Tower: Horizon 177, [this is] Seattle Tower, you’re No. 2 following a heavy Boeing 767, short final. Wind 130 at 8. Runway 16 Right. Cleared to land. (Pause). All right, we’ve got an earthquake. Everybody hold on, folks. (Pause). Attention all aircraft in Seattle. We have a huge earthquake going on. The tower is collapsing. I say again. The tower is falling apart. Hang on everybody. (Pause) OK, we got a huge earthquake going on in Seattle. Everybody be careful out there, all right?
Pilot: American 27 heavy, we’re about to turn final.
Tower: All right, everybody on Seattle Tower, I want you to use extreme caution. The tower windows have collapsed here. Asiana 272 heavy, turn left here, hold short on 16 Left, remain on this frequency. And Horizon 301, I want you to turn left, and I want you to go to the ramp, and remain on this frequency. All the windows are gone from the tower but two.
Pilot: This is 301, we’re turning off here at—
Tower: All right, I want everyone to pay attention here, because I don’t know what’s working and what’s not. All the windows are gone.
Sea-Tac tower operator Brian Schimpf during the 2001 Nisqually Earthquake
1. Commotion in the Ocean
The Juan de Fuca Plate is entirely oceanic (Figures 2-7), with thin crust made up of basalt. No part of it is above sea level. The crust is nowhere more than a few tens of millions of years old, which means that it is relatively shallow, weak, and hot. Its weakness means that it is subject to internal deformation where it interacts with the continental edge of North America. At its northern and southern ends, where the spreading center is closest to the base of the continent, and the oceanic crust is youngest, the weak oceanic plate is being actively deformed internally, deformation that is marked by frequent earthquakes (Figure 4-14). These seismically active regions are generally referred to as separate plates, the Explorer Plate off Vancouver Island and the Gorda Plate off northern California (Figure 5-1). The Juan de Fuca Plate between its northern and southern ends has few earthquakes, indicating that internal deformation is less important there.
The fact that the Juan de Fuca Plate is completely oceanic means that we are not able to measure its displacement rates directly but instead must rely on indirect geophysical evidence. All permanent seismic stations are onshore, resulting in considerable inaccuracy in locating earthquakes on the plate. However, in recent years, the declassification of the U.S. Navy’s hydrophone detection system has allowed scientists of the National Oceanic and Atmospheric Administration (NOAA) in Newport, Oregon, to study earthquakes using seismic waves (T-phase waves) that are transmitted through ocean water rather than through the crust beneath the ocean. They have been able to improve greatly the accuracy and detection threshold for earthquakes far from shore (Figure 4-14).
Mapping of the distribution of earthquakes shows that the spreading centers, the Juan de Fuca, Gorda, and Explorer ridges, generate low-level seismicity related to the movement of magma that rises to the surface and forms new oceanic crust. These earthquakes are small, most of them too small to be detected by ordinary seismographs onshore, although they are monitored through the SOSUS hydrophone detection system.
On the other hand, the Gorda Plate is cut by large strike-slip faults that rupture frequently to cause earthquakes (Figures 2-4, 5-2). The Gorda Plate west of Arcata, California, sustained an earthquake of M 7.3-7.6 on January 31, 1922, that was felt in Oregon and Nevada, and as far south as San Jose, California. Another earthquake of M 6.9-7.4 thirty miles west of Trinidad, California, on November 8, 1980, destroyed a bridge, liquefied the sand bar at Big Lagoon, and caused six injuries and $1.75 million in damage. In 1991, the Gorda Plate was shaken by an earthquake of M 6.9 on July 12, another of M 6.3 on August 16, and the largest one of M 7.1 on August 17, three hours after a crustal earthquake onshore. On April 26, 1992, one day after the M 7.1 Cape Mendocino Earthquake on the Cascadia Subduction Zone, two aftershocks of M 6.0 and M 6.5 struck the Gorda Plate twelve and eight miles, respectively, offshore. One of these aftershocks trashed the commercial district of the small town of Scotia. These were the largest of hundreds of aftershocks of the Cape Mendocino earthquake in the Gorda Plate, complicating the problem of whether that earthquake was mainly a subduction-zone earthquake or a Gorda Plate earthquake. Except for the 1980 earthquake and the two Petrolia aftershocks, these Gorda Plate earthquakes were far enough offshore that intensities on the coast did not exceed V or VI.
The Gorda Plate has accounted for more damaging historical earthquakes in northern California than any other source, including the Cascadia Subduction Zone and the North American Plate. However, it is incapable of producing earthquakes in the M 8 to 9 range, such as those expected on the Cascadia Subduction Zone.
The Explorer Plate off Vancouver Island is also shaken by frequent earthquakes (see Appendix A). But, unlike Gorda Plate earthquakes, these are far enough from populated areas that they do no damage and in some cases are not even felt onshore.
2. Offshore Transform Faults:
The Northwest’s Answer to the San Andreas Fault
In Chapter 2, we considered two types of plate boundaries: ocean ridges or spreading centers, where new oceanic lithosphere is created as plates move away from each other, and subduction zones, where oceanic lithosphere is recycled back into the interior of the Earth as plates move toward each other. The Juan de Fuca and Gorda ridges are examples of spreading centers, and the Cascadia Subduction Zone is an example of two plates converging (Figures 2-6 and 5-1). We also considered a third type of plate boundary where plates neither converge nor diverge but instead move past each other without destroying or creating lithosphere. These are called transform faults because they transform plate motion between two spreading centers. They involve the entire lithosphere and not just the Earth’s upper crust.
The San Andreas Fault is a transform fault in which continental rocks of the North America Plate move past continental rocks of the Pacific Plate (Figure 2-8, top diagram, and animation, Figure 2-9). Transform faults in the Pacific Northwest, on the other hand, are found on the deep ocean floor, where they form linear topographic features called fracture zones. The Blanco Fracture Zone separates the Juan de Fuca and the Gorda ridges, and the Sovanco Fracture Zone separates the Juan de Fuca and the Pacific plates (Figure 5-1). The Mendocino Fracture Zone separates the Gorda and Pacific plates and is the northwest continuation of the San Andreas Fault. These are typical transform faults. The grinding of one plate past the other causes many earthquakes on these fracture zones. Barring the next subduction-zone earthquake, they and the interiors of the Gorda and Explorer plates have the highest instrumental seismicity in the Pacific Northwest, onshore or offshore. Large earthquakes on the Mendocino and Blanco fracture zones are felt frequently every year in northern California and southern Oregon.
At first glance, the Blanco Fracture Zone resembles a left-lateral strike-slip fault because of the apparent left offset of the Juan de Fuca and the Gorda ridges (Figure 5-1). But this apparent left offset would only be true if these ridges had once been a continuous unbroken ridge that was later separated along the Blanco Fracture Zone. This is not the case. Remember that the Juan de Fuca Plate is moving away from the Pacific Plate at these spreading centers. Imagine yourself standing on the Pacific Plate looking northward across the Blanco Fracture Zone at the Juan de Fuca Plate. The Juan de Fuca Plate moves from left to right along the Blanco Fracture Zone with respect to your position on the Pacific Plate. This means that the transform fault on the Blanco Fracture Zone is a right-lateral, not a left-lateral, fault.
As another thought experiment, imagine two jigsaw puzzle pieces that lock together by a tab that projects from one piece into the other. Now pull the pieces slowly apart. They are difficult to separate because the sides of the tab resist being pulled apart. In the same way, the Pacific and Juan de Fuca plates are being pulled apart, with molten rock welling up along the spreading centers as the plates are separated. Along the Blanco Transform Fault, the crustal plates push past each other, generating friction and producing earthquakes. These earthquakes could be as large as magnitude 7 or even larger, but probably not 8. The crust is too warm and therefore too weak to generate such large earthquakes. Accordingly, despite the high instrumental seismicity on the Blanco Transform Fault, including many earthquakes felt onshore, it does not constitute a major hazard to communities along the coast, in part because the earthquakes are many miles offshore, and in part because these offshore earthquakes are not large enough.
Earthquakes on the Mendocino Transform Fault are frequent. The first recorded major earthquake was felt on May 9, 1878, causing chimneys to fall in Petrolia, California, at the Triple Junction (Appendix A). A larger earthquake, of M 6.5-7.3, struck close to Cape Mendocino on January 22, 1923, resulting in intensities of VIII and damage to buildings in Petrolia. Other earthquakes include a magnitude 6 in 1922 and smaller earthquakes in 1932, 1936, and 1951. Other earthquakes with magnitudes greater than 6 struck in 1954, 1960, and 1984. The 1984 earthquake of M 6.6, 166 miles west of the coast, was felt from Oregon to San Francisco, but it produced intensities of V or less because of its great distance from shore. On September 1, 1994, an earthquake of M 6.9-7.2 struck the Mendocino Transform Fault 88 miles offshore, the largest earthquake to strike the United States that year, larger even than the Northridge Earthquake of the preceding January. Because it was so far offshore, it did no damage, but it was felt from southern Oregon to Marin County, California.
Like the Blanco and Mendocino faults, the San Andreas Fault is also a transform fault, separating the Gorda Rise from a spreading center in the Gulf of California of northwest Mexico (Figure 2-8, top diagram; Fig. 2-9 animation). The offshore transform faults differ from the San Andreas in involving relatively hot oceanic crust and mantle, whereas the San Andreas cuts across colder continental crust for most of its length. For this reason, the San Andreas generates significantly larger earthquakes than does the Blanco, up to at least M 7.9. So, fortunately for the Pacific Northwest, the Blanco and Mendocino are the weaker relatives; they generate many earthquakes, but no giant ones.
Two transform faults lie off the coast of Vancouver Island: the Sovanco Fracture Zone that separates the Explorer Plate and the Pacific Plate, and the Nootka Fracture Zone that separates the Explorer Plate and the Juan de Fuca Plate (Figures 2-8, 5-1). Like the Blanco, these fracture zones are characterized by high seismicity, but are not believed to generate very large earthquakes. In the next chapter, we will consider the possible relation between the oceanic Nootka Fracture Zone and two large historical earthquakes in continental crust of central Vancouver Island.
Northwest of the Explorer Plate, the Pacific Plate grinds against the North America Plate along the Queen Charlotte Fault, located at the base of the continental slope. On August 22, 1948, this fault was the source of an earthquake of M 8.1, larger than any historical earthquake on the west coast of the United States south of Alaska. On October 27, 2012, this fault was the source of the Haida Gwaii earthquake of M 7.8 (using the First Nations name, Haida Gwaii, for the Queen Charlotte Islands). This fault had a focal mechanism of reverse faulting rather than the expected strike slip. These earthquakes are evidence that the Queen Charlotte Fault poses a hazard to the thinly populated coast of British Columbia north of Vancouver Island, including the Queen Charlotte Islands. However, the region is so thinly populated that it is not considered as a major earthquake threat in Canada.
3. Slab Earthquakes in the Juan de Fuca Plate Beneath the Continent: Puget Sound Region
The greatest amount of seismicity generated by the Juan de Fuca Plate itself (not including the Explorer and the Gorda plates) is beneath western Washington, where it is being subducted beneath North America (Figures 3-21, 5-1). These are called slab earthquakes or Benioff zone earthquakes. Most of the damage and loss of life in the Pacific Northwest has been as a result of these earthquakes, including the largest known historical shocks to strike either Washington or Oregon.
The first of these, on April 13, 1949, really should have been no great surprise. The southwesternmost Puget Sound region had been struck by earthquakes on November 13, 1939 (M 5.5-5.9), and on February 15, 1946 (M 6.3). Both were slab earthquakes, and both had produced intensities as high as VII, which meant minor damage and collapse of chimneys. The 1949 earthquake of M 7.1 struck the southern Puget Sound region just before noon on April 13. Strong shaking lasted about thirty seconds. Most people were at work, getting ready to go to lunch. Most schools were on vacation, which turned out to be a blessing because of the collapse of many unreinforced brick school buildings. The epicenter was between Olympia and Fort Lewis, and the high-intensity damage zone extended from Rainier, Oregon, on the Columbia River, north to Seattle (Figures 5-3 to 5-5). The earthquake was felt from Vancouver, B.C., to Klamath Falls and Roseburg, Oregon. A sidewalk clock outside a jewelry store at 1323 Third Avenue in Seattle stopped at the moment of the earthquake: 11:56.
Eleven-year-old Marvin Klegman was killed, and two other children were injured by falling bricks as they played outside the Lowell School in Tacoma. Jack Roller was killed when part of the Castle Rock School building collapsed on him. Five students and two teachers were injured at Adna School 10 miles west of Centralia. One little girl was critically injured as she left her second-grade classroom. Tons of bricks fell from the Lafayette School building in Seattle, but school was not in session, and children were playing in the schoolyard far from the building. The Lafayette School was one of ten Washington schools condemned after the earthquake. The auditorium collapsed at Puyallup High School (Figure 5-3), but no one was in it at the time. Part of the Boys Training School at Chehalis crumpled and fell, injuring two boys.
There were many narrow escapes. Freda Leaf, seventy-one, jumped into the Duwamish River but was rescued by a neighbor, D. V. Heacock. Part of the roof of the Busy Bee Restaurant on Second Avenue in Seattle fell in, and the patrons headed for the exit. The proprietor, George Pappas, immediately saw the danger and ordered the bartender, a big man named Bill Given, to block the exit. Moments later, tons of bricks cascaded onto the sidewalk in front of the restaurant. Water spilled out of an old water tower at the reservoir at Roosevelt Way and East 86th Street; a few minutes before, painters working at the tower had knocked off for lunch. At the Tacoma Narrows Bridge, under repair at the time, a twenty-three-ton steel saddle mounted to hold up a suspension cable dislodged and plunged off the bridge and through a scow on the water below, injuring two people. In Olympia, the Old State Building (Figure 5-4) and the State Insurance Building were the worst hit. Governor Arthur Langlie and his assistant, Dick Everest, were in their offices in Olympia and were showered with falling plaster.