Cascadia's Fault

by Jerry Thompson

  As the evening dragged on he tinkered with numerical codes he had developed to simulate the behavior of the massive waves that had smashed across Okushiri Island off the coast of Japan back in 1993 and was only half listening to CNN in the background. Suddenly a news bulletin caught his attention. There had been a seismic shock in the Indian Ocean. “The earthquake was small at first,” Titov recalled. “But it got bigger in front of my eyes.” Later reports said the magnitude could be 8 or even higher.

  Then came news that a tsunami had been generated off the coast of Sumatra and that people had been killed as far away as India. Titov was now riveted to the screen. “I saw the first reports of the deaths from the tsunami—the reports from CNN that fifteen people died in India. In India,” he repeated, astonished. “It was thousands of kilometers away from the source. So that definitely gave me some idea that it’s a huge event.” By now it was late night in North America—Christmas night—and December 26 in Sumatra and India. Titov relaunched his model and started looking for data from the Indian Ocean that might allow him to re-create the tsunami while it was still happening.

  Titov’s boss, Eddie Bernard, remembered the first call in the wee hours of the morning. “I was in bed and the phone rang and ABC News said, ‘Could you tell us about the tsunami?’ And I said, ‘Uh, I’ll have to go check my internet and email messages and I’ll get back to you.’” By the time Bernard joined Titov at the lab, the picture had changed radically for the worse.

  “Vasily went to work on the model and Shirley, my wife, and I answered the phones,” said Bernard. “The telephone calls were coming in, many more phone calls than we could ever address,” Bernard confessed. “I mean I was doing interviews in Australia, radio stations in India—all over the world—London. And we were trying to provide some graphical information to the broadcast media at the same time we were trying to educate people about what was actually happening. Because, you see, in the Indian Ocean they had never seen or experienced anything like this and many people didn’t even know what the word tsunami meant.”

  Titov, meantime, was hoping the codes he’d used for Japan and the North Pacific would work just as well for the coast of Sumatra, where the tsunami had been generated. Fortunately some research had been done relatively recently on the offshore subduction zones in that area and bathymetric grids were available for much of the Indian Ocean. There was very little information, however, about how big the first wave had been.

  “The only data we had available was a tide gauge record on Cocos Island,” Eddie Bernard explained, “which is south of the source in the Indian Ocean.” Titov used the reading from the Cocos Island gauge as a proxy for a deep-ocean gauge, estimated the tsunami height, and punched the number into his computer as a starting point for the model to begin running a simulation. “Vasily was able to take that information and invert it into his models and get a fairly accurate representation of the tsunami,” said Bernard.

  Titov’s math turned out to be pretty much bang on. When they ran the simulation it revealed things about wave behavior that no one had seen before. It showed, for example, that a tsunami could bounce off one island and hit the back side of another with even greater force.

  Titov, standing in front of a floor-to-ceiling map of the world’s oceans spread across the wall of a large conference room wall at PMEL, pointed to Sri Lanka, thousands of miles northwest of Sumatra. “So it came to Sri Lanka and hit hard.” Titov’s finger traced the leading edge of the tsunami to the southeastern beaches of Sri Lanka, where the first pulses crashed ashore with deadly force.

  “On the back side, it was protected,” Titov continued, pointing to the lee side of the island between India and Sri Lanka. “It was sort of shielded from the first wave. But then the wave bounced off the Maldives,” his finger followed the path across to the next neighboring island chain, southwest of Sri Lanka. “It reflected from here [the Maldives] and then hit the backside of Sri Lanka with much stronger waves.”

  And there was stunning home video footage to prove the point. A mound of water slammed against the seawall at a luxury resort, shot a geyser of white spume into the sky, and then surged across the pool deck, sweeping away everything in its path. On the backside of Sri Lanka.

  Titov’s model also illustrated how the train of monstrous swells would turn corners around continents and eventually hit beaches on the opposite side of the planet. Big mountain ranges at the bottom of the Indian Ocean steered the on-rushing tsunami in new directions, according to Titov. He shifted his hand to the southern Indian Ocean and pointed to an undersea ridge. “You don’t see them on the surface, but the wave does see them.”

  In a computer lab down the hall, he showed me how his model had replicated the motion. As the tsunami approached a ridge of very large undersea mountains, the wave began to bend and change trajectory. “The wave feels the shallow water and it slows down over this ridge,” Titov explained. The tsunami scraped along the edge of the mountain range and friction slowed the left side of the wave down. The right side—still in deep water—continued to move at a higher speed. The difference in speeds from one edge to the other caused the wave to turn. In essence the undersea mountains became a wave guide, warping the deadly swells in a new direction.

  But only up to a point. “If the mountain range turns sharply,” Titov continued, “the wave would not turn. It will leave the guide.” That’s exactly what happened at the bottom of Africa. The tsunami veered off a mid-ocean ridge and rebounded like an eight ball to the corner pocket—around the tip of the African continent, crossing from the Indian Ocean into the South Atlantic. An example of “very interesting physics,” Titov said, with barely controlled enthusiasm.

  Roaring past the Cape of Good Hope, the sea monster moved up and across the Atlantic, coming ashore and leaving footprints on the beaches of South America. “A tsunami generated in the Indian Ocean—in Sumatra, half the world around—turned out to be a meter in Brazil,” Bernard pointed out, with a shrug of amazement. “Fortunately it was at low tide, so it didn’t do any damage.” Sumatra’s biggest wave even showed up on tide gauges as far away as Halifax, Nova Scotia.

  The computer model predicted that the tsunami would circle the entire planet, and physical data from beaches and harbors around the globe confirmed the prediction. Then a bit of luck added another layer of confirmation. Two weeks after the tsunami, scientists at NASA notified Eddie Bernard and Vasily Titov that one of their satellites just happened to be overhead precisely when the tsunami was crossing the Indian Ocean.

  “This tsunami was big enough in the open ocean—about forty centimeters—that [the satellite] actually detected it,” said Bernard. Forty centimeters (15 inches) didn’t sound big to me until Bernard reminded me that most of the wave was beneath the surface, reaching all the way to the bottom. The ocean was several miles deep and yet the wave still lifted the entire surface of the sea another forty centimeters. From that perspective it was an enormous mountain of rolling water. “And that is exactly the kind of data that we needed to verify our model,” Titov said.

  “It was an incredible match,” Bernard continued. “So we were extremely happy.” He grinned and then added, “Vasily was of course beside himself because this was the first time we had seen an open ocean tsunami of this size. And to model it correctly was quite satisfying.”

  While it might sound callous to be satisfied about successfully creating a digital clone of a killer tsunami in a computer, what Titov and Bernard and dozens of other researchers around the world learned from the Indian Ocean will no doubt save lives the next time this happens. When it does, detection systems now in development will be tripped as soon as the tsunami begins to move. With the knowledge gained from putting Titov’s model to the test, it should be possible to say with some degree of certainty how that next monster will behave, what communities will be at risk, and how far up the beach the waves are likely to reach.

  As Eddie Bernard put it, “Scientifically we have a wealth of new informa
tion—I mean, unprecedented information—that will guide us and improve everything we do in this whole field. But socially I think the most important thing that’s happened is it’s raised the awareness of tsunamis through the whole world. People take tsunamis seriously now.”

  To underline his point, Bernard told us about the first real-time application of NOAA’s updated tsunami warning program in the aftermath of Sumatra. On November 15, 2006, a Kuril Island earthquake in the North Pacific generated a wave train big enough to create damage over a long distance. The swells began moving east across the Pacific, tripping the alarms on the deep-ocean warning buoys built and deployed by NOAA for just this purpose.

  When one of the pressure sensors anchored on the bottom of the Pacific detected the extra weight of a larger than normal wave passing overhead, it transmitted a signal to warning centers in Alaska and in Hawaii. Now, instead of knowing only that an undersea earthquake had occurred, NOAA personnel knew for certain that swells of a potentially dangerous size had been generated.

  Computers immediately ingested the data and spat out a prediction about where the waves would go and how big they would be when they got there. As the leading edge of the tsunami pounded across the Pacific from the Kuril Islands toward Crescent City, California, the computer produced a forecast. “There were ten waves,” said Bernard, “and the model picked this up and actually replicated it before the tsunami arrived in Crescent City. It predicted that number nine wave would be the biggest. And guess what—number nine wave was the biggest.”

  Although none of them was as large or as vicious as the killer from Sumatra, the power of the moving water stunned those who saw it first-hand. Grady Harris, a grizzled and weather-beaten fisherman we met on the docks at Crescent City, told us he had heard the warning from NOAA and desperately tried to get his fishing boat out of the harbor and into the relative safety of deeper water. He made it just outside the breakwater, then got caught in a twisting torrent of seawater.

  “In forty years on the ocean I’ve never seen that kind of a situation before,” said Harris. “It was like trying to drive the boat in a washing machine. It just turned the boat sideways, turned it—spinning it around.” He shook his head and stared across the dock, reliving the moment. “The awesome power of the water . . .”

  Harbormaster Richard Young saw the surge boiling through the harbor entrance. He and all the others standing on the docks that day were mesmerized. “Water came in so fast on that larger wave that it actually went over the top of the concrete floats.”

  “It broke up the docks and scattered boats all over the harbor,” said Harris. “There was several sailboats sittin’ up on the dock from just the force of the wave.”

  To put this relatively small tsunami into context, the largest surge was only six feet high (1.8 m) from peak to trough. Not even as high as some of the regular storm tides in Crescent City harbor. The difference, according to Richard Young, was the speed—the manic rush—of the incoming waves. “The fact is that we have six-foot and eight-foot and ten-foot tidal changes here all the time with no damage,” he said. “The difference is that they [the tides] happen in six hours instead of ten minutes.” The damage estimate for this small, non-fatal tsunami was nearly $10 million.

  On the positive side, NOAA’s ocean warning buoys and the new tsunami models, refined and upgraded by the Indian Ocean experience, had given scientists reason to believe they’d made a major breakthrough. Now when an undersea earthquake sets off seismic alarms, there will be information about wave generation as well. Emergency planners like Stephanie Fritts and Sheriff Benning won’t have to play guessing games with nature, wondering whether or not to issue evacuation orders to the citizens of Pacific County.

  Vasily Titov extended the thought. “We cannot say when the next big earthquake is going to happen, but from the moment a tsunami is generated our models can actually tell you pretty well what happens next. How high the tsunami wave is going to be at the coastline, how big the impact is going to be at a particular location. The only thing we have to know for that is the actual measurement of the tsunami wave.”

  The success of the tsunami models, even though they’re “never going to be perfect,” he said, “makes you feel that—gosh, all this mathematics that you learned in high school and the math at university can actually pay off and save lives. That’s a pretty amazing feeling. Dry mathematics applied correctly—it can save lives.”

  Chris Goldfinger, the marine geologist from Oregon State University, agreed with Titov. While still at sea off the coast of Sumatra, collecting mud-core samples in order to study how this earthquake happened, he couldn’t avoid thinking of home. With a mud core sliced open on his laboratory workbench and the sea gently heaving beneath the hull of the Roger Revelle, Goldfinger drew the connection.

  “I have to admit mud is not very exciting to look at. It just looks like some sand and some mud. But now every time I look at these cores, I see that giant breaking wave . . . It sort of brings home what these things really are,” he said. “Even though we work in the theory of all this, to actually see that for real was just stunning, and horrifying ... I think everyone’s pretty mindful of the reason we’re here,” he continued, with members of the Sumatra science team looking on, “and that maybe some of this research might help in some way.”

  He paused for a moment and then tied the two stories together. “The same thing applies to Cascadia. I live there. And every time I drive to the coast, I see towns that are not long from now going to be under water from the next tsunami . . . The Cascadia Subduction Zone earthquake and the tsunami that’ll come with it will be virtually identical to the one in 2004 in Sumatra. It’ll dwarf 1906 [in San Francisco]. And Katrina. It’ll be many dozens of Katrinas all at once. Coastal towns from northern California to Canada will be virtually wiped out. And there’ll be significant damage in all the coastal cities along there as well—all at the same time.”

  Goldfinger agreed that knowing what Cascadia’s fault is capable of had utterly changed his perspective about living on the edge of a continent. “It’s a little hard to go to the beach and just hang out there and enjoy it.”

  Garry Rogers at the Geological Survey of Canada told me that what happened in Sumatra should have come as no surprise—and yet it did. “To all the scientists it was obvious that that’s the kind of thing that happens,” Rogers said, matter-of-factly. “It was perhaps more severe in terms of death toll than most of us would have thought. But what it did, I think, to the general public—and what we tried to translate to the general public—is that Sumatra is Cascadia. Those zones are the same size; what happened in Sumatra is what happened in Cascadia in 1700, and many times before that.”

  Now Rogers’ voice was rising. “That’s what we’re talking about, big shaking. It damages a lot of buildings. And then a tsunami comes in on the shore. And we need to be able as a society to deal with that situation.”

  CHAPTER 22

  The Next Wave: Thinking the Unthinkable

  To me the good news is that people living on the edge of North America are finally beginning to respond. Outport communities on the west coast are taking advantage of new flood maps that show how high up the beach Cascadia’s wave is likely to reach, what parts of town will be inundated, and where the safe zones on high ground should be located. Tsunami evacuation routes have been mapped and signs posted. People are going to night classes to learn what they need to know in order to survive. Evacuation drills are being held. And in some cases schools, hospitals, and other vital public buildings are being reinforced or relocated.

  Scientists from at least half a dozen universities in the United States and Canada are creating or updating numerical models that use detailed bathymetric maps of the local harbors and offshore terrain to predict much more precisely how far uphill and inland the turbulent muck is likely to travel. They’re doing on a local level what Vasily Titov’s model did on a global scale. Computer simulations of Cascadia’s tsunami have been generate
d for the city of Victoria and for the fishing village of Ucluelet on the west coast of Vancouver Island, for Cannon Beach and Seaside on the Oregon coast, and dozens of others are in development.

  Not every community in harm’s way has a computer model to map the inundation zone. At least not yet. Some towns such as Port Alberni, which was hammered by the Alaska tsunami of 1964, don’t have detailed bathymetric grids to work with because hydrographic survey ships are expensive to operate and government research budgets have been slashed. Rather than wait for public funding to catch up with grim reality, Port Alberni’s emergency planning team took matters into their own hands—literally.

  Knowing from experience how waves coming in from the coast can get amplified by the narrow canyon walls of a fjord like the Alberni Inlet, local planners asked the experts at the Pacific Geoscience Centre for their best guess about the height of a seawater pulse coming from Cascadia’s fault. Then they took out a standard topographic map of the downtown core along with a red marking pen.

  Bob Harper, the head of emergency planning for the city of Port Alberni at the time, walked me through the exercise. “The best advice that we’ve received so far from the scientists is—because of the funneling effect in the Alberni Inlet—that we can expect somewhere in the twenty-meter range of inundation,” said Harper. “So that’s twenty meters high . . . Not twenty meters in, but twenty meters up.”

  “That means a sixty-foot wall of water?” I asked, trying to imagine the downtown waterfront.

  “Yes,” he said quietly. A technician working with Harper began tracing the contour in bright red ink across the heart of downtown. “There’s a band of residences along the riverside here,” his hand swept across the map following the contour around the harbor. It was clear that most of the central business district, all of the industrial waterfront, the pulp, paper, and lumber mills, would be inundated.