Cascadia's Fault

by Jerry Thompson

  Sixty miles (100 km) north in McKinleyville, Professor Lori Dengler of Humboldt State University was getting her family ready for a day of hiking at Patrick’s Point State Park, on the beach a little farther up the coast. “Suddenly the ground started to jiggle a little bit,” she said, “and then it started to jiggle a lot more strongly.” With her very next breath the instincts of a geologist kicked in. “Fortunately by that time I had actually developed a habit of starting to count the duration of an earthquake. It’s a very good habit to get into with—one, two,” she recounted the beats. “And by the time we got up to about seventy-five, I knew that my plans for the day were completely shot. There was no way we were going on a picnic.”

  Back in Ferndale the videotape showed piles of splintered gingerbread trim, cornices, and tons of old brick that had crashed to the ground in heaps of rubble and bunting. Thirty-six homes in this Victorian tourist town were seriously damaged. A dozen others twisted off their foundations and collapsed. Forty businesses in a four-block stretch were damaged, putting 80 percent of the town’s economic engine on life support.

  Pipes broke, gas mains ruptured, and fires started. In nearby Petrolia, population one hundred, the bay door of the fire hall got jammed during the initial shock and was stuck in the closed position. It took several volunteer firefighters considerable time and effort to pry it open before their pumper could respond to the now out-of-control blazes. The post office, a century-old general store, and a gas station burned to the ground. Landslides and rockfalls blocked roads and railway tracks.

  As the main shock died away, Lori Dengler made her way outside to begin the next phase of her research. “I lay down on the driveway so that I could feel all the aftershocks,” she confessed, apparently unfazed by what this must have looked like to the neighbors. “Earthquakes are really quite delightful if you are in a completely safe place. And I have a very open driveway with no big trees around, so I just lay there for about ten minutes, sort of feeling the music of the spheres—quite literally.” She could tell from the duration that the jolt had been at least magnitude 6, possibly higher, and knew it was time to get to work.

  Finding the focal point and calculating the strength of the jolt would take several days but a preliminary investigation showed that a nearly horizontal fault began splitting apart six miles (9.5 km) north of Petrolia and seven miles (11 km) underground in a magnitude 7.1 rupture. Little more than twelve hours later, at forty-two minutes past midnight, another quake, magnitude 6.6, centered fifteen miles (24 km) offshore from Cape Mendocino and thirteen miles (21 km) below the surface struck the same general area, causing additional damage. Less than four hours after the second rip it happened again—another 6.6 deep-sea jolt off the Lost Coast. Three strong quakes within fifteen hours. Ferndale, Petrolia, Honeydew, Scotia, Rio Dell—all the small towns at the southern end of Humboldt County took a beating.

  The main shock was felt as far south as San Francisco, as far east as Reno and Carson City, Nevada, and across most of southern Oregon. In Sacramento, 202 miles (325 km) to the southeast, a curious thing happened. Lori Dengler did some checking and found that most people living at ground level felt almost nothing. The farther up they were in apartment blocks and condo towers, however, the stronger the motion tended to be.

  “If they were in the sixteenth, seventeenth, eighteenth floor of a high-rise building, not only did they feel it—they felt it so sharply that it drove them to run down the stairs and evacuate the building,” she said. “By the time you get above the twentieth floor of some of these buildings, more than half the folks evacuated.” The astonishing thing was that no physical damage had been done in Sacramento, yet the kind of ground motion generated by the undersea fault off Cape Mendocino was able to travel a long distance inland and cause certain tall buildings to resonate with the frequency of the shockwaves. Shades of Mexico City.

  But it was Dengler’s cautionary note that really caught my attention. “This was an earthquake that was a thousand times weaker than what we’re talking about in terms of the amount of energy in a Cascadia earthquake,” she said. Put another way, whenever Cascadia does finally rip loose with a magnitude 9, the results will be off the scale. That’s simply the difference between magnitude 7 and magnitude 9.

  Several things did, however, make the Mendocino temblor different and indeed significant. It generated a small, three-foot (1 m) tsunami that hit nearby Crescent City, scene of so much destruction and a dozen deaths in 1964. And it lifted up a fourteen-mile (23 km) section of land along the beach at Cape Mendocino. “It turned out that the North American plate that we’re sitting on was shoved up and over the Gorda plate, which is subducting beneath us,” Dengler explained. In other words, this earthquake was apparently generated by tectonic movement along Cascadia’s fault. It was not just another small slip along one of the vertical cracks near the surface.

  With a total of ninety-eight people injured and a damage estimate of only $66 million, it may have looked to the outside world like a relatively insignificant lurch compared to the Loma Prieta event on the San Andreas in 1989. There were no dramatic helicopter shots of cars falling through a trap door on the Bay Bridge, no double-decker freeways crushing dozens of cars. But for those living in Humboldt County and those focused on Cascadia, this was a turning point in terms of putting the aseismic subduction argument to rest.

  Gary Carver’s memory of that deceptively sunny day was just as vivid. The shockwaves hit while he was driving to his office on the Humboldt State campus in Arcata. After the ground stopped moving he quickly rang home to make sure everything was okay there, then rounded up a bunch of graduate students and headed straight for Petrolia. By about noon the USGS had told them where to find the epicenter. They began scouring the countryside looking at landslides and other physical damage—without ever finding a surface fault.

  Because the ocean was high and rough that day Carver and company failed at first to see the quake’s most geologically significant wreckage. Several days after the jolt, however, the HSU crew stopped for lunch in a Petrolia café, where they overheard local residents in the next booth talking about how much the shoreline had changed. Offshore rocks normally submerged were now high and dry and the place “smelled like fish stew.”

  So Carver and his colleagues headed back toward the beach and sure enough a big swath of California landscape had been hoisted up sharply into new marine terraces. For the next several days the HSU team hiked the shoreline, documented the vertical displacement, and watched while acres and acres of shellfish clinging to rocks that once lay beneath the sea died and rotted.

  The seismic data showed an almost flat focal plane and the nearly horizontal motion of a thrust fault, agreeing with what Carver could plainly see on the beach—the upper plate had popped loose from the subducting ocean floor and ridden up over it. “It produced coseismic uplift,” he noted, “just like the ’64 one did. Except it was all in miniature.” In essence the continental plate along California’s western shore got massively and permanently deformed during the rupture, the same thing George Plafker had seen in both Alaska and Chile.

  “When you analyze exactly where that hypocenter or focus of where the earthquake was, it really coincides very closely with where we think that subduction zone interface is,” said Dengler, equivocating only slightly. “I mean there’s still some debate amongst the scientists as to whether it was really the main subduction zone or a subsidiary fault. And there’s also some debate as to really where the end of the subduction zone is and how complicated things get down there in that Triple Junction region. But I think we all agree that it was a thrust fault and it was clearly related to the subduction zone. And so it became really the first major earthquake to occur on the subduction zone or a very closely related fault.”

  “The Petrolia earthquake was a subduction zone earthquake because it broke on the subduction zone,” said Carver, equivocating not at all. “The boundary between the Gorda plate and the North America plate is a very l
ow-angle thrust fault. And it slipped and caused this uplift and subsidence of the coast and generated this earthquake,” he said. “It was a subduction zone earthquake, as far as I’m concerned.”

  The point of contention seems to be that the Gorda plate, which has broken off the southern end of the Juan de Fuca plate, seems to move and behave separately from the larger slab of the ocean floor and therefore might not be considered a part of the overall subduction zone. And to some extent Carver agreed with this view because evidence of several events found down in the Eel River valley showed radiocarbon dates for Gorda plate ruptures that were completely separate from the Juan de Fuca quakes.

  In other words, temblors like the 1992 Petrolia one seem to have happened several times before, with the Gorda plate breaking loose from the overriding continental plate on its own timetable. “It looks like there are some little end pieces that have a life of their own,” joked Carver, which to me sounded like yet more evidence for the decades of terror scenario.

  The interesting thing was that in Petrolia in 1992 there were three separate ruptures, and to me it looked as if the subduction zone had started coming apart underneath the oil town, working its way outward beyond Cape Mendocino under the ocean floor and toward the main subduction zone. “Once the fault started to unzip, why did it stop?” I asked. “Why didn’t it go all the way to Vancouver Island?”

  “Yep,” said Carver, “that’s a very good question. And I’ve been puzzling on that since ’92.”

  Lori Dengler agreed. “The 1992 earthquake ruptured the southernmost little corner of Cascadia—maybe fifteen miles [24 km] in length. So why did 1992 stop? It’s been a long time since the last event. There’s been a lot of strain put into the system, so why didn’t we have, on April 25th, 1992, a much larger earthquake? We don’t have a simple answer to that. Clearly it stopped because it didn’t have enough energy to make it through the bump or the asperity or the sticky spot that would have allowed it to go further. Does it mean that we’re closer to a larger rupture? Does it mean that we could have another little piece go? Well obviously we’re closer. Every day we get a little bit closer.”

  Proving that Cascadia’s fault had started coming apart was only part of the significance of the triple shock that morning in Petrolia. According to Lori Dengler the new data generated by that event caught the eyes of other scientists as well. “What was really important about that earthquake is that it brought two new communities onto the Cascadia bandwagon,” she said. “First it brought NOAA. Prior to 1992, the tsunami community really was not engaged in Cascadia.”

  The National Oceanic and Atmospheric Administration, or NOAA, got interested in Cascadia partly because of the small wave the Petrolia quake shot into the harbor in nearby Crescent City, California. Even though it didn’t cause much damage, it did revive painful memories of what had happened in 1964 and it was evidence that a tsunami could be generated by a tectonic source much, much closer to home. Prior to Petrolia the investigation of tsunami damage from Cascadia’s fault had been carried out by geologists and seismologists.

  Oceanographers and their math whiz colleagues who were attempting to create numerical models of tsunami wave behavior had been concentrating on waves crossing the Pacific from distant sources like Alaska, Japan, or Chile. Their goal was to predict what a given wave would do and devise a better warning system to alert the West Coast. Suddenly it seemed possible that large and damaging waves could be generated within twenty-five miles (40 km) of the beach all along the coastline of the Pacific Northwest.

  Seeing what happened off Cape Mendocino, “NOAA became very excited,” Dengler recalled, “particularly Eddie Bernard at the Pacific Marine Environmental Laboratory in Seattle. This was an event that really allowed him and his modelers to get their teeth into Cascadia and to actually model that tsunami.” Shortly after Petrolia, Eddie Bernard and his wave research team from PMEL were working with state geologists in California to put together a profile of what a larger Cascadia tsunami would look like. Combining a seismic shock and a killer wave in the same scenario had never been done before.

  The Petrolia story rang bells in the state capital as well, according to Dengler. “It brought the emergency management community into the picture. Prior to our event, I would say most emergency managers in the state of California weren’t really convinced that Cascadia was a problem. They had a really rapid conversion,” she declared. “And so we saw an incredible surge in momentum with NOAA and the emergency management community, which really culminated in the planning scenario for an 8.4 earthquake on the Cascadia Subduction Zone.”

  Dengler paused, looked up, anticipated my next question, then answered it all in the same breath. “Some people say, ‘Well, why an 8.4? Why does it stop at the California border?’Well, this was funded by the state of California. And so, it’s a great document, but it certainly has its limitations.” I took this to mean that the mandate for emergency planning by the governor’s Office of Emergency Services ends at the California state line. The larger scenario for a magnitude 9 catastrophe along the entire Cascadia margin would have to wait until other state, federal, and provincial governments were sufficiently motivated to get involved. For these other jurisdictions to the north, apparently the tipping point had not been reached yet. The good news was that awareness of Cascadia, along with a new sense of urgency, had now spilled across the boundaries from geology to the liquid sciences as well.

  Before Sumatra, very few people had seen a tsunami in action. Until those chilling home videos from Thailand and other fatally ruined vacation resorts were broadcast round the world, hardly anybody in the general public knew what a tsunami could do. Even the experts, oceanographers like Eddie Bernard at PMEL and top-ranked wave modelers like Vasily Titov, had only a theoretical appreciation of the beast they were dealing with. They understood the hydrodynamics, they could do the math and had seen photographs of damage done by waves in the distant past, but until Sumatra neither had seen the real thing in real time.

  Before Sumatra, the most recent and memorable tsunami had been the one triggered by the 1964 Alaska earthquake. “I was born in 1962,” Titov volunteered, “so I was two years old. I didn’t know anything about the tsunami personally. I knew everything from the [scientific] publications and from my studies.” Only a few people in affected communities worldwide had ever seen a killer wave and the tsunami scientist who had experienced one was an even greater rarity. Titov set out to solve this problem by capturing a wave in a computer.

  He wanted to master the mathematics—and the art—of digital water. If he could learn enough about fluid dynamics to reproduce the behavior of a wave with a numerical model in a computer, he and his colleagues might be able to improve the world’s tsunami warning systems and save lives. He recalls how hard it was before Sumatra to get people interested in or even concerned about this rare monster from the deep.

  “There was very little awareness in the larger society about the danger of tsunamis,” Titov said. “It was difficult to convey this message to society because the first question people would ask is, ‘When was the last big tsunami?’ And you say, ‘1964.’ It just doesn’t sound that convincing.” To many the threat seemed as farfetched as getting hit by an asteroid. The work remained an arcane specialty practiced by an elite group of gifted mathematicians who could have held their conventions in a phone booth.

  The study of wave mechanics had begun with work on hurricanes and typhoons about twenty-five years earlier. Hurricane science had decent funding because people saw the destructive power of these storms and their waves several times every year. Most of the world’s impression of tsunamis was based on scraps of grainy film footage shot decades ago in Hawaii or Japan or on badly faked waves in B-grade Hollywood disaster flicks.

  The émigré math whiz Vasily Titov, however, was destined to change all that and NOAA’s Eddie Bernard helped make it happen. Titov was one of the new wave of modelers Bernard assigned to the Cascadia problem not long after the Pe
trolia earthquake. “His models are—the way they convey so much information in such a short amount of time—can only be called art,” enthused Bernard, “because in science that’s not easy to do.”

  They first met in 1989 at an international tsunami conference held at the Novosibirsk Institute of Electrical Engineering in Russia—at the geometric center of Eurasia, the world’s largest continent. “I remember the banner,” said Bernard, picturing the slogan that adorned the meeting hall: “‘We are the furthest from any coast in the world, so this is the safest place from tsunamis in the world.’ And I think that’s true.” He laughed, enjoying the irony.

  Titov wanted the chance to work with state-of-the-art equipment to develop software that could anticipate the behavior of big waves. “Realizing how dangerous this phenomenon is, we definitely were working on the science of describing the tsunami with the ultimate goal of actually forecasting it,” Titov told me. Folding geology, oceanography, and hydrodynamics together in a package that could mine data from several sources at once and then create animated waves that accurately mimic nature in real time was a tall order on a shoestring budget.

  So he eventually moved to Seattle, where he joined Eddie Bernard’s research team at NOAA’s Pacific Marine Environmental Laboratory. Money and technology aside, the odds of getting something like that to work seemed as steep and improbable as forecasting earthquakes or asteroids. Even the best supercomputers back then were struggling to imitate the flow of water. Adding the complexities of gravity and friction across rough surfaces along the bottom, undersea mountain ranges that could steer a moving wave in a new direction, and the infinitely convoluted bathymetry of every harbor and beach—all of which would affect the movement of a tsunami—was a daunting prospect even for someone who loved math. Titov packed his kit and moved from the safest, most tsunami-free zone in the world to one of the most dangerous.