Princeton University - Council of Science & Technology

POPE PRIZE

Sleepless in Seattle

While leafing through reams of paper documenting past earthquakes, Kenji Satake, Kunihiko Shimazaki, Yoshinobu Tsuji, and Kazue Ueda, four Japanese geologists, noticed that on the night of January 26, 1700, a tsunami flooded homes and inundated rice paddies along a 1000 kilometer stretch of Japan's Pacific coast. Remarkably, all the waves crested between two and three meters. Intrigued, the geologists struggled to determine what generated this unusually long and uniformly high wave. What they found has made whole cities nervous.

At first, Satake and his group of geologists looked for a local earthquake source for the tsunami. But no records exist of a Japanese earthquake on that day. Moreover, any local earthquake would have created a tsunami that dissipated in height as it moved up or down the coast and away from the source region. Similarly, a near-shore typhoon or cyclone would create waves that dissipated in height as they moved up and down the coast. Only a very distant event could create such long and uniform waves. Yet, typhoons and cyclones could never be large enough to create waves that would travel such great distances and retain a height of one meter, much less three meters.

At a loss to explain the tsunami of 1700, the geologists put their reading glasses back on and scoured their history books. If they could determine the sources of any similarly long and high tsunamis in Japan's past, they would have a place to begin their investigation. They found three such tsunamis, all coming from similar sources: very and very large earthquakes. The 1960 magnitude 9.5 Chilean earthquake generated a tsunami that traveled 24 hours across the Pacific Ocean and killed 140 people in Japan as it washed fishing boats ashore and pounded homes and businesses. Weaker tsunamis from the 1964 Alaskan (magnitude 9.2) and 1952 Kamchatkan (magnitude 9.0) earthquakes also flooded the length of the Japanese Pacific coast. The Japanese geologists had found their answer but only part of it. They knew that a large and distant earthquake had generated the tsunami, but they did not know where this earthquake had

As a logical first step in their search for the earthquake, they looked to Chile, Alaska, and Kamchatka as possible source regions. South Americans had recorded earthquakes before, around, and after 1700, but they had no records of an earthquake in 1700. Similarly, settlers that moved into Kamchatka in the 1680s did not record an earthquake until 1737. Alaskan Inuits, a people with a strong oral tradition, have no tales of an earthquake around this time. If an earthquake had occurred in 1700 (and it would have been a large one), the inhabitants of these regions should have recorded it.

Yet, the geologists needed more than the absence of earthquakes in the historical records to rule out Chile, Alaska, or Kamchatka as source regions for the Japanese tsunami of 1700. They needed more concrete scientific evidence, and they found it in two phenomena of tsunami propagation: dissipation and directivity. Dissipation says that regions closer to the tsunami source will receive larger waves than regions farther away. Imagine the ripples that form when a rock is thrown into a pond. As the ripples move away from their source, they dissipate and get smaller. The same holds for tsunamis-they become smaller further away from their source. Directivity, the second phenomenon, says that earthquakes will produce the largest waves in the direction that they thrust. When one piece of the earth's crust pushes up over another piece, the thrusting block of crust pushes forward any water in the region. Anything in line with the direction of thrust will receive large waves. Anything off to the side will receive smaller waves. One can experience directivity by walking full circle around someone who is talking. The person's voice will be loudest when one stands directly in front of the person-in-line with the direction of the forward-thrusted air.

As it turns out, dissipation and directivity rule out Chile, Alaska, and Kamchatka as possible source regions. The Kamchatkan fault zone sits north of Japan and thrusts to the southeast. Due to dissipation, a Kamchatkan earthquake could not produce waves of uniform height along Japan's Pacific coast. The Kamchatkan earthquake of 1952 produced waves two meters high in northern Japan but only 0.2 meters high in southern Japan-the waves lost their energy. The Alaskan fault zone, far to Japan's northeast, thrusts to the southeast, away from Japan. Since the fault does not thrust towards Japan, it could not produce two- to three-meter high waves in Japan. The waves that lapped up against Japan's northern coast after the great 1964 Alaskan earthquake (magnitude 9.2) crested at only 0.3 meters in height. Geologists do not believe that an earthquake could occur in Alaska large enough to create three-meter high waves in Japan. A Chilean earthquake could not have produced the uniformly high Japanese tsunami of 1700 either. The 1960 Chilean earthquake did produce waves six meters high in Japan, but due to directivity and dissipation the waves ranged from the six meters in northern Japan to less than two meters in southern Japan. With these most likely contenders for the source region thus ruled out, Satake and his team of geologists had to look elsewhere.

Throughout the entire Pacific, only Chile, Alaska, and Kamchatka have experienced earthquakes large enough to generate the Japanese tsunami of 1700. Only one other region, the U.S. Pacific Northwest, has the right geologic setting-a subduction zone-to create such earthquakes. The theory of plate tectonics says that the surface of the earth is broken into many pieces such that it resembles a fractured eggshell. In constant but very slow motion (one to two inches per year), the plates crash against, slide along, and pull away from each other, building mountains, forming rifts, and causing very large earthquakes. In a subduction zone, two plates collide, and the plate composed of the lower density rock will thrust up over the plate of higher density rock. The high-density plate will sink down towards the middle of the earth.

In the Cascadia subduction zone of the Pacific Northwest, the North American continental plate thrusts up over the Juan de Fuca plate of the Pacific Ocean. The plates meet about 30 kilometers off the coast of British Columbia, Washington, and Oregon, and the zone of contact stretches over 1200 kilometers north to south. However, the Pacific Northwest has experienced only a few very minor (magnitude <5.5) earthquakes since Europeans first settled the region 200 years ago. To explain the lack of earthquakes, especially large ones, in such an apparently earthquake-prone environment, geologists hypothesize that the zone of contact between the two plates may be so hot that the rocks may have partially melted and now absorb the slip between the plates through ductile deformation. Earthquakes result from the brittle fracture of rocks, not from the ductile flowing of rocks. With their hypothesis about ductile slip and the 200-year dearth of earthquakes, geologists have no reason to believe that a great earthquake occurred in Cascadia in 1700.

Regardless, Satake and his fellow researchers felt that the tsunami had to have come from this region. Using a computer model of the Pacific Ocean, the ocean floor, and any Pacific coastal landmasses, they simulated tsunami propagation from Cascadia across the ocean to Japan. To their delight, they found that a magnitude 9 earthquake in the Pacific Northwest would produce a very long, two-meter high tsunami in Japan-just like the one of 26 January 1700.

The geologists had ruled out every other source for the tsunami-an earthquake had to have occurred in Cascadia and created the large wave. But if it had, physical evidence -of "' -c earthhquake should exist in the Cascadia region. Smiling, the geologists already knew that this evidence existed. In a 1995 letter to Nature, American geologists had presented research that involved radio-carbon and tree ring dating of vegetation that they believed had been killed by an earthquake. They found dead Sitka spruce, Pacific silverweed, and Lynby's sedge that had subsided and then been buried by a thick layer of sand. They believed that an earthquake caused the subsidence of the land and also generated a tsunami that deposited the sands on the sunken vegetation. Their dating experiments revealed that the vegetation died in the early 1700s. Satake et al's earthquake occurred in the early 1700s. Other American geologists have also found many layers of turbidites on the continental slope off the coast of the Pacific Northwest.

Turbidites form when the sediments from a submarine landslide collect in channels and solidify. They have a very jumbled and messy look, hence their name. In channels up and down the coast of the Pacific Northwest, the sequences of turbidites layers match each other nearly perfectly. This suggests that the landslides that formed the separate turbidites were triggered simultaneously, most likely by a great subduction zone earthquake. Geologists have determined that new turbidite sediments collect every 410 years.>The earthquakes that cause the landsliding must also occur every 410 years.

Satake and his partners had put all the scientific pieces of the puzzle together. Their tsunami data matched perfectly with the turbidite data, the subsidence data, and the radiocarbon and tree ring dating data. When joined, all the pieces painted a picture of a great earthquake on a winter's night in January of 1700. But the geologists wanted more life and color in that picture.

Flipping through the notebooks and journals of anthropologists, Satake and his team came across legends and myths of the Yuroks, the Makahs, and the Pachena Bay people-Pacific Northwest Native Americans-that greatly resembled the scenario they had developed. The Yuroks of northern California and southern Oregon believe in the god Earthquake whose every step makes the ground shake. They speak of Earthquake going north and making the ground "quake and quake again and quake again." With each step and each quake, the ground sinks and allows water to flow "all over." The Makahs of Cape Flattery, Washington, remember when the ocean receded for four days straight while building up into a large wave that crested over four hundred meters high and left only the mountaintops dry. The few survivors had put their belongings in canoes and floated off with the wave. They found themselves deposited together tens of miles from their homes when the wave subsided. In this new place, they reestablished their village, and they still live there today. The Pachena Bay people, also of Cape Flattery, recount a winter night when the land shook. Crawling out of their beds, they were barely awake and could not get to their canoes before a big wave smashed into the beach and drowned them all.

Although the legends and myths contain some exaggerations, such as the four hundred-meter high wave, the geologists saw that they lent both color and credence to the picture they had painted. The legends and the geologists' scenario both talk of the quaking and sinking of the ground and the flooding caused by large waves. Since anthropologists have dated these legends to between 1600 and 1800 AD, they match well the scenario for a Cascadia earthquake in 1700.

If we accept that the last C&,;cadia earthquake occurred in 1700, and that Cascadia earthquakes come every 410 years, and that these Cascadia earthquakes are large devastating earthquakes, we have to worry about Seattle and Portland and every small ,own in between. We know that the ground shakes in San Francisco and in Los Angeles, and we prepare for it. If we do not prepare for earthquakes elsewhere, the ground may shake where we least expect it.

References

Satake et al. "Time and size of a giant earthquake in Cascadia inferred from Japanese tsunami records of January 1700." Nature, Vol. 379, 18 January 1996, pp.246-248.

Clague, J.J. "Early historical and ethnographical accounts of large earthquakes and tsunamis on western Vancouver Island, British Columbia." Current Research of the Geological Society of Canada, 1995-A, 47-50.

Heaton, T. and Hartzell, S. "Earthquake Hazards on the Cascadia Subduction Zone." Science, Vol. 236, 10 April 1987, pp. 162-168.

Kroeber, A.L. Yurok Myths. University of California Press, Berkeley, 1976.

McCaffrey, R. and C. Goldfinger. "Forearc Deformation and Great Subduction Earthquakes: Implications for Cascadia, Offshore Earthquake Potential." Science, Vol. 267, 10 February 1995, pp. 856-858.Nelson et al. "Radiocarbon evidence for extensive plate-boundary rupture about 300 years ago at the Cascadia. subduction zone." Nature, Vol. 378, 23 November 1995, pp. 371-374.