Cascadia's Fault Read online

  Now, with the big trees gone, the ground was nearly naked and rock formations were easier to see. Still, Carver and Stephens had to hike for miles and miles along steep switchbacks, mapping and following a web of fractures from the mountains all the way west and downhill to the intertidal and beach zones along the coast. They noticed a distinctive angularity, what they called a “rhombohedral fracture” pattern, which to a non-geologist’s untrained eye would look like nothing more than “tiny little cracks in the sand” of cutbanks sliced through the wilderness by road builders for the logging crews.

  This rhombohedral pattern, Carver explained, was how faults propagate through unconsolidated sand deposits. This was loosely packed sand left behind when this part of the coast was under water, a wedge of ocean sediment that had been shoved against the continent and now stood well above the high-tide line. “Instead of a nice, clean, one-plane fault in which two pieces of the earth’s crust move past each other, it becomes hundreds or thousands of little tiny faults all closely spaced together,” he elaborated. “I thought this was really neat.”

  Each of the fractures Carver and Stephens found cut through geologically young terrain, suggesting the cracks were relatively recent. That meant whatever tectonic force had caused the fractures might still be an ongoing threat. Carver was pretty sure the rhombohedral fractures had been caused by plate convergence and compression. While none of the individual cracks had a huge amount of movement, taken as a whole the offset was significant.

  “We realized that these little tiny fractures we were seeing in many places were parts of faults that had very large amounts of displacement on them,” said Carver. The displacement added up to several miles in total. “And again—you’re sittin’right there on the edge of the mapped subduction zone and you see those big folds in young sediment,” said Carver, “and you can’t help but think that that subduction zone is still active.” All of this within a few miles of the nuclear plant at Humboldt Bay.

  That’s when Carver decided to fly north to Alaska for a first-hand look at what had happened there. He needed to “see what a big earthquake looked like” in all its mangled glory so that he could better understand what he was seeing on the ground in California.

  CHAPTER 7

  Proving the Doubters Wrong: The Chile Connection

  Gary Carver spent an entire summer in Alaska looking at the aftermath of subduction. He flew the entire length of the ’64 rupture, every mile of broken shoreline. He also met George Plafker, who was more convinced than ever that the primary fault that had caused the beaches and bays to heave, buckle, and subside could not have been vertical. Plafker had recently returned from Chile and was eager to tell anyone who’d listen that the two biggest earthquakes in recorded history had caused exactly the same kinds of physical damage to the landscape.

  Not only that, but some very prominent senior scientists were apparently coming around to Plafker’s point of view. Frank Press, who had so famously disagreed about the angle of the fault, sat in the audience at the 1968 meeting of the American Geophysical Union in Washington, DC, and listened to Plafker’s presentation of his paper on Alaska. The main theme of that year’s convention was “The New Plate Tectonics,” and here was Plafker telling the science establishment that the Alaska quake had been caused by two huge slices of the earth’s crust converging almost horizontally, getting stuck together, and then snapping apart.

  Plafker told me that after the speech Press cornered him and unloaded. “He came up and he was real mad,” Plafker recalled. “He said, ‘You know, I’ve written a lot papers and I’ve seldom been proven wrong. But you did it to me this time!’ He told me I had caught him in the biggest mistake he made in his career,” said Plafker. “His views on the mechanism of this earthquake had changed and he was man enough to say so.”

  But not everyone was convinced. Clarence Allen, another senior scientist who’d heard Plafker’s talk, still needed convincing. And he threw down a challenge that Plafker simply could not resist. How, he asked, could this underthrusting of the ocean floor be happening only in Alaska? Did Plafker think that’s what happened in the Chile earthquake as well? Plafker said yes—even though he didn’t know for sure—and so Allen arranged the funding necessary to send him south.

  Plafker spent two months scouring the Chilean coastline by car along the mainland and by chartered boat in the islands of the southern archipelago, measuring areas of heaved-up and down-dropped land. “There again in Chile, in the southern part, vegetation grows right down to the shorelines,” he said. “You could see the effects of subsidence from the drowned and dead trees and brush.” He found a zone of “tectonic warping, including both uplift and subsidence,” that was 125 miles (200 km) wide and roughly 625 miles (1,000 km) long. It affected an area of at least 50,000 square miles (130,000 km2) in southern Chile.

  The two-day series of temblors in 1960 had included two main shocks, thirty-three hours apart, along with fifty-six large aftershocks. The sequence of ruptures and the tsunamis they triggered killed 2,000 people in Chile and 230 more in Japan, Hawaii, and the Philippine Islands.

  The main finding of Plafker’s paper, however, refuted the previous conclusion that Chile’s wreckage had been caused by a nearly vertical strike-slip fault (like the San Andreas) because it was based on “incorrect” data that were “clearly incompatible” with his newer evidence of tectonic movement. He wrote that the Chilean main shock “resulted from a complex rupture on a major thrust fault or zone of thrusting roughly 1,000 km long that dips at a moderate angle from the continental slope beneath the continental margin.” In other words, the fault was more horizontal than it was vertical, just like in Alaska.

  Plafker estimated that to cause such widespread upheaval and deformation of the landscape, there must have been at least 65 feet (20 m) and perhaps as much as 130 feet (40 m) of horizontal slip once the fault broke. This might seem “surprisingly large,” he wrote, but “not excessive” if compared to the horizontal thrust of roughly 65 feet he had seen in the 1964 Alaska earthquake. The bottom line appeared to be that the events had both been caused by the same process: two pieces of the earth’s crust crashing together.

  So his gambit in Chile had been a success; he was able to prove plate convergence. “It was very straightforward, once you know what you’re looking for,” he said, putting the apparent mistakes of the first scientists on the scene in Chile into some kind of context, “but you know, it’s just like anything else. If you’ve done it once before, it’s a cinch. And if you haven’t, you don’t know what to do.”

  Nevertheless, when Plafker met Gary Carver a few years later and they compared notes on Alaska, Chile, and northern California, the similarities were hard to miss. “It’s pretty clear to me,” said Plafker, “that the southern end of Cascadia is very much like the eastern end of the Aleutian Arc and the area where the ’64 earthquake occurred. We have the same type of continental margin.”

  When I asked Gary Carver why he thought it took so long for most geologists to come around to the view that Cascadia was a threat, he could remember clearly one paper—written by Masataka Ando of the U.S. Geological Survey and Emery Balazs of the National Geodetic Survey—that stood out. They believed the subduction zone had “foundered” and was no longer active. “They related the idea that the rise [the Juan de Fuca Ridge where the sea floor was spreading apart] was so close to the trench that the plate was too hot to go down,” Carver explained. “It couldn’t go down, so therefore, subduction had stalled.” This would presumably explain why there had been no large subduction earthquakes in all of recorded history.

  But Carver was now infected by Plafker’s enthusiasm. He was sure those cracks in the sandstone meant something significant. Back at work in California he and Tom Stephens continued their research on the fracture zones along the northern California coast. They mapped each individual rupture and gave it a name—the Big Lagoon, Trinidad, McKinleyville, Mad River, and Fickle Hill faults. “As far as I know,” said
Carver, this was “the first recognition of the existence of large, active thrust faults north of the Mendocino Triple Junction.” It was also the first onshore evidence of tectonic motion on the southern end of the Cascadia Subduction Zone.

  The discovery of unknown crustal cracks on logging roads in the hills behind Arcata made geology a hot topic for students and other scientists working in the area. One of those drawn to a series of talks that Gary Carver gave in 1974 was Tom Collins, a geologist working for the U.S. Forest Service, based in Eureka at the Six Rivers National Forest office. When Collins saw slides of the “rhombohedral fractures” and heard Carver speculate about the relation between the faults in the hills and the big subduction zone offshore, he wanted to find out more about it.

  Collins went exploring on his own. He knew about the Little Salmon fault, which had been partially mapped back in 1953 by a local geologist named Bud Ogle, and perhaps because it was the one closest to where he worked in Eureka, Collins decided to have a closer look. Across Highway 101 from the nuclear power plant, he wandered into a recently excavated sand quarry at the base of Humboldt Hill. There he discovered, completely by accident, more of those rhombohedral fractures that Gary Carver had talked about in his lectures.

  A day or two later Collins phoned Carver who agreed to join him at the sand pit for a quick recon. “We recognized the Little Salmon fault extended further north than Ogle had mapped,” said Carver. Here again “young material” had been torn, meaning the shockwaves that had caused those distinctive fractures had occurred not so long ago in geological time. Even more worrisome, it looked like the crack probably continued right underneath the highway and onto the 143-acre (58 ha) site where PG&E had built the Humboldt Bay reactor.

  So Collins wrote up his discovery and, as a concerned citizen, sent it to the Nuclear Regulatory Commission (the new name for the Atomic Energy Commission) in Washington. It was the first in a long and increasingly political chain of events that galvanized local antinuclear activists who had formed the Redwood Alliance to do battle with PG&E. It also, as an unintended consequence, accelerated the scientific research that would finally confirm the true nature of the Cascadia Subduction Zone.

  A magnitude 5.2 earthquake shook the town of Ferndale on June 7, 1975, causing repeat damage to a town that had barely survived the pounding of 1906. The shockwaves also hit Humboldt Bay to the north of Ferndale, and in the aftermath fresh cracks were discovered in the concrete pavement of the road leading into the nuclear reactor site. A team of engineers from the University of California at Berkeley was called out to study the “ground motions and structural response” at the power station. The concrete caisson, with walls four feet (1.3 m) thick and an outside diameter of 60 feet (18.3 m), dug 85 feet (26 m) into the ground, appeared to be okay. But PG&E decided to err on the side of caution and ordered a thorough examination just in case.

  The Berkeley report confirmed that there had been no significant damage to the reactor. The summary page, however, spoke volumes. “The regulatory requirements led to an adequate but not excessively conservative margin of safety based on the motions recorded in this event” (my emphasis). In other words, PG&E had followed all the rules and nothing bad had happened this time, but if there was any chance of larger earthquakes, then all bets were off.

  Roughly a year later, in July 1976, when the reactor was shut down for routine refueling, the seismic safety questions were red-flagged by the Nuclear Regulatory Commission. The NRC decided to keep the plant closed until the Little Salmon fault and the new system of fractures discovered by Gary Carver and Tom Stephens could be checked and the seismic hazard issues dealt with.

  PG&E hired several consulting firms to conduct field studies to find out whether any of the faults were still active. A sixteen-station array of seismographs was installed in the surrounding mountains and along the northern California coast to get a more detailed picture of all the tectonic motion. In addition, the NRC decided to send its own team of scientists into the field to follow up on the work done by Carver and Stephens.

  They created a timeline of earthquakes in the region. With backhoes they dug trenches across the Little Salmon and Mad River faults for close-up looks at where and how often the various layers of soil and rock below ground had been torn apart. Taking samples of woody debris, dead plants, and the remains of tiny sea creatures contained in the layers disrupted by quakes, they used radiocarbon dating to figure out when the ruptures had happened.

  In the fall of 1980 the geologists concluded that the Little Salmon fault was indeed active and that it probably ran underneath or very close beside the reactor. The bottom line according to Woodward-Clyde Consultants, hired by PG&E, was that the seismic issues could be dealt with but the job would be neither cheap nor easy.

  At this point two other factors may have entered the equation for Pacific Gas and Electric. On March 28, 1979, while the Woodward-Clyde team was still documenting the gritty details of the Mad River area and how it might affect the reactor at Humboldt Bay, things went alarmingly wrong at a nuclear power station called Three Mile Island in Pennsylvania. A relief valve got stuck open, allowing large amounts of radioactive coolant to be released into the atmosphere. The reactor core overheated and barely survived a partial meltdown.

  The accident, while not as catastrophic as it might have been, helped turn the tide of public opinion against nuclear power. By some masterstroke of luck or serendipity, a Hollywood movie called The China Syndrome had been released only twelve days before the Three Mile Island accident. The eerily prescient film became an instant box office hit and probably did much to seal the fate of nuclear power in the United States. After months of investigation and analysis, the Nuclear Regulatory Commission issued a new set of far more stringent safety rules that would apply to all reactors, including the one at Humboldt Bay.

  Add to this the legal, political, and financial implications of California’s own new seismic zoning law, the Alquist-Priolo Earthquake Fault Zoning Act, which was passed in the aftermath of the Sylmar temblor, and the job of retrofitting and upgrading the reactor at Humboldt Bay became too expensive to be economically feasible for PG&E. Four years later the utility applied for permission to decommission the reactor permanently.

  In the aftermath, an official report to the U.S. Geological Survey described the twenty-five-mile (40 km) Little Salmon fault as “part of a broad, compressional fold and thrust belt developed in the accretionary wedge above the Cascadia subduction zone.” An accretionary wedge is formed by the sediment and pieces of seafloor crust piled up in a trench where two tectonic plates collide. Think of the North American continent drifting west like a snowplow across the sea floor, scraping up muck and compressing it into rock.

  In most cases the wedge is found under water. From Vancouver Island all the way south to the Oregon–California border, this folded and buckled sedimentary wedge is piled up against the continental shelf dozens of miles offshore, where it’s difficult and expensive for scientists to study. Only in northern California was it piled up right in plain sight and on dry ground. The towns of Eureka and Arcata were built on top of it, which is why Gary Carver and others at Humboldt State University were able to draw such a revealing picture of what Cascadia’s fault was actually doing. They took advantage of a unique geological setting to make an important discovery.

  If the Little Salmon fault was active, then the Gorda plate—which had caused the cracks—had to be active as well, pushing its way underneath California while North America plowed west. The subduction along Cascadia’s fault had not “foundered,” and the plates had not stopped moving. At least that was the conclusion I drew from reading the science papers and from interviewing both Plafker and Carver.

  Taken as a whole, the Humboldt Bay power project had a significant but unintended consequence. Building a reactor on top of a crack in the crust—a crack directly related to the Cascadia Subduction Zone just offshore—generated the new science that provided the first physical evidence that t
he northern section of the California and Pacific Northwest coast faced the same kind of tectonic disaster as the ones that happened in Alaska and in Chile.

  If I’d been a journalist in California back in the 1970s, I like to think I would have turned this story into headline news. But the immediate impact of these discoveries confirming continental drift was almost nil. The story of Cascadia’s fault got lost in the controversy over nuclear power. Fortunately the scientists on the ground knew they were working on significant stuff and refused to quit.

  It was the heady, meaningful kind of research that made it an exciting time to be a geologist—especially in the Pacific Northwest. Frank Press may have changed his mind, but many others in the science community still refused to buy the new geology. Even when the top half of a mountain in southern Washington State exploded, only a handful of researchers recognized the distinct sound of Cascadia’s smoking gun.

  CHAPTER 8

  Mount St. Helens: Cascadia’s Smoking Gun?

  Even though geologists and volcanologists saw it coming, there was no way to prepare for the impact of watching a mountain explode at close range. Mount St. Helens—roughly ninety miles (145 km) south of Seattle and fifty miles (80 km) northeast of Portland—blew steam and dust for two months as a bulge of hot rock sprouted like a giant goiter on its north face. At the same time, the ground trembled and shook. People in downtown Portland turned the prelude into a spectator sport.

  Government officials issued repeated warnings to evacuate the hills and valleys around the volcano as the frequency of tremors began to increase. Almost everybody did leave, except for an eighty-three-year-old recluse named Harry Truman who had lived in the woods near the mountain for more than fifty years and decided to stay close to his cabin. The media fell in love with him, a tragic hero in the making. A thirty-year-old volcanologist named David Johnston was collecting data until the very last minute. His final words, “Vancouver! Vancouver! This is it!” were shouted into a walkie-talkie and received at the USGS volcano observatory in Vancouver, Washington, across the Columbia River from Portland, only moments before the eruption. Neither man was seen again.