"Tsunami" is the Japanese term meaning wave in the harbor. As such it is most descriptive of the observed phenomenon frequently referred to as tidal wave or seismic sea wave, with both of these terms having misleading connotations with respect to the mechanism of generation.

In South America, the term "maremoto" is frequently used. However the use of the word "tsunami" is most commonly accepted by scientists and by most of the lay public in Pacific basin countries. For the TWS, tsunamis can be categorized as local, regional, or Pacific-wide, with those terms being used to describe the extent of potential destruction relative to the tsunami source area. Local tsunamis will often be associated with tsunami generation by submarine or subaerial landslides or volcanic explosions.

An example would be the awesome local tsunami of July 9, 1958, at Lituya Bay, Alaska, where wave run-up exceeded 485 meters but the destruction was confined to a very limited area. Regional tsunamis are by far the most common. Destruction may be limited in areal extent either because the energy released was not sufficient to generate a destructive Pacific-wide tsunami, or because the geomorphology of the source area limited the destructive potential of the tsunami.

Examples of some recent tsunamis are shown in the table below:

Pacific-wide tsunamis are much less frequent, but of far greater destructive potential in that waves are not only larger initially, but in transit across the Pacific basin, many distant coastal areas are subject to destructive impact. For example, the tsunami of May 22, 1960, spread death and destruction across the Pacific from Chile to Hawai`i, Japan, and the Philippines.

A tsunami is a system of gravity waves formed in the sea as a result of a large-scale disturbance of sea level over a short duration of time. In the process of sea level returning to equilibrium through a series of oscillations, waves are generated which propagate outward from the source region. A tsunami can be generated by submarine volcanic eruptions, by displacement of submarine sediments, by coastal landslides into a bay or harbor, by meteor impact, or by vertical displacement of the earth's crust along a zone of fracture which underlies or borders the ocean floor. The latter is by far the most frequent cause of tsunamis and for all practical purposes the primary cause of tsunamis capable of propagation across an ocean basin. The rupture of the earth's crust will also generate a major earthquake which can be detected and measured by seismic instrumentation throughout the world. However, not all major coastal or near-coastal earthquakes produce tsunamis. At present, there is no operational method to determine if a tsunami has been generated except to note the occurrence and epicenter of the earthquake and then detect the arrival of the characteristic waves at a network of tide stations.

When a major earthquake occurs, the resultant energy released into the earth will propagate over a wide range of frequencies and velocities. Even though the earth movements discernible to the viewer may be confined to the general region of the earthquake origin, the various seismic wave phases propagating throughout the earth result in small, but measurable, ground motion which can be detected by a seismometer. A seismograph then provides a visual record of the ground motion at that station.

For the purposes of the Tsunami Warning System, consideration is given to three significant seismic wave phases. The first, the P-wave, is a compressional wave traveling through the earth's interior at a velocity varying from approximately 8.0 km/second near the crust-mantle interface to about 13.5 km/second at the mantle-core interface. As such it is the first seismic phase to be recorded at any one seismic station and is the first indication that a distant earthquake has occurred. The location of the earthquake can be determined by assuming the "best fit" of the pattern of P-wave arrivals at several stations compared to a standard table of P-wave arrival times for various distances and depths of earthquake focus or, in the case of local earthquakes in or near the limits of a relatively small area seismic station network, compared to the calculated arrivals based on a local crustal seismic velocity model.

The second seismic phase of importance is the S-wave, or Secondary wave. This phase travels through the earth's interior as a shear wave, following approximately the same travel path as the P-wave but at a reduced velocity varying from approximately 6.7 km/second near the crust-mantle interface to about 8.0 km/second near the core. These seismic wave phases are classified as body waves due to their propagation through the earth's interior. In addition to providing a location, body waves are useful in determining the size of an earthquake, especially when the eathquake's focus is deep within the earth.

The third set of seismic phases to be considered are the surface waves resulting from ground displacements propagating outward along the surface of the earth. These are observed at a seismic station as local or regional surface waves and are the basis for measuring magnitude on the Richter scale. This is a logarithmic scale devised by Charles Richter to use the amplitude of the trace recorded on a seismograph and the distance from the epicenter to assign a somewhat consistent indication of size to a particular earthquake as measured at different stations. Beno Gutenberg extended the Richter scale to include distant Love and Raleigh surface waves. Though it is a logarithmic scale to the base 10, this is merely a reference to the Richter scale value being incremented as a logarithmic function of the trace deflection as recorded on the seismograph and the distance of the station from the epicenter. The actual energy released for each increment of the Richter scale is a factor of 32. Thus a magnitude 7.0 earthquake will release 32 times as much energy as a magnitude 6.0, and the energy release for a magnitude 8.0 is more than 1000 times greater than a 6.0.

Tsunamis travel outward in all directions from the generating area, with the direction of the main energy propagation generally being orthogonal to the direction of the earthquake fracture zone. Their speed depends on the depth of water, so that the waves undergo accelerations and decelerations in passing over an ocean bottom of varying depth. In the deep and open ocean, they travel at speeds of 500 to 1,000 kilometers per hour (300 to 600 miles per hour). The distance between successive crests can be as much as 500 to 650 kilometers (300 to 400 miles); however, in the open ocean, the height of the waves may be no more than 30 to 60 centimeters (1 or 2 feet), and the waves pass unnoticed. Variations in tsunami propagation result when the propagation impulse is stronger in one direction than in others because of the orientation or dimensions of the generating area and where regional topographic features modify both the wave form and rate of advance. The tsunamis are waveform extends through the entire water column from sea surface to the ocean bottom. It is this characteristic that accounts for the great amount of energy transmitted by a tsunami.

The successive waves of a tsunami in the deep sea have such great length and so little height they are not visually recognizable from a surface vessel or from an airplane. The passing waves produce only a gentle rise and fall of the sea surface. During the April 1946 tsunami at Hawai`i, ships standing off the coasts observed tremendous waves breaking on shore but did not detect any change in sea level at their offshore locations.

Upon reaching shallower water, the speed of the advancing wave diminishes, its wave length decreases, and its height may increase greatly, owing to the piling up of water. Configuration of the coastline, shape of the ocean floor, and character of the advancing waves play an important role in the destruction wrought by tsunamis along any coast, whether near the generating area or thousands of kilometers from it. Consequently, detection of relatively small tsunamis at any locality warrants immediate reporting -- through the TWS -- to spread the alarm to all coastal localities of approaching potentially dangerous waves.

At present, detection of tsunamis is possible only near shore where the shoaling effect can be observed. The first visible indication of an approaching tsunami is often a recession of water caused by the trough preceding an advancing wave. Any withdrawal of the sea, therefore, should be considered a warning of an approaching wave. A rise in water level also may be the first event. Tide-gauge records of the Chilean tsunami of May 22, 1960, generally showed a rise in water level as the first indication of this tsunami. This rise amounted to about one-half the amplitude of the following decrease in water level. Under certain conditions, the crest of an advancing wave can overtake the preceding trough while some distance offshore. This causes the wave to proceed shoreward as a bore -- a wave with a churning front.

The force and destructive effects of tsunamis should not be underestimated. At some places, the advancing turbulent front is the most destructive part of the wave. Where the rise is quiet, the outflow of water to the sea between crests may be rapid and destructive, sweeping all before it and undermining roads, buildings, and other works of man with its swift currents. Ships, unless moved away from shore, can be thrown against breakwaters, wharves, and other craft, or washed ashore and left grounded during withdrawals of the sea.

In the shallow waters of bays and harbors, a tsunami frequently will initiate seiching. If the tsunami period is related closely to that of the bay, the seiche is amplified by the succeeding waves. Under these circumstances, maximum wave activity often is observed much later than the arrival of the first wave.

A tsunami is not one wave, but a series of waves. The time that elapses between passage of successive wave crests at a given point usually is from 10 to 45 minutes. Oscillations of destructive proportions may continue for several hours, and several days may pass before the sea returns to its normal state.

During the 101-year period from 1900 to 2001, 796 tsunamis were observed or recorded in the Pacific Ocean according to the Tsunami Laboritory in Novosibirsk. 117 caused casualties and damage most near the source only; at least nine caused widespread destruction throughout the Pacific. The greatest number of tsunamis during any 1 year was 19 in 1938, but all were minor and caused no damage. There was no single year of the period that was free of tsunamis.

17 percent of the total tsunamis were generated in or near Japan. The distribution of tsunami generation in other areas is as follows: South America, 15 percent: New Guinea Solomon Islands, 13 percent; Indonesia, 11 percent: Kuril Islands and Kamchatka, 10 percent; Mexico and Central America, 10 percent; Philippines, 9 percent; New Zealand and Tonga, 7 percent; Alaska and West Coasts of Canada and the United States, 7 percent; and Hawai`i, 3 percent.

source: http://www.prh.noaa.gov/ptwc/abouttsunamis.htm