NATURAL AND HUMAN INDUCED HAZARDS – Vol. II Tsunamis ... · NATURAL AND HUMAN INDUCED HAZARDS ... Mitigation of Tsunami Hazards ... NATURAL AND HUMAN INDUCED HAZARDS – Vol. II

UNESCO – EOLS

S

SAMPLE C

HAPTERS

NATURAL AND HUMAN INDUCED HAZARDS – Vol. II – Tsunamis - Shuto N.

©Encyclopedia of Life Support Systems (EOLSS)

TSUNAMIS Shuto N. Iwate Prefectural University, Japan Keywords: earthquake, land slides, volcanic action, fault parameters, long-wave theory, Boussinesq equation, shallow-water theory, Green’s formula, shoaling effect, focusing effect, resonance effect, dispersion effect, bores, fires, hazard maps

Contents

1. Introduction 2. Causes of Tsunamis 3. Hydrodynamics of Tsunami from Generation to Coastal Effects 3.1. Generation 3.2. Propagation 3.3. Tsunamis in the Shallow Sea 3.4. Tsunamis in Nearshore and Runup on Land 3.5. Sound Generated by Approaching Tsunamis 4. Damages Caused by Tsunamis 4.1. Magnitude and Intensity of Tsunami 4.2. Loss of Human Lives 4.3. Damage to Houses 4.4. Damage to Fishing Boats 4.5. Fire Caused by Tsunamis 4.6. Damage to Aquaculture Rafts 5. Mitigation of Tsunami Hazards 5.1. Sea Walls and Coastal Dikes 5.2. Tsunami Breakwaters 5.3. Effectiveness of Tsunami Control Forests 5.4. Movement of Residence to Tsunami-Free High Ground 5.5. Tsunami-Resistant Building Zone 5.6. Tsunami Forecasting 5.7. Tsunami Hazard Maps 5.8. Continuation of Tsunami Disaster Culture 6. Needs in the Near Future Glossary Bibliography

Summary

Tsunamis are often triggered by submarine earthquakes. In order to estimate an initial

UNESCO – EOLS

S

SAMPLE C

HAPTERS



profile of a tsunami from an earthquake’s fault parameters, the elastic dislocation model is usually used with the assumption that the slip occurs uniformly on the fault plane in a semi-infinite elastic body. Actual tsunami is sometimes quite different from this assumed model profile. Efforts have been made to fill this gap. In the sea deeper than 200 m, a tsunami can be modeled reasonably with the linear long wave theory. In the shallow sea, nonlinear effects become nonnegligible and the fully nonlinear shallow-water theory is used. Tsunamis increase their height in the shallow sea, due to the shoaling, focusing, resonance, and dispersion effects. Tsunami “magnitude” is a measure to express the total energy of a tsunami, while tsunami “intensity” is a measure to express the strength of a tsunami at a given location. Damages to human lives and properties are given as a function of tsunami intensity. Fires induced by a tsunami often give devastating damages, especially when they are associated with oil spill. Tsunami countermeasures consist of three parts: hardware such as sea walls, city planning such as relocation of residence and introduction of tsunami-resistant building zones, and software such as forecasting, evacuation drills, and continuation of disaster culture. 1. Introduction Earthquakes are the primary cause of most tsunamis. But tsunamis are a more infrequent phenomenon than are earthquakes. A weak earthquake does not generate a tsunami. A great earthquake with the hypocenter at a deep location does not generate a tsunami. Neither does an earthquake caused by strike-slip fault movement generate a tsunami. The recurrence interval of huge tsunamis is usually much longer than a life span of human beings. Even a small tsunami occurs very infrequently. Difficulties in understanding tsunamis come from this infrequent occurrence. It is difficult to understand the whole picture of a tsunami. Each tsunami is quite different from another. Each tsunami behaves in quite different movements at different places, corresponding to different topography. Tide records provide quantitative time-series data, but there are two problems. One is the distribution and number of tide stations. Compared to the spatial extent of a tsunami, tide stations are often too sparse. In addition, tsunami data taken by tide gages are biased and deteriorated by the hydraulic characteristics of the tide well. High-frequency components of the water-surface fluctuations are cut off or filtered. Other kinds of data, more rich in number, are tsunami traces measured in the post-tsunami survey. They usually show the highest water level made by the tsunami but can not suggest any hydrodynamics or time histories of the tsunami runup. They are “fossils of tsunami." Hence, difficulties in understanding tsunamis also come from the lack of reliable data. Once it has occurred, a tsunami gives devastating effects on the coastal community. The last and best way to save human lives is an early evacuation guided by forecasting and warning. The second best is to make a coastal city tsunami resistant. It is quite difficult for

UNESCO – EOLS

S

SAMPLE C

HAPTERS



coastal residents to continue to be alert to tsunami hazards based on former experiences. At such places as the west coast of the United States where there are no written records at all about huge tsunamis some 300 years ago, except for one proof (i.e., sediments transported by the tsunami), it is impossible for the residents to learn of the former event. If a newly developed coastal resort community, without prior tsunami records, finds one day that they are located in the tsunami risk area in front of the subduction zone, how can they prepare for possible tsunamis? During these 30 years, the areas of tsunami science and engineering made big progress, assisted by advancement in seismology and computer science. Seismology made it possible to estimate the initial profile of tsunamis from fault parameters determined by seismic information. The nonlinear shallow-water theory is used to solve for tsunamis on complicated topography with the aid of a big high-speed computer. The technique thus developed is now applied to practical defense works in many fields, forecasting, hazard maps, design of defense structures, and so on. 2. Causes of Tsunamis The sea surface always fluctuates. There are three causes for wave motions: meteorological, astronomical, and the rest. Waves generated by winds become swells after they leave the wind field. Typical wave periods of wind waves and swells are 3 to 30 seconds. Moving atmospheric low pressure together with the winds accompanying it causes the sea-surface rise, called the storm surge. A typical storm surge period is several hours to half a day. Tides are more-regular oscillations caused by the pulling gravitational forces of the moon and the sun. The fundamental period of tides caused by the moon is 12 hours and 25 minutes. Water waves generated by other causes are tsunamis. Its period ranges from a few minutes to two hours. The first person who recorded a tsunami and thought that the tsunami was generated by an earthquake was Thucydides, a Greek historian. In summer of 426 BC during the Peloponnesian war, the sea receded after an earthquake, and then a huge wave hit the city of Orobiae at the northwestern coast of Euboea Island. A part of the city subsided into the sea. The island of Atlanta at the other side of the strait was also hit by the tsunami. A part of Athenian fortifications was swept away. Thucydides considered that the full force of the earthquake drew the seawater from the shore and then the sea suddenly swept back again even more violently. Most of the causes of tsunamis are submarine earthquakes. However, not the ground shaking but the vertical sea-bottom deformation generates a tsunami. The greater the earthquake is, the larger the vertical displacement of sea bottom is; hence the greater the tsunami that is generated. An earthquake that triggers a tsunami is often called the tsunamigenic earthquake.

Comment [GB1]: Please provide a

specific time span. Since 1970?

Comment [GB2]: And what swould the

rest be? Sounds like there are two causes

that you know of.

UNESCO – EOLS

S

SAMPLE C

HAPTERS



There are exceptions for this rule. Much larger tsunamis than expected from the earthquakes’ seismic waves can be generated. This type is called the tsunami earthquake. After several weak earthquakes on 15 June 1896, another weak earthquake was felt along the shore of the Sanriku District, Japan, at 1930 (local time). No one paid special attention to this earthquake and tried to evacuate. Half an hour later, a giant tsunami hit the coast and claimed more than 22 000 human lives. The highest runup was nearly 40 m. This was a typical tsunami earthquake. The mechanism of the tsunami earthquake in the Sanriku coast was thought to be a large but slow rupture of the earthquake fault. A huge tsunami can cause damage to remote places after travelling across the ocean. This is called a distant tsunami, remote-source tsunami, or far-field tsunami. At 1911, on 22 May (GMT) 1960, an earthquake of Ms = 8.5 or Mw = 9.5 struck off the Chilean coast. A giant tsunami was generated and hit the Chilean Coast first as a local tsunami or a near-field tsunami. Its maximum runup height was estimated as high as 20–25 m. The tsunami spread over the Pacific Ocean, gave damage to Hawaii, then concentrated toward Japan at the antipode of the tsunami source after a 22.5-hour journey. The whole coast on the Pacific Ocean side of Japan, more than 3000 km long, were affected and damaged. Its tsunami height was 3–6 m. Landslides can also generate tsunamis. Lituya Bay in Alaska repeatedly experienced huge local tsunamis in 1958, 1936, 1899, 1874, 1853–1854, and also probably in 1900. This bay is about 11 km long, 1 km wide, and 160 m deep. On 10 July 1958, an earthquake caused 30 × 106 m3 of rocks weighing 90 × 106 tons to slide from the northern shore from an average height of 600 m with the dimensions of 700 m to 900 m and an average thickness of 90 m. The slide forced the water surge up to the height of 520 m on the opposite shore. Then, the water ran down into the bay to form a huge tsunami higher than 30 m in the bay. Another mechanism is volcanic actions. In May 1883, the volcanic eruption of Krakatau in the Sunda Strait between Java and Sumatra Islands, Indonesia began. On 27 August, a giant tsunami was generated. Because of thickly falling ashes, no one could see the tsunami offshore. When coastal residents noticed the white cap of the tsunami, the tsunami was just in front of them. There was no time for evacuation. It claimed more than 36 000 human lives. Its maximum runup was higher than 30 m. The comparison of the topographies before and after the eruption indicates that two-thirds of the original Krakatau island was blown away, leaving an area 200 m deep and about 10 km wide. Among several generation mechanisms proposed, caldera formation is the most probable. 3. Hydrodynamics of Tsunami from Generation to Coastal Effects

3.1. Generation

UNESCO – EOLS

S

SAMPLE C

HAPTERS



A fault movement generates shaking (earthquake) and displacement of ground. If a submarine fault is shallower than 80 km, the displacement might penetrate to the surface of sea bottom. In order to imagine the generation of a giant tsunami, let us suppose that an area of sea bottom, a few hundred kilometers long and several tens to 100 kilometers wide, moves vertically by several meters within 100 seconds. The water a few kilometers thick above the area has no time to flow outward and consequently the water surface will show the same vertical displacement as the sea bottom. In this way a tsunami is born. No person could and/or will be able to measure the tsunami initial profile, but one can estimate it by calculation or, very rarely, by direct measurement of the sea bottom displacement. The most popular way of estimating an initial tsunami profile is to assume that the sea bottom displacement is a result of the fault movement in a semi-infinite, elastic, homogeneous body. A fault movement is described by its location including its depth, geometrical characteristics (strike, dip, and slip angles of the fault plane), physical characteristics (length, width, and dislocation of the fault plane) and dynamic characteristics (rupture direction, rupture velocity, and rise time of the fault movement). With fault parameters (except for dynamic characteristics), the static displacement of sea bottom can be computed. The assumption of a homogeneous movement in a fault plane leads us to a simple tsunami profile: one crest and one trough in the area of the generation. Efforts have been made to obtain more realistic initial profiles. If a rupture process is well recorded and analyzed, plural fault planes, if they exist, can be determined with fault parameters for each plane. Heterogeneous movement in a fault plane is estimated from inversion of seismic data or of tsunami data with geodetic data. In place of the semi-infinite body model, the multilayered model can be introduced and numerically solved. The introduction of heterogeneity changes the simple initial profile (one crest and one trough) to a complicated one. A tsunami initial profile determined with seismic data alone usually does not explain the tsunami and total tsunami energy. It is necessary to modify this first solution with results of tsunami simulation. First, compute a tsunami for the initial profile determined from seismic data and output the tsunami heights along the 200 m water-depth contour near the shore. Compare these results with the measured runup heights averaged in the interval of about 15 km along the open coast. The latter is usually two to three times the former, according to accumulated experiences in numerical simulation. Adjust the initial height to satisfy this condition. This method is widely used to determine the initial condition. There is no accurate measurement of the vertical displacement of sea bottom surface caused by a submarine earthquake except for the case of the 1964 Great Alaska earthquake. The deformation of ground was reconstructed by the displacement on islands and by comparison with a pre-earthquake topography. Along the direction normal to the

UNESCO – EOLS

S

SAMPLE C

HAPTERS



long axis of the deformation area, the vertical displacement (or the two-dimensional tsunami profile) shows a gentle wavy shape 450 km long with one crest and one trough. Its trough-to-crest wave height was about 6 m. Near its crest, there was a sharp rise about 6 m high and about 30 km wide at its base. This rise was concluded as a result of a subfault developed in the accretionary prism. It is impossible to detect this kind of subfault from seismic information at present, although its contribution to tsunami heights is important. There are several efforts to estimate tsunamis generated by landslides, submarine landslides, volcanic action, and others. Most of them are carried out based upon the measured tsunami data. Geodetic data alone are insufficient in accuracy to estimate the generation mechanism because the movements of landslides are the key factor to determine the tsunami generation efficiency. -

-

-

TO ACCESS ALL THE 20 PAGES OF THIS CHAPTER,

Please register For One Year Free Trial Access at: http://www.eolss.net/EolssRegister/SindRegister.aspx

Using the Preferential Code: SA9253 Enjoy this provision during the UN Decade of Education for

Sustainable Development.

Bibliography

Bernard E. (1999). Tsunami. Natural Disaster Management (J. Ingleton, ed.), pp. 58–60. Leicester, UK: Tudor Rose Holdings Limited. [This is a part of a presentation to commemorate the International Decade for Natural Disaster Reduction. A way to establish the tsunami-resistant community is proposed.] Iida K. (1963). Magnitude, energy, and generation mechanisms of tsunamis and a catalogue of earthquakes associated with tsunamis. IUGG Monograph No.24. pp.7–18. [The Imamura scale for tsunami magnitude m ranging from 0 to 4 is extended to the Imamura-Iida scale m ranging from –1 to 4.]

Kajiura K. (1970). Tsunami source, energy and the directivity of wave radiation. Bulletin of the Earthquake Research Institute, University of Tokyo 48, 835–869. [This paper makes clear the importance of the linear dispersion term in the long wave theory, in case of tsunami propagation over Earth. The limit of application of the linear long-wave theory is determined by a parameter including the distance of propagation, the size of initial profile, and the water depth. A diagram is given to indicate the range of the validity of the

Deleted: 3.2. Propagation¶

When generated, a tsunami has a

wavelength several tens of kilometers long

that is much longer than the water depth, at

most a few kilometers. For example, the

average water depth in the Pacific is about

4.2 km. This water wave is categorized as

long waves, for which the hydrostatic

water pressure is a good first-order

approximation. The initial height of the

tsunami several meters high is quite small

compared to the water depth and the

wavelength. This water waves belongs to

small amplitude waves, for which the

linear wave theory is applicable.¶

¶

A near-field tsunami in the ocean deeper

than approximately 200 m can be analyzed

with the linear long-wave theory. The

velocity of its energy propagation is the

same as its phase velocity that is the square

root of the product of the gravitational

acceleration and the water depth for the

first-order approximation. Every

component of different frequencies is

approximated to propagate with the same

velocity. A linear long wave propagating

on the water of constant depth, therefore,

does not change its wave profile.¶

¶

Comment [GB3]: Please spell out the

publisher's name and provide the

publisher's location.

... [1]

UNESCO – EOLS

S

SAMPLE C

HAPTERS



nondispersion assumption.]

Kanamori H. (1972). Mechanism of tsunami earthquakes. Physics of Earth and Planetary Interiors 6, 246–259. [A tsunamigenic earthquake follows the rule that the bigger an earthquake is, the bigger the tsunami is generated. This paper introduces the concept of the tsunami earthquake that generates a great tsunami even if the earthquake is small.]

Okada Y. (1985). Surface deformation due to shear and tensile faults in a halfspace. Bulletin of the Seismological Society of America 75, 1135–1154. [A complete suite of closed analytical expressions is presented for the surface displacements, strains, and tilts due to inclined shear and tensile faults in a half-space. Misprints in Mansinha-Smylie (1971) are corrected in Appendix.]

Shuto N. (1987). Effectiveness and limit of tsunami control forests. Coastal Engineering in Japan, Japan Society of Civil Engineers 30, 1–19. [Based upon data collected in Japan, diagrams are obtained to estimate effectiveness of coastal forests on reducing tsunami energy, in terms of number of trees, thickness of undergrowth, and tsunami height.]

Shuto N. (1991). Historical changes in characteristics of tsunami disasters. Natural Disaster Reduction and Civil Engineering, Japan Society of Civil Engineers, pp.77-86. [A simple estimate of the burnt area is given in terms of the volume of oil stored in the tsunami-risk zone. Damages to boats and houses as well as loss of human lives are also discussed.]

Shuto N. (1993). Tsunami intensity and disasters. Advances in Natural and Technological Hazards Research, Vol.1, pp.197-216. Dordrecht, Boston, London: Kluwer Academic Publishers. [In terms of the tsunami intensity, a measure of the local strength of tsunami, wave profiles of near-field tsunami and degree of damages are estimated.]

Shuto N. (1997). A natural warning of tsunami arrival. Advances in Natural and Technological Hazards Research, Vol.9, pp.157-173. Dordrecht, Boston, London: Kluwer Academic Publishers. [Abnormal sounds induced by tsunamis are classified and analyzed quantitatively in relation to the wave profiles and topographical condition.]

Simkin T. and R.S. Fiske. (1983). Krakatau 1883, Eruption and its Effects. 464 pp. Washington, DC: Smithonian Institution Press. [A thorough compilation of documents of the 1883 event in Sunda Straits, Indonesia.]

Tatehata (1997). The new tsunami warning system of the Japan Meteorological Agency, Advances in Natural and technological Hazards research, Vol.9, pp.175-188. Dordrecht, Boston, London: Kluwer Academic Publishers. [This paper gives a brief explanation of the JMA’s new warning system using a database built with the aid of numerical simulation for more than 100 000 cases.]

Thucydides. (for example): History of The Peloponnesian War, in Penguin Classics, p.247. [The first document of tsunami. The author recognized that the precedent earthquake was the cause of the tsunami.]

Yoshioka S., M. Hashimoto, and K. Hirahara (1989). Displacement fields due to the 1946 Nankaido earthquake in a laterally inhomogeneous structure with the subducting Philippine Sea Plate—A three-dimensional finite element approach. Techtonophysics 159, 121–136. [Layered models (vertically inhomogeneous model) are used to calculate the static displacement of sea bottom in place of the

Comment [GB4]: Please provide the

editor's name and the publisher's location.


editor's name (unless this is a periodical).




author's first initial.



Comment [GB9]: Please provide a

publication date, and the publisher's

location.

UNESCO – EOLS

S

SAMPLE C

HAPTERS



conventional Mansinha-Smylie method that assumes a semi-infinite, homogeneous, elastic body.]

UNESCO – EOLS

S

SAMPLE C

HAPTERS

Page 6: [1] Deleted Guest43 6/9/2011 12:49:00 PM

3.2. Propagation

When generated, a tsunami has a wavelength several tens of kilometers long that is much longer than the water depth, at most a few kilometers. For example, the average water depth in the Pacific is about 4.2 km. This water wave is categorized as long waves, for which the hydrostatic water pressure is a good first-order approximation. The initial height of the tsunami several meters high is quite small compared to the water depth and the wavelength. This water waves belongs to small amplitude waves, for which the linear wave theory is applicable. A near-field tsunami in the ocean deeper than approximately 200 m can be analyzed with the linear long-wave theory. The velocity of its energy propagation is the same as its phase velocity that is the square root of the product of the gravitational acceleration and the water depth for the first-order approximation. Every component of different frequencies is approximated to propagate with the same velocity. A linear long wave propagating on the water of constant depth, therefore, does not change its wave profile. For a higher-order approximation, the phase velocity is influenced by the dispersion effect depending upon frequency. Wave components of different frequencies propagate with different velocities. This difference, although very small, results in a non-negligible deformation in wave profile, if the travel time becomes long as in case of a far-field tsunami. A parameter pa is used to judge whether the dispersion effect should be included or not: pa = (6h/R)1/3(a/h) where h is the water depth, a the horizontal dimension of the tsunami source, and R the distance from the source. If pa < 4, the dispersion effect should not be neglected. Under this condition, the linearized Boussinesq equation that includes the first-order effect of the phase dispersion should be used. The equation should also be modified by including the Coriolis effects and expressed with the spherical coordinates. For a huge tsunami, such as the 1960 Chilean tsunami, the Pacific Ocean behaves like a small pond. Average water depth of the Pacific Ocean, 4.2 km, gives the tsunami a propagation velocity faster than 730 km hr–1. The tsunami traveled 17 000 km from the source off Chilean coast to Japan within 23 hours. It started toward Japan with the crest at its front but arrived at Japan with a big ebb. The first crest became unrecognizably small and the following trough began to grow near the Hawaiian Islands. This change was the

UNESCO – EOLS

S

SAMPLE C

HAPTERS

result of the dispersion due to the Coriolis effect. The energy of the 1996 Irian Jaya tsunami was effectively transported to Japan although Japan is not located on the major direction of initial tsunami energy radiation. This is explained by the existence of the south Honshu ridge that acted as an effective wave guide. The shallower the water is, the slower is the wave propagation. Waves change their propagation direction toward the shallower ridge crest. This refraction causes concentrating and effectively transporting tsunami energy along an oceanic ridge. Similar effects can be expected on the continental shelf, which is another wave guide. Tsunamis that enter the sea on the continental shelf are refracted toward the shore, reflected from the shore to the sea, and refracted again toward the shore. Typical oscillation characteristics of tsunamis thus propagating along a sea ridge or a continental shelf as edge waves are the beat in the time history that the tsunami wave height gradually increases and the highest wave arrives later. For the farther tsunami source, the later the highest wave appears.

3.3. Tsunamis in the Shallow Sea

The tsunami energy propagates in relation to the water depth. The rate of energy transmission (the product of wave energy per unit sea-surface area and the energy propagation velocity) is constant, if there is no energy loss due, for example, to sea-bottom friction. The shallower the water depth is, the slower the velocity is, and therefore, the higher the wave height becomes. This is the shoaling effect. Consider a bay with a broad entrance and narrowing towards inland. Tsunamis are usually much longer than the bay’s length. Tsunami energy concentrates toward the head of the narrow bay, hence tsunami height increases. This is the focusing effect. The similar effect occurs between two wave rays. (The ray is the trajectory of wave propagation.) The linear long wave theory gives a simple relation of the Green formula, among wave height H, water depth h, and width of the bay or normal distance between two wave rays b, Hh1/4b1/2 = constant. Another important amplification mechanism is the resonance. If an external force is applied to the water in a bay and then removed, the water begins to oscillate with its natural period and the oscillation gradually diminishes due to energy dissipation. The fundamental period of the natural oscillation, T, for a rectangular shaped bay is given by T = 4l/(gh)1/2,

UNESCO – EOLS

S

SAMPLE C

HAPTERS

where l and h are the length and water depth of the bay, and g the gravitational acceleration. If a tsunami having the same period as the natural period enters a bay, energy is stored and the resonance occurs in the bay. Three waves in succession are enough to establish the resonance for tsunamis. Even if a tsunami starts with a simple initial profile of one crest and one trough, the tsunami shows very complicated movements on the shore, because of refraction, reflection, and diffraction caused by the topography. In 1983, the Nihonkai Chubu (Middle Japan Sea) earthquake tsunami hit the north Akita coast, which has a smooth shoreline 55 km long. The relatively straight coast is bounded at the both ends by a peninsula and a rocky cliff projecting normal to the coast. Residents near the ends recognized only one or two waves within half an hour after the earthquake. But the tsunami repeatedly hit other places along the coast. A fisherman on his boat near the center of the coast experienced seven waves during two and half hours after the first wave. This was caused by the complicate refraction-reflection effects of the coast. In the shallow sea, the tsunami height increases and becomes comparable with the water depth. To compute the tsunami in the shallow sea, the fully nonlinear shallow-water theory including the effect of bottom friction is usually used. The theory still assumes the hydrostatic pressure but takes the finiteness of the wave amplitude into consideration. The first-order linear phase velocity is modified: the total water depth including the change in water surface elevation is used in place of the water depth in the linear phase velocity. The second-order phase velocity includes the effect of water surface elevation. In the shallow sea, this effect makes the higher portion of the wave proceed faster. The frontal wave slope becomes steeper. If the water particle velocity at the front exceeds the local phase velocity, the water projects into the air. Consequently, a breaking bore is formed. Under some conditions, the third-order phase velocity including the effect of surface curvature (dispersion effect) becomes important. This phase velocity that decreases at the crest and increases at the trough prevents the front from breaking. Then, a short-period wave train begins to form at the front, which leads to the formation of an undular bore or a cnoidal bore. This process is called the soliton fission. The undular bore is often observed in rivers and shallow seas with a very gentle bottom slope. When a single positive long wave of permanent form (i.e., a solitary wave) enters the shallower region such as a continental shelf, the wave is often transformed into a series of shorter solitary waves. This phenomenon is called the soliton fission. The higher-order nonlinear theory is required to model the soliton fission. There is no theory applicable to the two-dimensional soliton fission, at present.

UNESCO – EOLS

S

SAMPLE C

HAPTERS

3.4. Tsunamis in Nearshore and Runup on Land

In the case of a far-field tsunami that usually has a longer period than one hour, rise and fall of the sea water surface is similar to a rapid astronomical tide. In rivers, undular bores often develop at the front. In case of a near-field tsunami of 5–10 minute period, the wave profile is classified into four types, depending upon bottom slope, tsunami height, and number of waves. Type I is for a tsunami of gentle wave steepness on a steep bottom slope. It is described as “the water level made a slow rise like a tide,” “the tsunami was like a rapid tide,” or “the tsunami quietly advanced shoreward and rose suddenly at the breakwater.” Some of them have the breaking short waves at the front. Type II is characterized by “a rapid growth near the shoreline, although the tsunami is not recognized in the offing." This occurs on a relatively steep bottom slope. It is described as “the water level rapidly swells near the shore,” “the water swells from the bottom,” or “the water level is raised by a train of short waves, the succeeding waves overtaking and lying upon the preceding one.” Some of them, if they are high, show spilling breaking near the crest. Type III is for tsunamis on a relatively gentle bottom slope. In the offing, they were recognized as “a tsunami like a bank, like a stretched curtain or like a wall with splash at the crest." Half of them are accompanied with spilling breaker, even though they are not so high. Type IV always shows a plunging breaking. The second, third, or later waves of relatively low height (the smallest is about 2 m) may show this breaking when they meet the receding current of the preceding waves. The first wave higher than 7 m may show plunging breaking. A tsunami is a transitional wave. It was observed that the tsunami height increased from the first wave to the second or third waves and that the third wave was often the highest wave. Correspondingly, the runup height caused by the third wave was usually the highest. In case of the Nihonkai-Chubu earthquake tsunami, the second wave was the highest, higher than 20 m in the offing. Its runup was, however, much smaller than that of the small first wave, because the second wave met the receding current of the first wave and was stopped. An obliquely incident breaking bore forms an edge bore, which propagates sometimes following the ordinary refraction law and sometimes neglecting the topography. A small difference in the side boundary condition introduces a big difference in wave profile, suggesting also a big difference in wave force. No theory and no simulation method applicable to this edge bore exist at present.

3.5. Sound Generated by Approaching Tsunamis

UNESCO – EOLS

S

SAMPLE C

HAPTERS

Many witnesses said that they heard abnormal sound generated by tsunamis. A study on near-field tsunamis revealed a relationship among the tsunami height, topographical condition, and sound characteristics. When a tsunami, forming a plunging breaker higher than 5 m, hits coastal cliffs, a loud sound like a thunder is generated and is heard at distant places. It is expressed as a distant thunder, distant explosion, or distant firing of a cannon. When a tsunami higher than 2.5 m with a front in the form of a spilling breaker proceeds in the shallow water, a continuous sound like a locomotive, a group of heavy trucks, or an approaching storm, is heard in the area. These sounds can be used as the signal of a coming tsunami. 4. Damage Caused by Tsunamis

4.1. Magnitude and Intensity of Tsunami

On preparing a historical tsunami catalogue in Japan, a relative scale m of tsunami strength was established on the basis of available evidence (though poor in descriptions) on the tsunami height and the length of the coastline where tsunami activity was significant. The scale is known as the Imamura-Iida magnitude scale and is described in the following way: m = –1 Minor tsunamis having a wave height <50 cm m = 0 Tsunami having a wave height on the order of 1 m and causing no appreciable damage m = 1 Tsunami having a wave height on the order of 2 m causing damage to houses along the coast or to ships washed ashore m = 2 Tsunami having a wave height on the order of 4–6 m causing the destruction of some houses and considerable loss of life m = 3 Tsunami having a wave height of 10–20 m and producing a damaged area of about 400 km in length along the coast m = 4 The severest tsunami, having a maximum wave height of >30 m and producing a large damaged area of >500 km in length along the coast. Another more practical and quantitative tsunami scale is the Hatori scale mH based upon measured data: mH= 2.7 log H + 2.7 log D – 4.3

UNESCO – EOLS

S

SAMPLE C

HAPTERS

where H is the average inundation height in meters above mean water level and the maximum double amplitude observed by tide-gauges, and D is the epicentral distance in kilometers measured along the shortest oceanic path. This formula is applicable to the case D < 2000 km. This tsunami magnitude scale can be estimated at intervals of 0.5, and was designed to match the Imamura-Iida scale m. This magnitude scale is closely related to the potential energy of the tsunamis. Each change of one unit of magnitude implies a change of 2.24 times in the tsunami height and five times in the tsunami energy. In order to find relationships between local tsunami strength and the degree of damage based upon the data in historical documents, the local tsunami height H in meters or the tsunami intensity i is used: i = log2H. The definition of H varies for different subjects. For damage to fishing boats, H is the tsunami crest height above ground level at the shoreline. For damage to individual houses as well as for effectiveness of and damage to coastal forests, H is the inundation height at the site. For damage to aquaculture rafts, H is the maximum tsunami crest height above the mean sea water level at the raft location. For the percentage of damaged houses in a coastal village, H is replaced by the maximum runup height in the village HR, and the corresponding intensity is denoted as iR.

4.2. Loss of Human Lives

Loss of human lives depends on the actions of persons on the coast when an earthquake occurs. At night on 12 July 1993, a giant tsunami hit the fishing village of Aonae, Japan, 4 or 5 minutes after an earthquake. Most villagers immediately began to evacuate to higher ground when they felt the earthquake. This action saved the lives of 85 % of the at-risk population. In contrast, a tsunami that hit Warapu, Papua New Guinea, at night on 17 July 1998, claimed the lives of more than half of the at-risk population. Most of residents did not know what would come after the earthquake and did not try to use the precious 20 minutes of time between the earthquake and the tsunami arrival. According to the historical records, loss of human lives begins when the tsunami height exceeds 2 m (or i = 1). Here, the tsunami height is defined as the rise of water surface due to the tsunami above sea water level. The percentage of death increases with the tsunami height, but with a large scattering of more than two order of magnitude difference, reflecting the difference by the reaction of residents. The tsunami height of 2 m does not mean the thickness of water flow on land. When the 1983 Nihonkai-Chubu earthquake

UNESCO – EOLS

S

SAMPLE C

HAPTERS

tsunami hit the Jusan-ko, Aomori, Japan, there were nine persons enjoying fishing on the sandy beach. During the evacuation, all of them were caught by the shallow water flow of 70 cm thick, and three of them lost their lives.

4.3. Damage to Houses

The damage percentage of houses RHD in a given village is defined as follows. RHD = (a + 0.5b)/(a + b + c) where a is the number of houses washed away and completely destroyed, b is the number of houses partially damaged, and c is the number of houses flooded. In case of a village of wooden houses on low land, damage begins at iR = 1 (or runup height = 2 m). At iR = 2, RHD=0.5 (i.e., 50% of houses in the flooded area are demolished). At iR = 3, all the houses in the flooded area are demolished. Wooden houses are weak. On an average, a wooden house is completely destroyed if the tsunami height above ground exceeds 2 m. Not only the tsunami force but also impact of such materials as fishing boats, lumber, and houses transported by tsunamis are destructive. Reinforced-concrete buildings are strong, as is mentioned in many historical documents. There is only one example of a destroyed concrete building. That was the lighthouse 18 m high on ground 10 m high at Scotch Cap on Unimak Island by the 1946 Aleutian tsunami. The crest height of the tsunami was estimated as high as 30 m above sea water level or the tsunami height above ground was 20 m. When the tsunami height is lower than 5 m above the ground level, reinforced-concrete buildings are resistant enough to protect weak (e.g., wooden) houses behind them. There is no example for the cases between the tsunami height of 5 m and 20 m.

4.4. Damage to Fishing Boats

Similarly to the damage percentage of houses, the damage percentage of fishing boats is defined as follows: RBD = (a + b + 0.5c +0.25d)/(a + b + c + d + e) where a, b, c, d, and e are the number of boats washed away, seriously damaged, half damaged, slightly damaged, and not damaged, respectively. Roughly speaking, at i = 1, fishing boats begin to be damaged. At i = 2, RBD = 0.5. At i = 3, RBD = 1. On an average, bigger boats are safer than small boats.

UNESCO – EOLS

S

SAMPLE C

HAPTERS

4.5. Fire Caused by Tsunamis

At night on 27 January 1700, the village of Miyako on the Sanriku coast, Japan, was suddenly hit by an unusual tide without feeling any earthquake. A fire started from a house overturned by the tide, and as a consequence, 21 houses were burnt. This is the first record of fire caused by tsunamis. This tsunami was possibly originated from the earthquake at the Cascadia subduction zone off the western coast of North America. If an earthquake or a tsunami damages oil tanks, and if the oil spread by the tsunami catches fire, the consequence is devastating. There are four examples for such disasters; Seward, Valdez, and Whittier in Alaska in the 1964 Great Alaska earthquake and Niigata in Japan in the 1964 Niigata earthquake. During the first stage of oil spread, the gravitational force is the dominant driving force. This is the gravity-inertia regime in the momentum balance. Then, the gravity-viscous regime appears where the viscous force becomes the major resisting force. Finally, the surface tension takes a role of the major driving force: this is the surface-tension-viscous regime. It was found that the burned area corresponds well to the boundary between the second and third regimes. An estimate of the burned area AB in square meters thus found is given by AB = 320V where V is the volume of spilled oil in kiloliters. A numerical method was developed to simulate the spread of oil caused by tsunamis. This method gives the area and the position of spilled oil for the first and second regimes, and is sufficiently used to estimate the burned area.

4.6. Damage to Aquaculture Rafts

Damages to rafts for culture pearls is classified as follows. “Damaged” means that a raft is washed away from its original position, destroyed, or sunk, and more than 70% of the mother pearl shells are lost. “Partially damaged” means that the raft is washed away, collides with other rafts and 20–30% of the mother pearl shells are lost. “Undamaged” means that, even if the raft is moved, it is not washed away and the mother shells are safe. Damages to the rafts do not depend upon the change in water level but on the strength of current induced by tsunamis. If the maximum velocity does not exceed 1 m s–

1, the rafts are safe even for a tsunami higher than 3 m. In shallow bays and channels, a tsunami 1 m high could induce the current of 1 m s–1, hence, damages to the rafts occur.

UNESCO – EOLS

S

SAMPLE C

HAPTERS

5. Mitigation of Tsunami Hazards Tsunami countermeasures consist of three parts: structures, city planning, and systems. Sea walls and coastal dikes are the typical defense structures. Tsunami breakwaters and tsunami gates are constructed, but not many. Coastal forests are also classified into the defense structures. The major items in the city planning are movement of residences to “tsunami-free” high ground and establishment of the tsunami-resistant building zone. The tsunami prevention system consists of warning, evacuation, public education, drills, inheritance of disaster culture, and the relief operation system after disaster.

5.1. Sea Walls and Coastal Dikes

The direct method to stop the tsunami invasion onto land is sea walls and coastal dikes. Their crest height is often determined by taking the tsunami trace heights in the past. At the design, the following change should be carefully examined. If this old tsunami strikes again, its height may become higher than the old trace, because of the reflection effects from the structures at the shore. There is no rational method of the stability analysis for the section of structures, because tsunami forces on structures are not well understood. Massive structures are usually believed resistant enough based on the experiences in the past. The most important is the protection at the front and the rear toes of structures against erosion by overflowing water.

5.2. Tsunami Breakwaters

In contrast to normal breakwaters, tsunami breakwaters are constructed at the mouth of a bay, where the water is deep and tsunami forces are less than that along the shore. A tsunami enters the bay through a narrow gap at the tsunami breakwaters, then spreads over the wide-water area inside the bay and decreases its height in the bay. At the same time, the period of the natural oscillation becomes half of the original natural period because of the partial enclosure provided by the breakwater, thus changing the resonance characteristics of the bay.

5.3. Effectiveness of Tsunami Control Forests

There are two contradictory opinions of the effectiveness of coastal forests on the reduction of tsunami energy. Affirmative views assert that the forests are effective, because (1) it stops driftwood and other floating materials, (2) it reduces water flow velocity and inundation height, (3) it provides a life-saving means by catching persons carried off by tsunamis, and (4) it collects wind-blown sands and raises dunes, which act as a natural barrier against tsunamis. A negative opinion is that the forests may be ineffective against a giant tsunami, and at worst, trees could become destructive floating

UNESCO – EOLS

S

SAMPLE C

HAPTERS

objects to houses, if they were broken down by the tsunami. A quantitative answer was found in the case of coastal forests of pine trees in Japan. Effectiveness and limitations are expressed in terms of the summed tree diameter and the density of undergrowth. Roughly speaking, the control forests are undamaged and are effective to stop boats and others, regardless of the number of trees and the thickness of the undergrowth, if the tsunami height above ground is less than 3 m.

5.4. Movement of Residence to Tsunami-Free High Ground

A tsunami-defense planner needs to identify where the “tsunami-free” ground is. Similar to the idea in the flood-plain control, the tsunami-hazardous area must be determined such as the area inundated by a tsunami that occurs on the average of once every 100 years. The statistical study of the occurrence of tsunamis is, however, a formidable task, because of the infrequency of tsunamis at a given site. Numbers of reliable data are usually insufficient for its extreme value statistics. Therefore, the highest tsunami trace in the past or the highest tsunami runup computed for the biggest tsunami in the past is usually used to determine the tsunami-flood area. In Japan, tsunamis that may be generated by the largest earthquake expected from seismotectonics are also taken into consideration.

5.5. Tsunami-Resistant Building Zone

As was often noticed in the post-tsunami field surveys in the past, reinforced-concrete buildings were strong enough to resist tsunami forces. Even though windows and doors were broken, they withstood and worked to stop boats and other materials transported by tsunamis, thus effectively protecting weak wooden houses behind them. This fact introduces the idea of the tsunami-resistant building zone along the shoreline. Based on experience, reinforced-concrete buildings can withstand tsunami heights of up to 5 m above ground. Due to lack of data, we cannot answer whether or not a reinforced-concrete building can survive beyond this limit. There is not enough knowledge of tsunami forces and impacts of floating materials.

5.6. Tsunami Forecasting

The best way to save human lives is by early evacuation based on forecasting and warning. An earthquake that makes you unable to stand by yourself on a beach is a natural warning of a tsunami. Leave the beach as soon as possible and climb to ground higher than 20 m. This is the rule to save lives from the danger of the near-field tsunami generated by tsunamigenic earthquake, when we are on a beach where no forecasting is available.

UNESCO – EOLS

S

SAMPLE C

HAPTERS

Many countries have their own tsunami forecasting and warning systems. Most of them use the empirical relationships between earthquake magnitude and tsunami magnitude. The National Oceanic and Atmospheric Administration (NOAA) of the United States has deep-ocean seismographs, the records of which are transmitted to the headquarters via satellite. The magnitude of a tsunami is determined based upon the empirical knowledge, then the forecasting is disseminated via satellite to the areas in danger. Deep-ocean tsunami gauges are also used to assist this judgement. In the forecasting system in French Polynesia, the mantle magnitude is used in place of the conventional earthquake magnitude. The former is considered better than the latter, because it will not miss a tsunami earthquake. In order to obtain the mantle magnitude, a seismograph that can detect lower frequency components than the widely used conventional seismographs is necessary. Japan Meteorological Agency (JMA) began tsunami forecasting in 1952, based upon the empirical relationships among the distance to epicenter, amplitude of seismic waves, and magnitude of tsunamis. The judgement was qualitative and the forecasting information did not satisfy the need for coastal residents. The new quantitative forecasting began in April 1999. JMA constructed a tsunami database by carrying out more than 100 000 cases of tsunami numerical simulations. Once an earthquake occurs and its fault parameters are determined, the detailed tsunami information is derived from this database within a few seconds. The forecasting is disseminated through plural routes, including satellites, and reaches coastal residents within 4 or 5 minutes after an earthquake. After the 1946 Aleutian tsunami caused serious damages to the islands of Hawaii, the Pacific Tsunami Warning Center (PTWC) was established, not only for the United States but also for the entire Pacific Rim region. Since then, the international cooperation in tsunami forecasting has steadily progressed. PTWC is now acting as the national tsunami warning center for the United States, as the regional tsunami warning center for the State of Hawaii, and as the operational center for the Tsunami Warning System of the Pacific (TWS).

5.7. Tsunami Hazard Maps

To recognize the tsunami danger, hazard maps are useful. For a site where tsunami data are available, the hazard map is made based on the data of the past tsunamis. In order to include the topographical and social changes after the previous events, numerical simulations are utilized. First, the old topography is used to establish the tsunami initial profile. The established initial tsunami profile should be able to reproduce well the old

UNESCO – EOLS

S

SAMPLE C

HAPTERS

tsunami records. Then the computation with the present topography is carried out for this initial profile to determine the area of tsunami inundation. For a site where no or very little tsunami data is available, the numerical simulations are used to draw a tsunami hazard map. In this case, the validity and accuracy of the tsunami initial profile is the most important and, at the same time, difficult issue.

5.8. Continuation of Tsunami Disaster Culture

Tsunami occurs quite infrequently. Its recurrence interval is usually longer than a life span of human beings. In December 1992, a huge earthquake shook the coastal area of Flores Island, Indonesia, and a tsunami was generated. No resident tried to evacuate and more than 1000 lives were lost. Coastal residents did not know what would come after the huge earthquake, because this earthquake was the only natural disaster that the residents had experienced since 1900. On the contrary, in another example, residents were safe although they had not experienced a tsunami. Coastal residents in Vanuatu left for higher grounds soon after the ground shaking was over in November 1999, and no lives were lost by the tsunami that followed the earthquake. They obtained the necessary knowledge in 1998 from a TV report that showed the Papua New Guinea tsunami. It is quite difficult to pass on previous experiences to future generations, although they are precious and effective to save human lives. A survey in Japan revealed the following tendency about how humans can easily forget the severe experience. Within 8 years after a big disaster, the principal demands of the victims and their neighbors are to make every effort to see that the disaster never occurs again. Ten years passed, and many people begin to forget the severe experience. After 15 years passed, 40% of the victims think that they are now safe. Thirty or forty years passed, and the disaster remains only as a faint memory. One hundred years passed, and the memory of the disaster disappears. Then the next tragedy will occur again. 6. Needs in the Near Future The most urgent subject in research as well as in practical application is the improvement of the method to determine tsunami initial profiles with seismic data alone. If this is done, the initial condition is well fixed and the accuracy of tsunami forecasting is much increased. At present, an initial profile determined from fault parameters needs to be adjusted sometimes twice or more in order to explain the actual tsunami. This difference might be due to heterogeneity of the fault movement, existence of the subfaults, dynamic movement of the fault, etc. One of the requirements to solve this problem is the development of observation

UNESCO – EOLS

S

SAMPLE C

HAPTERS

networks in the ocean. Ocean-bottom seismographs near tsunami sources are indispensable to understand the details of fault movements. Deep-ocean tsunami gauges will capture tsunami at or just after their birth. New technologies such as satellite photometry are desirable to obtain a two-dimensional tsunami profile. With these data, further development in tsunami research becomes possible. The public education is a crucial key to save human lives. It is very rare for any person to experience large tsunamis more than once in his/her lifetime. If we have a simple knowledge of the fact that a tsunami will come after an earthquake, and if we behaves wisely to climb up to a high ground, our lives will be saved. Such knowledge should be transferred to the future generation and to coastal residents in all of the tsunami-risk areas. The nature of infrequency of tsunami occurrence makes it difficult for tsunami-defense planners, especially for a country where reliable tsunami data in the past are unavailable. The planners can expect the assistance of the international cooperation such as the TIME (Tsunami Inundation Modeling Exchange) project of IUGG (International Union of Geodesy and Geophysics) and IOC (Intergovernmental Oceanography Committee of UNESCO). During IDNDR (International Decade of Natural Disaster Reduction), more than 90 tsunami hazard maps in 11 countries were made by the TIME project. Glossary Boussinesq equation: The combination of the depth-averaged conservation of

mass and linear momentum with the assumption of near-hydrostatic pressure field but including the lowest-order correction due to wave motion.

Bore: A waveform, which has a sudden increase in water level with the steep front face, being covered with a strongly turbulent region; a quasisteady broken wave.

Breakwater: An offshore barrier to mitigate the wave (tsunami) effects in the enclosed areas behind, such as a harbor and a bay.

Dispersion effect: Evolution in waveform owing to the propagation-speed differences by different wave frequencies and amplitudes.

Fault parameters: Parameters that characterize a fault rupture, such as the seismic moment, fault size, fault slip pattern and direction, and rupture duration.

Green’s formula: A prediction model for wave height variations, as the waves propagate in the shallow water; the formula is based on the conservation of energy flux in the wave propagation

UNESCO – EOLS

S

SAMPLE C

HAPTERS

direction. Linear long wave theory:

The shallow-water theory with the assumptions of infinitesimal wave amplitudes and velocities.

Shallow-water theory: A water-wave theory based on the depth-averaged conservation of mass and linear momentum with the assumptions of hydrostatic pressure field and uniform horizontal velocity over the depth.

Tsunami earthquake: An earthquake, with mild ground shake, that generates disproportionately large tsunamis.

Sea wall: A wall or embankment made along the shore to protect the areas behind it from wave (tsunami) actions.

Shoaling effect: Change in wave height due to the water depth. Tsunami intensity: The empirical parameter to express the local effect of

tsunami. Tsunami magnitude: The empirical parameter to express the size of tsunami. Tsunamigenic earthquake:

An earthquake that generates measurable tsunamis.

NATURAL AND HUMAN INDUCED HAZARDS – Vol. II Tsunamis ... · NATURAL AND HUMAN INDUCED HAZARDS ... Mitigation of Tsunami Hazards ... NATURAL AND HUMAN INDUCED HAZARDS – Vol. II

Documents