How Alaska’s quake changed the Northwest’s outlook on disaster
Downtown Anchorage in the aftermath of the earthquake. Credit: U.S. Army
Fifty years ago, on March 27, 1964, the Pacific plate scraped under the North American plate beneath Alaska's Prince William Sound, triggering the second-largest earthquake ever recorded. In Anchorage, the Turnagain Heights neighborhood slid into the salt water of Knik Arm. A great wave triggered by landslides destroyed the town of Valdez — after the ground on which the town stood had already liquified — and a tsunami triggered by the quake itself killed 122 people, streteching as far south as Crescent City, California.
But the 9.2 magnitude quake itself was in Alaska; in most people's minds, tsunamis and big quakes were things that happened somewhere else, not the Pacific Northwest. Plate tectonics wasn't something that happened at all.
Obviously, a lot has changed. Valdez rebuilt across the harbor on higher ground that now holds the oil port from which the ill-fated Exxon Valdez set sail just 25 years ago in 1989. Low-lying coastal communities now sport tsunami escape route signs, although that is less reassuring than it might be if you're somewhere with no high ground. And it has become commonplace to explain earthquakes as the result of tectonic plates grinding together.
Even then, the idea wasn't new. A German scientist named Alfred Wegener had introduced it before World War One. But his theory and subsequent elaborations of it, attracted relatively little attention until the 1960s. University of Washington research professor of earth and space sciences emeritus Steven Malone explains that the Alaska quake was the first to be seen in the intellectual context of plate tectonics. It took place "sort of at the dawn of plate tectonics theory." And a lot of people didn't want to see it in that context.
"There was tremendous resistance to having that recognized as a subduction earthquake," says William Steele, director of outreach and information services at the UW's Pacific Northwest Seismic Network.
The year 1964 wasn't exactly the stone age. NASA had shot John Glenn into earth orbit two years before. And yet modern seismic monitoring was very new. Malone says that the Alaska quake was "the largest eqarthquake at that time that was recorded by [relatively] modern seismographs." There are better instruments now, but "the difference between [what existed in 1964] and not having any records for big earthquakes was huge." Malone says the resulting data "pointed us in a lot of new directions."
People had a lot to think about. "For the first time, the United States [outside Hawaii or the Aleutians] got clobbered," Steel says. It got people to reevaluate their assumptions. Until then, "I don't think they really thought that it could happen here."
In places, the Alaskan coastline sank 6 feet. The trees that had grown there above the high tide line were flooded. Eventually, they died.
By the late 1980s, researchers had found similar signs of subsidence in estuaries along the Washington coast. Gray stumps made it clear that whole forests of cedar had expired there at some point. Did their ghostly presence signal a quake and tsunami somewhere back in Washington/s history? And if so, when?
The trees didn't die in the tsunami. They died because the quake had fundamentally altered the land on which they grew. "What killed the trees in Alaska, and at Cascadia as well," explains U.S. Geological Survey research scientist and UW affiliate professor Brian Atwater, "was the lowering of land — the subsidence that allowed post-earthquake tides to cover the forest floors too often for the trees to survive."
The evidence of a great Cascadia earthquake and wave had been there all along, but until recently, no one had known how to interpret it. “The oral traditions of Cascadia’s native peoples . . . tell of flooding from the sea,” Atwater and colleagues write in "The Orphan Tsunami of 1700." People who lived along the coast “faced horrific tsunamis, like the one implied by the story of a sea flood that swept canoes into trees. Survivors then watched tides relentlessly cover their subsided, bayside fishing camps. Several archaeological sites tell wordlessly of the waves and tides that overran them. Each lies buried beneath tidal mud.”
The idea that a big quake had centered in the Pacific off Washington wasn't big news in Japan. "Japanese earthquake historians began recognizing written records of the 1700 tsunami in the 1930s or early 1940s," Atwater explains. "They confirmed these in the 1960s and found still more [records] in the 1980s." From the first, "they recognized that the tsunami probably was of distant origin."
In the 1980s, Japanese researchers found that a rice ship bound for Nakaminato, the Honshu port from which inland waterways led up to the old imperial capital at Edo, had gone down in big waves. The ship had been carrying 28 metric tons of rice, and local government had looked into the sinking. (The rice had been packed in bales, like those that appear in Hokusai sketches on the shoulders or backs of husky porters.) The Japanese researchers therefore had a precise date for the sinking. Was the all-but-spent tsumani that sank the rice boat the same one that had drowned trees along the Washington coast?
The trees themselves held the answer. "Tree rings dated by radiocarbon methods in the early 1990s assigned the most recent subsidence along the Cascadia coast to the decades between 1680 and 1720," Atwater says. "Tree rings dated by ring-width pattern matching narrowed that window, for trees along about 100 km of coast, to the dormant months between the 1699 and 1700 growing seasons." The Japanese records indicated a date in January, 1700. Bingo.
The Japanese data made it clear: the earthquake that dropped the Washington coast below the old high tide line had been huge, probably about the same size as the 1964 Alaska quake. "Evidence of flooding and damage in Japan helped geophysicists bracket the estimated earthquake size between magnitude 8.7 and magnitude 9.2," Atwater explains.
That makes it one of the strongest quakes ever. Subsequent research suggests quakes of roughly that magnitude happen every 300 to 600 years, and yes, if you do the math, 1700 was more than 300 years ago —–— hence the popular conviction that sooner or later, the Cascadia fault is likely to generate The Big One.
That day in 1964 when another big one hit Prince William Sound, 1200 miles to the southeast, Seattle's brand-new Space Needle moved visibly. Maybe it was trying to tell us something.