Solutions to Coastal Disasters 2008: Tsunamis

Tsunamis pose a significant threat to coastlines around the world. Effective tsunami warnings, issued by NOAA's Tsunami Warning Centers (TWCs) depend on reliable and accurate observation systems, including both coastal water level stations and offshore buoys, such as the Deep-ocean Assessment and Reporting of Tsunamis (DART^TM) buoy network. The United States National Oceanic and Atmospheric Administration (NOAA) funds several water level observation networks in support of operational tsunami warning capability. These observing systems include NOAA-funded stations operated by the University of Hawaii Sea Level Center, stations operated by the NOAA Tsunami Warning Centers (TWCs), the NOAA-operated DART^TM network, and the stations of the NOAA National Water Level Observation Network (NWLON). Each of these networks provides a critical component of an effective tsunami warning system. This paper describes the NOAA National Ocean Service (NOS), Center for Operational Oceanographic Products and Services (CO-OPS) contribution to tsunami warning, through operation of the NWLON.

Almost 3 years after the extreme 26 December 2004 tsunami event, which devastated the Banda Aceh region of Sumatra, certain coastal protection works have been planned and implemented. They comprise mainly of seawalls, breakwaters, and planting forest trees and mangroves as a wave-absorbing greenbelt. Numerical modeling of this catastrophic event was established to investigate the processes of tsunami propagation and inundation flows over the Banda Aceh coastal region, identify zones that had escaped devastation (as identified during the post event field survey) and assess possible types of coastal protection along the affected coasts. Results of the modeling showed that for protection against such a 2004-type event, huge structures ~ 15 m high would be required, but are not economically feasible. However, mature mangroves stands may offer reasonable protection, and this option shows greater promise as an affordable solution.

The Integrated Tsunami Research and Information System (ITRIS) has been developed and is being maintained at the Novosibirsk Tsunami Laboratory (NTL) as a joint project with the World Agency for Planetary Monitoring and Earthquake Risk Reduction (WAPMERR). The concept of the ITRIS is based on the integration of historical data, numerical models, processing and analyzing tools along with supporting mapping software. These components are embedded inside a specially developed GIS-type graphic shell for easy data retrieval, visualization and processing. The shell operates on Windows PC platforms with no additional co-located software required. The current version of the built-in tsunami database covers the period from 2000 B.C. to present time and contains 2250 entries in the event catalog and almost 9500 run-up heights. The set of the built-in numerical models includes a subroutine for fast calculation of Tsunami Travel Time (TTT) charts and numerical code for calculation of tsunami generation and propagation in an ocean with a real bathymetry within non-linear shallow water model. The approach to the long-term tsunami hazard assessment, implemented in the ITRIS software, is based on the statistical analysis of historical run-up observations. For a particular coastal area, this type of analysis can be interactively made using a special pop-up menu and dialog windows. For areas that are not provided with sufficient number of historical data, another, so called, scenario approach can be used. It is based on application of numerical models for calculation of synthetic tsunami catalog for pre-established set of earthquake models with parameters derived from analysis of seismotectonic features of the area under study.

It is expected that many coastal regions in Japan will be attacked by huge tsunami in near future. Therefore, tsunami evacuation plans are considered in Japan. In this paper, tsunami disaster education was carried out for elementary school children and their parents. The purpose of this education was to encourage development of a family plan for tsunami evacuation. At first, children and their parents learned about tsunamis and how to evacuate safety during tsunamis. Second, they searched for safety tsunami evacuation places near their houses. Third, they made a tsunami evacuation map and developed the family evacuation plan based on this map. Through this education program, children not only got the knowledge about tsunami but also got the wisdom about tsunami evacuation.

Tsunami education and response are critical life-saving aspects for near-source tsunamis; there is little time for buoys to detect the generation of a tsunami and interpret the signal and/or for communities to issue evacuation messages. The 1 April 2007 magnitude Ms 8.1 earthquake off the New Georgia Group in the Solomon Islands generated a tsunami that killed 52. A reconnaissance team deployed within one week investigated 65 coastal settlements on 13 remote Islands, interviewed survivors and made surveys of both the tectonic response and the inundation, finding tectonic uplift up to 3.6 m, subsidence down to — 1.5m, locally focused run-up heights of 12 m, local flow depths of 5 m. Such an event would be expected to have caused a large loss of life, and did in fact cause extensive damage to low-lying buildings. However, the ancestral heritage "run to high ground after an earthquake" passed on to younger generations by survivors of smaller historic tsunamis triggered an immediate spontaneous self evacuation that likely spared many lives.

More than two hundred thousand people were killed in the Indian Ocean tsunami disaster. The deaths that may be due to drowning are discussed through reference to past experiences. A fault tree is presented to summarize the physical danger posed by tsunami waves to the human body, and the personal danger is discussed based on the tree. The stability of the human body against wave front collision and tsunami currents is discussed and threshold conditions are proposed. The time for evacuation is usually very limited and the risk of encountering a tsunami during evacuation is often high. Walking in inundated areas during a tsunami attack is especially dangerous. Vertical evacuation is highly recommended.

This paper summarizes guidance on siting concepts, design concepts, performance objectives, design loads, and load combinations that should be considered in locating and designing tsunami vertical evacuation structures. This work is the result of a three-year effort, known as the ATC-64 Project, funded by the Federal Emergency Management Agency and the National Oceanic and Atmospheric Administration. The purpose of the project is to develop guidance for designing such structures in order to provide refuge from tsunami inundation in communities where there is not enough time to evacuate horizontally out of the inundation zone.

The tsunami wave forces on buildings are experimentally investigated for the effects of openings. The beach condition at Khao Lak, Thailand, which suffered great loss in the 2004 Indian Ocean tsunami, is simulated in a wave flume. Solitary waves are generated by an abrupt release of water from an elevated tank. Based on the pressure profile at various instants of time, a simplified pressure profile is proposed for the maximum tsunami force acting on the structure without openings. The influence of openings on the pressures on the front and back of the upstream panel is examined.

The problem of the tsunami wave runup on a beach of constant slope is discussed in the framework of rigorous solutions of the one-dimensional nonlinear shallow-water theory. The key and novel moment here is the analysis of influence of initial wave shape on extreme (maximal) characteristics of the wave on a beach (runup and drawdown heights, runup and drawdown velocities and breaking parameter)

Tsunami-induced local scour around a land-based structure with a square cross-section on a sand foundation was investigated with both laboratory experiments and numerical simulations. An isolated long wave was adopted as an incident tsunami wave. In the hydraulic model experiments, it was found that the maximum scour depth at the seaward corner of the structure was sensitive to the relative overtopping height of the tsunami wave. In the numerical simulations, it was revealed that the local scour around the seaward corner of the structure could be evaluated with not only flow velocities on the foundation but also effective stresses near the surface of the foundation because restricting force between sand particles declined due to a decrease in the effective stresses.

Being the most populous area in the Aleutians, Unalaska is considered as one of Alaskan coastal communities with high potential for tsunamis. As part of NOAA's Short-term Inundation Forecast for Tsunami (SIFT), a Stand-by Inundation Model (SIM) based on the MOST model is applied in this study to assess the tsunami impact for Unalaska. The model validation using historical tsunami events show excellent agreement between the model computation and observations, which gives rise to the accuracy of the inundation model. This study provides inclusive tsunami impact assessment for Unalaska, AK subject to a total of 2681 distant and local tsunamis scenarios in the Pacific at different level of earthquake magnitude M_w 7.5, 7.8, 8.2, 8.7 and 9.3. This study also investigates the impact caused by the hypothetical scenarios initiated in Unalaska gap and Shumagin gap at different level of earthquake magnitude M_w 7.5, 8.0, 8.5, and 9.0. The computational maximum tsunami runup suggests the current definition of Tsunami Safe Zone in Unalaska, areas above 50ft (~15m), is conservative.

The 1992 Flores Tsunami off of Babi Island, Indonesia caused extensive damage and took more than 1,500 human lives. Also, the 1993 Hokkaido Tsunami killed about 300 people in Japan and caused large run-up heights on Okushiri Island. Both tsunami events had attracted much attention of scientists because unexpected large run-up heights had been observed around Babi and Okushiri Islands, respectively. In the southern sea of the Korean Peninsula, many islands are scattered. These islands are very vulnerable to unexpected tsunami attacks. In this study, a numerical model is employed to investigate behaviors of tsunamis around circular islands.

Hawaìi has a long, though sporadic, history of deadly tsunami attacks, and since the 1946 disaster the state has developed increasingly sophisticated and effective mitigation strategies. The evolution and operation of these strategies is described. Tsunamis will no longer be Hawaìi's deadliest natural hazard.

Education and outreach related to coastal hazards has been identified as an increasing need for local and state government agencies in the U.S. Pacific Islands. To address this need, emergency managers, coastal zone managers, and other decision makers are seeking to leverage advances in information technology (IT) to improve the way hazard risk information is communicated to and used by the public. In response, the National Oceanic and Atmospheric Administration (NOAA) Pacific Services Center (PSC) developed an online tool that can be easily applied by state and local governments for enhancing the public's awareness of coastal hazard risks and educating them on appropriate preparedness and response actions. This tool, the Hazard Education and Awareness Tool (HEAT), has recently been successfully applied in the State of Hawaii for enhancing tsunami awareness and preparedness. To view how HEAT has been applied in Hawaii, go to: www.csc.noaa.gov/psc/tsunami

An earthquake in the Aleutian Islands on April 1, 1946 generated a large tsunami that destroyed the waterfront in Hilo, Hawaii. Evidence from oblique aerial photographs taken during the tsunami show that the wartime air bases at Kahuku (NE Oahu) were also devastated. The photographs demonstrate that: wide-spread inundation occurred with the sea flooding 1.6 km inland destroying many wooden structures and lifting large barracks buildings off their foundations, and collapsing several. Vegetation was damaged by the waves. Sands from Kahuku beach and coastal sand dunes were dispersed throughout the inundation zone. A scenario is presented that outlines the erosive and deposition processes involved with tsunami runup and drain back. The information discussed here will be used to recommend a broad building set back in this area, in order to avoid similar damage in the future.

Crescent City, California is particularly vulnerable to tsunamis. Since 1938, 24 tsunamis have been recorded and three have caused significant damage, most recently in November 2006. The largest and most deadly of these, the tsunami generated by the 1964 Alaskan earthquake, was discussed in detail by Dengler and Magoon at the previous Solutions conference. This paper discusses a ‘new' source of data for assessing the risk to Crescent City, i.e. digitized high-speed gauge records from the second most damaging tsunami that followed the May 22, 1960 Chilean earthquake. These newly digitized records will allow spectral and statistical analyses so that the evolution of spectral characteristics with time can be studied and questions of tsunami resonance and the partitioning of the energy into wave types can be better understood.

Humboldt County on California's North Coast is at risk from tsunamis generated locally from faults associated with the Cascadia subduction zone (CSZ), other regional fault systems, and from distant sources elsewhere in the Pacific. The Redwood Coast Tsunami Work Group (RCTWG), an organization of representatives from government agencies, tribes, service groups, academia and the private sector from the three northern coastal California counties, was formed in 1996 to coordinate and promote tsunami hazard awareness and mitigation. The RCTWG and its member agencies have sponsored a variety of projects including education/outreach products and programs, tsunami hazard mapping, and signage and siren planning. In 2007, Humboldt County was the first region in the country to participate in a tsunami training exercise at FEMA's Emmitsburg training facility and the first area in California to conduct a full-scale tsunami evacuation drill. Two Humboldt County communities were recognized as TsunamiReady by the National Weather Service in 2007. Six assessment surveys from 1993 to 2006 have tracked preparedness actions and personal awareness of earthquake and tsunami hazards in the county.

On November 15, 2006 the harbor at Crescent City in Del Norte County, California was hit by a series of tsunami surges generated by the M_w 8.3 Kuril Islands earthquake. This earthquake produced the largest Kuril tsunami to reach Crescent City since 1933 when the tide gauge was installed. The largest amplitude surge, 1.76 m peak to trough, was the sixth cycle, arriving over two hours after the first wave arrived. The tsunami caused an estimated $9.2 million in losses to the small boat basin, damaging the floating docks and several boats. The event highlighted the vulnerability of harbors from a relatively modest tsunami, problems with the tsunami warning system for a marginal event, and the particular vulnerability of the Crescent City harbor area. We look at how the tsunami warning system and human response can exacerbate or reduce the impacts of tsunamis.

We model tsunami runup and inundation along the Northern California Coast caused by hypothetical earthquakes on the Cascadia Subduction Zone (CSZ). The CSZ forms the boundary between the Pacific and North American tectonic plates and extends from Northern California to Southern British Columbia and is believed capable of producing great earthquakes with moment magnitudes (M_w) 9.0 or greater. A suite of CSZ rupture scenarios based on the geological characteristics of the southern part of the CSZ were simulated with the tsunami propagation and inundation model MOST. CSZ inundation projections were compared and related to historical accounts and modeled data for two teletsunamis at Crescent City: the tsunami of 1 April 1964 that caused 12 deaths and more than 17 million in damages to Northern California, and the 15 November 2006 Kuril Islands tsunami that caused nearly 10 million in damages to the Crescent City harbor. Our modeling suggests that near field tsunamis from the CSZ present a hazard greater than the 1964 event for the northernmost part of the State. However, south of Shelter Cove tsunamis generated by great earthquakes along the Alaska — Aleutians subduction zone are potentially more damaging.

When propagating in deep water, tsunami waves are known to show no change in the profile of the horizontal particle velocity. However, the water depth in which grain motion at the bottom actually is initiated by a passing tsunami wave has remained undetermined. Based upon linear wave theory and a parametrization of the Shields curve, equations are derived and solved, using an iterative scheme, to illustrate the transport capability of the 2004 Sumatra tsunami. The focus is on waves in the deep water where tsunami waves behave linearly and on non-cohesive sediment grains. According to the analytical results, even the 2004 Sumatra tsunami in 2004 was incapable of moving fine sand in water deeper than 985 m in the Bay of Bengal and 335 m in the Indian and Pacific oceans. It is concluded that earthquake-generated tsunami waves are not able to generate deep-water deposits that can be correlated on a marine-basin wide scale. Only tsunami waves maintaining amplitudes well above 2 m, such as waves generated by meteoritic impacts, have the potential to move sediment and generate tsunami deposits throughout entire oceanic basins. These considerations give a lower boundary of generating tsunami deposits. Tsunami deposits near-shore and onshore are more prone to reworking, for example to due storms. A boundary water depth is defined as the depth below which tsunami deposits cannot be reworked by storm waves. To define this water depth, it is assumed that the wave base coincides with the sediment surface and the waves behave linearly. As an example, tide-gauge records are used from about 50 km offshore from Brisbane, Australia, of a Category 1 Cyclone (March 2–4, 2004). The Brisbane storm wave reached the maximum wave height of 14.3 m. Wave periods varied between 6 s and 14 s and the computed maximum wavelength was 237 m. Assuming the highest wave amplitude during the 2004 March storm at Brisbane as such a boundary wave, the water depth below which preservation of tsunami deposits is most likely greater than 65 m.

A trench survey for large tsunami traces was conducted in Nemuro lowland area along the Kuril subduction zone, eastern Hokkaido. According to our stratigraphic methods such as tephrochronology and AMS14C dating, we identified 15 tsunami sands in marsh deposits and their recurrence interval was estimated almost 300–350 years. In fact, these traces involve more layers than other large tsunami traces extending between Tokachi coast and Kiritappu marsh. It has been concluded that it is necessary to estimate other tsunami sources of the southern part of Kuril Islands as well, other than tsunami sources caused by multi-segment of interplate earthquake (probably M 8.6) such as off Tokachi and Nemuro areas. According to our sedimentological analyses, these tsunami sands derived only from the coastal area, and although they do not show clear graded bedding, they commonly have convolution structures and erosional bases and include peat clasts and internal bed forms such as plane beds, dunes, and current ripples, reflecting bedload transportation.

Sidescan sonar and sub-bottom reflection surveys off Shikabe town, along the Pacific coast, show the extent and volume of the submarine debris-avalanche deposit caused by the AD 1640 eruption of the Hokkaido-Komagatake volcano. The avalanche deposit extends from the subaerial part of the Shikabe lobe to as far as 20 km seaward. The subaqueous deposit, which reaches to 80 m of water depth, is 15 km wide and covers 126 km² of the seafloor. Hummocks of the subaqueous deposit decrease in height and width with distance from the source. The margin of the deposit lacks hummocks. The ratio of the collapse height to the traveled distance (H/L) is 0.06, suggesting that the debris avalanche of origin was more mobile than those of most subaerial debris-avalanche deposits. The subaqueous volume of the avalanche deposits, estimated by extrapolating the pre-eruptive topography of the surrounding area, ranges from 0.92 to 1.20 km³.

Evidence exists for at least one large tsunami affecting northern New Zealand around the 15^th Century. Numerical modelling of potential volcanic and seismic sources in the Southwest Pacific was used to constrain potential source regions and assess likely frequency-magnitude relationships. A volcanic source is considered the most credible mechanism, suggesting a lower frequency of occurrence and longer response time than assumed for the alternative seismic source.

An incomplete catalogue of tsunamis in the written record hinders tsunami risk assessment. Tsunami deposits, hard evidence of tsunami, can be used to extend the written record. The two primary factors in tsunami risk, tsunami frequency and magnitude, can be addressed through field and modeling studies of tsunami deposits. Recent research has increased the utility of tsunami deposits in tsunami risk assessment by improving the ability to identify tsunami deposits and developing models to determine tsunami magnitude from deposit characteristics.