Cascadia's Fault

by Jerry Thompson

  NOAA, the USGS, FEMA, and the five Pacific states that funded the project realized early on that if and when the network were finally built to its full length, it would still be too small. With buoys this far apart, some smaller but nonetheless destructive tsunamis could slip through the gaps undetected. This floating line of defense between the “tsunamigenic zones” and the vulnerable coastal communities of the Pacific would be permeable.

  The engineers and scientists at PMEL went to work on a new generation of satellite communications technology that would work in both directions and on demand. If a moderate-size earthquake ruptured somewhere under the sea and the seismometers picked up the signal, the on-duty crew at the warning centers ought to be able to send out a signal to wake up the tsunameter buoys and get a reading instantly on all the waves coming across the system—even those slightly below the 1.2-inch (3 cm) threshold.

  This way they could spot smaller tsunamis immediately and issue—or cancel—warnings based on real data rather than “an absence of triggered data.” So the PMEL engineering team continued to plug gaps in the existing buoy system. At the same time, Vasily Titov and his colleagues were fine-tuning their tsunami model software so that it could use the incoming data stream to forecast what the waves would do when they finally made landfall.

  Eddie Bernard knew that roughly 900,000 people would be at risk from a fifty-foot (15 m) tsunami, which is what the computer models said might happen if Cascadia ripped apart. Inundation maps were being drawn for California, but work on the Oregon and Washington coastlines had barely begun.

  The weakness in the system continued to be the strategic location of the detection equipment. Because the deep-ocean buoys were placed far enough out in the Pacific to give western North America and Hawaii plenty of warning for a tsunami from Japan, the sophisticated new technology is too far away to be of use in a local rupture of Cascadia’s fault. The buoys are anchored out beyond the subduction zone, so Cascadia’s waves would hit the beaches of the Pacific Northwest at almost exactly the same moment that the pressure detectors picked up the signal at mid-ocean and sounded the alarm.

  Bottom line: if you’re on a beach and the ground starts shaking—and especially if that shaking lasts more than one minute—it’s probably a subduction earthquake and there probably will be a tsunami. The shaking is all the warning you’re going to get. Head for higher ground immediately and don’t wait for any official notification.

  Defining Cascadia’s zone gave scientists a more accurate sense of what they were dealing with. Building the prototypes for a deep-ocean tsunami alarm system in the Pacific gave emergency responders a way to make better decisions about whether to evacuate coastal communities when a distant fault ripped and heaved the ocean floor. But the potential for megathrust quakes closer to home remained a subject of debate, and the implications of giant waves generated not far off the West Coast had still not sunk in. An air of unreality and deniability hung over the whole business. It would stay that way until somebody could pin a specific date and magnitude on Cascadia’s last great rupture.

  CHAPTER 16

  Cracks, Missing Rings, and Native Voices: Closing In on a Killer Quake

  Long before Chris Goldfinger sailed the Indian Ocean in search of mud cores from the Sumatra 2004 earthquake, he dropped a fish off the Oregon coast and found Elvis. He and Bruce Applegate, both graduate students at OSU in the late summer of 1989, went to sea in a research ship called Wecoma using side-scan sonar to take state-of-the-art pictures of the ocean floor. The ship towed a chirping metal “fish” at the end of a long cable thousands of feet beneath the sea surface, pinging sound waves off the bottom to create a digital map that looked as realistic as aerial photos showing the terrain of the ocean floor in stunning detail.

  Gliding across the wide, flat abyssal plain, the fish kept chirping, sound waves echoed back, and a strange new picture emerged from the deep. In Living with Earthquakes in the Pacific Northwest, Bob Yeats described Goldfinger’s discovery of a fault that had cracked the floor of the sea channel and buckled the sediments into a low hill. The onscreen sonar image looked remarkably like a man with a guitar, so inevitably it became known as Elvis. Later, of course, the fractured sea lump was formally named the Wecoma fault after the university’s research ship.

  Over the next weeks and months the OSU team discovered nine more strike-slip fractures off the Washington and Oregon coast: cracks that penetrated both the Juan de Fuca plate and the overriding continental plate. The ocean floor, at the point where it dives beneath the continent, was buckled, crushed, and deformed into cracks and folds very much like the mangled terrain Gary Carver and his colleagues had found onshore around Humboldt Bay, just down the coast in California.

  This was literally and figuratively the cutting edge—the point of impact between two tectonic plates. For Chris Goldfinger the bottom line in this wealth of data was that the newfound fractures and deformations in the crust might be telling us something about the width of the locked zone and also about the kinds of rough spots—the asperities—that the down-going oceanic plate could get stuck on, preventing the entire subduction zone from rupturing all at once.

  Goldfinger and another colleague, Robert McCaffrey, published their findings in Science on February 10, 1995, concluding that a series of “smaller” earthquakes—perhaps magnitude 8s along the subduction zone or even magnitude 7s in these newfound cracks in the upper plate—could account for pretty much all the Cascadia geologic evidence to date. Since nobody really knew how big a quake had to be to drown the tide marshes that Brian Atwater had found, since nobody really knew how big a shockwave had to be to trigger the offshore landslides that John Adams had written about, it was entirely possible that smaller ruptures could have done all the damage discovered on this coast.

  With a fractured and buckled outer edge, the North America plate might be incapable of magnitude 9s simply because it couldn’t build up and store enough strain for a long enough period to generate a full-zone rupture. That was the “good news.” The decades of terror scenario seemed to Goldfinger and McCaffrey more likely than a magnitude 9 apocalypse. They suggested that the seismic hazard and public safety implications of Cascadia’s fault did not look quite as daunting as they had before.

  In the fall of 1995, however, an international team of mud, marsh, and sand diggers thought enough evidence had accumulated to suggest quite the opposite—that a magnitude 9, full-length rupture had occurred along Cascadia’s fault. And they were willing to speculate that it happened roughly three hundred years ago. A dozen scientists from federal, state, provincial, and university research labs on both sides of the Canada–U.S. border got together and jointly published a summary of all their separate bits and pieces of evidence for Cascadia’s most recent quake.

  From John Clague and Peter Bobrowsky’s samples of dead plants from sunken marshes that had been quickly covered by sheets of sand left behind by tsunami waves sweeping across the western beaches of Vancouver Island near Tofino and Ucluelet, to Gary Carver’s similar evidence of drowned trees in northern California, the picture looked remarkably consistent all the way down the coast. What Brian Atwater had found in estuaries along the Washington shore, Alan Nelson and his USGS colleagues had found in Oregon. The dozen scientists spent considerable effort—including eighty-five new radiocarbon-dated samples—to obtain the most accurate timeline possible. They found that all the ghost forests and marsh plants had been killed at roughly the same time as the land dropped down and was covered by tsunami sand—roughly three centuries ago.

  Given the long distance between Tofino, British Columbia, and Humboldt County, California, the dozen “marsh jerks” (as they jokingly called themselves after the jerky spikes in a sawtooth graph denoting the quake-induced sinking of land) said it was all Cascadia’s fault. The plate boundary was the only fault common to all the far-flung sites. They reported their findings in a paper published in Nature in November 1995. But because there was still reasonable doubt about
the exact dates, they still had to equivocate about whether all these events had occurred at exactly the same moment. The paper concluded that a single magnitude 9 earthquake, “or a series of lesser earthquakes,” had ruptured most of the length of the Cascadia Subduction Zone “between the late 1600s and early 1800s, and probably in the early 1700s.”

  Any one of these studies viewed in isolation might not have been enough to convince the most stubborn skeptics. Taken as a whole, however, and seeing that they all said basically the same thing, this united front of twelve top-level scientists looked like a critical mass. Having to equivocate a bit by including the phrase “or a series of lesser earthquakes” no doubt rankled those who wanted to make the clearest, least ambiguous statement possible. But counting the slow decay of carbon atoms to find out when something happened hundreds of years ago was just too imprecise. And there were many examples elsewhere of several earthquakes occurring in series, several years apart. So this remained an unsolved mystery and a real debate.

  What they still needed was a more precise date and some way to show, convincingly, how big that mega-shockwave had been. If they could say that an earthquake happened in a specific year, or better yet on a specific day, the whole thing would become more real, more believable not only to skeptics in the science community but to elected officials, emergency planners, and the people who live within striking distance of Cascadia’s fault.

  In all the mucking about along the coast, scientists had become increasingly aware that people had in fact been living on the edge of Cascadia’s rupture zone when it tore itself apart the last time. Beneath a thick layer of tidal mud in the Nehalem and Salmon River estuaries of northern Oregon, Rick Minor of Heritage Research Associates in Eugene and Wendy Grant of the USGS found fire pits full of charcoal and woven cedar mats. Near Willapa Bay in Washington State, fishing weirs and cobblestones modified by fire were found buried in submerged shorelines. Trying to figure out when the campfires had been doused or when coastal villages and their Aboriginal residents had been shaken from their sleep or perhaps drowned by killer waves turned out to be exceedingly difficult.

  Anthropologists had been collecting flood and disaster stories from tribal elders since the 1850s but it was only in the 1980s that they began to compare notes with geologists, who found the oral histories intriguing and frustrating at the same time. Judge James Swan, who had lived among the Makah people of Neah Bay, Washington, for a time in the late 1860s published the family history of Billy Balch, a Makah leader, who told him about a catastrophic flood that had turned Cape Flattery into an island. The Balch story began “a long time ago” although “not at a very remote period,” when the water of the Pacific flowed like a swollen river through the swamp and prairie between his village and Neah Bay.

  This, of course, was the same tide marsh where Brian Atwater found his first evidence of land subsidence caused by large seismic ruptures. Balch’s story does not mention the ground shaking, so perhaps the flooding he talks about was caused by a distant tsunami from across the Pacific. In any case, Balch told Swan that after the initial flooding, the ocean began to recede and left Neah Bay dry for four days. It then rose again “without any swell or waves” and submerged the whole of Cape Flattery.

  “As the water rose, those who had canoes put their effects into them and floated off with the current which set strong to the north,” Swan paraphrased Balch. “Some drifted one way and some another; and when the waters again resumed their accustomed level, a portion of the tribe found themselves beyond Nootka [on Vancouver Island] where their descendants now reside . . . Many canoes came down in the trees and were destroyed, and numerous lives were lost.”

  “Could this be an account of a great tsunami?” asked Tom Heaton of the USGS in a paper he wrote with Parke Snavely in October 1985. “Great tsunamis may have periods of tens of minutes to hours [between waves] but 4 days is without precedent. If the event is real, then it is apparent that the effects must have been substantial. However, it seems incredulous that any tsunami could have overtopped the entire Cape Flattery region since the highest elevation of the cape exceeds 400m.”

  Careful not to dismiss the story as myth, Heaton and Snavely tried to imagine what sort of geological event might explain it. The tsunami from the 1964 Alaska quake, which did so much damage to Port Alberni and Crescent City, delivered to Neah Bay a wave that was only 4.3 feet (1.3 m) above the tide, so a tsunami that could completely submerge the entire cape seemed a remote possibility. Then they wondered—what if the land behind Neah Bay had been hoisted up during a seismic event, draining sea water from the bay and leaving the land temporarily dry? Heaton cautiously wrote that “crustal deformation associated with a nearby subduction earthquake could explain the uplift. However, this and any other conjectures about the significance of this legend are purely speculative.”

  If there was a quake, why didn’t Balch mention the shaking when he told the story to Swan? Living as close as they did to the subduction zone, the Makah people must have experienced the worst imaginable ground motion. This doesn’t sound like the kind of detail a person could easily forget. Heaton and Snavely found enough inconsistencies to consider that the story might be “entirely fictional.” In the end, however, they decided it was “noteworthy that such a report exists for a region where there is growing concern that large subduction earthquakes and subsequent tsunamis may be a real possibility.”

  Farther down the coast, along the beaches of southern Oregon and northern California, Deborah Carver gathered stories told nearly a hundred years ago by Wiyot, Yurok, Tolowa, and Chetco people. From six different villages along at least two hundred miles (320 km) of coastline came similar cultural memories of a violent rupture that struck at night followed by many aftershocks and tsunami waves that rolled in for hours. Many people were killed as villages were inundated or swept away. By the light of day the survivors found places where the ground had liquefied and slumped. In a Yurok village near what is now Redwood National Park, south of Crescent City, Carver learned that one purpose for the ceremonial “jumping dance” was to repair or relevel the earth after an earthquake.

  As geologists continued their investigations, more and more stories like this came to light, including one that I heard in Pachena Bay on the west side of Vancouver Island. Seismologist Garry Rogers of the Pacific Geoscience Centre had told me about a reference he’d found in the provincial archives to an earthquake that rocked villages along the west coast of the island in the middle of a winter’s night long ago. It didn’t take long to find an elder who knew the details, a descendant of the people who once lived in Pachena Bay. The year was 1998 and I was filming another documentary for the CBC called Quake Hunters.

  Crossing the rocky spine of the island westbound, I thought about living on the edge of the known world two or three centuries ago when the earth tore itself apart. Wild, remote, and spectacular were words that came to mind. Not for the faint of heart. Since getting from the sheltered eastern shore at Nanaimo to the wild west coast presents a challenge even today, imagine what it was like back then.

  After a fast and curvy two-lane tour along the postcard edge of Cameron Lake and through the ancient cedar tunnel of Cathedral Grove, the slow grind up and down the switchbacks of Mount Arrowsmith and the deeply rutted trail beyond Port Alberni became more work than fun. Potholed and scattered with fallen rock, the unpaved washboard logging road that runs parallel to the fjord out to the fishing village of Bamfield and Pachena Bay was a challenge. But once the camera started rolling, we knew the ride had been worth the effort. We were closing in on a story about Cascadia’s last big quake.

  Around a crackling fire in a big stone hearth in a house just above the beach sat half a dozen young children, mesmerized by one of their elders. Outside, pounding surf and a fine mist of rain washed against the dark green shore. Sitting on a spattered windowsill, his back to the churning sea, Robert Dennis, chief councilor of the Huu-ay-aht Nation, waited until all the young eyes were trained
on his weather-beaten face and then began his story in a low, quiet voice that forced the children to sit forward and listen carefully.

  “The story that I’m about to tell happened a long, long time ago and was told to us by Chief Louis Nookmis,” said Dennis, grandson of Nookmis, the man who originally told this story to an anthropologist back in 1964. “Some of you are directly related to that man,” the chief added, nodding and calling the names of several children at his feet.

  Like the ancient cedars of the ghost forest in Willapa Bay, the Huu-ay-aht people felt a violent shudder in the earth one dark and stormy winter night many generations ago. “They had gone to bed,” said Chief Dennis, “just like any other normal night. Had gone to bed, gone to sleep. And they were awoken during the night when the earth began to shake.” He paused, letting the words sink in. “The earth shook. Startled the people. They woke up.” He shrugged very slightly. “Thinking everything was over, they just relaxed. A little while later, the water came in real fast. Swept their homes away. Swept everything away.” The children were transfixed. “The water just came too fast. They didn’t have time to go to their canoes. So all the people that were living there were drowned. They were all wiped out.”

  Chief Dennis let the thought sink in and then explained how it was that the story itself survived the death of an entire village. The modern fishing village of Bamfield is next door to the rebuilt Native community where Chief Dennis and his people now live. On slightly higher ground, however, in the trees behind the beach, was another settlement of people who watched what happened that horrible night long ago and who did live to tell the tale.

  “They who lived at ‘House-Up-Against-Hill’ the wave did not reach because they were on high ground . . . Because of that they came out alive. They did not drift out to sea with the others.” This according to the 1964 transcript of the story told by Chief Nookmis. For scientists eager to follow the story back in time, the original translation was a bit garbled, making it impossible to figure out when this disaster might have happened. Years later Robert Dennis would help organize a more thorough translation that narrowed the timeframe somewhat. The newer, more detailed version placed the event some time between 1640 and 1740, the same period in which geologists had pegged the last megathrust earthquake on Cascadia’s fault.