After Atwater’s discovery at Willapa Bay, other scientists found evidence of marshes buried by sudden subsidence accompanying earthquakes at South Slough near Coos Bay in southern Oregon, at Salmon River near Lincoln City, Oregon, at Nehalem Bay and Netarts Bay in northern Oregon, at the mouth of the Copalis River in Washington, and at Port Alberni and Ucluelet on the Pacific coast of Vancouver Island (Figure 4-12). Carbon from buried soils and from drowned tree trunks was sent to radiocarbon labs for dating. The result: the youngest marsh burial occurred about three hundred years ago at nearly all sites along the Cascadia Subduction Zone from British Columbia to southern Oregon. If this was caused by a single earthquake, as the similarity in radiocarbon ages would suggest, that earthquake would have a moment magnitude (Mw) of 9, close to the size of the great Alaskan earthquake of 1964. It would rank among the largest ever recorded.
A common saying among geologists is that what has happened, can happen. If the earthquake three hundred years ago was a magnitude 9, the next subduction-zone earthquake could also be a 9. If this were to happen, what would be the impact on our society?
All of western Washington and Oregon, southwesternmost British Columbia, and Del Norte and Humboldt counties in north-coastal, California would be devastated by a magnitude 9 earthquake, so that emergency response teams would have to come from inland cities or from central and southern California. Intense shaking from a magnitude 9 event would last two to three minutes or longer; a magnitude 8 event would have strong shaking for about half that time. A building might survive strong shaking lasting a minute, but not two or three times as long. For comparison, the strong shaking for the Kobe and Northridge earthquakes each lasted less than thirty seconds. Some of the shaking during these smaller earthquakes was as strong as that expected for a great subduction-zone earthquake; it just didn’t last as long.
This shaking would trigger landslides throughout the Coast Range, Olympic Mountains, and Vancouver Island, in Puget Sound and the Willamette Valley, and even on the continental slope, where landslides could trigger tsunamis. For even a magnitude 8 event, large sand bars like those at Long Beach, Washington, or at the mouth of Siletz Bay, Oregon, could become unstable, as would low-lying islands in the tidal reaches of the lower Columbia River. The Pacific coastline would drop permanently, as shown in Figures 4-10 and 4-11, as much as two to four feet, inundating low-lying areas such as Coos Bay, Yaquina Bay, Siletz Bay, Tillamook Bay, Cannon Beach, Seaside, and Astoria, Oregon, and Long Beach and Grays Harbor in Washington.
Seismic sea waves, or tsunamis, could be as high as thirty to forty feet with a magnitude 9 earthquake, but less than half that with an 8. Fifteen to thirty minutes after the mainshock had died away, the first of several tsunami waves would strike. In some cases, the water would first rush out to sea, exposing the seafloor (never before seen as dry land). However, a short time later, a wall of water would rush inland, sweeping the sand from barrier bars inland, overwhelming beach houses, bayfront boutiques, and restaurants as far as several blocks away from the sea. These destructive waves would be repeated several times.
The mainshock would be followed by aftershocks, some with magnitudes greater than 7, large earthquakes in their own right. These aftershocks would continue at a diminishing rate for many years. For a magnitude 8 earthquake, aftershocks would affect a limited part of the Pacific Northwest perhaps two to three hundred miles long, but for a magnitude 9 event, the entire Northwest from Vancouver Island to northern California would be shaken.
Because a magnitude 9 earthquake would devastate such a large area, it would have catastrophic effects on the economy of the Northwest, the ability of government to serve the people, and the ability of insurance companies to pay their claims. The economic effects of a magnitude 8 event would be great, but not as cataclysmic as those of a magnitude 9 because a much smaller area would be affected. If a magnitude 8 earthquake originated west of the mouth of the Columbia River, it would severely damage the Portland metropolitan area but would have lesser effects on the cities of Puget Sound or the southern Willamette Valley. Emergency response teams from those areas could come to the aid of Portland and adjacent communities in Oregon and Washington. There would be less damage, fewer insurance claims, less destructive effects on the overall economy of the United States and Canada than from a magnitude 9 earthquake. I return to the effects on society in a later chapter.
A diagram illustrating the buildup and release of strain in the next great Cascadia earthquake is shown as Figure 4-14.
How do we learn whether the last earthquake was a magnitude 8 or a 9? Radiocarbon dates can provide accuracy to within a few decades, which is not proof that all the marshes and estuaries were buried at the same time from Vancouver Island to southern Oregon. In southwest Japan, the Nankai Subduction Zone broke in two magnitude-8 earthquakes, one in 1944, while Japan was in the throes of World War II, and one in 1946, when the country was trying to rebuild after the end of the war. If these earthquakes had not been recorded historically, radiocarbon dating could not have provided evidence that these were two separate earthquakes; the numbers could just as likely have documented one great earthquake rather than the two that actually occurred. Gary Carver of Humboldt State University points out the dilemma: one gigantic earthquake (“instant of catastrophe”) versus a series of smaller ones (“decade of terror”) about three hundred years ago.
Tree-ring dating can get closer to a true date than radiocarbon dating can (Figure 4-15. Gordon Jacoby of Columbia University and Dave Yamaguchi of the University of Washington compared the pattern of growth rings of trees killed in several estuaries in southwest Washington. Variations in the growth patterns of trees from year to year, related to unusual wet seasons or drought years, allowed these scientists to use radiocarbon dating to conclude that trees in four of these estuaries were inundated some time between August 1699 and May 1700, strong evidence that the estuaries were down-dropped at the same time by an earthquake of magnitude greater than 8. However, at that time, these correlations had not been extended north to Vancouver Island or south to California, which would strengthen the case for a single magnitude 9 earthquake.
Clifton Mitchell and Ray Weldon of the University of Oregon studied re-levelings of U. S. Highway 101 along the coast from Crescent City, California, to Neah Bay, Washington, taking advantage of a more accurate leveling survey carried out after John Adams had published his results. They found that over the past fifty years, southern Oregon, the mouth of the Columbia River, and northwest Washington have been rising at about an inch or more every ten years (Figure 4-16, map on the left side). But the central Oregon coast around Newport and the area around Grays Harbor, Washington, are hardly uplifting at all. This suggested to them that only some parts of the Cascadia Subduction Zone are building up the elastic strain. Imagine irregular hang-ups or strong points (called asperities) along the subduction zone that concentrate all the strain and localize the uplift, separated by other regions where strain is not accumulating. The zones of little or no strain around Newport and Grays Harbor could have terminated the rupture, preventing it from shearing off the next asperity to the north or south. This line of reasoning supported the “decade of terror” hypothesis of several smaller earthquakes rather than one humongous one.
But Roy Hyndman and Kelin Wang of the Pacific Geoscience Centre at Sidney, B. C., argued that the earthquake is more likely to be a 9 rather than an 8. Using temperature estimates in the crust on Vancouver Island and offshore, they calculated which parts of the subduction zone would be stuck and which parts would slide freely due to higher temperature at greater depth. They also measured the changes in leveling lines across Vancouver Island and the Georgia Strait. They compared their leveling data with the uplift of the land with respect to sea level, taking advantage of the fact that they could use three coastlines: both sides of Vancouver Island and the mainland coast northwest of Vancouver. Hyndman and Wang calculated where the brittle-ductile transition would be, together with a deeper transition zone that would be brittle under rapid strain and ductile under slow strain (like Silly Putty). This can be seen in Figure 2-1, except that the brittle-ductile transition would be along the subduction-zone fault itself. Their model predicted that the next great earthquake would rupture the entire subduction zone from Canada to California, a magnitude 9 rather than an 8 (Figure 4-16, right-hand map).
Chris Goldfinger had long been an advocate of the smaller-earthquake hypothesis. However, his study of the Holocene turbidites convinced him otherwise. There is the same number of turbidites in Rogue River submarine canyon off southern Oregon as there are in the submarine canyons off the coast of Washington, which provided support for an earthquake of magnitude 9. However, further dating of turbidites showed that some turbidites were limited to the southern part of the subduction zone, evidence that some of the subduction-zone earthquakes were limited to the southern part of the subduction zone. This means that the recurrence interval estimated for subduction-zone earthquakes in southern Oregon and northern California would be shorter than for magnitude 9 earthquakes rupturing the entire subduction zone. Goldfinger suggests that the average recurrence interval for smaller southern Cascadia earthquakes has already exceeded the historical record.
Additional evidence for the most recent earthquake came from Japan.