Most earthquakes pop in the Earth’s uppermost layer, the crust, where rocks are cold and brittle. Quakes like Nisqually originate thirty to sixty miles down, where temperatures are approaching two thousand degrees Fahrenheit and rock is becoming more like Silly Putty than porcelain. At the same time, the weight of those miles of crust and mantle presses down so hard that any cracks in the rock are sealed shut. How could a fault possibly slip and generate an earthquake under those conditions?
“It was a big controversy,” Kirby said. “One of the most famous geophysicists in Great Britain declared that deep earthquakes were not possible.”
As seismometers improved, the data became harder to dismiss. By the late 1930s, Japanese seismologist Kiyoo Wadati had documented the locations of more than a dozen deep earthquakes in Japan. Some originated four hundred miles or more below decks. Victor Hugo Benioff, a Caltech colleague of Charles Richter’s, took a similar inventory of deep earthquakes around the world. The dots on Benioff’s map clustered along the edges of continents, many positioned under volcanic arcs like the Cascades. The bands of deep seismicity were angled like ramps descending from the coastline into the planet’s bowels. Scientists named them Wadati-Benioff zones.
It remained for plate tectonics to make sense of the pattern. An understanding of what happens when oceanic and continental plates collide was one of the last pieces of the tectonic puzzle to fall into place. Scientists had already figured out ocean rifts and seafloor spreading by the time they hit on subduction. The realization that the seafloor sank under the continents and melted down in the Earth’s interior furnace set mental lightbulbs flashing among the handful of scientists who were curious about deep earthquakes. Kirby’s thesis adviser was one of them, and he urged his protégé to help unravel the mystery.
As seismologists looked closer, they saw that all deep quakes occur within subducting slabs. The reason is temperature. As a slab of ocean floor like the Juan de Fuca Plate dives into the Earth, it keeps its cool on a geologic temperature scale. Even one hundred miles or more below the surface, the rocks remain stiff enough to snap. Down where Nisqually-type quakes are born, the subducting Juan de Fuca Plate is about a thousand degrees Fahrenheit, half as hot as the surrounding rock. The still-rigid slab is under enormous strain as it’s shoved, bent, stretched, and crunched beneath North America. Imagine trying to bend and twist a marble countertop. No wonder the rocks might crack.
CASCADIA EARTHQUAKE SOURCES
The process of subduction is responsible for all three types of earthquakes that occur in the Pacific Northwest: deep quakes, shallow crustal quakes, and megaquakes. (image credits 8.1)
But the fact that the subducting slab is cool enough to crack isn’t enough to explain deep earthquakes. Even if the rocks in the Juan de Fuca Plate were riddled with cracks, thirty miles of the Earth’s mass pressing down from above would pinch those cracks—or faults—tightly shut. There’s no way the faults should be able to slip and generate earthquakes under pressures that intense. And yet they do. Something must be greasing the skids, and Kirby and other scientists think it’s probably water.
Kirby knew from the types of lab experiments he conducted early in his career that if you heat rock and squeeze it hard enough, it can morph into something else. Pressure and heat can also wring water from crystals in the rock, like squeezing a wet sponge. Consider basalt, the black volcanic rock that dominates the seafloor. When shoved down a subduction zone, it pulls a Cinderella act. The result is a drop-dead gorgeous stone called eclogite (ek-lo-jite), laced with green and studded with blood-red garnets. A very lucky rockhound might find a chunk of eclogite that shot to the surface through the kind of volcanic pipes that also transport diamonds. During basalt’s high-pressure metamorphosis into eclogite, water is ejected. If that liquid accumulates in existing cracks or weak spots in the rock, it could serve as a wedge to pry open the fault, and enough of a lubricant to allow the fault to slip.
There’s no shortage of water-rich minerals in oceanic slabs like the Juan de Fuca Plate. Kirby saw it for himself on an expedition off the southwest coast of Africa. Using a ship-mounted rig, scientists drilled nearly a third of a mile into the seafloor. The cylinders of dark rock they pulled up were shot through with cracks. Lining the cracks was a type of emerald-colored mineral called greenschist that’s loaded with water. “You could see it from across the room,” Kirby recalled. “That really rang a bell with me.”
The mineral-laden cracks struck him as triggers, cocked to set off future earthquakes. As the plate subducts deeper into the earth, the water would be squeezed out of the minerals. “So the water is right where you need it to reactivate these faults,” Kirby said. The observation inspired him to return to the lab, where he conducted experiments using ice as a stand-in for oceanic crust to show the types of fractures that might result. Later, a team of British scientists went even further, subjecting rock cores the size of a pencil lead to searing temperatures and some of the highest pressures ever generated in a lab. Water rushed out of the crystals and the rock fractured with a pop. “You can hear it,” Kirby said. “It’s like creating a mini-earthquake.”
The water idea is the best anyone has come up with to explain what makes the Northwest’s deep quakes tick, but there’s no way to prove it. The study of seismic goings-on deep underground is a classic example of what’s sometimes called black-box geology. Another word for it is geophysics. Author John McPhee captured the tension in the early 1980s between the old-school field geologists whose numbers were dwindling and the new breed who spent most of their time sitting in front of computers. In Annals of the Former World one geologist groused, “The name of the game now is ‘modeling.’ A lot of it, I can’t see for sour owl shit.”
A hint of that sentiment remains, though it mostly takes the form of good-natured banter between the few field geologists left and the geophysicists who far outnumber them on university faculties. The debate over the Cascadia Subduction Zone’s ability to generate big quakes pitted field geologists like Brian Atwater and others against geophysicists whose models and calculations suggested the fault was dead. There was a saying going around back then that could have come from McPhee’s hard-rock man: “When geology and geophysics clash, throw geophysics in the trash.” In other words, believe the evidence in front of your eyes. Most geoscientists would agree on that point, Kirby included. “My own view is that there’s no substitute for being able to actually lay your hands on the fault and look at the rocks,” he said, wistfully.
But until someone invents a machine to descend into the Earth à la Jules Verne, Kirby and his colleagues aren’t going to get their hands on the rocks involved in deep quakes. Instead they pore over fuzzy gravity images, break rocks in the lab, and build mathematical models to extrapolate those results to what goes on in a Nisqually-style quake. Seismology is a type of geophysics that uses seismic waves, both natural and man-made (as in the Kingdome implosion), to probe the Earth. The seismic signals from deep earthquakes proved to be a particularly powerful tool. Much of what scientists know about the planet’s inner layers—the mantle and the core—came from analyzing the ways waves from deep quakes reverberate and ricochet through the Earth.
Kirby coined the term intraslab to describe the deep quakes common in the Pacific Northwest. The name caught on, but there hasn’t been a stampede of scientists eager to study them. In addition to being out of reach, the quakes don’t leave geologic fingerprints that researchers can follow. Intraslab quakes don’t rupture the ground like quakes on shallow faults, so there aren’t any scarps to show up on lidar. Nor do deep quakes uplift beaches or drown forests, like subduction zone quakes. The Nisqually quake was the region’s first since GPS and satellite sensors came on the scene, and the instruments detected just a half-inch slump at the epicenter. With no way to burrow into the distant past, scientists have little more than a hundred years of history to rely on as an indicator of what the future holds.
Since the start of the twentieth centur
y, the Northwest has been rattled by eighteen quakes known or suspected to have deep roots. Quakes from the pre-seismogram era were included on the list if newspaper accounts didn’t mention aftershocks. The lack of significant aftershocks is another mystery about deep quakes, but it’s a feature they seem to share around the world.
Eight of the suspects on the list had estimated magnitudes of 5.5 or more. A quake in 1909 severed underwater telegraph cables and cracked concrete walls near the Canadian border. Mr. Ed. R. Novak, a rubber salesman whose residence was Seattle’s Lincoln Hotel, was moved to compose a poem about his experience. Published in the Seattle Daily Times, it read, in part:
“Just then, Ha! Ha! I felt a feeling. Was I drunk or sober?
And gazing on my eggs I found that they had been turned over.”
In 1939 a tremor estimated at magnitude 6.2 shook the ground so hard that water mains broke in Seattle and cornices toppled from buildings. “I was never so scared in my life,” Washington Governor Clarence D. Martin admitted. When a more powerful quake struck in 1946, UW Professor G. E. Goodspeed assured a rattled citizenry that “no more strong shocks will occur for five or six years.” The biggest quake in the state’s history arrived three years later, in 1949.
The following sixteen years were a period of relative seismic peace. By the time Carol Davis experienced her second major earthquake, in 1965, she was married and working as a schoolteacher. On the morning of April 29, Davis was in her home economics class at Renton Junior High School south of Seattle when she felt the familiar, sickening motion. “It scared the living daylights out of the students,” she recalled. Bricks fell at the school, but no one was injured.
Elsewhere across the region, six people died. A fifty-thousand-gallon wooden water tank toppled seven stories from a flour mill on Seattle’s Harbor Island, killing one man and injuring another. A huge landslide ripped down Mount Si, east of Seattle. The shaking toppled power lines near Everett and knocked out circuits at Grand Coulee Dam, nearly two hundred miles from the epicenter. The Seattle Times published a picture of a worker up to his knees in suds at the Rainier Brewery, where fifteen thousand gallons of beer spilled from tanks and broken pipes.
It’s no more than an interesting coincidence that the last three major quakes all struck in late February or April. But their geographical proximity is probably no accident. The 2001 and 1949 quakes seem to have originated from almost the same spot under the Nisqually delta. The 1965 quake was centered a bit farther north. Every significant deep quake in the past century had its epicenter somewhere under the Puget Sound basin or Georgia Strait.
“Why do they occur there?” Kirby asked. “That’s a very good question and we don’t have many clues.”
Creager, the UW professor, thinks the answer lies in simple geometry. Along most of its length, the Northwest coastline is as straight as a ruler. The exception is the seaward jog that begins at the central Washington coast and continues along the southern end of Vancouver Island. The subduction zone, which parallels the coast, seems to have a hard time negotiating that curve.
Like a tablecloth draped over a corner, it bunches up. But a basalt slab doesn’t pleat as easily as linen. “It creates a huge strain, geologically speaking,” Creager said. That stretch of seafloor is also slightly older and colder than segments to the north and south, which could make the plate there even less pliable. “That may be why we get more earthquakes in this corner.”
The absence of deep quakes elsewhere in the Northwest in recent times doesn’t mean those places get a free pass. The quakes might strike more frequently where the subduction zone is squeezed into a corner, but there’s no obvious reason why they couldn’t strike in Oregon and southern Washington as well. “I think we’re just in a quiet period,” Creager said. Come back in another hundred years and the record will probably include deep quakes throughout the region.
Time alone will answer the size question, as well. Theory suggests deep quakes bigger than about 7.2 aren’t likely in the Northwest. It’s a matter of space, Creager explained. Most of the region’s deep earthquakes seem to originate in the brittle crust of the subducting Juan de Fuca Plate. That crust is only about five miles thick, which isn’t big enough to contain a rupture of the size it would take to generate a giant quake. In places like Japan and Bolivia, where the biggest intraslab quakes occur, the subducting plates are older and thicker.
But Creager isn’t convinced that deep quakes as big as magnitude 8 couldn’t happen in the Northwest. There’s a lot scientists don’t know about the phenomenon, including whether the deep quakes under Puget Sound are strictly confined to the subducting crust. Maybe ruptures can propagate through both the crust and mantle of the subducting plate—which would allow for bigger quakes. Creager agrees that it’s highly speculative, but he thinks it’s worth considering.
Even if Creager’s theory isn’t right, the region could still be in for some very nasty deep quakes in the future, said Craig Weaver, the regional USGS seismology chief. A magnitude 7.2 would be four times more powerful than anything the Northwest has experienced in more than fifty years. Not only would the ground shake harder—it would shake longer. The Nisqually quake lasted about forty seconds. Another five to ten seconds of shaking, and many of Seattle’s old brick buildings probably would have collapsed, along with the rickety Alaskan Way viaduct along the city’s waterfront.
Nisqually was Carol Davis’s most serene seismic experience. From the chairlift at Crystal Mountain ski area, she noticed something odd about the treetops. “They were dancing, waving back and forth,” she recalled. The chair stopped for about ten minutes. “I said to my friend, ‘I’ll bet that was an earthquake.’ ” At the top of the hill, a message board confirmed Davis’s hunch. The road down the mountain was blocked for hours by boulders and fallen trees. Back at home Davis’s broken bric-a-brac were the only casualties.
USGS seismologist Bob Norris’s day got off to a more explosive start. He was bumping along a gravel road on Seattle’s Harbor Island, a sprawling industrial complex that sits on what used to be the Duwamish River Delta. Fill shakes harder than any other type of ground, so the USGS had installed a strong-motion seismometer near a tank farm in the former tide flat. Norris was there to download data from the instrument.
When the earthquake hit, he thought he had run over something. Norris stopped the truck. “I went through several seconds of confusion because the truck was still rocking sharply,” he recalled. In most parts of town, the strongest motion lasted about twenty seconds. But the ground under Norris’s truck kept pounding like a jackhammer. All he could do was hold on and hope his head didn’t smash into the windshield. As the motion eased, he watched two-hundred-foot cargo cranes flexing “like huge steel giraffes trying to dance.”
What happened next was something even the seasoned earthquake scientist had never seen. Almost ten minutes after the shaking stopped, water started pouring from the ground, then coalesced into a muddy geyser. Norris at first thought the quake had broken a water pipe. But what he was witnessing was liquefaction. When sandy soils take a seismic pounding, they turn to a watery slop. Some of that water is ejected from the ground under pressure. By the time Norris left, the swirling pool was fifty feet across and growing by the second.
Liquefaction in built-up tide flats, waterfronts, and along river valleys is a serious cause of damage in major quakes, undermining structures and heaving up buried utility lines. In the days and weeks that followed the Nisqually quake, scientists fanned out across the region to catalog liquefaction effects. They found sinkholes, buildings with lopsided foundations, and cracks in airport runways.
Scientifically, the quake was a bonanza. One group of researchers examined stream gauges and discovered that water levels spiked in many areas, presumably because the quake compacted the soil and squeezed out groundwater. Another team followed the trail of broken chimneys to identify neighborhoods that shook harder than others. Spectacular sand volcanoes, like the one Norris saw, erupted
around Lake Sammamish east of Seattle, as buried sand liquified by the shaking burst to the surface like toothpaste shot from a tube. Worms surfaced, too—as many as one per square foot in the most jostled areas. Every evening scientists and building engineers gathered at the University of Washington to drink beer and talk about what they’d seen. “It was a pretty intense ten days that followed the quake,” said UW seismologist Bill Steele.
The most sobering insights emerged later, when seismologists analyzed patterns of shaking and seismic waves. The UW had just installed more than two dozen sturdy new instruments that wouldn’t be knocked off scale by strong ground motions. Nisqually was the first big quake in the Northwest to be so thoroughly documented. The payoff was unequivocal proof that the five-mile-deep basin underlying Seattle and its environs is bad news.
During the Nisqually quake, the basin trapped and amplified seismic waves even more than scientists had expected. The ground in the basin shook longer, and it shook up to three times harder than elsewhere. The effect will come into play in every type of future quake, whether deep, shallow, or from the subduction zone. “There’s always going to be that extra kick from the basin,” Weaver said.
The USGS calculates 80 percent odds that another deep quake will strike the Puget Sound region in the next fifty years. That’s as close to a sure bet as seismology can offer. What scientists have no way of estimating is whether the next one will be a relative lamb, like Nisqually, or whether it might set a state record. “People have a tendency to think, I made it through 1965 and 2001 so my building must be ready for the next one,” Weaver said. “They don’t appreciate the fact that there’s a lot of variability in these things.”
Full-Rip 9.0: The Next Big Earthquake in the Pacific Northwest Page 15
