University of Hawai'i University Relations Media & Publications Honolulu, HI 96822	(808) 956-8856 Telephone (808) 956-3441 Facsimile ur@hawaii.edu E-Mail
For Immediate Release:	September 7, 2000
Contact: Gerard J. Fryer, Associate Professor, Hawai'i Institute of Geophysics & Planetology, University of Hawai'i at Manoa, (808) 956-7875, gerard@hawaii.edu Costas E. Synolakis, Professor of Civil and Environmental Engineering, University of Southern California, (213) 740-0613 costas@mizar.usc.edu Emile A. Okal, Professor of Geological , Sciences, Northwestern University, (847)491-3194, emile@earth.northwestern.edu

Scientists puzzle over 54-year-old tsunami; size of waves underestimated

Years after a high-tech warning system was set up to protect Pacific nations from tsunamis, scientists are still wondering about the devastating tsunami of April 1, 1946, the reason the system was set up in the first place. Gerard Fryer of the University of Hawai'i, Costas Synolakis of the University of Southern California and Emile Okal of Northwestern University have just completed a survey of the tsunami's effects in the Marquesas Islands, the northernmost group in French Polynesia.

"Because of the warning system it's unlikely that another 1946 can hit us unawares," Fryer says. "But these measurements in the Marquesas show that we might seriously underestimate the size of the waves. We have to work out exactly what happened in 1946 so that the next time something like this happens, emergency managers around the Pacific can give the appropriate warnings."

Okal adds, "The one thing we know for certain is that there will be a next time."

The tsunami that killed 167 people stemmed from a magnitude 7 earthquake that shook Alaska's Aleutian Islands April 1, 1946. In the wilderness of the Aleutians, such earthquakes are seldom a problem, but this one was unusual. Waves more than 100 feet high killed five people in Alaska, then raced across the Pacific to kill 159 people in Hawai'i, one in California and two in French Polynesia. The tsunami has remained an enigma ever since: The waves were far too large for a magnitude 7 earthquake. The scientists hoped that by getting information from the Marquesas they could somehow solve the enigma.

"We were amazed by what we found," Fryer says. "The waves of 1946 were far larger in the Marquesas than they had been in Hawai'i, even though the Marquesas are a lot farther away. We had read vague reports of waves running up on shores in the Marquesas to a height of 30 feet at one or two restricted locations. What we actually found was the waves averaged 20 feet, and reached up phenomenally high ­ as much as 65 feet ­ in narrow valleys. It looks like a very narrow beam of extremely high waves was projected across the Pacific. The largest waves just missed Hawai'i but hit the Marquesas dead center."

The scientists' findings take them back to the drawing board:

"We know the whole thing started with an earthquake in the Aleutians, but an earthquake alone cannot explain these waves," Fryer says. "We are pretty confident that the earthquake shook loose a huge submarine landslide into the Aleutian Trench; we even have a good idea of where the landslide was and how big it was.

"But when we build a numerical model of the tsunami generated either by the landslide or by the earthquake, we get waves in the Marquesas that are too small, and we fail to reproduce the leading depression that was observed. Perhaps the tsunami was generated by some combination of landslide and earthquake, but it's clear that we don't yet know enough about the source."

"Fortunately," Synolakis says, "the tsunami of 1946 was so large and affected so much of the Pacific that we should be able to get similar information from other locations. We already have information from Hawai'i, from the west coasts of North and South America and from Japan. Now we hope to get a broad enough spread of information to be able to work backward from the observed wave heights to figure out exactly what the source was. Once we figure out the source, we can better understand what risks it poses for the Pacific Coast of the U.S. and all other Pacific coastlines."

To get the information the team went from village to village by four-wheel-drive truck, boat and helicopter to interview eyewitnesses and to look for evidence of tsunami damage. "Everywhere we went we were received graciously by silver-haired men and women, some bed ridden, some in wheelchairs, but all happy to relate their experiences," Fryer says.

What emerged was a surprising and very consistent story: In the early afternoon of April 1, 1946, the ocean retreated, exposing rocks usually covered even at low tide. Hissing, the ocean came back and flooded ashore. Then the ocean retreated a great distance, as much as 100 yards in some bays, before roaring back faster and much higher. This time, when the water receded, it carried with it trees, homes and livestock. Most people had already fled to high ground. The next wave, the third, was immense. Then the phenomenon slowly died away, although the sea was still slowly oscillating at sunset. The next morning coasts were littered with branches, coconuts, broken remains of boats and houses and the bodies of thousands of stranded fish. There had been two deaths: A woman and her baby were drowned at Tahauku on the island of Hiva Oa, where a village was washed away by a 50-foot rise in sea level that flooded more than half a mile inland.

"The population of the Marquesas is concentrated in the valley bottoms along the shoreline, exactly where the waves were the largest," Okal says. "But the tsunami arrived in the middle of the day, when people were up and about; it began as a fall in sea level, and it made a lot of noise. Even so, they were lucky: The first wave was not the largest. The people had time to escape, and they took heed of the warning signs."

Synolakis adds: "It is also clear from their stories that the waves did not break. There was a rapid flooding of the land, but nobody was overtaken by a wall of water."

Sometimes the team was shown hard evidence, such as a wrecked church or a coral boulder left by the waves; more often they had to work with information like "The water got just beyond the big mango tree next to the river." If the stories seemed credible, the team would measure the maximum height reached on shore using standard surveying gear and would also calculate the distance from shore using the global positioning system.

Fryer explains the interest in the Marquesas: "Most Pacific islands are protected behind offshore reefs," he says. "The Hawaiian Islands and the Marquesas are unusual in having almost no offshore reefs, so they are very prone to tsunami attack. We already have abundant information about 1946 for the Hawaiian Islands; the Marquesas were the obvious next step."

Even so, the Marquesas survey happened by accident. "Last year, we made a quick-response survey of the southern island of Fatu Hiva to investigate a local tsunami caused by collapse of a cliff," Fryer says. "While we were there, we asked about other tsunamis. We were amazed at how much people remembered about 1946, so we decided we should come back to collect that information."

Synolakis was intrigued by the fact that the people in the Marquesas have a special word for a tsunami, taitoko. Interestingly, while all the older people knew this word, few of the younger did. "It is telling," he says, "that most of the surviving eyewitnesses were children when the 1946 tsunami hit and had heard of tsunamis from their parents, but since no large tsunami have hit since then, they did not pass this information and knowledge to their children. "

"We came back this year hoping to get one or two measurements from each of three islands," Okal adds. "But everyone was so helpful and cooperative that we ended up with more than 40 measurements from 25 coastal villages on five islands. Together with measurements we made on Fatu Hiva last year, we now have a clear understanding of how the tsunami of 1946 affected each of the six populated islands of the Marquesas.

"Now we have more questions," Fryer says. "Big waves should have hit Pitcairn and Easter Islands; how big were they? The British Antarctic Survey tells us a hut was destroyed by the waves. How big were the waves in Antarctica? We have more work to do."

Fryer, Okal and Synolakis were accompanied on the survey by Gerard Guille and Philippe Heinrich of the French Atomic Energy Commission (CEA) and by Daniel Rousseau of USC. The survey was funded by CEA and by the U.S. National Science Foundation.

Note to editors: Photographs of the survey are available. Videos of the interviews (mainly in Marquesan and French) are available. For maps and digital pix, please refer to www.usc.edu/dept/tsunamis.

-UH-