Tsunami Research at the End of a Critical Decade

The Member States of the United Nations unanimously proclaimed the International Decade for Natural Disaster Reduction (IDNDR) by UN resolution 46/182 on 22 December 1989. The same resolution adopted an IDNDR International Framework of Action for 1990–99 with the objective of stimulating concerted international action, especially in developing countries, to reduce the loss of life, property damage, and social economic disruption caused by natural disasters. The Decade was established on the basic understanding that sufficient scientific and technical knowledge already exists which, with more extensive application, could save thousands of lives and prevent millions of dollars in property losses from natural and similar disasters.

Tsunamis are a series of ocean waves generated by abrupt, large disturbances of the ocean surface such as earthquakes, volcanic eruptions, landslides, slumps, and meteor impacts. These waves can engulf a coastal community within minutes of their birth and cause loss of life, catastrophic destruction to structures and infrastructure, and severe erosion of the shoreline by hours of repeated attack of waves many minutes apart (Figure 1). Human suffering during tsunami flooding can be enormous: people are swept along with other debris in the tsunami-induced currents at speeds up to 60 km/hr, resulting in drowning due to multiple injuries like broken bones, lacerations, abrasions, punctures, and crushed body cavities. Following the hours of tsunami attack, survivors may suffer from exposure to the environment; untreated shock may lead to gangrene, exacerbating the injuries and leading to more deaths. Since 1850, tsunamis in the Pacific have caused the death of over 120,000 coastal residents. Tsunamis are a major hazard to coastal residents in earthquake-prone regions (Table 1).

The theoretical relationship between tsunami intensity I (on the Soloviev-Imamura scale) and the moment magnitude of an earthquake M _w , has been obtained (I = 3.55M _w − 27.1) by Chubarov and Gusiakov (1985). This relationship was used to calculate the expected tsunami intensity for 293 Pacific tsunamigenic earthquakes with known moment-magnitude M _w . The present study introduces the formal classification of these earthquakes on the basis of their ΔI parameter, that is the difference between observed and expected tsunami intensity. Based on the ΔI value, all events are divided into three groups: “red” (ΔI>1), “green” (−1≤ΔI≤1), and “blue” (ΔI<−1). The geographical distribution of events in these groups shows their clear correlation with climatic and circum-continental zonation in oceanic sedimentation, as described by Lisitsyn (1974). Specifically, the equatorial humid zone, characterized by the highest rate of oceanic sedimentation, is clearly indicated by an increased level of “red” tsunamigenic earthquakes. The circum-continental zonation is clearly expressed by the fact that all tsunamigenic events that occurred in this century in the East China, the Yellow, the Japan, the Okhotsk and the Bering Seas belong to the “red” group. On the other hand, all major submarine earthquakes that occurred in such remote subduction zones as Guam, Tonga, and New Zealand are designated as “green” or “blue”. Despite their large M _w values (greater than 7.9), these events generated very minor tsunamis with run-up heights less than 1 meter. The present study indicates that earthquake-induced disturbances of bottom sediments that result in submarine slumping can be a significant factor in the tsunami generation mechanism. Therefore, the potential slumping process should be taken into account in operational tsunami warning as well as in coastal tsunami zoning.

The scouring mechanism due to tsunami run-up around a vertical cylinder was investigated experimentally using a large-scale tsunami experimental facility. A variety of solitary waves were generated and allowed to run up onto a uniformly sloping sandy beach. The scouring process around the cylinder on the beach was optically recorded using CCD video cameras installed inside of the cylinder. In addition, the water-surface elevation, flow velocity, and pore pressure around the cylinder were measured with electronic sensors. We surveyed the topographic changes before and after the run-up. It was found that the pore-pressure field around the cylinder is the key factor in understanding tsunami scouring mechanism.

Mitigation is the translation of direct and interactive vulnerabilities into public policies to reduce risks; such mitigation policies tend to be multi-faceted. Communities vulnerable to tsunami impacts generally have administrative structures and planning procedures that can be adapted to guide land use and building practices in the hazard zone. To be effective, however, these policies and regulations must be based on a clear understanding of local tsunami effects, including both direct and interactive characteristics.

The “Golden 24” rule for disasters says that the greatest percentage of survivors of a natural disaster are saved in the first 24 hours after it occurs. This rule is a common and important guide for saving as many human lives as possible after a big disaster occurs. According to Kawata (1996), 18,000 persons were rescued in the case of the Great Hanshin-Awaji Earthquake in Kobe, Japan, in 1995. Among them, 15,000 persons were rescued by neighbors soon after the earthquake, and the ratio of persons saved alive was 80 %. Others were saved by the fire fighting rescue teams and by the Self-Defense Army. On the first day, the number of survivors was larger than the number of dead among the persons saved. On the second and third days, the ratio of persons saved alive was about 30%. On the fourth day and after, there was almost no hope to rescue any person alive.

This paper describes a Research and Development activity to develop tsunami forecasting tools for the Pacific Disaster Center (PDC). The activity included analytical and numerical sensitivity studies of tsunami wave characteristics offshore of Hawaii, for ranges of earthquake source parameters in the Alaska-Aleutian Subduction Zone (AASZ). This region is a major source of destructive tsunamis that strike Hawaii. A set of tsunami numerical simulation scenarios was designed to be the basis for the sensitivity analysis. The simulation results have been stored as an online database with a World-Wide Web interface. A database user can very quickly obtain a model prediction of the tsunami offshore wave heights at chosen locations for a wide variety of AASZ earthquake scenarios. This database provides a comprehensive offshore tsunami forecasting tool for hazard mitigation managers.

One of the most catastrophic tsunamis in history occurred on November 5, 1952 near the Nothern Kuril Islands. Unfortunately, most of the information related to this catastrophe was classified as secret and top secret in the USSR for many years. Until recently, this tsunami is not well-known. This report contains a part of the results of a search made in the different recently declassified Russian archives and also information given by eye-witnesses.

Combined sedimentological and field studies can be used to investigate the impacts of past tsunami. In this paper, sedimentary deposits from Livadia, Astypalaea Island, Greece are described which are interpreted as being associated with the July 9, 1956 Aegean tsunami. While the tsunami run-up indicated by these deposits is broadly consistent with previously reported values, topographic evidence and anecdotal eyewitness accounts from nearby Astypalaea Town indicate that previously reported run-up elevations were over-estimated. This observation is consistent with recent modeling data for the 1956 tsunami, and illustrates the need for systematic investigations of the effects of past tsunami events within the Aegean region. These investigations should involve sedimentological and other field-based surveys and modeling studies coupled with critical (re)evaluation of contemporary reports and other historical records. This is a necessary first step for providing the proper basis on which reliable assessments of tsunami risk can be made and hazard zone maps, coastal protection and emergency and disaster management plans developed.

The Kythira strait constitutes a complex transform-extensional deformation and rotation in the Western Hellenic Arc characterized by high seismicity. Historical documents, such as descriptions, chronicles, memoires, and diaries have been combined with archaeological evidence to compile a catalogue of earthquakes and tsunamis reported in the area of Kythira from the antiquity to 1910 inclusive. This attempt revealed earthquake events that remained unknown so far in the seismological literature. For some already known events the times of occurrences were corrected and/or their macroseismic fields were better defined. The seismic potential in the Kythira strait is exceptionally high as is reflected in the historical seismicity of the area. Apart from the 66 AD and 365 AD large earthquakes, and the questionable event of 800, at least ten strong (M_S ≥ 6.0) earthquakes occurred from 1750 to 1910 with a mean recurrence of about 18±18 years. As for the tsunami potential, excluding the questionable wave of 800, at least five strong tsunamis were observed from the 1st century A.D. onwards. Assuming that the tsunami data are complete only from the beginning of the 17^th century, we conclude that the mean frequency of strong tsunamis is one per 130 years. Of special seismological, archaeological and historical interest is the supposedly seismic destruction of Skandia, the ancient harbour of Kythira, in association with the large 365 AD and 800 AD earthquakes and tsunamis. A future interdisciplinary research effort could cast a new light to this working hypothesis.

Among the Italian regions, the Messina Straits are one of the most exposed to large tsunami attacks. Tsunami catalogues (e.g. Tinti and Maramai, 1996) report several disastrous events hitting the Straits generated both locally, such as the December 28 1908 tsunami, and in the adjacent regions (e.g. the tsunami following the 1693 eastern Sicily earthquake to the south, and the 1783 event generated in the Tyrrhenian Calabria to the north). In this contribution we will put our attention on the December 28, 1908 earthquake generated tsunami, which was the last catastrophic event of this kind to have hit the Italian coasts. The parent earthquake had an estimated M≅7.2 magnitude and produced catastrophic effects in an area as large as 6000 km² (Boschi et al., 1995): Figure 1 shows the region that suffered the highest damage. The two most important towns facing the Straits, Messina and Reggio Calabria (stations 35 and 8 in Figure 1, respectively), were completely destroyed, and very severe damage was produced in all southern Calabria and in the northeastern part of Sicily. The estimated total number of victims was around 80000, 2000 of which produced by the tsunami that followed the earthquake (Boschi et al., 1995). The initial water movement observed along both sides of the Straits was a significant retreat, followed by a violent sea attack that struck the coasts with at least three big waves (Tinti and Maramai, 1996). In some places the impact of the water waves was so violent that the rubble of the buildings destroyed by the earthquake was completely swept away (Boschi et al., 1995).

Runup and inundation data on the sand spits of Sissano lagoon are described with discussions of flow state, current velocity, and degree of damage to houses. Sand erosion data on the sand spits are also described with discussions of their relation to the inundation depth and the current velocity. Laboratory experiments were carried out to confirm the flow state and to discuss effects of vegetation and so on.

The relatively small July 17, 1998 Papua New Guinea earthquake produced very high tsunami amplitudes, localized along a short length of coast near the earthquake source. This prompted speculation regarding possible tsunami source scenarios, including seismic bottom deformation, submarine and/or subaerial slumping, or combinations of each. The MOST numerical model was used to simulate scenarios in which the source was assumed to be either co-seismic bottom deformation or a submarine landslide, modeled as viscous sediment flow. Sources of various sizes at different locations were assumed, and the computed runup estimates were compared with measurements obtained by the International Tsunami Survey Team (ITST). Although distinctions in the tsunami runup dynamics for different tsunami sources are apparent, it was found that both pure landslide and pure bottom deformation scenarios could produce results that satisfactorily matched the observed runup heights. Thus, in the case of the PNG tsunami, and probably many other events as well, runup values alone are insufficient to distinguish between co-seismic, landslide, or combined source mechanisms. Additional information, such as current velocities, sediment deposition and scouring, number of waves and direct measurements in the source area, are necessary for accurate tsunami source determination.

A model for tsunamis caused by a sequential process from landsliding to debris flow is proposed and applied to the cases of the 1741 Oshima-oshima volcanic tsunami in Japan and the 1998 Aitape tsunami in Papua New Guinea. In both cases the observed tsunami heights were locally very large. A numerical model combining a circular-arc slip for sliding with two-layer flow model for debris flow is developed to simulate tstmamis due to land/underwater sliding. Stability conditions for CFL, open boundary, and splash of debris flow into seawater are discussed. One stability condition of the numerical model is that ∆x/∆t = max(C₁, C₂) should be less than unity. A stable numerical model of the splash of debris flow is obtained by introducing a drag force at the front and artificial viscosity. The model could well reproduce the distribution of tsunami heights along the coast in the southwestern part of Hokkaido, Japan. The numerical result for the 1998 Papua-New Guinea tsunami suggests a sediment slump 10–20 km offshore the lagoon as an additional source important for recreating the reported tsunami behavior.

The generation mechanism of the Papua New Guinea (PNG) Tsunami in 1998 is investigated through a numerical simulation. Tsunami amplification effect in the shallow water region is also evaluated. There is still now ambiguity about the generation mechanism due to the lack of the exact bathymetry data in the shallow water region. But it is quite certain that, based only on the earthquake, the behavior of the PNG tsunami cannot be fully explained. The earthquake, followed by landslides or slumps triggered by the original earthquake, was the most probable cause of the tsunami. These kinds of tsunamis have occurred in the past and caused heavy damages. To minimize damages, a tsunami warning concept that measures tsunami directly is needed. The newly invented laser tsunami meter is one of the candidates for use in such a warning system.

A three-dimensional, shallow-water numerical model for a viscous landslide with full slide-wave interaction (Kulikov et al.; 1996; Fine et al., 1998) has been modified to include the subaerial component of the landslide. The model is used to simulate the November 3, 1994 tsunami in Skagway, Alaska generated by collapse of the PARN Dock. Results show that the dock slide moved down the steep (30–35°) slope of Taiya Inlet and was guided along the trough at the base of the slope, consistent with geomorphological findings. The leading tsunami wave, propagating in front of the advancing slide, impacted the Alaska State Ferry Terminal and the NOAA tide gauge site as a positive wave (crest), consistent with the tide gauge record and with the results of laboratory modelling by Raichlen et al. (1996). Computed wave heights for the PARN Dock failure (13 m at the Ferry Terminal, 7.7 m at the tide gauge site, and 1.3 in the Small Boat Harbor) agree closely with the tide gauge record and eyewitness accounts. The computed 3.0 min period for the fundamental long-wave mode for Skagway Harbor is nearly identical to the observed period. Estimates of the Q-factor (Q≈24) are comparable to observed values (Q≈21), suggesting significant tsunami energy retention in the harbour. Energy loss appears to be through radiation damping rather than from frictional effects. A detailed examination of the slide motion and associated tsunami waves in the vicinity of the PARN Dock reveals that, in the first few seconds, a “wall of water” would have formed opposite the dock and that the floating Ferry Terminal would have been impacted 15 to 20 s after onset of the event, consistent with eyewitness accounts. The floating debris observed at the still-standing northern portion of the dock was apparently carried alongshore by a secondary wave crest originating near the collapsed southern part of the dock.