In the late 1970s, Dragert helped pioneer the use of sensitive gravity meters to detect changes in elevation. The measurements were finicky, and the scientists had to run every transect at least eight times with multiple instruments. “Boy, it was labor intensive,” Dragert recalled. After all that work, it was often impossible to tell if the gravity fluctuations represented real land-level changes or just the movement of groundwater.
A breakthrough arrived in the 1980s with a method called laser ranging. The logistics weren’t any easier, but the technique allowed scientists to measure the distances between mountaintops with near-pinpoint accuracy. Helicopters deposited the scientists and their bulky gear on summits across central Vancouver Island. Sometimes teetering on the brink of a chasm, the researchers set up bread-box-size lasers and banks of parabolic reflectors. They fired off light pulses and timed how long it took the beams to bounce back and forth between peaks.
As the “met” man in charge of collecting meteorological data, Dragert had one of the cushier jobs. While his colleagues hunkered on the mountains, he shuttled back and forth in a helicopter, measuring air temperature, barometric pressure, and humidity. All those factors affect the speed of light and had to be carefully factored into the calculations. At least the weather was always good. The scientists couldn’t run their surveys in the rain.
Collecting enough data took an entire field season, but when the scientists finally laid out their results against conventional surveys from the 1930s, there was no doubt about it. The mountaintops were being squeezed together in a northeast direction.
The Geological Survey of Canada is tiny compared with the USGS. But Dragert and others at the agency’s West Coast outpost in Sidney, British Columbia, were well ahead of their American colleagues in accepting Cascadia’s menace. Tuzo Wilson, discoverer of the Juan de Fuca Plate, was Canadian. John Adams, who was the first to point out that coastal ranges in Washington and Oregon were tilting like those near other active subduction zones, works for the Canadian Survey.
Several Canadians collaborated with the team that discovered magnetic zebra stripes on the bottom of the Pacific, proving that fresh seafloor was oozing up from the depths. Enthusiasm for geology’s new paradigm spread quickly at the Canadian Survey, helped along by the organization’s small size, recalled Garry Rogers, longtime leader of the Sidney group.
Oceanographers rubbed shoulders with geologists, whose offices were next door to seismologists, geodesists, and computer modelers. Everyone gathered for coffee and bounced around ideas. Years before Brian Atwater started digging in marshes, Rogers was sounding the alarm in British Columbia about possible megaquakes. Atwater’s evidence was the smoking gun the Canadians had been waiting for. “Brian had a tough sell in the U.S., but we were totally convinced right away,” Rogers said.
Canada’s pioneering investment in GPS added to the growing evidence. By 1999, when Dragert noticed another strange blip in some of his measurements, the regional network had expanded to fourteen stations. “Lo and behold, I was looking at the Albert Head station and I thought, ‘Hmmm. This looks like what I saw in 1994,’ ” Dragert recalled. “I immediately checked all of the other sites.” Seven contiguous stations arrayed in a broad band from southern Vancouver Island through Washington’s Olympic Peninsula had sidled seaward a few millimeters. The stations outside that band hadn’t budged. “I said, ‘Either somebody is trying to fool me by pushing over every one of those monuments, or we’ve got something here.’ ”
Dragert spent nearly a year reanalyzing the data to assure himself it wasn’t instrumental glitches or some artifact of the data analysis technique. In 2001 he and two other colleagues set the earthquake world abuzz when they published their interpretation. Deep underground, they argued, the subduction zone that seemed zipped up tight was slipping a tiny bit. The movement had started under Seattle and migrated up to Vancouver Island over about a two-week period. Then it stopped and all the GPS stations reverted to their normal trajectories.
A lot of geologists were suspicious of Dragert’s claim. The persnickety Canadian might trust GPS to detect such teeny shifts, but many others weren’t so sure. Another line of evidence soon won over the doubters. At a conference in New Zealand in 2002, Dragert chatted with a Japanese scientist who had reported tiny tremors emanating from within the Nankai Trough subduction zone off Honshu’s southeast coast. The tremor seemed to migrate over several days, very much like the slip on Cascadia, and at about the same rate. “I’ve often been credited with an uncanny grasp of the obvious,” Dragert deadpanned, “and this ability now came into play.” As soon as he got back to Sidney, he started quizzing his colleagues to see if they had seen similar vibrations.
Rogers was the center’s senior seismologist. He told Dragert he’d noticed odd vibrations on occasion but wrote them off as unexplained background noise. By this time Dragert and others had compiled a list of multiple slip events over the past several years. Rogers picked a date and went to the room where he stored seismograms in the same flat boxes the photosensitive paper came in. “I found the box, opened up the month, and I flipped to the day,” Rogers said. “And there was this funny-looking noise I was talking about.”
He tore down to Dragert’s office. “It’s there! It’s there at exactly the same time,” he said. Dragert gave him another date; Rogers raced back and pulled out another box. “Every fifteen minutes I’d come tearing back to Herb’s office with another match,” Rogers recalled. “It was a very exciting afternoon, the kind that makes the hair stand up on the back of your neck.” It was as if after listening to human chests for centuries doctors suddenly discovered the heart can beat out a salsa rhythm.
The subduction zone wasn’t just slipping—it was chattering under its breath as it did so. Scientists coined the term “silent quake” to describe the process. Stretched over a period of two weeks or more, the subduction zone was unleashing as much energy as a Nisqually-size earthquake. But the motions were so slow and faint that only GPS and the most sensitive seismometers could detect them. Even more remarkable, scientists at Central Washington University reported that silent quakes seemed to occur on a regular schedule, every fourteen to fifteen months. The term Dragert came up with to describe them is “episodic tremor and slip.”
A process that releases energy sounds like a good thing, a way to defuse the seismic time bomb. But Dragert and Rogers could see at once that the silent quakes had just the opposite effect. Instead of bleeding off stress, every episode bumps it up and nudges the fault closer to catastrophic failure. “The megathrust earthquake is probably more likely during or immediately after one of these slip events,” Dragert said. “It’s like tightening a guitar string. It’s more likely to break at the time you’re adding that extra bit of stress.”
The subduction zone’s complex geometry accounts for the counterintuitive effect. The portion of the fault that’s slipping is twenty to thirty miles down, in a region where the plates are growing increasingly hot and pliable. In Washington that section lines up about midway under the Olympic Peninsula. Scientists have deployed arrays of special seismometers there to better listen in on the faint murmurs.
But the fault’s danger zone—the brittle, locked portion that will rupture in a megaquake—is shallower and closer to the coast. That section doesn’t budge at all during the silent quakes. So every fourteen months when the subducting plate slips a tiny bit deeper into the Earth, the motion ratchets up the strain on the locked portion. “It’s adding another straw to the camel’s back,” Rogers said.
The discovery was a watershed for earthquake scientists. One of the questions they started asking was whether silent earthquakes could represent something geologists have sought for decades: a harbinger that a big quake could be on the way.
What the world wants most from seismology is what the discipline has never been able to deliver. The quest to predict earthquakes has been such a resounding failure that most seismologists shun the subject like vampires shun sunlight.
There’s no shortage of people, some crackpots, some serious, who claim the power to foretell based on electrical discharges from rocks, phases of the moon, or radon gas emissions. “It is a holy grail, and it’s just a magnet for crappy science,” said Tim Melbourne, a geology professor at Central Washington University (CWU) in Ellensburg.
The theories are seductive because many of them have a semblance of validity. The weight of tides and the moon’s pull can tug on faults. The most thorough analyses do find a tiny correlation between tides and earthquake rates. The problem with all precursors, however, is lack of consistency. Many big quakes announce their arrival with a burst of foreshocks. Many don’t. There’s no way to tell the difference until after the fact.
Melbourne was newly arrived at the CWU campus when Dragert discovered Cascadia’s slow slip. Swept up in the excitement, he felt a frisson of trepidation every time a new event started. “There was a sense of, Oh my God! The fault is slipping,” Melbourne recalled. But the scientists could all see it would be tricky to use the phenomenon to time the next megaquake.
Few experts doubt that a silent quake could be enough to knock Cascadia over the edge. Historic records hint that Chile’s 1960 megaquake and several big quakes in Japan were preceded by similar slip. Japan’s 3/11 quake was, too, though scientists are still trying to make sense of the complex pattern of foreshocks, slip, and deformation that unfolded in the days before the great earthquake. “The slip almost certainly loaded the region of the fault that subsequently ruptured,” Melbourne said. “It’s about as close to a smoking gun as you could ever hope to see.”
By one estimate the odds of a Cascadia megaquake are thirty to one hundred times higher during slip events. But what should people do with that information? Even when the subduction zone is slipping and chattering, the odds of a megaquake remain small—about one in four thousand, according to the same analysis. It’s been more than three hundred years since Cascadia last ruptured, and none of the silent quakes in the intervening years has yet triggered a repeat. Each silent quake may pile on another straw, but scientists have no way of knowing how many straws it will take to break the camel’s back.
The Geological Survey of Canada used to post notices on its website when a silent quake was under way and pass on the information to emergency managers. The routine practice led to a nationwide scare when the Globe and Mail in Toronto published a story under the headline “Scientists predict monster quake.” In those days Dragert and his colleagues thought the patch under southern Vancouver Island and the Olympic Peninsula was the only part of the subduction zone where tremor and slip occurred on a repeating basis. Since then silent quakes have turned up under Central Oregon and Northern California, returning on one- and two-year cycles. So nearly a third of the time, some segment of the fault is slipping. The megaquake could be triggered in any of those regions. “The whole idea that the subduction zone is most likely to fail when you add an increment of stress is still probably right,” Rogers said. “It’s just not terribly useful because it’s happening so often.”
But the discovery of silent quakes helped galvanize a new era of investigation in the Northwest. Dragert’s four GPS stations have proliferated into an array of nearly five hundred across the region. Instead of providing measurements once a day, the devices now stream data in real time to CWU, where Melbourne oversees the network. Scientists around the world can spot unusual patterns as soon as they appear. The instantaneous GPS reveals changes as small as two centimeters. When Melbourne and his team spend more time crunching the numbers to factor out noise, the resolution zooms down to one millimeter, less than four hundredths of an inch.
More than 500 GPS stations track the way the Northwest is being shoved and squeezed, setting the stage for future earthquakes. (image credits 13.1)
With that level of precision and coverage, GPS has emerged as one of the most powerful tools in the earthquake arsenal. From his office Melbourne can track the region’s slow march to the northeast. On average the Northwest moves about half an inch a year. But the GPS observations clearly show that the motion isn’t uniform.
The town of Forks, on the Washington coast, travels a full inch. Across the Cascades in Central Washington the rate dwindles to an eighth of an inch a year. More than any seismogram, GPS illustrates the inevitability of earthquakes. The motion never ceases. The pressure never stops building on the subduction zone, the Seattle Fault, the Tacoma Fault, the South Whidbey Island Fault. Yearly increments are insignificant but they add up, and add up, and add up. Since the 1700 megaquake, the coast has moved more than twenty-five feet.
The GPS network also provides a new way to tackle old problems. Scientists still can’t say for sure how close the Cascadia rupture will come to Seattle and the region’s urban corridor. The closer it gets, the more destructive it will be. Earlier research put the locked portion of the fault—the danger zone—well offshore. But GPS measurements, as well as the slip and vibrations under the Olympic Peninsula during silent quakes, suggest that the locked portion of the fault could extend nearly fifty miles closer to Seattle.
For the next generation of Cascadia research, the action is shifting to the seafloor. Several ambitious experiments are under way or in the works to wire up the Juan de Fuca Plate. There’s an outside chance the effort could hit the jackpot and detect signs of stirring before the next big quake. More likely, it will succeed in peeling away a few more layers of mystery to give the region a clearer picture of the threat. “The new technology we’re tapping into now is unbelievable,” said University of Washington marine geologist John Delaney, one of the masterminds behind what he calls underwater observatories.
The inspiration came to him in a Seattle bar more than twenty years ago. Delaney was grousing with a fellow scientist about the limitations of deep-sea research. One of Delaney’s interests is black smokers, fantastical chimneys that form where fresh magma superheats seawater. But opportunities to dive to the seafloor in a submersible like Alvin are rare and costly. “You might have twenty hours a year to look out the window at unbelievably complicated systems—black smokers, tube worm beds, big fractures,” Delaney said, recalling his booze-fueled lament. “And then you come back two years later when you get another grant, and everything is changed and you don’t know why or when.”
Delaney’s friend pointed out that telephone companies had recently laid the first submarine fiber-optic cables from North America to Ireland. Delaney agreed that was cool, then thought a moment. “There ought to be a way to use that for science,” he said. He grabbed a cocktail napkin and sketched out a fiber-optic loop snaking from the coast across the seafloor. The cable could deliver power to instruments on the ocean bottom and transmit data to scientists on land. Most important, the observations would be continuous.
Delaney spent several years expanding and refining the concept. He named it NEPTUNE and patched together an unwieldy string of words to fit the acronym. Then he started pitching the idea to anyone who would listen. Eventually, the National Science Foundation agreed to fund the project and put Delaney and his UW colleagues in charge of building it.
But when the region’s first underwater observatory started operating in 2009, it wasn’t an American initiative. The five-hundred-mile-long loop of cable starts in Port Alberni on Vancouver Island, and crosses the continental shelf. Branches thread down submarine canyons and cross over the buried seam where the Juan de Fuca Plate meets the continent. The cables extend across the pitted expanse of the abyssal plain before rising onto the Juan de Fuca Ridge, where the steady upwelling of molten rock replenishes the tectonic plate and sets it in motion.
Spanning the entire tectonic plate was always central to Delaney’s plans. Canada got on board early and avoided the bureaucratic and budget battles that delayed Delaney’s progress at home. Cable for the first American observatory was finally laid off the Oregon coast in 2011. The footprint looks a lot like Delaney’s napkin sketch, with cable extending from Pacific City to Axial Seamount, the
most active underwater volcano in the north Pacific. The first instruments will be installed in 2013 and full operations are scheduled to start in 2014.
Both observatories are multipurpose. Cameras allow biologists and school kids alike to spy on crabs, cod, and octopuses. Oceanographic instruments track temperatures and salinity and tackle climate change questions. But getting a better handle on Cascadia is a prime objective. “The great subduction zone earthquakes of the world occur in the ocean,” said Kate Moran, director of NEPTUNE Canada. “For the first time now, we’re putting very, very sensitive instruments close to the location where these earthquakes take place.”
Underwater observatories off the Northwest coast are taking studies of subduction zone quakes to the source. NEPTUNE Canada and the National Science Foundation-funded Regional Scale Nodes off Oregon include seismometers and instruments to measure seafloor deformation and will expand in the future. (image credits 13.2)
Permanent seafloor seismometers can detect tiny quakes invisible to land-based instruments. A grid of ocean pressure sensors deployed off Vancouver Island is already helping assess the size of incoming tsunamis and fine-tune warnings. Similar instruments will sit on the seafloor off Oregon. Delaney’s not satisfied, though. He’s already pushing for the next big thing: a dense web of sensors to run the length and width of the subduction zone and take its pulse constantly.
Lack of good seafloor monitoring cost Japan dearly during the 3/11 quake. Pressure sensors on the seafloor would have allowed scientists to quickly size up both the earthquake and the tsunami by estimating displacement. Instead, warnings based solely on seismometers underestimated the disaster in the crucial first minutes.
Full-Rip 9.0: The Next Big Earthquake in the Pacific Northwest Page 25