn’t think of it.”
Atwater assembled a team of strong backs to wrestle roots from the ground in the summer of 1996. They targeted Yamaguchi’s star specimens—the trees with the most distinctive ring patterns—and dug like badgers on steroids. “It was a treasure hunt,” recalled Boyd Benson, a UW student at the time. The “nice, juicy roots” were five feet or more underground. Benson excavated pits around the tree bases deep enough to swallow a man. Atwater scrambled into the holes with his chain saw and carved out root chunks. The swampy ground was always wet, and it was a race to extract roots before the incoming tide flooded the pits.
“The hole would be filling up and there would be a rooster tail of wood chips and water flying into the air as Brian cut the roots,” Benson recalled.
By the end of each session, the men were so muddy that they jumped in the river to clean off before paddling back. Some of the roots were two feet in diameter. The scientists polished them until they gleamed. “When you’re done, they look like pieces of art,” Benson said.
To establish the time sequence, Yamaguchi had to match the bar codes in the trunks to the bar codes in the roots, something that hadn’t been done in the Northwest. It sounds straightforward, but dendrochronologists can go cross-eyed working with roots. The rings are lopsided and so tiny that the slightest error in measurement can throw the sequence off kilter. Add in the regional curse of abundant rainfall that blurs ring patterns and Yamaguchi was nervous. “This was new territory for tree-ring dating,” he said.
Benson packed the sections in cardboard boxes and loaded them and himself onto an Amtrak train headed for Lamont-Doherty Earth Observatory, perched on a bluff overlooking the Hudson River. The tree ring lab there is one of the world’s best. Benson worked with lab cofounder Gordon Jacoby to scrutinize the wood slabs with an apparatus that measures rings down to one-millionth of a meter—less than the thickness of a hair. Benson handed off the measurements to Yamaguchi, who fired up his bar code–matching program and held his breath. “We were asking these trees, ‘When did you die?’ ” Yamaguchi said. “They told us.”
Seven out of the eight cedars Yamaguchi examined had laid down their last ring in the warm months of 1699. By the start of the next growing season, in May 1700, all were dead. The single outlier limped along until 1708 because one root extended like a snorkel into higher ground, providing a temporary lifeline that delayed the tree’s demise.
The results drew a bull’s-eye around January 1700. There was no longer any question about the source of Japan’s orphan tsunami or about the size of the parent quake.
It was a doozy.
CHAPTER 4:
A VIEW FROM THE SEA
THE BOUNDARIES BETWEEN TECTONIC PLATES make good postcards. Picture the Himalayas lit by a setting sun or Andean spires piercing the sky. The rift valleys in Iceland and Africa, where plates are pulling apart, offer vast panoramas of fissures and volcanic cones.
The Cascadia Subduction Zone has produced calendar-worthy vistas, too—think Olympic Mountains. But the actual boundary where the seafloor and continent collide isn’t much to look at, according to Oregon State University researcher Chris Goldfinger, one of the few people who has been there. “All you see is mud and a few fish going by,” said Oregon State researcher Chris Goldfinger.
In 1990 Goldfinger dove to the subduction zone aboard the submersible Alvin. The descent took nearly three hours with the lights turned off to preserve the vessel’s batteries. When Alvin settled onto the abyssal plain two miles down, the pilot switched on the high beams. “I expected it to be smooth and flat,” Goldfinger recalled. “But it’s very lumpy and full of pits.” Ghostly white octopuses hunkered in the pits, which were rimmed with the bones of their victims. Albino sharks wove in and out of view.
Goldfinger was just a student, thrilled to visit the submarine world that was the focus of his doctoral research. Crammed into Alvin’s 82-inch spherical cockpit, the crew was the first to visit the subduction zone that was causing so much consternation. The pilot asked the young scientist what to expect.
“I don’t know,” Goldfinger replied. “But if you keep going east, we’ll hit North America sooner or later.”
Many subduction zones are marked by a deep trench. Cascadia’s is filled to the brim with sediment from the Columbia River. There’s actually a knoll on the seafloor where mud scraped off the nose of the descending plate piles up. Gliding eastward, Alvin bumped into that bank and a cloud of sediment enveloped the craft. The pilot ascended, then motored forward. Another oomph into the mud. “This must be it,” Goldfinger told the pilot. “The plate boundary.”
He filmed his close encounter with Cascadia and offered the footage to documentary makers. No one wanted it. “It’s just so boring,” Goldfinger said. Even when Alvin finally navigated up and out of the mud foothills and encountered a cliff at the edge of the continental shelf, it didn’t make for exciting images. “It’s like crawling up to the base of El Capitan in the middle of the night with a flashlight and saying, ‘I wonder what that is?’ You can’t get far enough back to really see what you’re seeing.”
Goldfinger’s underwater adventures since then have been more illuminating. Picking up where Atwater’s landlocked history of Cascadia quakes left off, his exploration of marine geology has extended the record of the
region’s tumultuous past back almost ten thousand years ago. The result is the longest earthquake record for any subduction zone.
Cascadia may look dull through a submarine porthole, but Goldfinger discovered it has probably unleashed quakes even more powerful than magnitude 9. In ocean sediments he found hints that quakes may come in clusters. And he’s also unearthed evidence that some parts of the subduction zone snap much more frequently than Atwater found—every 250 years or so. If Goldfinger is right, the odds are higher than one in three that a great quake will hit within the next fifty years.
One of the few things Goldfinger and Atwater have in common is that they both work with mud. Professionally, the two men are often each other’s biggest critics, and their approaches could hardly be more different. Atwater drives a beater and paddles a patched canoe. Goldfinger practices a Lamborghini brand of science. A single Alvin dive costs $70,000. Those are rare, but he routinely mounts months-long research cruises with price tags of $1.5 million or more. That’s the kind of muscle it takes to wrench secrets from the seafloor.
It was almost by accident that Goldfinger discovered the ocean bottom can record earthquakes with as much—or more—sensitivity than Atwater’s marshes. The mechanism was so simple he found it laughable at first: big earthquakes can shake loose landslides underwater just as on land. “I thought, no way would that work,” he recalled. In fact, he mounted his first oceangoing expedition in 1999 to prove that it wouldn’t.
The seafloor is such a messy place that Goldfinger assumed it would be impossible to sort out earthquake-triggered landslides from slides and slumps caused by gravity or the churning of storms. He didn’t doubt the power of earthquakes to shake things up underwater. That was established in November 1929, when a magnitude 7.2 earthquake struck near Newfoundland’s Grand Banks. Damage from the quake itself was minor. But over the next several hours, a dozen of the trans-Atlantic cables that tethered North America to Europe blinked out one by one. When repair crews fished them up from the seafloor, they found the wire bundles gnawed to bits, as if by giant crabs.
Scientists studied the cable company’s records and the underwater topography and concluded the culprit was a turbidity current, a supercharged landslide where sediment and water form a slurry that barrels down the continental slope with more force than an avalanche. At speeds up to 40 mph, the churning mass traveled 450 miles—more than the distance from New York to North Carolina—and tore through every cable in its path.
It took almost seven decades for researchers to make the connection between Cascadia and what happened off Newfoundland, but science wasn’t standing still all that time. In the 1960s a small group at Oregon State University (OSU) started pulling up cores from the seafloor, not realizing the haul included the first hard evidence of giant quakes off the Pacific Northwest.
“People ask, ‘How could you not have understood?’ ” said Gary Griggs, who was part of the OSU team. “You’ve got to transport yourself back to 1965. There was no plate tectonics. There was no Cascadia Subduction Zone. We were just flying by the seat of our pants.”
Griggs was twenty-one that year when he landed a graduate slot at OSU. A surfer from Southern California, Griggs figured the fledgling field of marine geology would let him earn a living while hanging out near the beach. His arrival in Corvallis was perfectly timed to catch a wave of exploration fueled by the Cold War obsession with submarines. The Northwest seafloor was still largely unknown, and an OSU oceanographer got federal funding for some of the first expeditions. Griggs started out processing mud cores extracted from the ocean bottom during previous trips then collected his own during several cruises.
“It was exciting being out there, bringing up the cores and coming back to the lab to open them up,” recalled Griggs, now director of the UC Santa Cruz Institute of Marine Sciences. “It was like going through a history book that nobody had ever seen before.” Deep-sea coring was a young science. Pulling up a single thirty-foot tube of mud took more than half a day, and bad luck or poor technique could scramble the contents. In the lab Griggs used a circular saw to slice the cores into ten-foot lengths and a guitar string to split them down the middle. A pattern jumped out almost immediately. Layers of ordinary gray clay were interspersed with jumbled-looking olive green bands that Griggs recognized as the signature of underwater landslides. In some cores he counted as many as twenty-one of these green layers.
With multiple cores he could trace the paths the landslides followed across the seafloor and estimate their size, which was staggering. Some were taller than a thirty-story building as they hurtled down submarine canyons. Griggs calculated that an average turbidity current swept up enough sediment and sand to bury Seattle under a layer seven feet thick. The slurries took two days to roll down the channels and spill out on the seafloor a thousand miles from their source. “That’s an enormous event.”
The landslide layers, also called turbidites, seemed oddly uniform: they were about the same thickness and evenly spaced. Luckily for Griggs each of his cores also contained a natural clock that provided a time reference—and posed a puzzle.
About 7,700 years ago, a volcano called Mount Mazama in southern Oregon exploded, creating what is now Crater Lake National Park. Ash flew as far as Nebraska. In almost all of Griggs’s cores, those distinctive ash particles first showed up in the thirteenth green layer from the top, which meant the thirteenth landslide must have occurred roughly seven thousand years ago. Griggs couldn’t figure out why the layers were so consistent from place to place, but he realized he could use the time window to calculate a rough recurrence interval. Something seemed to be kicking off the massive slides every five hundred years or so. But what? Griggs could think of only two options: storms or earthquakes. The latter seemed almost too outrageous to mention. Griggs noted the possibility in his thesis, then moved on. The report sat on the shelf for nearly twenty years.
It was show-and-tell day on the RV Melville in 1999, and Goldfinger was laying out cores in the ship’s main lab. It’s a tradition on research cruises for the scientists to host a briefing for the crew members who keep the ship running but might not know much about the research itself. The afternoon’s presentation would also be the first chance for Goldfinger and fellow cruise leader Hans Nelson to see all their cores side by side.
The scientists had been at sea for several weeks, but they hadn’t had time to reflect on their findings. Seagoing research trips are mostly a blur of work orchestrated to wring the maximum amount of data from every moment of vessel time. “It’s like being on a factory ship,” Goldfinger explained. This factory’s mission was to extract and process mud cores.
It had been thirty years since Griggs did his work, but the coring process wasn’t much changed. Technicians slowly lowered a steel tube topped with a three-ton lead weight until it hung suspended about sixty feet above the seafloor. With the pull of a lever, the apparatus was allowed to free-fall. The lead weight propelled the tube into the mud like a giant lawn dart. The science team worked around the clock in twelve-hour shifts, pulling up two cores a day when nothing went wrong. As soon as a core was fished from the water, scientists and grad students manned their posts on an assembly line. First they ran the four-inch-wide core through a multisensor scanner to measure the density and magnetic properties of the mud. Next, they sliced the core down the middle and chopped it into four-foot sections. The final step was the most time-consuming. Millimeter by millimeter the scientists diagrammed, described, and measured every layer of sand, silt, pebbles, and clay.
A diagram of a core extracted from the seafloor off the Pacific Northwest shows a history of Cascadia earthquakes recorded in thick, sandy layers formed by quake-triggered landslides, or turbidites, and separated by thin bands of sediment. The thirteenth layer contains ash from the eruption that formed Crater Lake about 7,700 years ago. (image credits 4.1)
“You basically walk off the boat with a complete set of data,” Goldfinger said. Today he sails with a crew of forty. Bu
t the 1999 cruise was his first time in charge, and he had been able to lure only about a dozen students. Everyone was frazzled. “You’d work sixteen hours, then fall into bed with only a vague idea of what happened that day,” Goldfinger recalled.
Still, he couldn’t shake his amazement that the National Science Foundation had decided to gamble on a very junior professor. He had swung for the bleachers in his grant application, requesting thirty days of ship time and enough money to bring back fifty cores. The agency said yes. He and Nelson planned to sample sediment from the entire length of Cascadia, a much bigger area than anyone had tackled before. They were confident they could put the lie to claims being made by a geologist named John Adams.
Adams was one of the first researchers to reexamine road surveys and report that the Northwest coast seemed to be tilting ominously. A few years later, the Canadian government scientist had decided to take another look at Griggs’s cores in the light of modern plate tectonics. Why, he asked, would cores collected near the California-Oregon border show the same sequence of thirteen landslides as cores collected off the Washington coast? Storms couldn’t explain it because they don’t churn up such a big area. Adams concluded that giant earthquakes were the only forces capable of shaking lose simultaneous slides along such a long stretch of coast.
Goldfinger didn’t buy it. He and Nelson suspected Adams of cherry-picking cores with thirteen layers to support his argument. They were confident samples from a wider area would blow Adams out of the water.
But on show-and-tell day, as the two scientists surveyed the twenty-five cores they’d collected so far on the Melville, they started feeling queasy. Most of the cores looked like mirror images of each other, with the same number of layers. “We could see where the Mazama ash was, and we could count to thirteen,” Goldfinger recalled. “Our whole hypothesis was going down the tubes.”
Full-Rip 9.0: The Next Big Earthquake in the Pacific Northwest Page 7