Full-Rip 9.0: The Next Big Earthquake in the Pacific Northwest

by Sandi Doughton

  In 1996 Bainbridge Island wanted a high-resolution topo map for land-use planning. The arrangements fell to the Kitsap Public Utility District. The PUD operated a handful of water systems and had a geographer on staff who knew his way around the latest digital mapping technology. Greg Berghoff estimated it would cost more than a quarter of a million dollars to make a map the old-fashioned way, with a team of surveyors.

  He decided to check out a local start-up that was pioneering a system called Light Detection and Ranging (lidar). The company offered to do the job for less than $70,000. Developed by NASA and originally used to measure things like cloud height and ice caps, lidar was new to the commercial world. The company’s plan was to fly grids across the island, shooting thousands of laser pulses at the ground every second and calculating elevations based on how quickly the pulses bounced back.

  Most of the signals bounce off treetops or roofs, but about one in ten strikes the ground. The magic is performed by computer programs that weed out ricochets and save the true surface data. In a lucky coincidence, the Bainbridge company flew its survey just after a snowstorm knocked off the last of the leaves, which made for even crisper images.

  The finished map was so sharp it was as if a giant hand had peeled back the island’s skin of vegetation. North-south scorings chiseled by glaciers popped out in stark relief. Watershed boundaries jumped off the page. Berghoff also noticed an odd east-west line cutting across a hill not far from Restoration Point. At first he figured it must be a road cut. But the lidar data were fine-grained enough to show that one side was higher than the other. “I saw that offset, and I thought, ‘Uh oh. If that thing isn’t a data error, it might be a fault,’ ” he recalled.

  This lidar image of the southeastern tip of Bainbridge Island strips away vegetation and reveals the Toe Jam Hill Fault, which is closely associated with the Seattle Fault. Also visible is the bench around Restoration Point, uplifted in the Seattle Fault quake about 1,100 years ago. (image credits 6.1)

  Berghoff pinned up the map in the PUD lobby where a local geology buff did a double take when he saw it. The next thing Berghoff knew, he was shaking hands with Craig Weaver, leader of the USGS earthquake team in Seattle.

  Weaver and his colleagues were so jazzed by their first look at the Bainbridge lidar that they hopped a ferry as soon as they could and bushwhacked through devil’s club and blackberry vines to the exact GPS coordinates on a promontory called Toe Jam Hill. What they found looked very much like the scarp on Mount Hood—which is to say that it didn’t look like much at all.

  “You can’t see the forest for the trees, and you can’t see the fault when you’re standing out in the brush,” Madin, of Oregon’s geology department, explained at his trench site in 2011. Fault scarps in Washington and Oregon are also hard to spot because they’re usually not very striking. On the San Andreas Fault, millions of years of motion and erosion have etched out a gash that’s visible from the space station. Northwest faults may be just as ancient, but glaciers wiped the slate clean less than twenty thousand years ago, then dumped a thick layer of sand and gravel on top.

  Geologists get excited by scarps for several reasons, Madin said. They reveal the presence of faults. They prove that at least one earthquake broke the surface. The length of the scarp and the amount of offset are rough yardsticks for quake size. But most important, scarps provide a portal into the geologic past.

  “That’s where digging comes in,” Madin said, watching as a backhoe struggled to dislodge a rock the size of a dishwasher. “The scarp tells us the fault is here and that it moved sometime in the recent past. What we want to nail down now is, ‘When were the earthquakes? How big, and how often?’ ”

  Those are the same questions the USGS team posed in 1996 about the newly-discovered fault on Toe Jam Hill.

  With the lidar image for leverage, the USGS pried loose enough funding to mount a major assault on the Northwest’s first urban scarp. “It was a big deal,” recalled Sherrod, who had recently landed a full-time job with the USGS. “There was a lot of interest and we wanted to document it well.” A team of two dozen scientists converged on Bainbridge Island, working on and off for two years. They excavated five trenches.

  For his project on Mount Hood, Madin had two weeks and one full-time helper. The backhoe was a loaner, normally used to dig up broken water mains, and seriously overmatched by some of the boulders it bit into. “I work for a state agency,” he said with a shrug. “We always have to bootleg it.”

  Despite the shoestring, the approach is the same. When one side of a fault pops up or down during an earthquake, the result is a kind of stair-step on the ground surface. Dirt, sticks, and leaves tumble off the high side of the step and accumulate at the base. Over time, the scarp is buried under a fresh layer of soil. Then a new quake splits the surface and the cycle starts over.

  When geologists dig into a fault, they scrutinize the walls of the trench for soil layers broken or offset by past quakes. They also look for buried pockets of sticks, leaves, and other organic material, hoping for something well-preserved enough for radiocarbon dating.

  What sounds straightforward in theory is usually a muddle in the field, Madin said. He scrambled into the ten-foot-deep trench where the dusty walls proved his point. A uniform tan, they revealed no obvious layers, no clear lines of displacement. Madin’s crew whisked the sides of the trench with push brooms to remove loose dirt. He fired up a smoke-belching leaf blower and followed behind for a final polish. After several hours of labor, a few faint features emerged.

  Embedded near the middle of the trench wall was a patch of volcanic rocks. Instead of lying flat, as they were formed, the rocks were tipped on edge. “That’s evidence of some kind of violent shoving in the past,” Madin said. He poked his finger into air pockets in the soil, which he explained also indicated a time when the ground churned. “It would be nice to have distinct layers where one side is three feet higher than the other, but that’s only in textbooks. In the real world, there’s a lot of scratching your head and trying to make out subtle differences.”

  Compared to sunken marshes and drowned forests, fault trenches are often blurrier windows into the past. In the Northwest, it’s unusual to find evidence of more than two or three quakes and harder to interpret radiocarbon results. “About the best you can get is a general sense of the rate of recurrence,” Madin said, settling down in the dust to probe the trench wall for bits of charcoal or wood.

  On Bainbridge Island, the USGS team discovered that the Toe Jam Hill Fault is an offshoot of the Seattle Fault, so closely linked that both fired off together 1,100 years ago. The trenches revealed that the fault has been rocked by three or four quakes in the past 2,500 years, some of which shoved the ground up by as much as fifteen feet. It was the first evidence that shallow quakes strike the Puget Sound area on a regular basis, and the first proof that some of those quakes actually rupture the ground.

  “It was a really important find,” Sherrod said. The Bainbridge trenches also marked the beginning of an era of breakneck discovery that is still going strong more than a decade later.

  After its first trenching success, the USGS was eager to collect wall-to-wall lidar coverage of the Seattle Fault and the Puget Sound basin. Some of the early funding came from NASA, which was equally eager to see its technology commercialized. By the late 1990s, the USGS’s fault-finding mission was launched.

  There were already hints that the Seattle Fault wasn’t the only player in the region. Seismic data collected by oil companies in the 1960s and newer magnetic and gravity surveys all suggested the shadowy outlines of several possible faults. Lidar confirmed past earthquake on one candidate after another.

  There’s a fault that slices through Tacoma. Another passes near the state capital and is sometimes called the Legislature’s Fault. The Saddle Mountain Fault skirts the eastern edge of the Olympics. The Devil’s Mountain Fault cuts a path from the tip of Vancouver Island to the foothills of the Cascades. A broad avenue
of subterranean cracks passes south of Everett. Two of the newest additions to the list lurk along the Canadian border, near Bellingham. Interspersed between the faults are basins like the one that underlies Seattle—many of which will experience an extra boost of shaking in future quakes.

  As the faults kept coming and trenching proved most of them dangerous, the USGS hosted briefings for communities across the region. “Each time it was a big deal,” Weaver recalled. “We would have these workshops, three hundred people would show up, and we would explain what we’d found on the newest fault. At some point the engineers got it.”

  A modern fault map shows that it’s hard to find a place where an earthquake-phobe could feel cozy. The few blank spots are mostly areas where geologists haven’t looked yet, like southwestern Washington. Sherrod has probably spent more time in the earthquake trenches than anyone else in the Northwest, and he has come to view the hazard regionally rather than fault-by-fault.

  Since the 2001 Nisqually earthquake, scientists have discovered nearly a dozen shallow faults that split the Puget Sound area. Not shown in this map are two recently discovered faults near Bellingham. (image credits 6.2)

  “Draw a box from the Canadian border to Olympia, and from the Olympic Mountains to the foot of the Cascades,” he said. “Anywhere in that box, we need to be thinking we could have a pretty big, shallow earthquake.” And no matter which specific fault breaks the next time, everybody in that box is going to feel it.

  Toe Jam Hill was Sherrod’s first scarp excavation, but it wasn’t his first experience rooting around in the ground. As a college student in his home state of Virginia, he worked on dozens of archaeological digs. The most memorable was at the site of a Civil War field hospital, where his team unearthed surgical instruments and pits filled with amputated arms and legs.

  But Sherrod’s professional career started out on a much finer scale. He earned a master’s degree in micropaleontology, the study of fossil diatoms, dinoflagellates, and other creatures invisible to the naked eye. When consulting on construction projects proved unsatisfying, Sherrod applied for the doctoral program at the University of Washington, but got turned down. By the time he decided to take another run at admission, the Seattle Fault studies were under way. Robert Bucknam and his UW collaborators needed someone to analyze marine fossils from uplifted beach terraces. Sherrod was in.

  He took to trenching like a beaver takes to dam building. Plain-spoken, with a shaved head and a wrestler’s build, Sherrod loves the logistical challenge of cobbling together projects and reconstructing history. He’s wired to think in three dimensions—four, if you factor in a geologic time scale so vast it leaves most people woozy. “I can’t explain it,” he said. “It’s one of those things that you’ve either got, or you don’t.” Sherrod’s other passions include riding his Harley and catching big fish.

  As the lidar surveys started turning up one fault after another, Sherrod and two of his USGS colleagues developed an assembly-line approach to identify and prioritize the most promising ones. Lidar remained the cornerstone, and Ralph Haugerud was its virtuoso. A former hard-rock geologist, Haugerud carved out a new niche when the market for geologic maps dried up along with the American mining industry. He wrote one of the earliest computer programs to sort through millions of lidar points and discard all but the true ground hits.

  As a boy Haugerud decorated his room with topo maps. As a man he spent decades traversing the West on foot, by horseback, and behind the wheel of pickup trucks, honing an eye for landforms that he applies to lidar images. Rivers, gullies, and landslides can all masquerade as fault scarps, and Haugerud is expert at identifying the impostors and zeroing in on the real thing.

  Not every scarp is worth paying attention to, either. “There are lots of faults out there, but if they stopped moving twenty million or fifty million years ago, who cares?” Haugerud asked. The bar is set by the Nuclear Regulatory Commission, which considers a fault active if it has ruptured within the last ten thousand years. By that criteria most scarps west of the Cascades are automatically suspect because of the recent passage of the glaciers.

  Haugerud hands off the likely candidates to Rick Blakely, a geophysicist out of the USGS Menlo Park office. Blakely’s forté is airborne magnetic surveys. Different types of rock have different magnetic properties, so the measurements can reveal places where layers broke and shifted in the past. The aeromag validates the existence of a fault and is often able to pick out segments that don’t show up on the lidar.

  But the technique is temperamental. The airplane that carries the instruments has to be covered in coils to snuff out its own magnetic field. It also needs to fly as low as possible. In remote areas pilots skim along two hundred feet above the ground. When the terrain allows, Blakely fine-tunes his data by getting even closer, strapping a magnetometer and antenna on his back and walking the length of a scarp.

  The newest aeromag technology is so good at finding faults that Weaver decided to test it near his home east of Seattle. Overlaying the magnetic data on a street map, he drove to the spot where the aeromag indicated the road crossed a fault. Sure enough, the pavement took a dip. “The trees are maybe six feet higher on one side than the other,” Weaver said. “I was stunned.”

  Trenching is the last step in the assembly line. Only when the lidar and aeromag agree that a fault is real and probably ruptured in the not-too-distant past will Sherrod call in the earth-moving equipment. In nearly fifty excavations, he’s fine-tuned the routine. He and his crew can be in and out in two weeks. Still, trenching is expensive, and he can’t afford to waste money on a dry hole.

  Even before Sherrod lines up excavators, he has to convince landowners to let him tear up their property. The owners of an environmental education center on Bainbridge Island welcomed the geologists and used the trench as a teaching tool. But not surprisingly, some folks aren’t delighted to find out there’s a fault running through their backyard. One couple okayed digging as long as Sherrod promised not to tell the press. Others turn their backs the instant they hear the word government. “I’ve had folks who would love to take a gun to us,” Sherrod said, chuckling.

  Along the Seattle Fault and its offshoots, Sherrod and his team excavated ten trenches, including several on Bainbridge Island and one near a submerged forest in Lake Sammamish, east of Seattle. What the geologists found was a mix of good and bad news. Best was the discovery that the 900 AD quake seems to be one of a kind—so far. Nothing else in the record wrenched the ground so hard or had such far-reaching effects. But the Seattle Fault is no geologic slacker: Sherrod and his USGS colleagues uncovered evidence of three to four significant quakes in the past 2,500 years.

  The numbers are uncertain enough that geologists still don’t agree on how frequently Seattle might get slammed by shallow earthquakes. A consensus conference in 2004 came to the unhelpful conclusion that the Seattle Fault could snap every five hundred years—or every five thousand. “That’s the range from ‘oh shit’ to ‘who cares,’ ” said Haugerud. “It’s frustrating that repeat time on the Seattle Fault remains so ambiguous.”

  That’s one of the reasons Sherrod prefers to pool the data from all the Puget Sound faults and trenches. Although individual faults may go several thousand years between quakes, collectively Western Washington seems to get hit with a large, shallow quake every thousand years or so. The really big quakes, like the one that hit in 900 AD, are much rarer, striking perhaps every 3,000 years or more.

  With the last Seattle Fault quake 1,100 years ago, the thousand year recurrence interval isn’t comforting. The Seattle Fault scenario that estimated more than 1,600 fatalities wasn’t based on the magnitude 7-plus quake that shoved up Restoration and Alki points. It was a more modest magnitude 6.7, the variety that averages suggest could be due any time in Western Washington.

  Geologists still don’t know what to make of the fact that almost every fault in the region seems to have popped off roughly in concert with the massive 900 AD upheaval o
n the Seattle Fault. The Tacoma Fault ripped around then. So did Toe Jam Hill, Olympia, Devil’s Mountain, and the Little River Fault, which skirts the northern edge of the Olympic Peninsula.

  Radiocarbon dates are so imprecise there’s no way to tell whether the quakes hit all at once, over a period of decades, or over more than a century. But the syncopation is unnerving. Add in a possible quake on the Cascadia Subduction Zone that occurred at roughly the same time, along with an eruption of Mount Rainier, and you’ve got what geologists call “Puget mayhem.” It’s something they talk about when they’re shooting the breeze in a bar. After the second round, some even speculate that the next subduction zone quake could set off a similar domino effect.

  “It sounds like a horrible period,” Haugerud said. “I think most of us are a little bit hesitant to say there’s a cause-and-effect relationship.” He paused, then raised his eyebrows. “But, my. That’s quite a coincidence.”

  The seismic domino effect could even extend beyond the Northwest. Chris Goldfinger’s seafloor cores show that some Cascadia megaquakes were so closely followed by quakes on the San Andreas Fault that he suspects triggering. Several volcanoes around the world have erupted so hot on the heels of major earthquakes that scientists don’t doubt cause and effect. In 1975 Kilauea began spewing lava half an hour after a magnitude 7.5 quake under the volcano’s south flank. Cordon Caulle in central Chile sent a column of ash soaring nearly five miles into the sky two days after the record-breaking 1960 subduction zone quake.

  The link between Rainier’s eruption about a thousand years ago and the 900 AD quake is less clear. Massive debris flows that roared off the volcano and ran all the way to Puget Sound suggest Rainier may have let loose before the earth shook, though by how much is hard to tell. David Hill, a USGS researcher based in Menlo Park, wonders if a spike in eruptive activity throughout the Cascades in the 1800s might have been set in motion by the 1700 Cascadia quake. He wouldn’t be surprised to see volcanic rumblings the next time the subduction zone rips. “The Earth does amazing things,” he said. “My own feeling is that there are lots of things going on that are related that we don’t quite appreciate.”