TsunamiChapter - K. Fuji collection, Pacific Tsunami...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: K. Fuji collection, Pacific Tsunami Museum. In 1946, a large tsunami arrived unexpectedly in Hilo Bay, Hawaii, nearly 4,000 kilometers from the Alaskan earthquake that caused it. Note the people and large truck at left. Harbor Waves TSUNAMI Tsunami, the Japanese name for “harbor waves,” are so named because the waves rise highest where they are focused into bays or harbors. They are also called seismic sea waves because they are most frequently caused by ocean-floor earthquakes ( Figures 5-1 and 5-2). Note that the term tsunami is both singular and plural. The colloquial use of “tidal wave” is not appropriate because tsunami have nothing to do with tides. The Great Wave 4 Sumatra Tsunami: December 2004 At 7:59 A.M. on December 26, 2004, a giant earthquake, the fourth largest in the world since 1900, shook the gently northeast-dipping subduction zone just west of northern Sumatra, Indonesia ( Figure 5-1). The Indian Plate is moving northeast at 6 centimeters per year relative to the Burma Plate. In the ten years preceding this event, there were forty events larger than magnitude 5.5 in the area, but none generated tsunami. In the last 200 or so years, several other earthquakes larger than magnitude 8 have generated moderate- 4 99 W3215.indb 99 2/10/05 2:52:53 PM INDIA BURMA INDIAN PLATE THAILAND Kenji Satake, Geological Survey of Japan, AIST. SRI LANKA K. Sataki (a) Digital Globe, satellite image. SUMATRA AUSTRALIAN PLATE (b) FIGURE 5-1. (a) Tsunami times (in shown 4 zone, which spans fromwave-front travel epicenter hours) arethroughemanating from the rupture the earthquake (red star) the area of after- Digital Globe satellite images. shocks (red dots). (b) A broad offshore beach is exposed at Kalutara, Sri Lanka on December 26, 2004, as the first wave of the tsunami drains back to the ocean. The red dotted line is the normal beach edge. (a) (b) FIGURE (a) The northern part of Banda Aceh, Sumatra, on June 23, 2004, before 4 (b)5-2. same area on December 28, 2004, after the tsunami. Note that virtually allthe tsunami. The of the buildings were swept off the heavily populated island. The heavy rock riprap along the north coast of the island before the tsunami remains only in scattered patches afterward. A large part of the island south of the riprap has disappeared, as has part of the southern edge of the island between the two bridges, where closely packed buildings were built on piers in the bay. sized tsunami that have killed as many as a few hundred people. Paleoseismic studies show that giant events occur in the region on an average of once every 230 years. The subduction boundary had been locked for hundreds of years, causing the overriding Burma Plate to slowly bulge like a bent stick; it finally slipped to cause a magnitude 9 earthquake. Given the size of the earthquake, offset on the thrust plane was some 15 meters, with the sea floor rising several meters. The subduction zone broke suddenly, extending north over approximately 1,200 kilometers of its length ( Figure 5-1a), or the entire length of the segment between the India and Burma Plates; that presumably relieved much 4 100 W3215.indb 100 2/10/05 2:52:55 PM of the strain in that zone. The Australia Plate–Burma Plate section of the boundary immediately to the southeast did not break. It may do so in a future event as it did in an estimated magnitude 9.3 quake in 1833. The consequences of this disaster were tragic. Sumatra, Thailand, and nearby countries are mountainous, with huge populations living near sea level along the coasts. That made them vulnerable to tsunami waves that rose to 10 meters and more above normal sea level. Although the tsunami waves in the open ocean are often only a half meter or so high, they drag on the bottom in shallower water, causing them to slow and rise into much larger waves. Given that most homes and other buildings average only 3 meters high, only buildings such as some coastal tourist hotels remained above the incoming waves. Although tsunami waves can break onshore like normal waves (see the opening chapter photograph), they can flow inland for more than a kilometer in fairly flat regions, becoming a torrent that sweeps up everything in its path: cars, people, buildings, and fragments of their crushed remains. Even strong swimmers have little chance of survival because of impacts from all of the churning debris. As the wave recedes into the trough before the next wave, the onshore water and its debris flowing back offshore are almost as fast and dangerous to those caught in it as the initial tsunami ( Figure 5-1b). Because the time between tsunami waves is often more than a half hour, the wave trough is well offshore; people and debris are carried out to sea. With the retreat of the first wave, the first survivors felt that the danger had passed, only to be overwhelmed by the next still larger wave. Hours later, debris and bodies washed up on beaches, in some places accumulating like driftwood. The dead included not only locals but many foreigners vacationing in the region’s warm weather. As efforts continued to find survivors, other threats loomed. Bodies began decomposing in the tropical heat. Concern quickly shifted to the danger of contaminated water, cholera, typhoid, hepatitis A, and dysentery. Compounding the nightmare were pools of stagnant water that can foster breeding of mosquitoes that may carry malaria and dengue fever. Often-futile attempts were made to identify the dead before burial, using fingerprints or merely by posting photos. Then there was the problem of burying the dead in ground that was completely saturated with water. Some people who survived the waves died from infected gashes, lacerations, broken bones, and other wounds as both antibiotics and health care workers were in short supply. The magnitude 9 earthquake shook violently for as much as eight minutes. For people nearby in Sumatra, the back-and-forth distance of shaking, with accelerations greater than that of a falling elevator, made it impossible to stand or run while poorly reinforced buildings collapsed around them. The sudden rise of the ocean floor generated a huge wave that moved outward at speeds of more than 700 kilometers per hour; it reached nearby shores within 15 minutes. Those who were living on low- 4 http://earthscience.brookscole.com/hyndman W3215.indb 101 lying coastal areas had little or no warning of the incoming wave. Most people were preoccupied with the earthquake, and few were aware of even the possibility of tsunami. For some who did not happen to be looking out to sea, the first indication was apparently a roaring sound similar to fastapproaching locomotives. Elsewhere there was no sound as the sea rose. For the vast majority of the people, there was no official warning. A tsunami warning network around the Pacific Ocean monitors large earthquakes and then transmits warnings to twenty-six participating countries of the possibility of tsunami generation and arrival time. In the Indian Ocean, however, there is no such warning network. The Indian Ocean continental margins do not have active subduction zones, except along the southwest coasts of Sumatra and Java. A warning system would not have been able to save most of the lives in the most devastated region of Sumatra because the time between the earthquake and wave arrival was short; however, it could have saved many lives in more distant locations such as Sri Lanka and India. Although the massive earthquake was recorded worldwide, people along the affected coasts were not notified of the possibility of major tsunami. The reasons were several. The Pacific Tsunami Warning Center in Hawaii alerted member countries around the Pacific and tried to contact some countries around the Indian Ocean a tsunami might have been generated. Because tsunami in the Indian Ocean are infrequent, no notification framework was in place to rapidly disseminate the information between or within countries. Compounding the problem was the lack of knowledge, even among officials, that a large earthquake could generate large tsunami. On the other hand, a ten-year-old girl who had recently learned in class about tsunami, saw the sea recede before the first wave and yelled to those around her to run uphill. A dock worker on a remote Indian island had seen a National Geographic special on tsunami, felt the earthquake, and ran to warn nearby communities that giant waves were coming. Together, these two individuals saved more than 1,500 lives. Clearly, knowledge of hazard processes can save lives. An official in West Sumatra recorded the earthquake and spent more than an hour unsuccessfully trying to contact his national center in Jakarta. An official in Jakarta later sent e-mail notices to other agencies but did not call them. A seismologist in Australia sent a warning to the national emergency system and to Australia’s embassies overseas but not to foreign governments because of concern for breaking diplomatic rules. Officials in Thailand had up to an hour’s notice but apparently failed to disseminate the warning. Among the public, few people had any knowledge of tsunami or that earthquakes can produce them. Ironically, the country’s chief meteorologist, now retired, had warned in the summer of 1998 that the country was due for a tsunami. Fearing a disaster for the tourist economy, government officials labeled him crazy and dangerous. He is now considered a local hero. TSUNAMI 101 2/10/05 2:53:04 PM Although scientists have expressed concern about the lack of a warning system in the Indian Ocean, most officials in Thailand and Malaysia viewed tsunami as a Pacific Ocean problem and the tens of millions of dollars it would cost to set up a network left it a low priority in a region with limited finances. In addition, it was Sunday and the day after Christmas, so few people would have been at work. Compounding the problem was the time delay in determining the size of the earthquake. The location of the earthquake could be determined quickly and automatically from the arrival times of seismic waves from several locations. The magnitude was apparently large, initially estimated by Indonesian authorities as magnitude 6.6, a size that would not generate a significant tsunami. However, because the magnitude of giant earthquakes is determined by the amplitude of the surface waves and such large earthquakes have lower frequency waves, it often takes more than an hour to determine the magnitude. By that time, it would have been too late because waves had already battered Sumatra. The first reports in northern Sumatra indicated that the earthquake severely damaged bridges and knocked out electric power and telephone service. Buildings were heavily damaged. People ran into the streets in panic. Smaller earthquakes quickly followed farther north along the subduction zone. A short time later, tsunami waves 5 meters high struck northernmost Sumatra, wiping out 25 square kilometers of the provincial capital of Banda Aceh ( Figure 5-2). Locally, the wave swept inland as far as 8 kilometers; it had a 24-meter-high run up on one hill almost a kilometer inland. In less than two hours, the first of several tsunami waves crashed into western Thailand, the east coast of Sri Lanka, and shortly thereafter the east coast of India ( Figure 5-1a); seven hours later, it reached Somalia on the coast of Africa. In Sri Lanka, a coastal train carrying 1,000 passengers was washed off the tracks into a local swamp. More than 800 bodies have been recovered. By 4 P.M. local time, it was apparent that early reports drastically underestimated the level of destruction. Indonesia reported 150 deaths, Thailand 55, Malaysia 8, India 1,000, and Sri Lanka 500. Many villages were completely washed away, leaving no one to identify or bury the bodies. In the following days, the official death count rose rapidly. By January 13, more than 283,000 people were presumed dead and tens of thousands more remained missing. Even with many hours between the earthquake and the first waves, hundreds of people died in Somalia on the northeast African coast. At least 31,000 died in Sri Lanka, 10,750 in India, and 5,400 in Thailand. In Indonesia, at least 230,000 are dead or missing. In Banda Aceh alone, 30,000 bodies may yet remain in the area in which no buildings were left standing. Relief organizations were overwhelmed by the unprecedented scale of the disaster encompassing eleven countries. Affluent countries around the globe quickly 4 4 pledged millions of dollars in aid, in the form of food and water, medical and technical help, and relief funds. With almost everyone affected by the disaster, people came together to help one another. Desperation, however, led to some fights over relief food and water. Many people in the region refused to go near the beaches, fearing that the many large aftershocks could generate more tsunami. Five million people in the region lost their homes; hundreds of thousands of survivors huddled in makeshift shelters. This tsunami was the most devastating natural disaster of its kind on record. Chile Tsunami: May 1960 The largest earthquake in the historic record (magnitude MW = 9.5) was on the subduction zone along the coast of Chile on May 22, 1960. Fifteen minutes after the earthquake, the sea rose rapidly by 4.5 meters (14.7 feet). Fifty-two minutes later, an 8-meter-high second tsunami arrived at 200 kilometers (124 miles) per hour, crushing boats and coastal buildings. A third slower wave was 10.7 meters high. More than 2,000 people died. In Maullín, Chile, the tsunami washed away houses on low ground or carried them off their foundations. Some of those houses were carried more than a kilometer inland; others were demolished or washed out to sea. Many people wisely ran for higher ground. Some who ran back for valuables were not so lucky. One group survived by climbing to the loft of a barn; several others climbed trees. One person in a tree watched water rise to his waist. One farmer who watched his house on a river floodplain collapse later found 10 centimeters of sand covering his fields ( Figure 5-3). Forests on low ground dropped abruptly below sea level, permitting saltwater to flow in and kill the trees. Fifteen hours later, as predicted, the tsunami reached Hawaii. Coastal warning sirens sounded at 8 :30 P.M. When the 9 P.M. news from Tahiti reported that waves there were only 1 meter high, many Hawaiians relaxed. Few people in Hawaii realized that well-developed reefs protect the Tahitian islands. Warned hours earlier by radio and sirens, a third of the people in low areas of Hilo evacuated; others did not because some previous warnings involved small tsunami that caused little damage. The first wave just after midnight was little more than 1 meter high ( Figure 5-4). Many people thought the danger was over and returned to Hilo. At 1:04 A.M., a low rumbling sound like that of a distant train became louder and louder, followed by crashing and crunching as buildings collapsed in the 4-meter high, nearly vertical wall of the largest wave; 282 people were badly injured and sixty-one died, all in Hilo, including sightseers who went to the shore to see the tsunami. The waves destroyed water mains, sewage systems, homes, and busi- 4 4 102 W3215.indb 102 2/10/05 2:53:04 PM FIGURE 5-3. A 4 material fromtsunami wave swept the nearby beach Brian Atwater, USGS. near Maullín, Chile, in 1960. It left a sand layer over a soil horizon on a farmer’s field. Midnight 14 12:30 A.M. 1:00 A.M. Water-level measurement 1:30 A.M. FIGURE 5-4. This shows the 4 waves in Hilo,tide gauge record 1960, foltsunami Hawaii, May 23, 2:00 A.M. Wave that devastated part of Hilo lowing the Chilean earthquake. In this case, the first wave is relatively low, followed by successively higher waves to more than 4 meters above the low tide that preceded the tsunami. After the first couple of waves, their wavelength and frequency increased. 4 12 Wave height (ft) 8 6 First wave of tsunami 2 4 1 2 ? 0 −2 USGS. −4 −6 Wave height (m) 3 10 0 −1 Approximate low tide level −2 nesses. Most of the deaths were avoidable; people heard the warnings but misinterpreted the severity of the hazard. Hilo is the most vulnerable location on the Hawaiian Islands. Although it has a particularly good harbor, Hilo Bay also focuses the damage ( Figure 5-5). Even tsunami waves that come from the southeast refract in the shallower waters around the island to focus their maximum height and energy in the bay ( Figure 5-6). The bay also has the unfortunate form that as a first tsunami wave drains back offshore, it reinforces the incoming second wave that arrives about a half hour later. Nine hours after the tsunami hit Hawaii, it reached the island of Honshu, Japan. Although its wave height decreased to only 4.5 meters, 185 people died, 122 of them on Honshu. Following a few unusual waves up to 1 meter high, the first sign of the tsunami was retreat of the sea accompanying a rapid 1.5-meter drop in sea level ( Figure 5-7). Then the first tsunami wave arrived more than 4 meters above the 4 4 http://earthscience.brookscole.com/hyndman W3215.indb 103 USGS photo. 4 FIGURE 5-5. The 1960 4the head of Hilo Bay. tsunami destroyed the waterfront area at TSUNAMI 103 2/10/05 2:53:05 PM Tsunami path Hilo Bay N Wailuku River bridge Electric power plant 0 KAUAI Flooded area 0 0.5 miles 0 0.5 kilometers Noon 20 6:00 P.M. Midnight 100 miles 0 100 kilometers OAHU MOLOKAI Honolulu MAUI HAWAII Pacific Hilo Ocean FIGURE 4 area of5-6. This lowlying downtown Hilo was destroyed by the Chilean tsunami of 1960. 6:00 A.M. 6 3 10 Waves noticed by fireman High tide High tide Low tide 0 Chilean tsunami previous low. It was that withdrawal, followed immediately by rapid rise, that caught people off guard and drowned many of them. Five waves over six hours culminated in a huge wave more than 5 meters high that disabled the tide gauge and further record of the tsunami (right edge of Figure 5-7). Note that the highest wave was far from the first and that the waves can be an hour or more apart. Tsunami Generation When the subject of tsunami comes up, most people immediately think of earthquakes. Actually, tsunami are also generated by a variety of other mechanisms that cause sudden displacement of water. These include volcanic eruptions, landslides, rockfalls, and volcano flank collapse. We consider these causes individually. 104 W3215.indb 104 0 Water level (m) Water level (ft) International Tsunami Information Center. Tide gauge disabled FIGURE 5-7. 4 recorded a The tide gauge at Onagawa, Japan, dramatic drop in sea level as the May 23–24, 1960, Chilean tsunami arrived. Such a drop provides as much tsunami warning as a rapid rise in sea level. Earthquake-Generated Tsunami Most tsunami are generated during shallow-focus underwater earthquakes associated with the sudden rise or fall of the seafloor, which displaces a large volume of water. Earthquake tsunami occur most commonly by displacement of the ocean bottom on a reverse-movement subduction-zone fault and occasionally on a normal fault. Strike-slip earthquakes seldom generate tsunami because they do not displace much water. Waves with short periods ( Figure 5-8) form with small earthquakes, and tsunami waves with long periods form with larger earthquakes. It is the long-period earthquakes that displace the most water and create the largest waves. Subduction-zone earthquakes off Japan, Kamchatka, Aleutian Islands and Gulf of Alaska, Mexico, Peru, and Chile are the most frequent culprits. The subduction zone off the 4 CHAPTER 5 2/10/05 2:53:07 PM FIGURE 5-8. The is the time between the 4two successive wavewave periodexample, from wave crestpassage of crests—for 1 to wave Wave period 1 crest 2. 2 Tsunami waves spread Earthquake starts tsunami Slow distortion USGS. Locked (a) Stuck area ruptures, releasing energy in an earthquake Between earthquakes (b) During an earthquake (c) Minutes later FIGURE 5-9. A subduction-zone the leading 4 displacing a huge volume of earthquake snaps a tsunami. edge of the continent up and forward, water to produce and 5-10). See the Cases in Point for Anchorage, Alaska, and Hokkaido, Japan, and the introductory discussion of the Sumatra tsunami. Waves move out both ways 3 Sea surface Sea surface resurges pulled down Sea surface 2 1b Ocean Reverse fault FIGURE 5-10. create a 4 is generated byThe sequence of events thatfault in thetsunami that a subsea reverse or thrust ocean floor are: (1a) Seafloor snaps up, pushing water up with it; (1b) sea surface drops to form a trough; (2) displaced water resurges to form wave crest; and (3) gravity restores water level to its equilibrium position, sending waves out in both directions. coast of Washington and Oregon is like a tightly drawn bow waiting to be unleashed. Major tsunami somewhere around the Pacific Ocean occur roughly once a decade. A 30-meter high wave forms somewhere around the Pacific Ocean, on average, once every twenty years. The leading edge of the continent, overlying the descending oceanic crust at a subduction zone, is typically “locked” for many years before breaking loose in a large earthquake ( Figure 5-9a). While locked, the overlying continental edge is pulled downward, causing upward flexure of the overlying plate (Figure 5-9b). When the locked zone finally breaks loose in an earthquake, the leading edge of the continent snaps oceanward and up, commonly over a considerable length parallel to the coast. That moves a lot of water and causes a tsunami ( Figures 5-9c 4 4 http://earthscience.brookscole.com/hyndman W3215.indb 105 Volcano-Generated Tsunami 1a Tsunami are also caused by volcanic processes that displace large volumes of water. Possibilities include collapse of a near-sea-level caldera that pulls down a large volume of water from the surrounding sea. Water is also driven upward by fast-moving ash flows or submarine volcanic explosions into a large body of water. More than one of these mechanisms can occur at an individual volcano. One of the most infamous and catastrophic events involving a volcanogenerated tsunami was at Krakatau Volcano in 1883. See “Case in Point: Krakatau,” page 111. Tsunami from Fast-Moving Landslides or Rockfalls When major fast-moving rockfalls or landslides enter the ocean, they can displace immense amounts of water and generate tsunami. At first thought, the height of the tsunami might be expected to depend primarily on the volume of the mass that displaces water. However, a more important parameter is the height of fall. A striking example was the Lituya Bay, Alaska, tsunami in 1958. It happened when a large cliff fell into a coastal fjord to cause the highest tsunami in historic record (see “Case in Point: Lituya Bay, Alaska,” p. 109). That event killed only two people, but the potential may exist for a similar but more catastrophic event as more people move into similar environments (see “Case in Point: Glacier Bay, Alaska,” p. 110). TSUNAMI 105 2/10/05 2:53:08 PM CASE IN POINT Anchorage, Alaska, 1964 Effects Close to the Epicenter The giant magnitude 8.6 (MW = 9.2) Anchorage, Alaska, earthquake on March 27, 1964, showed large vertical effects of slip on the subduction zone. The slab sinking beneath the Aleutian oceanic trench slipped over a length of 1,000 kilometers and an area more than 300 kilometers wide. A strip 500 kilometers long by 150 kilometers wide of extremely shallow seafloor rose 10 meters above sea level and moved 19.5 meters seaward. Another belt onshore from the coast, fully comparable in length and more than 100 kilometers wide, sank as much as 2.3 meters. Low-lying coastal areas actually dropped below sea level. Twenty-seven years later, the same areas had slowly rebounded to again be above sea level ( Figure 5-11). The sudden change in seafloor elevation displaced the overlying water into giant tsunami waves that washed ashore on the Kenai Peninsula within nineteen minutes and onto Kodiak Island in thirty-four minutes ( Figures 5-12 and 5-13). The maximum tsunami wave run-up occurred where it funneled into Valdez Inlet, just west of the Kenai Peninsula, where the earthquake caused a submarine landslide. Of the 131 people killed in the earthquake, 122 drowned in the 61-meter waves that funneled into and devastated waterfront areas in Valdez and Seward. 1964 Smaller wave heights extended all the way to Crescent City, northern California, destroying much of the waterfront area. 4 4 Effects Far from the Epicenter (a) Brian Atwater photos, USGS. 1991 (b) FIGURE (a) Immediately earthquake of the coastal 4 flooded5-11. the coastal bulgeafter the AlaskaThis is the same1964, twenty-seven area when collapsed. (b) area years after the bulge again began to rise. 106 W3215.indb 106 Eight minutes after the Anchorage earthquake, an alarm sounded at the Honolulu Observatory. The location and magnitude of the earthquake were determined from seismograms within an hour. The California State Disaster Office received warnings of a possible “tidal wave” two hours after the earthquake. The county sheriff at Crescent City in northern California received the warning one and one-half hours later and notified people in low-lying coastal areas to evacuate. An hour after that, a 1.5-meter high wave reached Crescent City, amplified by the shallowing water near shore and narrowing of the harbor. According to the Del Norte Historical Society files, the curator of Battery Point Lighthouse, on a small rocky island just offshore, recalled that it was a clear moonlit night and the waves were clearly visible as they pitched into the town. The first wave carried giant logs, trees, and other debris, demolishing buildings and cars. The debris-laden wave receded as quickly as it arrived, leaving battered cars, houses, logs, and boats. The sea receded to a kilometer offshore. After a second wave came into the harbor, some people returned to clean up. A third and larger wave then washed inland more than 500 meters, drowning five people; it knocked out power and ignited a fire before the sea receded even farther. The fourth and largest CHAPTER 5 2/10/05 2:53:08 PM CASE IN POINT (continued) FIGURE A parking of cars 4 thrown 5-12. at the headlot fullbay like was about of a USGS. so many toys following the March 1964 Prince William Sound tsunami in Alaska. Large fishing boats joined them onshore. FIGURE A fishing 4 fuel5-13. restlargeshore in boat and crushed truck on Resur- National Geopysical Data Center, NOAA. rection Bay, Seward, Alaska, following the March 1964 Prince William Sound tsunami. wave was 6.3 meters high; it killed ten people who went back down to check their houses. That wave submerged the damaged Citizen’s Dock and lifted a big, loaded lumber barge, setting it down on top of the dock, crushing it. Fuel from ruptured tanks at the Texaco bulk plant spread to the fire and ignited. One after another the tanks exploded. Pieces of everything imaginable drained back offshore with the outgoing wave. The fifth wave was somewhat smaller. In all, the tsunami destroyed fifty-six blocks of the town. http://earthscience.brookscole.com/hyndman W3215.indb 107 ★ TSUNAMI 107 2/10/05 2:53:14 PM CASE IN POINT Hokkaido, Japan, 1993 On July 12, 1993, a large magnitude 7.8 earthquake in the Sea of Japan, off the west coast of Hokkaido in northern Japan, generated one of the largest tsunami in Japan’s recorded history. A large slab of ocean floor at a depth of 15 kilometers moved up on a thrust fault dipping gently to the east, sending a tsunami onshore within only two to five minutes. Average run-ups were 10 to 20 meters but reached as high as 30.6 meters near one small coastal village. The tsunami inundated the small island of Okushiri, killing almost 200 people and causing $600 million in property loss. The hardest hit community was Aonae, where the first wave from the north- east swept over a massive 4.5-meter breakwater to run up to heights of 3 to 7 meters throughout the town. A second wave ran up to 5 to 10 meters throughout the town. Nearby areas of coast, beyond the protection of the breakwater and a dune field, saw run-up heights of 10 to 20 meters. The damage in Aonae is apparent in the before-and-after images ( Figure 5-14a, b). The residential area and port facilities identified in Figure 5-14a were completely destroyed ( Figures 5-14b and 5-15). Fires were fueled by plentiful propane and kerosene used for heating, the flames fanned by strong northeast winds. 4 4 FIGURE 5-14. 4 This photo of (a) the small village of Aonae on Okushiri Island, west of the southwestern end of Hokkaido Island, Japan, was taken in 1976; compare the areas damaged in Figure 5-14b. (b) The small village of Aonae was heavily damaged by the 1993 tsunami. The smoke comes from tsunami-caused fires. Note that strong refraction carried the wave around the end of the island and resulted in major damage on the east side of the island. (a) (b) NOAA. ★ FIGURE 5-15 A 4 debris on the sidefishing boat and crushed fire truck lay among the of the island away from the incoming 1993 tsunami wave at Aonae, Okushiri Island, Japan. 108 W3215.indb 108 CHAPTER 5 2/10/05 2:53:23 PM CASE IN POINT Lituya Bay, Alaska, 1958 One of the most spectacular tsunami resulting from a landbased rockfall was in Lituya Bay, Alaska, on July 9, 1958. Lituya Bay, a deep fjord west of Juneau, Alaska, and at the west edge of Glacier Bay National Park, was the site of one of the highest tsunami run-ups ever recorded ( Figures 5-16 and 5-17). On July 9, 1958, 60 million cubic meters of rock and glacial ice, loosened by a nearby magnitude 7.5 earthquake on the Fairweather Fault, fell into the head of Lituya Bay. The displaced water created a wave 150 meters high, or the height of a fifty-story building. It surged to an incredible 524 meters over a nearby ridge and removed forest cover up 1874 wave 1936 wave 4 Donald Miller photo, USGS. 1958 wave FIGURE Bay trimlines from 4 previous5-18. This view of Lituyalargershowsthe 1958 event. Do two tsunami that were even than narrow, steep-sided inlets elsewhere show similar healed trimlines? Could they provide hazard information for future rockfall tsunami? Rockfall scarp to an average elevation of 33 meters and up to 152 meters over large areas ( Figures 5-16 and 5-17). Older trimlines in the forest and damaged tree rings of various ages document previous similar tsunami in 1936, 1874, and 1853 or 1854 ( Figure 5-18). This was a huge wave compared to common tsunami that may be 10 to 15 meters high; it swept through Lituya Bay at between 150 and 210 kilometers per hour. Although three fishing boats, with crews of two each, were in the bay at the time, only those on one boat died when their boat was swept into a rocky cliff. On another boat on the south side of the island in the center of the bay, Howard Ulrich and his seven-year-old son hung on as their boat was carried high over a submerged peninsula into another part of the bay. They were actually able to motor out of the bay the next day. On the third boat, Mr. and Mrs. William Swanson, anchored on the north side of the bay, were awakened as the breaking wave lifted their boat bow first and snapped the anchor chain. The boat was carried at a height of 25 meters above the tops of the highest trees, over the bay mouth bar, and out to the open sea. Their boat sank, but they were able to climb onto a deserted skiff and were rescued by another fishing boat two hours later. Examination of the forested shorelines of Lituya Bay, above the trimline of the 1958 event, shows two much higher trimlines produced by earlier tsunami. U.S. Geological Survey scientists examined trees at the level of these higher trimlines and found severe damage caused by the earlier events. Counting tree rings that had grown since then, they determined that the earlier tsunami occurred in 1936 and 1874 ( Figure 5-18). 4 Donald Miller photo, USGS. 4 FIGURE the 4 Alaska,5-16. A huge rockfall intowavehead of Lituya Bay, generated a giant tsunami that stripped Donald Miller photo, USGS. the forest and soil from a ridge. This view to the northeast shows the broad areas of forest that the tsunami swept from the fringes of the bay. The scarp left by the rockfall is visible at the head of the bay (arrow). 4 FIGURE 5-17. This 4the crest of a ridge 524photo details the tsunami damage at meters above the bay. http://earthscience.brookscole.com/hyndman W3215.indb 109 TSUNAMI 109 2/10/05 2:53:29 PM CASE IN POINT Glacier Bay, Alaska Potential Future Tsunami? Glacier Bay, 50 kilometers east of Lituya Bay, is in a similar spectacular environment. It is a deep fjord bounded by precipitous rock cliffs and glaciers. It lies between two major active strike-slip faults, the Fairweather Fault and the Denali Fault, each 50 or 60 kilometers away, and each capable of earthquakes of magnitudes greater than 7. Glacier Bay is a prominent destination for cruise ships touring from Seattle or Vancouver to Alaska, so a tsunami generated by a large landslide into the bay is a concern. Study by USGS geologists suggests that an unstable rockslide mass on the flank of a tributary inlet to Glacier Bay would generate waves with more than 100-meter run-up near the source and tens of meters within the inlet ( Figure 5-19a). In the deepwater channel of the western arm of Glacier Bay, the wave amplitude would decrease with distance out into the bay. Ships near the mouth of the tributary inlet could encounter a 10-meter wave only four minutes after the slide hit the water, then 20-meter high waves after twenty minutes. The waves would likely strike the cruise ships broadside as shown in Figure 5-19b. If the ships kept to this central channel, the largest waves would likely only be approximately 4 meters high. The response of a ship to waves near the mouth of the tributary inlet would depend on the wave height, the wave frequency relative to the ship’s rocking frequency, and the height of the lowest open areas on the ship. However, because the cruise operators are now aware of this risk, they can avoid the dangerous near-inlet waters. 4 (a) ~4m Glacier Bay Gerald F. Wieczorek, USGS. ~ 20 m (b) FIGURE (a) The tidal inlet landslide 4 the5-19.face from the new higher scarpmass next to Glacier Bay, Alaska, includes rock to below water; (b) Photo was taken from the apex of the slide with the tidal inlet in the foreground; two cruise ships in Glacier Bay are visible in mid-photo. Lituya Bay ★ ★ Glacier Bay 110 W3215.indb 110 CHAPTER 5 2/10/05 2:53:37 PM The three broad ridges that radiate outward from the top of the volcano spread slightly under their own enormous weight, producing rift zones along their crests. The volcano eventually breaks into three enormous segments that look on a map like a pie cut into three slices of approximately equal size. One or more of the three volcano segments may begin to move slowly seaward ( Figures 5-20 and 5-21). The rifts between the sediments provide easy passage to molten magma rising to the surface. Those rifts that become the sites of most of the eruptions also form weak vertical zones in the volcano. There is a long history of one or more of the volcano segments breaking loose to slide into the ocean, sometimes slowly but sometimes catastrophically. Studies of the ocean floor using side-scanning radar around the Hawaiian Islands reveal sixty-eight giant debris avalanche deposits, each more than 20 kilometers long. Some extend as far as 230 kilometers from their source and contain several thousand cubic kilometers of volcanic debris. Such slides must have moved rapidly to reach such a distance. At least some of those deposits are the remains of debris avalanches that raised giant tsunami waves that washed high onto the shores of the Hawaiian Islands. Boulders of CASE IN POINT Krakatau August 27, 1883 4 People in towns on the west coast of Java, in Indonesia, awoke to the sounds of Krakatau rumbling, 40 kilometers to the west. At 10:02 A.M., the mountain exploded in an enormous eruption. This was the climactic eruption of activity that had been going on for several months. Thirtyfive minutes later, a series of waves as high as 30 meters flattened the coastline of the Sunda Strait between Java and Sumatra, including its palm trees and houses. Only a few who happened to be looking out to sea saw the incoming wave in time to race upslope to safety. More than 35,000 people died. Studies of the distribution of pyroclastic flow deposits and seafloor materials in the Sunda Straits between Krakatau and the islands of Java and Sumatra show only pyroclastic flow deposits, suggesting that an enormous flow entered the sea to produce the tsunami. Computer simulations of three possible causes, however, suggest that two of the reasonable hypotheses fit the data poorly. They were (1) that the waves were caused by caldera collapse that pulled down a large mass of water or (2) that pyroclastic flows entering the sea displaced the water. The third hypothesis fits the data well. It suggests that seawater seeping into the volcano interacted with the molten magma to generate huge underwater explosions and upward displacement of a large volume of seawater. Major slumps Debris avalanches Maui PACIFIC OCEAN Pololu Kohala Mauna Kea Hualalai Tsunami from Volcano Flank Collapse GIANT PREHISTORIC EVENTS IN HAWAII The flanks http://earthscience.brookscole.com/hyndman W3215.indb 111 A ERZ Kilauea Alika Papa'u Hilina Gerald F. Wieczorek photo, USGS. of many major oceanic volcanoes, including those of the Hawaiian Islands in the Pacific Ocean and the Canary Islands in the Atlantic Ocean, apparently collapse on occasion and slide into the ocean. Hawaii’s volcanoes grow from the seafloor for 200,000 to 300,000 years before breaking sea level, then build a reasonably solid lava shield above sea level for a similar time. Mega-landsliding occurs near the end of the shield-building stage when the growth rate is fastest, heavy basalt load on top is greatest, and slopes are steepest and thus least stable. The lower part of each volcano, below sea level, consists largely of loose volcanic rubble formed when the erupting basalt chilled in seawater and broke into fragments. It has little mechanical strength. Above sea level, oceanic islands are built from heavy, reasonably solid basalt lava flows. Mauna Loa North Kona A' South Kona Kaalualu Punalu'u West Kalae East Kalae 0 50 kilometers FIGURE 5-20. This map of the island of Hawaii 4 slumps and debris avalanches formed by collapseshows the major of the island’s flanks. ERZ ϭ East Rift Zone. TSUNAMI 111 2/10/05 2:53:44 PM 4 4 Mauna Loa northeast rift zone Kaoiki fault system (little vertical slip at present) John Bortniak photo, NOAA. coral litter the flanks of some of the islands to elevations of more than 400 meters above sea level and more than 6 kilometers inland; the flanks of some of the islands have lost much or all of their soil to a similar elevation. It now seems clear that both the displaced boulders and the scrubbed slopes are evidence of monstrous waves that washed up the flanks of the islands as one of the enormous pie segments plunged into the ocean. The headscarps of such collapsed segments become gigantic coastal cliffs, some more than 2,000 meters high and among the highest cliffs in the world ( Figure 5-22). The landslides appear to occur during major eruptive cycles and have a recurrence interval of roughly 100,000 years. Headscarps of the slides are the giant “pali” or cliffs that mark one or more sides of each of the Hawaiian Islands. Despite the existence of such evidence, the frequency of these horrifying events remains unclear. If we can judge from the age of coral fragments washed onto the flanks of several islands, the most recent slide detached a large part of the island of Hawaii 105,000 years ago. That slide raised tsunami waves to elevations of as much as 326 meters on the island of Lanai. Mauna Loa, the gigantic volcano on the big island of Hawaii, the youngest and largest of the Hawaiian Islands, has collapsed repeatedly to the west. Two of these collapses were slumps and two were debris avalanches. Most were submarine collapses, though the head scarp of the North Kona slump grazes the west coast of Hawaii. Kilauea, the youngest and most active volcano in Hawaii, is now slumping. The Hilina slump on its south flank, 100 kilometers wide and 80 kilometers long, is moving seaward at 10 to 15 centimeters per year, sometimes suddenly ( Figure 5-23). On November 19, 1975, a big sector of the south flank of Kilauea volcano moved more than 7 meters seaward and dropped more than 3 meters during a magnitude 7.2 earthquake. The resulting relatively small tsunami FIGURE 5-22. Giant 4 big island of Hawaii. cliffs or pali amputate the lower slopes of the drowned two people nearby, destroyed coastal houses, and sank boats in Hilo Bay on the northeast side of the island. What this portends for further movement is not clear. Will the flank of the volcano continue to drop at unpredictable intervals or could it fail catastrophically? Hawaiian geologists wondered for years about blocks of coral and other shoreline materials strewn across the lower slopes of the islands. It now seems clear that huge tsunami formed when enormous masses of one of the islands collapsed into the ocean and the waves washed material up the slopes from the beach. Tsunami formed by island flank collapse are documented from tsunami deposits consisting of a cemented mix of fragments of limestone reef and basalt, in many places tens of meters above sea level. On Molokai, they left deposits 70 meters above sea level; on Lanai, they left blocks of coral as much as 326 meters above sea level. An eventual repetition of those events seems inevitable; it would kill much of the population of the Hawaiian Islands. Kilauea caldera Koa'e fault system Hilina and Holei Pali Mauna Loa lavas Sea level Kilauea lavas Oceanic crust Detachment fault at base of volcanoes USGS. Upper mantle A A' FIGURE 5-21. cross section volcano shows the prob4 failure surfacesThis northwest–southeastthe volcano’s flof Kilaueacross-section location as able that lead to collapse of anks. See line A–AЈ in map Figure 5-20. 112 W3215.indb 112 2/10/05 2:53:46 PM 155°15′ 155°9′ 155°3′ FIGURE 5-23. The south flank Kilauea 4 is slowly slumping seaward.ofThe arrows Volcano indicate directions and rates of movement as measured by Global Positioning Systems (GPS). 19°24′ 19°18′ 0 The potential for a future collapse of the flank of Kilauea volcano on the big island is emphasized in scarps forming in the 80-kilometer-long coastal area southeast of Kilauea. A slab almost 5 kilometers thick is sliding ocean-ward at 10 to 15 centimeters per year. If this huge area collapses into the sea suddenly, perhaps triggered by a large earthquake or major injection of magma, it would likely generate tsunami greater than 100 meters high. Many coastal communities in Hawaii would be obliterated with little warning. However, to put these numbers in perspective, if 100,000 people were killed in such an event every 100,000 years, the average would be one person per year. Although unimaginably catastrophic when it does happen, there are certainly greater dangers, on average, in one person’s lifetime. The danger is not limited to Hawaii. If the flank of Kilauea, now moving seaward, should fail catastrophically, it could generate a tsunami large enough to devastate coastal populations all around the Pacific Ocean. Those in Hawaii would have little warning. The Pacific coast of the Americas would get several hours. It remains to be seen how many people could be warned and how many of those would heed the warning. Certainly major urban centers such as San Francisco and Los Angeles could not be evacuated in time. We hope that the next event will not be any time soon, but we have no way of knowing. CANARY ISLANDS Like other large basaltic island volcanoes, Tenerife—in the Canary Islands off the northwest coast of Africa—shows evidence of repeated collapse of its volcano flanks. Tenerife reaches an elevation of 3,718 meters, almost as high as Mauna Loa. It is flanked by large-volume submarine debris deposits that left broad valleys on the volcano flanks. Lavas filling these valleys are as much as 590 meters thick and overlie volcanic rubble along http://earthscience.brookscole.com/hyndman W3215.indb 113 5 kilometers USGS. 5 cm/yr an inferred detachment surface that dips seaward at about 9 degrees. Collapse may have been initiated by subsidence of the 11- to 14-kilometer-wide summit caldera into its active magma chamber. The most recent caldera and island flank collapse was 170,000 years ago. That event carried a large debris avalanche from the northwest coast of Tenerife onto the ocean floor. It carried 1,000 cubic kilometers of debris, some of which moved 100 kilometers offshore. The much larger El Golfo debris avalanche detached 15,000 years ago from the northwest flank of El Hierro Island. It carried 400 cubic kilometers of debris as much as 600 kilometers offshore. An average interval of 100,000 years between collapse events on the Canary Islands may be long, but the consequences of such an event would be catastrophic. And the interval is merely an average. The next collapse could come at any time, and the giant tsunami caused by collapse would catastrophically inundate not only heavily populated coastal areas around the north Atlantic Ocean but also reach coastal Portugal in two hours, Great Britain in little more than three hours, and the east coasts of Canada and the United States in six to seven hours ( Figure 5-24). Because large populations live in low-lying coastal cities and on unprotected barrier islands along the coast, millions would be at risk. Even if warning were to reach endangered areas as much as six hours before arrival of the first wave, we know from experience with hurricanes, that evacuation would likely take much longer that that. Imagine hundreds of thousands of people trying to evacuate without a well-thought-out plan and in traffic that is heavy under normal circumstances. What about congestion on the single two-lane bridges that link most barrier islands to the mainland? How many would ignore the warning, 4 TSUNAMI 113 2/10/05 2:53:48 PM North America rs ou 6h Simplified from Simon Day and Stephen Ward. 9 ho ur s 3h ou rs tes inu 3m Canary Islands Africa South America FIGURE 5-24. A large landslide from La Palma, Canary Islands, 4 generate immense tsunami waves that would fan out into could the Atlantic Ocean. Computer simulations suggest that huge waves would reach the east coast of North America in six to seven hours. not realizing the level of danger? The death toll could be staggering. Similar situations are now known to exist on Reunion Island in the Indian Ocean, Etna in the Mediterranean Sea, and the Marquesas Islands in the Pacific Ocean. Recently discovered fractures along a 40-kilometer stretch of the continental shelf 100 miles off Virginia and North Carolina suggest the possibility of a future undersea landslide. Such a slide could generate a tsunami like one that occurred 18,000 years ago just south of those fractures. Tsunami-deposited sand layers have been discovered as well at several sites on islands west of Norway. That tsunami 8,150 years ago formed from the immense Storegga submarine slide on the continental slope off the coast of Norway. The tsunami carried sand, pieces of wood, and marine fossils onto peat as much as 20 meters above sea level. A much more recent but smaller subsea slide in 1998 generated a tsunami that killed 2,200 people in Papua New Guinea. Tsunami from Asteroid Impact Because anything that suddenly displaces a large volume of water would generate a large wave, the impact of a large asteroid into the ocean would generate large tsunami that would radiate outward from the impact site, much as happens with any other tsunami. The average frequency of such events is low, but a 1-kilometer asteroid falling into a 5-kilometer-deep ocean might generate a transient 3-kilometer-deep cavity. Collapse of the ocean cavity walls to refill the cavity will reach supersonic speeds to send a plume high into the atmosphere. Initial kilometer-high waves crest, break, and interfere with one another. Waves with widely varying frequency radiate outward. The behavior of such complex waves is not well understood, but they are thought to decrease fairly rapidly in amplitude away from the impact site. The different wave frequencies would, however, interfere and locally pile up on one another to cause immense run-ups at the shore. The chance of a 1-kilometer asteroid colliding with Earth are only once every million years, so such a hazard, though significant in scale, is not major in terms of human lifetimes. The chance of catastrophic tsunami from flank collapse of an oceanic island such as Hawaii or the Canary Islands is perhaps 10 times as great. Velocity and Height Tsunami wave velocities can be as high as 870 kilometers per hour. Because tsunami wave heights in the open ocean are small, and the average tsunami wavelength is 360 kilometers, slopes on the wave flanks are extremely gentle. The time between waves may be half an hour; thus, it takes that long for a ship to go from the wave trough to its crest and back to the trough. As a result, ships at sea hardly notice them. Water particles in wind-driven waves travel in a circular motion—the water does not travel with the wave (see Sidebar 5-1). That circular motion fades downward. Because waves touch bottom at depths less than approximately half their wavelength, tsunami waves drag bottom everywhere in the ocean. Their velocity does not depend on wave- Sidebar 5-1 The velocity of tsunami waves depends on the water depth and gravity. C = Ί๶ gD ๵ where C = velocity in meters per second D = depth in meters g = gravitational acceleration (9.8 m/sec.2) Thus, C = 3.13Ί๵ D ๵ For example, if D = 4,600 meters (deep ocean): C = 3.13Ί๶๶ = 3.13 × 67.8 meters per second 4,600 ๶ or 763 kilometers per hour (the speed of some jet aircraft!) If D = 100 meters (near shore): C = 3.13Ί๶ = 3.13 × 10 = 31.3 meters per second 100 ๶ or 112.7 kilometers per hour (the speed of freeway traffic) 114 W3215.indb 114 2/10/05 2:53:58 PM FIGURE 5-25. Both wind waves and tsunami 4 drag on bottom near shore, becoming shorter in waves Waves steepen, tighten, and lean forward Servicio hidrologíco y geophysica de la armada, Chile. wavelength and higher in amplitude before breaking at the shore. Fault FIGURE 5-26. The December 2004 4 leveled almost all of the homes in tsunami Tyler Clements, U.S. Navy photo. Banda Aceh, Sumatra, Indonesia. length as do wind-driven deepwater waves, but on water depth. As tsunami waves reach shallower water, such as on the continental shelf, they slow down and build in height because their circular motions at depth drag on the ocean bottom and slow down. Still closer to shore, they slow and build even higher before breaking on shore. The waves are forced to slow dramatically in shallower water close to shore, causing the waves to change to a much shorter wavelength with a dramatically greater height, Because the deep part of the wave is slowed the most, the crest of the wave rushes ahead and the wave begins to break near the shore ( Figure 5-25). Equally important is the height of the tsunami wave. A tsunami wave 3 meters high in the open ocean could shorten dramatically in shallow water to roughly one-sixth of its wavelength and rise to 6 times its open ocean height. The same wave in the open ocean could rise in shallow water to 4 http://earthscience.brookscole.com/hyndman W3215.indb 115 18 meters! On shore its run-up height could be even higher ( Figure 5-26), causing massive damage. At a velocity of 760 kilometers per hour and a wavelength of 200 kilometers, a wave would pass any point or arrive onshore every 360 kilometers (720 km/hr), or approximately every half hour. Needless to say, it would not be wise to go down to the beach to see the damage after the first wave had receded. Many tsunami deaths have resulted from people doing just that. What seems like a calm sea or a sea in retreat can be the trough before the next wave. Run-up heights, the height that a wave reaches onshore, varies depending on distance from the fault rupture and whether the wave strikes the open coast or a bay. For the largest earthquakes such as the subduction event in Alaska and the 1960 earthquake in Chile, run-up heights were generally 5 to 10 meters above normal tide level. Local run-up reached as high as 30 meters in Chile. Water levels can change 4 TSUNAMI 115 2/10/05 2:54:05 PM Wave in shallow water Wave in deep water Wave height rises dramatically ine rel o Sh FIGURE 5-28. A 1-meter-high the open ocean 4 in shallower water near shore,tsunami wave involume remains slows so if the wave the same, its wavelength shortens and its amplitude rises. Coastal Effects and Vulnerability Dragging on the continental shelf, a 1-meter open ocean tsunami wave may slow to 150 to 300 kilometers per hour and dramatically rise in height to 15 meters or more. Water piles up as it slows. The front of the wave slows first and the rear keeps coming. Because the volume of the wave remains the same, its height must rise dramatically ( Figure 5-28). Similarly, as the first wave slows, the following waves catch up and thus arrive more frequently. In harbors, tsunami have wave periods of ten to thirty-five minutes and may last for up to six hours ( Figure 5-4, p. 103). Areas most at risk are the low-lying parts of coastal towns, especially near the mouths of rivers and inlets that funnel the waves and dramatically raise their height. If waves arrive at high tide, their height is amplified. Sloshing back and forth from one side of a bay to the other can constructively interfere with one another to raise wave level. Because most coastal towns and seaports are located in bays, the damage resulting from these waves is enhanced. That is also the reason the Japanese call them tsunami—that is, harbor waves. Low-lying Pacific and Caribbean islands would seem likely to be extremely vulnerable to incoming tsunami waves, but some are actually less vulnerable than would be expected. Many are surrounded by offshore coral reefs that drop steeply into deep water. Thus tsunami waves are forced to break on the reef, providing some protection to the islands themselves. Tsunami warnings have now been perfected for far-field tsunami, or those far from the source that generated them. A world network of seismographs locates the epicenter of major earthquakes, and the topography of the Pacific Ocean floor is so well known that the travel time for a tsunami to reach a coastal location can be accurately calculated. In addition, an environmental satellite takes readings from tidal sensors along the coasts, and ocean bottom sensors detect ocean surface height as the waves radiate outward across the Pacific Ocean ( Figures 5-29 and 5-30). This 4 H. Helbush, National Geophysical Data Center. 4 FIGURE 5-27. sequence of shows the arrival 4 wave onto Thisbeach at Laiephotoson Oahu, Hawaii. of a tsunami the Point This tsunami was generated on March 9, 1957, as the result of a magnitude 8.6 (MW) earthquake that struck the Aleutian Islands of Alaska approximately 3,600 kilometers away. rapidly, as much as several meters in a few minutes. Run-up is typically about perpendicular to the orientation of the wave crest, but return flow drains downslope as controlled by local topography. The sequence of photos in Figure 5-27 shows a tsunami wave pushing onshore in Hawaii following an earthquake in Alaska in 1957. Driftwood, trees, and the remains of boats, houses, and cars commonly mark the upper limit of tsunami run-up. 4 4 116 W3215.indb 116 2/10/05 2:54:08 PM GOES Satellite T S AM UN T s U N A M I I 4 Transducers Signal flag Acoustic telemetry Glass ball flotation ~ 6000 m Transducer Bottom pressure recorder NOAA. information now permits prediction of tsunami arrival times at any coastal location around the Pacific Ocean within five minutes. Pacific tsunami warning centers are located at the National Weather Service Tsunami Warning Center in Honolulu, Hawaii, and the Alaska Tsunami Warning Center in Palmer, Alaska. Some low-lying areas such as parts of Hawaii are equipped with warning sirens mounted on high poles to warn people who are outdoors in coastal areas. Tsunamis are most likely to appear within a few minutes to several hours after an earthquake with major vertical motion of the seafloor, depending on the distance from the epicenter. A nearby earthquake will be felt but allow people little time to move to higher ground and no time for official warning. Tsunami warning signs in coastal Oregon suggest moving to higher ground if you feel an earthquake ( Figure 5-31). However, many of those coastal areas have no nearby hills. Quickly moving inland can still help because wave energy, height, and speed dissipate rapidly on land. An earthquake thousands of kilometers away will not be felt but may allow time for official warning. A wave reaching shore may either break on the beach or rush far up onto the beach in a steep front. Large tsunami waves can reach as much as 1.6 kilometers inland. Tsunami dangers include not only drowning in the incoming wave but also severe abrasion by being dragged along the ground at high speed, being thrown against solid objects, being carried back out to sea in the outgoing wave, and being hit by debris carried by the wave. Such debris can include boards and other fragments of houses, trees, cars, and boulders. Even when the wave slows as it drags on shallow bottom, it moves, for example, at 55 kilometers per hour, much too fast to outrun. Sensor anchor Anchor FIGURE 5-29. A pressure sensor on 4 in wave height because a higher the ocean floor detects changes wave puts more water and therefore more pressure above the sensor. The pressure sensor transmits a signal to a buoy floating at the surface and to the warning center via satellite. Alaska earthquake 1 2 3 4 5 6 7 8 9 11 Hawaiian Islands FIGURE 5-30. Tsunami 4 times across the Pacific travel 10 USGS. 15 14 13 12 11 10 9 8 http://earthscience.brookscole.com/hyndman W3215.indb 117 Chile earthquake 7 6 5 4 3 2 1 Ocean from the Chile, 1960, and Alaska, 1964, subduction zone earthquakes. Concentric arcs are travel time estimates in hours after each earthquake. From the Alaska earthquake, for example, the first tsunami wave reached Hawaii in approximately six and one-half hours. It would reach the north island of Japan after nine hours. TSUNAMI 117 2/10/05 2:54:14 PM Donald Hyndman photo. Hawaii, hit by disastrous tsunami in 1946 and 1960, saw tsunami run-up elevations from 1.5 to 6 meters above sea level. In Hilo, where the worst damage occurred, making the waterfront area at the head of the bay into a park has minimized future damage. Even though the source of the 1960 tsunami was in Chile, far to the southeast, and Hilo Bay faces northeast, refraction of the waves around the island left the head of the bay vulnerable to waves as much as 4 meters above sea level. Residents on an island coast opposite the direction of an incoming tsunami should not be comforted. On December 12, 1992, a magnitude MS 7.5 earthquake in Indonesia generated a tsunami in the Flores Sea. The southern coast of the small island of Babi, opposite the direction from which the waves came, was hit by 26-meter tsunami waves, twice as high as the northern coast. In this case, the waves reaching the northern coast split and refracted around the circular island, constructively interfering with one another on the opposite coast. More than 1,000 people died. The mound of water suddenly appearing at the sea surface, in response to a major event, generates a series of waves that may cross the whole Pacific Ocean. Because the initial mound of water oscillates up and down a few times before fading away, it generates a series of waves just like a stone thrown into a pond. The magnitude of a tsunami wave depends on the magnitude of the shallow-focus earthquake, area of the rupture zone, rate and volume displaced, sense of motion of the ocean floor, and depth of water above the rupture. The height of the tsunami wave is initially more or less equal to the vertical displacement of the ocean floor. Because the maximum fault offset is typically 15 meters in a giant earthquake, the maximum earthquake tsunami height in the open ocean is roughly 15 meters. FIGURE 5-31. 4 coast. This sign warns of potential tsunami along the Oregon Survivors of tsunamis often report an initial withdrawal of the sea with a hissing or roaring noise. In many cases, curious people drown when they explore the shoreline as the sea recedes before the first big wave or before subsequent waves. In Hilo in 1946, assuming the danger had past, people went out to see the wide, exposed beach with stranded boats and sea creatures. There they were caught in the second and larger wave. Adding to the danger is the fact that tsunami waves may continue for several hours and the first wave is often not the highest. In Hawaii in 1960, in spite of several hours of tsunami warning, people went down to the beach to watch the spectacular wave, only to be overwhelmed by it ( Figure 5-32). The most vulnerable parts of the United States are Hawaii and the Pacific coasts of Washington, Oregon, California, and Alaska. Locally generated subduction-zone earthquakes, landslides, and volcanic events in the Caribbean can affect Puerto Rico, the U.S. Virgin Islands, and other islands in the Caribbean. 4 Tsunami from Great Earthquakes in the Pacific Northwest Slabs of oceanic lithosphere sinking through an oceanic trench at subduction zone boundaries typically generate earthquakes from as deep as several hundred kilometers. Such a boundary undoubtedly exists offshore along the 1,200 kilometers between Cape Mendocino in northern California and southern British Columbia (B.C.) ( Figure 5-33), so the apparent absence of those deep earthquakes in the Pacific Northwest has worried geologists for years. Several lines of evidence now show that major earthquakes do indeed happen but at such long intervals that none have struck within the period of recorded Northwest history. Finally, in the 1980s, Brian Atwater of the U.S. Geological Survey found the geologic record of giant earthquakes in marshes at the heads of coastal inlets. It consists of a consistent and distinctive sequence of sedimentary layers. A bed of peat, consisting of partially decayed marsh plants that grew just above sea level, lies at the base of the sequence. Above the peat, lies a layer of sand notably lacking the sort of internal layering contained in most sand deposits. Above Pacific Tsunami Museum, Hilo. 4 FIGURE 5-32. People from a tsunami in 4 in 1946. The wave runvisible in hugecenter of theHilo, Hawaii, is the image behind the people. 118 W3215.indb 118 2/10/05 2:54:18 PM the sand is a layer of mud that contains the remains of seawater plants ( Figure 5-34). That sequence tells a simple story that begins with peat accumulating in a salt marsh barely above sea level. It appears that a large earthquake caused huge tsunami that rushed up on shore and into tidal inlets, carrying sand swept in from the continental shelf. The sand covered the old peat soils in low-lying ground inland from the bays as the salt marsh suddenly dropped as much as 2 meters below sea level. Then the mud, with fossil seaweed, accumulated on the sand ( Figure 5-35). The sequence of peat, sand, and mud is repeated over and over. In some cases, forests were drowned by the invading salt water or were snapped off by a huge wave ( Figure 5-36). Huge tsunami-flattened forests in low-lying coastal inlets are found all down the Pacific coast from British Columbia to southern Oregon. These stumps are now at and below sea level because the coastal bulge dropped during the earthquake. 4 4 4 (a) 130° 126° 122°W NORTH AMERICA PLATE Exp lo Rid rer ge Queen Charlotte Fault 52°N British Columbia EXPLORER PLATE Vancouver Victoria Nootka Fault 48° Seattle Jua Fuc n de aR idg e Washington Cascadia subduction zone Cascade volcanoes Portland JUAN DE FUCA PLATE Blanco Fault zone Tsunami sand 44° Oregon Peat 40° s 200 kilometers rea nd n A ault F Mendocino Fault 0 California (b) FIGURE 5-33. The Cascadia oceanic trench to the 4 San Andreas transform fault to the south dominate north andc the the Pacifi continental margin of the United States. (a) Seafloor topography. (b) Map of plate boundaries. http://earthscience.brookscole.com/hyndman W3215.indb 119 John Clague photo. GORDA PLATE Sa Gorda Ridge PACIFIC PLATE FIGURE 5-34. Tsunami sand from a megathrust earthquake 4 deposited in 1700 over dark brown peat in a British Columbia coastal marsh. The scale is in tenths of 1 meter. TSUNAMI 119 2/10/05 2:54:21 PM FIGURE 5-35. Simplified 4 showing tsunami sand sketch deposited immediately after a subduction earthquake when a tidal marsh suddenly drops below sea level. Tsunami Tidal mud Stump Tsunami sand Sea Level Peat marsh soil Peat marsh soil Before earthquake Just after earthquake Centuries after earthquake Donald Hyndman photo. Peat marsh soil FIGURE 5-36. This ancient Sitka spruce forest in the bay at 4 Neskowin, Oregon, was felled by a giant tsunami following the huge subduction zone earthquake of January 1700. Stumps of the giant trees punctuate low tide at this beach some 25 kilometers north of Lincoln City. The forest with trees as old as 2,000 years, was suddenly dropped into the surf during a megathrust earthquake and then felled by the huge tsunami that followed. Radiocarbon dating of leaves, twigs, and other organic matter in the buried soils at Willapa Bay, Washington, indicates seven of those giant events in the past 3,500 years, an average of one per 500 years. Elsewhere along the coast, the records show that twelve have occurred in the last 7,000 years since the eruption of Mount Mazama in Oregon deposited an ash layer on the seafloor at an average interval of 580 years. The intervals between them range from 300 to 900 years. The last one was some 300 years ago, so the next could come at any time. Those analyses indicate similar dates at most, though not all, sites all along the coast between Cape Mendocino and southern British Columbia. That probably means that the fault generally broke simultaneously along this entire 1,200-kilometer length of coast, an extremely long rupture that would likely correspond to an earthquake of about magnitude 9. Such an enormous earthquake offshore would surely start a wave large enough to cross the Pacific Ocean. The sand sheets were deposited at elevations to 18 meters above sea level. Tsunami of this size expose coastal communities to extreme danger. The larger cities of Seattle, Portland, and perhaps Vancouver would not be at significant tsunami risk from such a subduction zone earthquake because they are well up inlets or rivers; the waves would largely dissipate before reaching them. Communities on the open coast or smaller coastal bays, however, are in real danger. An earthquake near the coast could generate a tsunami wave that would reach the shore in less than twenty minutes, which would leave too little time for warning and evacuation of those in danger. Feeling an earthquake along the coast, people should immediately move inland to higher ground. The first indication along the coast of Oregon, Washington, or British Columbia may be an unexpected rise or fall of sea level. Shaking in such a major earthquake, with accelerations of at least 1 g, would make it difficult to stand. Strong motion would continue for several minutes, leaving little time to evacuate. Thus, the first defense for people in the area is to protect themselves during the earthquake—take cover from falling objects until the earthquake ends. Then immediately move inland and to higher ground, because the large tsunami generated by the sudden shift of the ocean floor will arrive at the west coast within fifteen to thirty minutes. If hills are available nearby, evacuation on foot may be preferable because of traffic jams and damaged roads. If trapped in a broad, flat area, a reinforced concrete building may offer some protection, but only as a last resort. Even climbing a sturdy tree has saved more than one person. The sudden drop of the coastal area will raise sea level compared with the land even more. Thus, the tsunami will rush ashore to higher levels than would otherwise be expected. Calculations suggest that a 7- to 8-meter tsunami will invade some coastal bays. The record of the last event indicates that waves were as high as 20 meters where they funneled into some inlets. 120 W3215.indb 120 2/10/05 2:54:33 PM Knowledge of when the last event happened would provide some indication of when to expect the next one. Radiocarbon dating of the peat and buried trees places the last of those events within a decade or two of the year 1700. In a separate analysis, careful counting of tree rings from killed and damaged trees indicates that the commotion happened shortly after the growing season of 1699. In a clever piece of sleuthing, geologists of the Geological Survey of Japan found old records with an account of a great wave 2 meters high that washed onto the coast of Japan at midnight on January 27, 1700. No historic record tells of an earthquake at about that time on other Pacificmargin subduction zones, Japan, Kamchatka, Alaska, or South America. That leaves the Northwest coast as the only plausible source. Correcting for the day change at the international date line and the time for a wave to cross the Pacific Ocean, the earthquake would have occurred on January 26, 1700, at approximately 9 P.M. Coastal Indians have oral traditions that tell of giant waves that swept away villages on a cold winter night. Archeologists have now found flooded and buried Indian villages strewn with debris. These many lines of data help confirm the timing of the last giant earthquake on the coast of the Pacific Northwest. It seems likely that the oceanic plate sinking through the trench off the Northwest coast is now stuck against the overriding continental plate. If so, the continental plate should bulge up; precise surveys confirm that expectation ( Figures 5-37 and 5-38). The locked zone is 50 to 100 kilometers off the coasts of Oregon, Washington, and southern British Uplift Between events Shortening CONTINENT OCEAN PLATE Locked Subsidence Extension Rupture ~100 kilometers Coast FIGURE 5-37. Denser oceanic plate a 4 As strain accumulates, a bulge risessinks in thesubduction zone. above sinking plate 4 while an area landward sinks. Those displacements reverse when the fault slips to cause an earthquake. 48° 0 0.3 1 2 British Columbia 48° British 2 Columbia 5 10 20 3 25 46° 3 h Trenc 2 Washington Washington 44° 44° 42° Vertical (mm/yr) 0 Oregon 42° Horizontal (mm/yr) California 100 kilometers 0 40° http://earthscience.brookscole.com/hyndman W3215.indb 121 Zone of maximum eastward movement of western edge of continent Oregon Oregon Zone of maximum uplift of bulging edge of continent Washington Washington Trench between the oceanic trench at the landward edge of the Juan de Fuca Plate and halfway to the coast. Convergence of the plates causes bulging of the edge of the North American Plate. Uplift rates are as high as 4 millimeters per year, and eastward transport is as high as 30 millimeters per year. Rates shown on both maps are in millimeters per year (mm/yr). 4 Trench FIGURE 5-38. The 4 subduction zone is locked 30 15 California 100 kilometers 40° 126°W 124° 126°W 124° TSUNAMI P. Flück, R. D. Hyndman, and K. Wang. 46° Roy Hyndman, Pacific Geoscience Center. EARTHQUAKE 121 2/10/05 2:54:35 PM Columbia. Just inland, the margin is now rising at a rate between 1 and 4 millimeters per year and shortening horizontally by as much as 3 centimeters per year. In these convergence zones, accumulating stress eventually breaks the bond between the sinking slab and the continental margin. Then the raised continental crust snaps back down along the coast and rises just offshore. The sudden drop of the coastal bulge and rise accompanying thrust fault movement generates a giant earthquake and a huge ocean wave. Computer numerical models estimate the heights of those waves as approximately 10 meters offshore. These heights would be amplified by a factor of two to three in some bays and inlets. Port Alberni, at the head of a long inlet on the west coast of Vancouver Island, B.C., in the 1964 Alaska earthquake, for example, had a run-up amplified by a factor of three compared with the open ocean. Although approximate, similar numbers are obtained from studies of onshore damage. The general pattern of ground movements described above is exactly like that in the Alaska earthquake of 1964. Even in Southern California, south of the Cascadia subduction zone, a near-field vertical-motion earthquake poses a potential problem. An earthquake on the Santa Catalina Fault offshore from Los Angeles would reach the community of Marina Del Rey, just north of the Los Angeles Airport, in only eight minutes. Given the large population and near sea-level terrain, the results could be tragic. Not all tsunami are in the Pacific Ocean. One in 1929 killed fifty-one people on the south coast of Newfoundland. On November 1, 1755, a series of large earthquakes in the Atlantic Ocean southwest of Lisbon, Portugal, wrecked the city and killed tens of thousands of people. The associated 10-meterhigh tsunami waves washed ashore, killing still more. Tsunami Hazard Mitigation Tsunami hazards can be mitigated by land use zoning that limits building to elevations above those potentially flooded and by engineering structures to resist erosion and scour. Coastal developments that orient streets and buildings perpendicular to the waves survive better that those that are aligned parallel to the shore. They limit debris impact and permit waves to penetrate without building higher. Landscaping with vegetation capable of resisting wave erosion and scour can help, as can trees that permit water to flow between them but slow the wave. But the trees need to be well rooted or they can themselves become missiles. A large ditch placed in front of houses can help reduce the level of the first wave, and may provide a little extra evacuation time. Surviving a Tsunami In summary, most tsunami are caused by earthquakes. ■ For a nearby subduction zone earthquake, you do not have much time before the first wave arrives, possibly ■ ■ ■ ■ fifteen to thirty minutes. You need to get to high ground or well inland immediately. A road heading directly inland is an escape route, but blocked roads and traffic jams are likely. Climb a nearby slope as far as possible, certainly higher than 30 meters. Do not return to the shore after the first wave. Although the sea may pull back offshore for a kilometer or more following that first wave, other even higher waves often arrive for several hours. Wait until officials provide an all clear signal before you return. Never go to the shore to watch a tsunami. Tsunami move extremely fast, and traffic jams in both directions are likely to require abandoning your vehicle where you least want to do so. Even without warning, an unexpected rise or fall of sea level may signal an approaching tsunami. Move quickly to high ground. Stay tuned to your radio or television. Tsunami waves appear much like ordinary breaking waves at the coast, except that their velocities are much greater and they are much larger. Some come in as high breaking waves, a high wall of water that destroys everything in its path. Others advance as a rapid rise of sea level, a swiftly flowing and rising “river” without much of a wave. Even those are extremely dangerous because they advance much faster than a person can run. Loose debris picked up as the waves advance act as battering rams that impact both structures and people. Even a strong swimmer caught in the swift current as the wave retreats will be swept out to sea. The Pacific Tsunami Warning System has two levels: a tsunami watch and a tsunami warning. A watch is issued when an earthquake of magnitude 7 or greater is detected somewhere around the Pacific Ocean. If a significant tsunami is identified, the watch is upgraded to a warning and civil defense officials order evacuation of low-lying areas that are in jeopardy. Tsunami Examples Some of the largest tsunami events on record include those shown in Table 5-1 (compiled from many sources). Seiches A seiche is a big wave in a large lake or enclosed bay that sways back and forth from one end of a basin to the other. The same thing happens in a bowl or bathtub if you move much water toward one end of the tub. Seiches form in larger bodies when water is disturbed by a large earthquake or landslide, a change in atmospheric pressure, or a storm surge. Seiches were first studied in Lake Geneva, Switzerland, in the 1700s, when people noticed that the water level at each end of the lake rose and fell almost a meter once an hour or so after a period of strong wind along the length of the 122 W3215.indb 122 2/10/05 2:54:36 PM Table 5-1 Examples of Large Tsunami Time and Place Cause Tsunami Arrival Site Height (m) Deaths B.C.1620 Santorini, Greece July 21, 365 A.D. Caldera collapse of ancestral Santorini volcano Earthquake in eastern Mediterranean Subduction earthquake (magnitude 8–9), coastal Washington and Oregon See pp. 118–122 Earthquake Eastern Mediterranean 6 Destroyed Minoan culture on Crete 50,000 in Alexandria Felled forests at heads of bays. Probably many deaths along coast. 30,000 + 20,000 in resulting fire 14 Jan. 26, 1700 Nov. 1, 1755 Lisbon, Portugal 1837 Chile Aug. 27, 1883 Krakatau, Indonesia June 15, 1896 Japan Mar. 2, 1933 Japan April 1, 1946 Unimak Island, Alaska July 9, 1958 Lituya Bay, Alaska May 22, 1960 Chile Greece, Egypt, Sicily West coast of Washington, Oregon (near field; would be 30–40 minute delay) Japan (far field) Lisbon 10 Hilo, Hawaii Volcano collapse See p. 111 Earthquake 6–36 Ͼ35,000 29 27,000 20 3,000 Aleutian Islands, Alaska; Hilo, Hawaii (far field) Waves 15 minutes apart Lituya Bay, Alaska 30 15 11 (Oahu) 33 to 524 159 (96 in Hilo) Coast of Chile; Hilo, Hawaii (far field) Honshu, Japan (far field) Anchorage and Seward, Alaska (near field; 30minute delay) Port Alberni, British Columbia Flores Island, Indonesia (near field) Masachapa (near field) 10.7 5.3 4.5 6 30 Ͼ2,000 61 in Hilo, Hawaii 122 in Honshu, Japan 125 26 Ͼ1,000 10 150 Okushiri, Japan (near field; 5-minute delay) Villages, north coast Papua, New Guinea (5–10 minute delay) Sumatra, Thailand, Sri Lanka, India, Somalia 11 200 12 Officially 2,134 (possibly 3,000) >10 >283,000 Sumatra and Java Earthquake Subduction earthquake Rockfall See p. 109 Subduction earthquake March 27, 1964 Prince William Sound, Alaska Subduction earthquake See pp. 106–107 Dec. 12, 1992 Indonesia Sept. 1, 1992 Nicaragua July 12, 1993 Hokkaido, Japan July 17, 1998 Papua New Guinea Magnitude 7.5 earthquake Dec. 26, 2004 Sumatra Subduction earthquake off northwest Sumatra See pp. 99–102 Earthquake Earthquake See p. 108 Undersea landslide triggered by earthquake lake. During a seiche, the displaced water moves the length of the lake, causing rise at the far end. That higher water sinks again to cause rise at the other end of the lake, and so on ( Figure 5-39). Because it takes time for movement of a large amount of water over a significant distance, larger basins have longer periods of oscillation. The same longer period, or lower frequency, is true of a longer pendulum, a child’s swing, or a larger earthquake. Their large wavelength and frequency of moving back and forth depend heavily on the basin size. 4 http://earthscience.brookscole.com/hyndman W3215.indb 123 18 2 Although tsunami have periods of approximately eight to eighty minutes in the open ocean, seiches in lakes or other more confined bodies of water typically have periods of less than ten minutes in small water bodies to several hours in larger ones. Wind-driven seiche effects are common in the Great Lakes. Strong westerly winds in mid-November 2003 caused a seiche with more than 4 meters of difference in water level from Toledo on the western end of Lake Erie to Buffalo on the eastern end of the lake ( Figure 5-40). Lake Erie 4 TSUNAMI 123 2/10/05 2:54:36 PM FIGURE 5-39. seiche in 4 by the size ofAthe basin.an enclosed basin oscillates back and forth at a frequency controlled FIGURE 5-40. Seiche levels in Lake Erie from 4 New York, to Toledo, Ohio, November 12– Buffalo, Lake Erie water level (meters above S.L.) 176 15, 2003. 175 Buffalo 174 173 Toledo 172 171 11/14/03 0 :0 12 00 0: 11/13/03 0 :0 12 00 0: 0 :0 12 00 0: 0 :0 12 00 0: 11/12/03 11/15/03 Date commonly experiences seiche effects because the winds tend to blow from west to east along the primary axis of the lake. The impact of such seiche events on coastal erosion is amplified because of large wind-driven waves. The winds that created the November 2003 seiche on Lake Erie gusted to more than 90 kilometers per hour, causing 3- to 5-meter waves on top of the seiche. Even hurricanes can cause pile up water. As the eye of Hurricane Frances passed close to Lake Okeechobee, Florida, in September 2004, it caused the south end of the lake to rise 3.6 meters higher than the north end. As the winds shifted later in the day, the north end of the lake rose to that height as the south end fell. Nearly 1,000 years ago, a large earthquake on the Seattle Fault deposited tsunami sand deposits along the shores of Puget Sound just north of Seattle. That earthquake also caused a landslide that submerged three areas of forest, drowned trees, and generated a seiche in Lake Washington, a narrow, 12-kilometer-long lake in the eastern part of Seattle. If a major earthquake were to occur on the Seattle Fault today, such a seiche would probably cause many deaths and severe damage to expensive housing around the lakeshore. The Potential for Giant Tsunami Tsunami in the historic record have been dramatic and sometimes catastrophic. Are even larger tsunami possible? Because tsunami waves form by sudden displacement of a large mass of water, they are generated by earthquakes, volcanic eruptions, and landslides underwater or into water. Earthquake-generated tsunami are most frequent, but their size in the open ocean is limited to the maximum displacement on an earthquake fault. Horizontal displacement underwater would not displace significant water. Extrapolation of its trend suggests that an earthquake of moment magnitude 8 from vertical displacement on a normal fault, the most likely type to displace significant water, could have a vertical offset of 15 meters. A thrustfault movement might have greater offset, but its gentler dip would likely cause a lesser vertical displacement of water. Because the tsunami wave height approximates the vertical displacement on a fault, the maximum wave height from an earthquake is a few tens of meters. As noted above, wave heights are amplified when waves are pushed into shallow water and bays. 124 W3215.indb 124 2/10/05 2:54:37 PM Tsunami generated by volcanic eruptions are occasionally catastrophic, but they are poorly understood; their maximum size is unknown. We do not know enough about the mechanism of water displacement from an underwater eruption to do much more than wildly speculate. Tsunami generated by landslides into water can be truly gigantic. The Lituya Bay tsunami of 1958 described earlier in this chapter created a 150-meter-high wall of water that surged more than 500 meters over a nearby ridge. The event began with a large earthquake on a nearby fault. The shock waves dislodged a large slab of rock and ice that fell into the bay to cause the tsunami. In this case, the earthquake was the initial culprit but not the direct cause of water displacement. The large size of a rock mass and the height from which it fell led to the tsunami’s great height. Gigantic submarine landslides in Hawaii and other oceanic volcanoes can be even more impressive. Occasionally, a huge slice of an island tens of kilometers wide collapses into the ocean, suddenly displacing thousands of cubic kilometers of water. The resultant tsunami, reviewed above, can be hundreds of meters high. None have happened in historic time, but it is only a matter of time. When KEY POINTS ✓ Tsunami have such long wavelengths that they always drag on bottom. Their velocity depends on water depth. Review Figure 5-25. ✓ Tsunami are caused by any large, rapid displacement of water, including earthquake offsets or volcanic eruptions underwater, landslides, and asteroid impacts into water. Review pp. 104 –105, 111–114; Figures 5-9 and 5-10. ✓ Tsunami, sometimes misnamed “tidal waves,” have nothing to do with tides. Review p. 99. ✓ Tsunami come as a series of waves, often tens of minutes apart. The largest waves are often the third or later to arrive. Review pp. 103 –104, 118; Figures 5-4 and 5-7. ✓ Tsunami can reach the coast within a few minutes from a nearby earthquake or many hours later from a distant quake. Review pp. 102–105, 116 –118. ✓ A subduction-zone earthquake can suddenly drop a low-lying coastal zone below sea level. Review p. 105; Figure 5-16. ✓ A volcano flank collapse that suddenly moves an enormous amount of water can generate giant tsunami that would be catastrophic for much of the East Coast of North America, especially lowlying coastal communities. Review pp. 113 –114; Figure 5-24. http://earthscience.brookscole.com/hyndman W3215.indb 125 it does happen, the low-elevation populations of Hawaii are in deep trouble. Presumably such a giant tsunami would also cross the Pacific to obliterate coastal communities of western North America, Japan, and elsewhere. Collapse of a flank of the Canary Islands, off the northwest coast of Africa, could generate a giant tsunami that would cross the Atlantic Ocean to obliterate coastal cities on the eastern coast of North America and perhaps those in western Europe. In these cases, we could have a catastrophe without needing coincidental overlapping events. Collapse on Reunion Island, a similar volcano in the Indian Ocean, could cause tsunami inundation and destruction of many coastal areas, including the dense sea-level populations of Bangladesh. Thus, it seems likely that a catastrophic tsunami, many times larger than any in historic time, is likely to come from the flank collapse of an oceanic volcano. Our geologic record of such events is clear enough to indicate that they have happened and will again. Limited evidence for collapses in Hawaii suggests an approximate recurrence interval of 100,000 years, but that is only a crude average. As noted throughout this chapter, however, the result could be truly cataclysmic. ✓ The impact of a large asteroid into the ocean would displace a huge amount of water and generate a massive tsunami. Review p. 114. ✓ Tsunami waves in the open ocean are low and far apart but move at velocities of several hundreds of kilometers per hour. They slow and build much higher in shallow water near the coast, especially in coastal bays. Review pp. 114 –116; Figure 5-28. ✓ Dangers from tsunami waves include drowning, impact from tsunami-carried debris, and severe abrasion from being dragged across the ground. Review p. 117. ✓ Tsunami from a Pacific coast subduction earthquake come every few hundred years and would come onshore within twenty minutes of the earthquake to destroy coastal communities, particularly those in bays and inlets. The safest areas are more than a kilometer inland and several tens of meters above sea level. Review pp. 118 –122. ✓ The record of subduction-zone tsunami is based on sand sheets over felled forests and marsh vegetation in coastal bays. Review pp. 118 –121; Figures 5-34 and 5-35. ✓ In between earthquakes, the leading edge of the continental plate slowly bulges upward before suddenly dropping during the earthquake. Review pp. 121–122; Figures 5-37 and 5-38. TSUNAMI 125 2/10/05 2:54:37 PM ✓ Danger signals for tsunami include a large earthquake and a rapid rise or fall of sea level. You can survive a tsunami by running upslope or driving directly inland immediately upon feeling an earthquake. Review p. 122. ✓ Seiches are back-and-forth swaying motions of water in an enclosed basin. ✓ Seiche surges in lakes are caused by strong winds that push up mounds of water as much as a few meters. Review pp. 122–124; Figures 5-39 and 5-40. IMPORTANT WORDS AND CONCEPTS Terms caldera collapse, p. 111 coastal bulge, p. 119 coastal inlets, p. 119 deepwater waves, p. 115 far-field tsunami, p. 116 giant debris avalanche, p.111 harbors, p. 116 narrowing of a harbor, p. 106 pali, p. 112 refract (waves), p. 103 run-up, p. 115 sand sheets, p. 120 seiche, p. 122 seismic sea wave, p. 99 shallowing water near shore, p. 106 submarine collapse, p. 112 submarine landslides, p. 125 trimline, p. 109 tsunami, p. 99 tsunami warning, p. 122 tsunami watch, p. 122 water displacement, p. 125 QUESTIONS FOR REVIEW 3. About how high are the largest earthquake-caused tsunami waves in the open ocean? 4. How does the height of a tsunami wave change as it enters a bay? Why? 5. How many tsunami waves are generated by one earthquake? 6. How fast do tsunami waves tend to move in the deep ocean? 7. Do tsunami speed up or slow down at the coast? Why? 8. Why is even the side of an island away from the source earthquake not safe from a tsunami? 9. For a subduction-zone earthquake off the coast of Oregon or Washington, how long would it take for a tsunami wave to first reach the coast? 10. Because the Atlantic coast experiences fewer large earthquakes, what specific other event could generate a large tsunami wave that would strike the Atlantic coast of North America? 11. What specific evidence is there for multiple tsunami events having struck coastal bays of Washington and Oregon? 12. What is a seiche? Explain what happens and what causes it. FURTHER READING Assess your understanding of this chapter’s topics with additional quizzing and conceptual-based problems at: http://earthscience.brookscole.com/hyndman. 1. What are three of the main causes of tsunami? 2. Of the three main types of fault movements—strike-slip faults, normal faults, and thrust faults—which can and which cannot cause tsunami? Why? 126 W3215.indb 126 CHAPTER 5 2/10/05 2:54:37 PM ...
View Full Document

This note was uploaded on 01/31/2012 for the course SCIE 027 taught by Professor Edward during the Fall '11 term at Broward College.

Ask a homework question - tutors are online