Cascadia's Fault (48 page)

The scientists who made those projections quickly admitted that “the historic record” was indeed where the trouble started.Two months after the Tohoku disaster, five prominent Japanese seismologists wrote a post-mortem (published in
Nature
) to explain what they had learned from the tons of new data gathered in the aftermath. One of the authors, James Mori of Kyoto University, was remarkably candid in his assessment: “The Tohoku earthquake came as a frightening and disheartening surprise to Japanese seismologists.”

Most experts had thought it should be possible to predict the locations and approximate magnitudes of these monstrous plate-boundary earthquakes. They gathered the existing geologic data and did the math. But it turned out that roughly four hundred years of historic records covered “too short a time period to be a reliable guide.” Nothing larger than a magnitude 7.5 had been seen since 1923. There was no written record of anything larger than an 8.5 since the seventeenth century. A long-term forecast for the Tohoku region issued by the Japanese government in 2002 estimated an 80 to 90 percent probability that the area would have an earthquake of magnitude 7.7 to 8.2 sometime in the next thirty years. The probability “of a magnitude 9 earthquake affecting a 400–500 kilometre [250–310 mile] area was not specifically mentioned,” wrote Takeshi Sagiya of Nagoya University. “As a member
of the working group involved in the evaluation, it is with great regret that I reflect on the causes of this failure.”

For several years there had been discussion and debate about how much pressure was building up along the subduction zone. Japan's land-based network of GPS tracking devices had revealed that the tectonic plates were converging at an overall rate of nearly four inches (9 cm) per year. A rather frightening speed when it comes to potential earthquake generation, especially when you think of four inches multiplied by hundreds of years. It seemed that there was more stress going
into
the rocks along the fault than there was coming
out
during earthquakes. The numbers didn't add up. The term “slip deficit” came into use. Slippage along the fault during all known earthquakes in historic time had not been enough to bleed away the huge amount of pressure that should be there if two giant slabs of the earth's crust were converging this quickly.

As far back as 1996, Yasutaka Ikeda of the University of Tokyo had warned that “the strain accumulated in the last 100 years at abnormally high rates is likely to be released by slip on the megathrust at the Japan Trench, which would produce big earthquakes with magnitude 8 or greater.” And
greater
is exactly what happened on March 11, 2011.

The energy released during the Tohoku quake was astonishing. “The amount of strain release is nearly an order of magnitude larger than what we have seen in other megathrust earthquakes,” wrote Hiroo Kanamori of Caltech in
Nature
. “The strain must have accumulated in this zone for nearly 1,000 years.” And clearly not all of it had been released by smaller quakes during that time.

Apparently the same thousand-year timeline was true of giant tsunamis. Waves as large or larger than Tohoku's had hit Japan's east coast before. Scientists from Tohoku University in Sendai, led by geologist Koji Minoura, dug trenches and took core samples of sand deposits from a tsunami triggered by a shockwave, now known as the Jogan earthquake, that occurred on July 13, 869. The waves penetrated more
than two miles (4 km) inland, so the Jogan quake produced a tsunami as large or larger than the one we saw in 2011.

But that's not all. The researchers found at least two more sand sheets buried underneath the Jogan layer, each of them about the same thickness as Jogan's and each representing another huge wave. Which means that the same astonishing release of seismic energy has happened three times in roughly three thousand years, according to the findings of Koji Minoura and his colleagues. Their paper, published back in 2001, concluded that large-scale tsunamis happen every 800 to 1,100 years. They pointed out that more than 1,100 years had passed since the Jogan event and that therefore “the possibility of a large tsunami striking the Sendai plain is high.” Thus, they sounded the alarm ten years before the Tohoku disaster of 2011. But like Ikeda's caution in 1996 about the likelihood of high stress values on the fault triggering large earthquakes, their words of warning somehow failed to reach the right people.

Calculations made by various agencies to forecast the next big quake and tsunami on Japan's east coast did not include Jogan or the other, more distant events of the prehistoric past. Perhaps with so many smaller jolts in recent centuries, those who did the math simply assumed they had enough data to figure out what was “most likely” to occur on a purely statistical basis. Smaller quakes do happen more often, and therefore they are thought more likely to occur.

Three main events in
recent
history—in 1896, 1933, and 1960—had convinced officials that their seawalls and tsunami gates would be strong enough and high enough to defend the coast. They had enough data—geologic and historical—to form an image of the maximum credible event (MCE), the earthquake and tsunami most likely to occur along this part of the coast. Something like a magnitude 8 must have seemed like the logical thing to plan for.

More than two dozen ancient tsunami marker stones had been found in northeastern Japan. One marker stone in the hills above Aneyoshi
Bay, showing how high the water from the Sanriku tsunami of 1896 had reached, had an inscription suggesting the village be rebuilt above that level. Considering that 22,000 people died in the 1896 tsunami, the survivors did decide to relocate their community above the high-water mark. And fortunately, in 2011 the highest Tohoku wave fell thirty-three feet (10 m) short of the 1896 marker. The people who had resettled above the stone were indeed safe. However, a survey by the ITST investigators showed that at least 20 percent of the other marker stones in the area had been overtopped or washed away.

While the Tohoku tsunami was the highest in more than a thousand years, the Jogan wave pushed seawater at least 650 feet (200 m) farther inland. So the Jogan event may have been even larger. The consensus about how enormous these waves could get was simply wrong: the most likely event was not the worst-case scenario.

“For people aware of paleo-tsunami work, it was not a huge surprise,” said Lori Dengler. “But that information had definitely not been incorporated into the thinking of either the seismology community in Japan or the mainstream hazard folks. And it certainly had not been included in any of the planning for nuclear power plants.”

Clearly some of the basic assumptions about what to expect from a subduction zone were wrong as well. Until Tohoku many seismologists thought old rocks like those along the Japan Trench were not likely to stick together for long periods of time. Only hot, fresh rocks in young subduction zones like Cascadia's fault were supposed to get stuck for hundreds of years and generate monster quakes. This is why computer models assumed the shallow part of the Japan Trench—where the fault comes to the surface of the sea floor—was not locked and therefore not likely to generate a monster shock. The realization that this idea is wrong has huge implications because the same assumption has been applied to many other subduction zones around the world.

“There are two big lessons from Tohoku for my nickel,” Chris Goldfinger observed. “The first is that you can't base your estimates of
hazard on short instrumental and historical records. Their giants come at thousand-year intervals, roughly, so even a thousand-year history was inadequate.

“The second lesson is that all previous seismological theories that relate simple things like plate age and subduction velocity to the maximum size of an earthquake . . . are now out the window. This is a very big deal, because these rather poorly justified models have been used to predict which systems globally are giant earthquake producers. And many regions were taken off the list on this basis. Tohoku was one of them. Now, all of them have to go back
on
the list and be reconsidered. Places like Java, Peru, New Zealand, et cetera. These are all places that haven't really worried about magnitude 9 earthquakes. Just as northeastern Japan didn't worry about them.”

If any of this sounds familiar, recall that here in North America, scientists for years underestimated the potential for monster quakes from Cascadia's fault for exactly the same reasons. They had not looked far enough back into the geologic past for evidence of much bigger ruptures. But the evidence was always there. And in another parallel with Japan's tsunami history, it turns out that Cascadia's fault has also left tsunami footprints that were much larger than all the others in recent history.

Several years ago, when he first showed me the mud cores at his lab in Corvallis, Oregon, Goldfinger drew my attention to a sample from the great Cascadia earthquake of 1700. A magnitude 9 megathrust jolt had triggered an underwater mudslide along the continental slope, which dumped a significant swath of turbulent sand and gunk on the deep-sea floor. It was plain to see, in the sliced-open core sample, a darkish gray-brown layer nearly six inches thick.

Then Goldfinger pointed to another sample that was huge by comparison. “What's this?” he asked rhetorically, with a nervous laugh. Evidently an even larger quake and tsunami happened roughly 5,800 years ago, and it may have had
three times
as much energy as the magnitude 9
of 1700. Worse yet, there was another giant just like it roughly 8,800 years ago.

As monstrous as the 1700 Cascadia quake was—it wiped out native villages along the West Coast and sent deadly waves all the way to Japan—it was not the worst-case scenario for North America. It was only an “average-size” disaster. The much larger mega-tsunamis from Cascadia's prehistoric past have so far been ignored by the experts who estimate risk and plan for the future—a carbon copy of what just happened in Japan.

“[The Tohoku quake]at first seems like a rogue earthquake,” said Goldfinger, “but in geology if you have something happen once, you immediately know that it has probably happened a thousand times. And you just didn't know about it.”

Well, now we do know. The next obvious question is, what can we do about it? How well can any nation defend itself against a “thousand-year event”? Robert Geller, an American geologist at the University of Tokyo, has commented that “no conceivable economically realistic countermeasures could have precluded damages from a magnitude 9 earthquake and the resulting tsunami.” But the civil engineers who inspected the damage agreed that well-built buildings saved lives. The seawalls may have failed, but even more people would have died if the barriers hadn't been there to slow the water down. Although devastation was widespread and the outcome tragic, strict construction standards did pay off.

Millions of people around the world saw huge high-rise towers in Tokyo swaying back and forth on live television—yet none of the towers collapsed. It was mostly smaller wood-frame houses that got swept away by the tsunami and destroyed. The EERI and ASCE reports noted that even in the worst-hit areas of northeastern Japan, the larger reinforced-concrete buildings that survived “did not appear to have significant structural damage from either the earthquake or tsunami . . . This provides some encouragement regarding the potential resilience of
larger modern buildings with robust seismic designs and scour/uplift-resistant foundations.”

The same was mostly true of the massive 2010 quake in Chile. Seismologist Garry Rogers of the Geological Survey of Canada described both recent events as “success posters for modern building codes.” The point he stressed, however, is that seismic regulations are primarily designed to save human life, not to guarantee that a building will survive without a scratch. “Yes, there was damage to some modern buildings in Chile (and one collapse), but building codes are ‘life safety' codes, not ‘no damage' codes,” Rogers explained. “They did the job they were supposed to do.”

In the aftermath of the 2010 Chile quake, Lori Dengler and many other scientists felt a sense of optimism. Most people there escaped to higher ground as soon as the shaking stopped. They knew what to do and did not wait for anyone to tell them; they just did it.

“Chile
got
it,” Dengler told me emphatically. Their story “was a success from almost any perspective. In Chile there were problems, but we all kind of felt like, yeah, they did a really good job. Engineering codes work. Education works.”

But in the next breath she added, “That confidence that a lot of us had for Chile definitely evaporated in Japan. It's very sad . . . The one thing I'm absolutely sure of is that the next major event will surprise us in yet another way we haven't really thought about.” Considering the parallels and implications for California, the Pacific Northwest, and the west coast of Canada, she added, “We need to be really humble in this business . . . That loss of confidence—it's significant . . . And it will happen to us.”

But it seems Professor Dengler's spirit never stays down for long. She's the one who told me the wonderful story about the people of Langi Village on Simeulue Island who managed to escape the Indian Ocean tsunami of 2004 because their oral history included a strong memory of the last time the earth shook. The village elders knew exactly
what to do. They moved every man, woman, and child to a prearranged sanctuary on higher ground. Not a single life was lost—even though they had only eight minutes before the first wave arrived.

Dengler writes an occasional column for the
Times-Standard
, a newspaper in Eureka on the northwest coast of California. After thinking about the risk that we who live in the shadow of Cascadia's fault must face, she came to pretty much the same conclusion I have. She wrote that “the North Coast is a special place and I feel extremely fortunate to live here. We have some of the highest volunteerism rates in the nation, we have an amazing entrepreneurial spirit, and we lead the nation in our tsunami preparedness efforts. It's not an accident or coincidence. It is because of our natural setting, not in spite of it. It creates self-reliance, and we should be thankful for that. If the Simeulue Islanders can do it, so can we.”