S53A-1008
Tsunami Vulnerability Assessment of Casablanca – Morocco, Pilot Study
In this study we present a preliminary evaluation of the tsunami vulnerability of Casablanca area (Morocco) and the expected inundation from a tsunami generated in the North East Atlantic area in the Gulf of Cadiz using a combination of numerical modeling and GIS tools. The study area is the intensively occupied area of the Casablanca Harbor, a location that may be considered as one of the most vulnerable areas to tsunami hazard, along Morocco Atlantic coast, due to its location close to the source of moderate and strong magnitude earthquakes, the extension of low flat areas close to the sea and the existence of dwellings characterized by unstructured constructions. To study vulnerability and inundation we considered geomorphologic and artificial elements as well as the existence of defensive constructions. The inundation zone was divided in for sub-zones according to their height above sea level; the buildings were classed according to: building material, number of floors, conditions of foundation soil. The study area presents a large variety of constructions:1 storey buildings of 2- 2.5 m height, houses and stores of the ancient medina and some modern buildings. The results are presented, using GIS and show a preliminary "picture" of built stock behaviour in case of tsunami for Casablanca. The results clearly demonstrate that the vulnerability to tsunami impact is not uniform within the inundation zone. This work was developed in the framework of NEAREST and TRANSFER projects, EU.
S53A-1009
Tsunami Hazard in Western Mediterranean: Preliminary Study of Scenarios for the Balearic
Within the framework of the TRANSFER European project, in order to mitigate tsunami hazard in Western Mediterranean, we aim at producing inundation maps for the bay of Palma (Balearic) using numerical modeling. To this end we build a set of scenarios based on realistic tectonic sources able to generate tsunamis. Such a study has been carried out in order to constrain the source(s) responsible for the destructive Djijelli (Algeria) earthquake of 1856 (estimated I0=VIII MSK). This event produced at least one reported inundation in the Balearic and flooded some sites along the Algerian coast over a wide area. In addition the Palma bay is part of bays which are inclined with rissaga phenomenon (atmospheric tsunami) which supposes particular possible site effects (high resonance for example). Available new bathymetric and seismic data along the North-Algerian margin allowed us to propose a compressive "en échelon" fault system offshore Djijelli as possible candidate for this event. Three faults, each one about 35 km long, have been defined in agreement with the isoseismal map of the quake, and have been tested alone or combined. We choose to model a Mw=7.2 earthquake generated tsunami. A co-seismic slip of 1 to 1.5 m with a pure compressive mechanism has been chosen consistent with a convergence rate of 6mm/y leading to a period of recurrence of about 200-500yr according to bibliography. Wave propagation and inundation models are calculated using a finite difference code on a set of imbricated grids derived from the Gebco 1' grid. Grids from Balearic bays (Palma for instance) are constructed after digitalization of nautical bathymetric charts, geo-referencing and krigging interpolation in order to reach a high level of details in harbours (8m resolution). Synthetic mareograms for the 1856 tsunami are then compared with witness historical data such as arrival times, wave polarity, amplitudes, etc. in order to validate the modeling and above all, the sources. Then the final inundation maps obtained from the set of scenarios should be further integrated in GIS.
S53A-1010
Characteristic Tsunami Events in the Area of Rhodes Island, East Hellenic Arc
The eastern segment of the Hellenic arc-trench (EHAT) is one of the most active in the Mediterranean region producing both large earthquakes and tsunamis. Therefore, it was selected as the master test-site of the EU TRANSFER tsunami research project. Previous research results indicates that EHAT produces characteristic shallow earthquakes of M¡Ö7.2. From historical documentary sources and instrumental records we conclude that in the last six centuries six characteristic earthquakes ruptured the EHAT: 1481, 1609, 1741, 1851, 1957. This implies that EHAT is ruptured by strong, shallow shocks as an average every 120±13 years, indicating a quasi- periodic than random process. However, the tsunami generation deviates from such a quasi-periodic pattern given that only the 1481, 1609 and 1741 earthquakes produced strong tsunamis, while the 1851 shock generated only a small-to-moderate local tsunami and the 1957 one was non-tsunamigenic. We introduce the concept of ¡°characteristic tsunami" to describe similar, strong tsunamis produced by earthquakes occurring in the same fault segment. Then, we conclude that only some characteristics earthquakes produce characteristic tsunamis while others do not. From the point of view of tsunami hazard assessment it is of great importance to estimate the probability that the next characteristic earthquake may generate a characteristic tsunami. To this goal we apply a conditional probabilitiy procedure tested recently in highly tsunamigenic zones of the Pacific Ocean.This is a contribution of the EU Research Project "TRANSFER", contract N.037058, FP6-2005-Global-4, Reduction of Tsunami Risks.
S53A-1011
Probabilistic Tsunami Hazard Analysis - Results for the Western United States
We have developed a series of probabilistic tsunami hazard maps for the coasts of western North America based on fault source characterizations of the circum-Pacific subduction zones as well as local offshore faults. The maps show the probabilistic offshore exceedance waveheights at 72, 475, 975 and 2475 year return periods, which are the return periods typically used in Probabilistic Seismic Hazard Analysis (PSHA). Our method follows along similar lines as (PSHA) which has become a standard practice in the evaluation and mitigation of seismic hazard in particular with respect to structures, infrastructure and lifelines. Its ability to condense complexities, variability and uncertainties of seismic activity into a manageable set of ground motion parameters greatly facilitates the planning and design of effective seismic resistant buildings and infrastructure. Because of the strong dependence of tsunami wave heights on bathymetry, we use a full waveform tsunami waveform computation in lieu of attenuation relations that are common in PSHA. By pre-computing and storing the tsunami waveforms at points along the coast generated for sets of subfaults that comprise larger earthquake faults, we can rapidly synthesize tsunami waveforms for any slip distribution on those faults by summing the individual weighted subfault tsunami waveforms. This Green's function summation provides accurate estimates of tsunami height for probabilistic calculations, where one typically integrates over thousands of earthquake scenarios. We have carried out tsunami hazard calculations for western North America and Hawaii based on a comprehensive source model around the Pacific Ocean including both subduction zone sources as well as local offshore faults. We will present the tsunami hazard maps and discuss how these results are used for probabilistic inundation mapping, including a follow-up inundation study of the San Francisco Bay area that is based on disaggregation results of the probabilistic analysis. We will also show that even for tsunami sources with large uncertainties (e.g. submarine landslides) a probabilistic framework can yield meaningful and consistent results.
S53A-1012
Tsunami Hazard in Crescent City, California from Kuril Islands earthquakes
On November 15, Crescent City in Del Norte County, California was hit by a series of tsunami surges generated by the M = 8.3 Kuril Islands earthquake causing an estimated 9.7 million (US dollars) in damages to the small boat basin. This was the first significant tsunami loss on US territory since the 1964 Alaska tsunami. The damage occurred nearly 8 hours after the official tsunami alert bulletins had been cancelled. The tsunami caused no flooding and did not exceed the ambient high tide level. All of the damage was caused by strong currents, estimated at 12 to 15 knots, causing the floating docks to be pinned against the pilings and water to flow over them. The event highlighted problems in warning criteria and communications for a marginal event with the potential for only localized impacts, the vulnerability of harbors from a relatively modest tsunami, and the particular exposure of the Crescent City harbor area to tsunamis. It also illustrated the poor understanding of local officials of the duration of tsunami hazard. As a result of the November tsunami, interim changes were made by WCATWC to address localized hazards in areas like Crescent City. On January 13, 2007 when a M = 8.1 earthquake occurred in the Kuril Islands, a formal procedure was in place for hourly conference calls between WCATWC, California State Office of Emergency Services officials, local weather Service Offices and local emergency officials, significantly improving the decision making process and the communication among the federal, state and local officials. Kuril Island tsunamis are relatively common at Crescent City. Since 1963, five tsunamis generated by Kuril Island earthquakes have been recorded on the Crescent City tide gauge, two with amplitudes greater than 0.5 m. We use the MOST model to simulate the 2006, 2007 and 1994 events and to examine the difference between damaging and non-damaging events at Crescent City. Small changes in the angle of the rupture zone results can result in a half meter difference in water heights. We also look at the contribution of fault segments along the Kuril subduction zone using the FACTS server to look at the potentially most damaging source regions for Crescent City. A similar-sized rupture as the November 15 event located further south along the Hokkaido – Honshu area of the subduction zone, is likely to produce a slightly larger amplitude signal with and even greater delay between the first wave arrivals and the largest waves.
S53A-1013
Tsunami Hazard Assessment and Inundation Maps for Crescent City
We model tsunami inundation and runup heights in Crescent City, California triggered by possible earthquakes on the Cascadia Subduction Zone (CSZ). The CSZ is believed capable of producing great earthquakes with magnitudes of Mw 9.0 or greater. We simulate plausible CSZ rupture scenarios and calculate inundation using MOST (Titov and Synolakis, 1998). We benchmark our CSZ inundation projections against mapped flooded areas and tide gauge data from the 1964 tsunami, which destroyed 29 city blocks, and also from the damaging 15 November 2006 Kuril Islands tsunami. Results suggest that inundation from CSZ tsunamis could extend over 3 km inland, twice as far as the limits of the 1964 inundation. Crescent City is most vulnerable to slip on the Gorda segment of the CSZ. Rupture of the northern or Juan De Fuca segment produces lower water heights than the 1964 event. At Crescent City, the tsunami surges a leading elevation wave just after the earthquake. Educational and preparedness for self-evacuation are essential to save lives. Titov V.V. and Synolakis, C.E., 1998, Numerical modeling of Tidal Wave Runup, J. Waterway, Port, Coast. Eng., vol 124 (4) 157-171. http://www.usc.edu/dept/tsunamis
S53A-1014
Overview of Collaborative Research to Assess Tsunami Hazard for Nuclear Plants on the Atlantic and Gulf Coasts
In response to the 2004 Indian Ocean Tsunami, the United States Nuclear Regulatory Commission (US NRC) initiated a long-term research program to improve understanding of tsunami hazard levels for nuclear power plants in the United States. In order to complement research focused on the Pacific Ocean by other organizations, the US NRC sponsored a collaborative research project with the United States Geological Survey (USGS) and the National Oceanic and Atmospheric Administration (NOAA) for the purpose of assessing tsunami hazard on the US Atlantic and Gulf Coasts. This research considers landslide and seismic tsunamigenic sources in both the near and the far fields. The first phase of this work, undertaken by the USGS and now nearly complete, consisted of collection, interpretation, and analysis of available offshore data, with significant effort focused on characterizing offshore near-field landslides and analyzing their tsunamigenic potential and properties. As part of the next phase of research, work will be undertaken to refine estimates of hazard on the Atlantic and Gulf coasts for a wide variety of sites. Field investigations will be undertaken by the USGS in key locations of interest which currently lack sufficient existing data. Simultaneously, the MOST tsunami model developed at NOAA will be enhanced to include landslide-based initiation mechanisms. The enhanced MOST model will then be used to investigate the tsunamigenic sources characterized by the USGS to achieve the long-term goals of the US NRC program, which include an estimation of deterministic tsunami hazard levels for the length of the Atlantic and Gulf Coasts. The potential for probabilistic tsunami hazard assessment will also be explored in the final phases of the program.
S53A-1015
Hazard from far-field tsunami at Hilo: Earthquakes from the Ring of Fire
Historical data and modeling are used to study tsunami hazard at Hilo, Hawaii. Hilo has one of the best historical tsunami record in the US. Considering the tsunami observations from the early eighteen hundreds until today reveals that the number of observed events per decade depends on the awareness of tsunami events. The awareness appears to be a function of the observation techniques such as seismometers and communication devices, as well as direct measurements. Three time periods can be identified, in which the number of observed events increases from one event per decade in the first period to 7.7 in the second, to 9.4 events per decade in the third one. A total of 89 events from far-field sources have been encountered. In contrast only 11 events have been observed with sources in the near field. To remove this historical observation bias from the hazard estimate, we have complimented the historical analysis with a modeling study. We have carried out modeling of 1476 individual earthquakes along the subduction zones of the Pacific Ocean in four different magnitude levels (7.5, 8.2, 8.7 and 9.3). The maximum run up and maximum peak at the tide gauge is plotted for the different magnitude levels to reveal sensitive and source areas of tsunami waves for Hilo and a linear scaling of both parameters for small, but non-linear scaling for larger earthquakes
S53A-1016
Potential Hazards of Tsunami Waves along the Chinese coast in the next 100 years
In the next 100 years the Chinese coast faces potentially non-negligible danger from tsunamogenic earthquakes originating at the neighboring subducting plate boundaries in the Phillipines and the Ryukyu Islands. There are significant differences in the bottom bathymetry between the South China Sea bordering the southern province of Guangdong and the East China Sea and Yellow Sea adjacent to the provinces of Zhejiang, Jiangsu, and Shandong. We have found that the linear shallow-water equations can be used to predict with good enough accuracy the travel time of tsunami waves in the South China Sea, but the nonlinear shallow-water equations must be used for the shallower seas next to the northern Chinese provinces. There are some differences in the travel time predictions between the linear and nonlinear theories for the Yellow Sea region. This difference is enough to make a difference in terms of warning. We will use our newly developed probability method, called the probabilistic forecast of tsunami hazards ( PFTH ) for predicting the danger of tsunami waves with a certain height of around 2 meters to impinge on the cities along the Chinese coast in the next century.We have used the Gutenberg-Richter relationship applied locally to each locale for evaluating the probability of the seismic risk for large earthquakes, greater than magnitude 7. We have only included the frequency of shallow large earthquakes to take place . For the southern cities of Hong Kong and Macau, we found that the probability of a 2 meter wave to hit these ports is around 10 % in the next century. Cities in Taiwan are less vulnerable than the large coastal cities on the Chinese mainland. The probability results for the northern cities of Shanghai and Qingdao are around a few per cent for smaller wave heights like one meter or so. But even these smaller waves can be damaging --
S53A-1017
Understanding Landslide Tsunami Hazard in Alaska Fjords for Tsunami Inundation Mapping
Several communities of the southern coast of Alaska are located in glacial fjords, which are fed by major rivers and creeks draining nearby glaciers and depositing sediments into the bays at a high rate. Sediment accumulation on the steep underwater slopes contributes to the landslide tsunami hazard in these communities. During the Great Alaska Earthquake of 1964, the majority of tsunami-related deaths was due to local landslide tsunamis that occurred almost immediately after the initial shaking, and without any warning signs. In these coastal communities, tsunami potential from tectonic and submarine landslide sources must be evaluated for comprehensive mapping of areas that are at risk for inundation. We are creating tsunami inundation maps for Seward, Alaska, in the scope of the National Tsunami Hazard Mitigation Program. Seward is a community located at the head of Resurrection Bay, which was hit hard by both tectonic and landslide-generated tsunami waves during the 1964 earthquake. The purpose of the project is long- term prediction of potential landslide-generated tsunamis in Resurrection Bay, and public education on landslide-related tsunami hazard. In order to construct tsunami inundation maps for Seward, we use an approach that combines modeling of the historical tsunami events of 1964 in Resurrection Bay for model verification, and assessing the landslide tsunami hazard by simulating hypothetical landslide scenarios and performing sensitivity analysis. To reconstruct the sequence of waves observed at Seward on March 27, 1964, we model tsunami waves caused by superposition of the local landslide-generated tsunamis and the major tectonic tsunami. Next we create hypothetical landslide scenarios that are based on the underwater sediment accumulation areas derived from the bathymetry difference maps. Numerical simulations yield runup heights, extent of maximum inundation for chosen tsunami scenarios, depths of inundation on dry land, and maximum velocity current distribution in inundation zones. We deliver modeling results to the community for local tsunami hazard planning and construction of evacuation maps.
S53A-1018
An underwater landslide or slump on an active submarine fault Ea possible source of a devastating tsunami?
A Mw 7.7 earthquake and subsequent large-scale tsunami occurred on 17th July 2006 off the southern coast of Java Island, Indonesia. A maximum of 7.7 m inundation height was recorded in Pangandaran on the southern coast of Java, according to the field survey just after the tsunami (Tsuji et al., 2006). However, since there are few residents who noticed the earthquake tremors, the earthquake may possibility be so-called "tsunami earthquake". The aftershocks from July to September occurred on the fore-arc area and their CMT solution suggests the predominant north-south tensile stress. R/V MIRAI passed the aftershock area in 2004 and 2005 with continuous multibeam bathymetric survey. The processed topographic map shows a lot of amphitheatres with the scale of 8-20 km along the fault scarps. Convexity landforms are located below the footwall of the amphitheatres, apparently the relics of an underwater landslide, and their relative elevation exceeds 1000 m in maximum. This area is characterised by the southeastern extension of Mentawai Fault and its associated minor faults ranging from the Sumatra fore-arc area. The observed amphitheatres and relics of underwater landslides are located along the active faults. Considering that the tsunami wave height distribution is concentrated on a specific narrow area compared with the scale of the mainshock, the mainshock possibly triggered underwater landslides on these amphitheatres and the slides generated a large-scale tsunami. The Ryukyu district was attacked by a M7-class earthquake followed by a devastating tsunami in 1771. 'The East Ishigaki Fault,' one of the across-arc active normal faults in this area was studied precisely by multibeam echo sounding. The study revealed a slump on the segmented active fault. The slump itself may generate a tsunami with the wave height of several meters and might cause the 1771 tsunami together with the faulting itself. Some recent geohazards also show that a landslide or a slump is associated with an identified active fault, for example, 17 Jul. 1998 Papua New Guinea Earthquake and Tsunami. This study is to compare the several similar cases mentioned above in order to discuss the liability to landlides on an active fault.
S53A-1019
New contributions to the debate on the cause of the January 11th, 1693 tsunami in eastern Sicily (Italy): earthquake or offshore landslide source (or may be both)?
Eastern Sicily is among the most exposed regions in Italy and in the whole Mediterranean to tsunami hazard and risk. The historical tsunamis recorded here were generally associated to moderate-to-large magnitude earthquakes. The largest tsunami documented in the area occurred on January 11th, 1693. It followed the highest-magnitude earthquake (7.4) of the Italian seismic history. The tsunami, whose first significant motion was a retreat along the entire eastern Sicily coastline, produced the most devastating effects at Augusta (15 meters run-up) and Catania, being relevant at Siracusa and Messina too. A lively debate exists on whether the earthquake was the only source of the tsunami, or other causes (such as submarine landslides, possibly triggered by the earthquake) contributed to the tsunami generation. In the framework of the EC funded project TRANSFER, we investigate both hypotheses, starting from suitable onshore and offshore faults as well as from offshore landslide bodies, and hence simulating numerically the ensuing tsunami and comparing the results with the available historical information. We base on the results obtained during recent offshore surveys, in particular the multichannel seismic survey MESC2001, carried out in year 2001 on board the R/V Urania of the Italian National Council of Researches (CNR), which mapped both active normal faults and a number of possible landslide bodies along the Hyblaean-Malta escarpment, the most prominent tectonic structure found just few kilometres offshore eastern Sicily. From the modelling point of view, the initial condition for the earthquake- generated tsunamis coincides with the vertical coseismic deformation of the seafloor. Instead, the landslide motion is simulated through the Lagrangian block model UBO-BLOCK2, developed at the University of Bologna. Finally, the finite-element code UBO-TSUFE, implemented by the same research team, is used to simulate the tsunami generation and propagation. The main conclusions are: 1) if the earthquake is postulated to be the only responsible for the tsunami, then the historical information can be reproduced only by assuming an offshore tectonic source; 2) taking into account the largest of the mapped landslides, we are able to reproduce quite satisfactorily both the first polarity and the size distribution of the tsunami; 3) we cannot rule out the idea that there was a concurrent contribution of the earthquake and of the landslide in generating the tsunami.
S53A-1020
A European effort towards the development of tools for tsunami hazard and risk assessment and mitigation, and tsunami early warning: the EC-funded TRANSFER project
TRANSFER (acronym for "Tsunami Risk ANd Strategies For the European Region") is a European Community funded project being coordinated by the University of Bologna (Italy) and involving 29 partners in Europe, Turkey and Israel. The main objectives of the project can be summarised as: 1) improving our understanding of tsunami processes in the Euro-Mediterranean region, 2) contributing to the tsunami hazard, vulnerability and risk assessment, 3) identifying the best strategies for reduction of tsunami risk, 4) focussing on the gaps and needs for the implementation of an efficient tsunami early warning system (TEWS) in the Euro-Mediterranean area, which is a high-priority task in consideration that no tsunami early warning system is today in place in the Euro- Mediterranean countries. This paper briefly outlines the results that were obtained in the first year of life of the project and the activities that are currently carried out and planned for the future. In particular, we will emphasize the efforts made so far in the following directions. 1) The improvement of existing numerical models for tsunami generation, propagation and impact, and the possible development of new ones. Existing numerical models have been already applied to selected benchmark problems. At the same time, the project is making an important effort in the development of standards for inundation maps in Europe. 2) The project Consortium has selected seven test areas in different countries facing the Mediterranean Sea and the eastern Atlantic Ocean, where innovative probabilistic and statistical approaches for tsunami hazard assessment, up-to-date and new methods to compute inundation maps are being and will be applied. For the same test areas, tsunami scenario approaches are being developed, vulnerability and risk assessed, prevention and mitigation measures defined also by the advice of end users that are organised in an End User Group. 3) A final key aspect is represented by the dissemination of the project data and results to the largest possible public. The two privileged means are and will be the project web site (http://www.transferproject.eu) and a web-based GIS database integrating existing data and new project data, ranging from tsunami catalogues to inventories of seismic and non-seismic sources, from topographies and bathymetries at different scales and resolutions to layers containing tsunami inundation maps. Hopefully, this paper will stimulate a discussion and an exchange of experiences with tsunami scientists from different regions of the world.
S53A-1021
Toward a Euro Mediterranean Tsunami Warning System: the Case of the February 12, 2007 Ml=6.1 Earthquake.
UNESCO has committed the scientific community for the creation of the North East Atlantic and Mediterranean Tsunami Warning System (NEAMTWS). We wonder: "Is the European Seismological community ready for running a Euro Mediterranean Tsunami Warning System?" In this study we use the February 12, 2007 Ml=6.1 earthquake, occurred close to the epicenter of the 1755 Great Lisbon one, as a case study to verify/discuss our actual capability of alerting for a hypothetical tsunami coming. Starting from the data availability and the available real-time tools, we emulate an automatic real-time processing to perform the source analyses needed for discriminating a tsunamigenic earthquake, and subsequently we evaluate the time shift between the release of results and the potential tsunami wavefront reaching the coasts.
S53A-1022
Progress in developing an Indian Ocean Tsunami Warning System (IOTWS)
Nearly three years following the devastating 2004 Indian Ocean tsunami, there is much progress to report on building a new Indian Ocean Tsunami Warning System (IOTWS) which will provide tsunami early warnings and framework for disaster management and response systems. To date, the IOTWS has utilized the leadership and technical expertise of many countries including Indonesia, Thailand, India, Sri Lanka, and the Maldives, together with assistance from international partners. Inter-agency cooperation has combined expertise in a broad range of disciplines to accomplish several goals including: 1) developing infrastructures for both real-time analysis of seismic data and rapid communication and warnings (including the upgrade of several Indonesian seismic and GPS stations), 2) land use planning and community preparation aimed at minimizing damage and loss of life from future disasters, and 3) international support for logistics, communications, training, management and administration. Throughout the implementation of the IOTWS, a primary focus was placed on "in-country capacity building," so that individual nations can be self-sustaining in their efforts. We believe that this has been accomplished through extensive training sessions, workshops and site visits.
S53A-1023
Derivative-based Sensitivity Analysis for Tsunami Scenario Computations in Support of the German Contribution to the Indonesian Tsunami Early Warning System for the Indian Ocean
The tsunami in December 2004 set the starting point for the establishment of a German-Indonesian Tsunami Early Warning System (GITEWS). The unstructured grid tsunami propagation and inundation model TsunAWI is under development at Alfred-Wegener-Institute (AWI), Bremerhaven. For either scenario computations as well as the early warning simulation support, input data are not exact but include measurement errors. In order to assess the error bounds of simulation results a rigorous sensitivity analysis has been conducted. Questions are: How sensitive are simulation results on input data like bathymetry and topography? What impact has the inexact determination of the rupture zone to the inundation in coastal areas? We present a derivative-based sensitivity analysis approach to test the relation of perturbed input data to output data.
S53A-1024
Post-Sumatra Enhancements at the Pacific Tsunami Warning Center
Following the tragic Indian Ocean Tsunami of 2004, the Richard Hagemeyer Pacific Tsunami Warning Center (PTWC) has dramatically enhanced its capabilities. With improved communications PTWC now ingests seismic data from almost all broadband stations of the Global Seismographic Network and will soon add many stations from the International Monitoring System. As data sources are increased PTWC's response time to any earthquake declines; for most earthquakes the center now gets out an initial message in about 12 minutes. With 24-hour staffing, that performance is maintained around the clock. Direct measurement of tsunamis has been improved through communications upgrades to coastal tide gauges by NOAA and other collaborators in the Pacific Tsunami Warning System, and by the NOAA deployment of DART instruments throughout the world's oceans. In addition to providing warnings for the Pacific (with the exception of Alaska and the west coasts of the U.S, and Canada, which are the responsibility of the West Coast and Alaska Tsunami Warning Center), PTWC also operates as an interim warning center for the Indian Ocean (a task performed in collaboration with the Japan Meteorological Agency) and the Caribbean. PTWC also operates as a local warning center for the State of Hawaii. In Hawaii, the installation of new seismometers again means a continuous reduction in PTWC's response times. Initial assessments of local earthquakes are routinely accomplished in less than five minutes, and the first message for the Kiholo Bay Earthquake of 2006 was issued in only three minutes. With the development of the Hawaii Integrated Seismographic Network, in collaboration with the U.S. Geological Survey, the goal is to reduce the time for tsunami warnings to under two minutes for any earthquake in the Hawaiian Islands. http://www.prh.noaa.gov/ptwc/
S53A-1025
West Coast and Alaska Tsunami Warning Center Global Earthquake Detection System
A necessary Tsunami Warning first step is fast and accurate earthquake location and magnitude. The West Coast and Alaska Tsunami Warning Center (WCATWC) has designed and implemented a seismic detection system that is capable of detecting sizeable earthquakes globally since its inauguration. This detection system has been enhanced and improved over the years of operation. Since the 2004 Sumatra Tsunami, WCATWC has taken particular steps to enhance its global seismic detection capability to lower the detection threshold and improve accuracy. These steps primarily include 24 hour staffing; enhanced seismic station coverage, improved wave energy detection, waveform spectral analysis and coherent phase association. In this talk, we summarize seismic detection capabilities improvements and the current status of WCATWC's detection methodology. We discuss the preliminary performance analysis results of our current seismic detection system.
S53A-1026
The West Coast/Alaska Tsunami Warning Center Forecast Model Project and Associated Sea Level Data Analysis Tools
Forecast (hydrodynamic) modeling is a necessary component in the refinement of tsunami warnings. Model results predict coastal regions likely to be impacted by either non-threatening or by destructive waves. They are used to predict the time of landfall and impact level, and play a role in hazard assessment. Model output data are integrated into the operational environment, where they are scaled and refined in real time with a new sea level analysis tool known as TideView. Over the last two years the West Coast/Alaska Tsunami Warning Center (WCATWC) forecast model project has grown to include use in hazard assessment on the Atlantic and Gulf coasts. It is also undergoing an upgrade to include inundation boundary conditions, increased use of nested finer meshes, more graphical output options for watch standers, and more flexible input of real time sea level data. The objective of this project is to provide warning center personnel with immediate, widespread tsunami amplitude forecasts based on pre-computed models and sea level observations. The TideView sea level data analysis software, developed at the WCATWC, enables warning center personnel to mesh observed tsunami information with the forecast model. The software also provides an easy method to compare observed waves with predicted tide and estimated tsunami arrival times, as well as digitally filter the tsunami signal. TideView interacts with the center's geographical information system and message generation software. This talk summarizes numerical and operational enhancements to the WCATWC forecast model and sea level analysis tools. Several examples from the Kuril Islands Tsunami of 11-15-06 will be given, along with a preview of new graphical outputs available to watch standers.
S53A-1027
The Catalog of Event Data of the Operational Deep-ocean Assessment and Reporting of Tsunamis (DART) Stations at the National Data Buoy Center
DART systems are a critical component of the tsunami warning system as they provide the only real-time, in situ, tsunami detection before landfall. DART systems consist of a surface buoy that serves as a position locater and communications transceiver and a Bottom Pressure Recorder (BPR) on the seafloor. The BPR records temperature and pressure at 15-second intervals to a memory card for later retrieval for analysis and use by tsunami researchers, but the BPRs are normally recovered only once every two years. The DART systems also transmit subsets of the data, converted to an estimation of the sea surface height, in near real-time for use by the tsunami warning community. These data are available on NDBC's webpages, http://www.ndbc.noaa.gov/dart.shtml. Although not of the resolution of the data recorded to the BPR memory card, the near real-time data have proven to be of value in research applications [1]. Of particular interest are the DART data associated with geophysical events. The DART BPR continuously compares the measured sea height with a predicted sea-height and when the difference exceeds a threshold value, the BPR goes into Event Mode. Event Mode provides an extended, more frequent near real-time reporting of the sea surface heights for tsunami detection. The BPR can go into Event Mode because of geophysical triggers, such as tsunamis or seismic activity, which may or may not be tsunamigenic. The BPR can also go into Event Mode during recovery of the BPR as it leaves the seafloor, or when manually triggered by the Tsunami Warning Centers in advance of an expected tsunami. On occasion, the BPR will go into Event Mode without any associated tsunami or seismic activity or human intervention and these are considered "False'' Events. Approximately one- third of all Events can be classified as "False". NDBC is responsible for the operations, maintenance, and data management of the DART stations. Each DART station has a webpage with a drop-down list of all Events. NDBC maintains the non-geophysical Events in order to maintain the continuity of the time series records. In 2007, NDBC compiled all DART Events that occurred while under NDBC's operational control and made an assessment on their validity. The NDBC analysts performed the assessment using the characteristics of the data time series, triggering criteria, and associated seismic events. The compilation and assessments are catalogued in a NDBC technical document. The Catalog also includes a listing of the one-hour, high-resolution data, retrieved remotely from the BPRs that are not available on the web pages. The Events are classified by their triggering mechanism and listed by station location and, for those Events associated with geophysical triggers, they are listed by their associated seismic events. The Catalog provides researchers with a valuable tool in locating, assessing, and applying near real-time DART data to tsunami research and will be updated following DART Events. A link to the published Catalog can be found on the NDBC DART website, http://www.ndbc.noaa.gov/dart.shtml. Reference: [1] Gower, J. and F. González (2006), U.S. Warning System Detected the Sumatra Tsunami, Eos Trans. AGU, 87(10), 105-112. http://www.ndbc.noaa.gov/dart.shtml
S53A-1028
Empirical Orthogonal Functions Based Tidal Forecast
Accurate forecasting of tides is essential for extracting tsunami signal from tsunami buoy / tide gage records. The existing methods used to predict tides are based on approximating tides with harmonic constituents derived from global tidal models and long data records. Determining the harmonic constants for every particular buoy at a particular location requires a record of the data at a given buoy that is optimally years long. This technique breaks down when applied to a recently deployed buoy. In this work, a method is developed to forecast the signal on a tide gage or tsunami buoy for the next day, given previous several day long record of the signal on the gage or buoy. For this paper the term 'day' refers to a lunar day which is 24 hours 48 minutes long, and the term 'tide' refers to a day long section of a record. If a record is a tsunami buoy (DART buoy) record, it is sampled with 15 min interval and a tide consists of M=99 readings. The forecasting technique presented here is based on decomposing consequent tides in a narrow sub-space of the M-dimensional space of day long basis functions in which points, corresponding to consequent tides, fall on a smooth curve. Decomposition coefficients for the next tide are obtained by extrapolating the curve one point forward. For most of Pacific DARTs this method provides tide estimate within a few centimeter precision, which is generally as accurate as an estimate done using harmonic constituents. This method also works and is being used in NOAA Center for Tsunami Research for forecasting newly deployed DARTs, starting as early as on fifth day after their deployment, before any tidal harmonic constants can be obtained. The basis functions, which are capable of enclosing 1-day long tide shapes into a narrow sub-space, are derived here as eigenvectors of matrix A × AT, where matrix A of size M × N has the N previously recorded tides as its columns. These vectors, also known as Empirical Orthogonal Functions (EOFs) of A, are therefore determined by tide auto-correlation function for an ensemble of N tides. To build a sub-space containing the next tide for a particular buoy, it is enough to use the N ~ 10 previous tides recorded by the buoy being forecasted. However, if a larger ensemble is used (N > 100), tides being chosen (not necessary one after another or in any regular order) from a several month long record, then the resulting EOF basis is capable of representing, with only few EOFs, tides distant in time and even in space, that is, on a different buoy (due to similar statistical properties of tides on most buoys). That last feature of the forecasting technique is important for forecasting new buoys with a little data to build an EOF basis. http://staff.washington.edu/etolkova/tealeaves.htm
S53A-1029
Tsunami Detection Systems for International Requirements
Results are presented regarding the first commercially available, fully operational, tsunami detection system to have passed stringent U.S. government testing requirements and to have successfully demonstrated its ability to detect an actual tsunami at sea. Spurred by the devastation of the December 26, 2004, Indian Ocean tsunami that killed more than 230,000 people, the private sector actively supported the Intergovernmental Oceanographic Commission's (IOC"s) efforts to develop a tsunami warning system and mitigation plan for the Indian Ocean region. As each country in the region developed its requirements, SAIC recognized that many of these underdeveloped countries would need significant technical assistance to fully execute their plans. With the original focus on data fusion, consequence assessment tools, and warning center architecture, it was quickly realized that the cornerstone of any tsunami warning system would be reliable tsunami detection buoys that could meet very stringent operational standards. Our goal was to leverage extensive experience in underwater surveillance and oceanographic sensing to produce an enhanced and reliable deep water sensor that could meet emerging international requirements. Like the NOAA Deep-ocean Assessment and Recording of Tsunamis (DART TM ) buoy, the SAIC Tsunami Buoy (STB) system consists of three subsystems: a surfaccommunications buoy subsystem, a bottom pressure recorder subsystem, and a buoy mooring subsystem. With the operational success that DART has demonstrated, SAIC decided to build and test to the same high standards. The tsunami detection buoy system measures small changes in the depth of the deep ocean caused by tsunami waves as they propagate past the sensor. This is accomplished by using an extremely sensitive bottom pressure sensor/recorder to measure very small changes in pressure as the waves move past the buoy system. The bottom pressure recorder component includes a processor with algorithms that recognize these characteristics, and then immediately alerts a tsunami warning center through the communications buoy when the processor senses one of these waves. In addition to the tsunami detection buoy system, an end-to-end tsunami warning system was developed that builds upon the country's existing disaster warning infrastructure. This warning system includes 1) components that receive, process, and analyze buoy, seismic and tide gauge data; 2) predictive tools and a consequence assessment tool set to provide decision support; 3) operation center design and implementation; and 4) tsunami buoy operations and maintenance support. The first buoy was deployed Oct. 25, 2006, approximately 200 nautical miles west of San Diego in 3,800 meters of water. Just three weeks later, it was put to the test during an actual tsunami event. On Nov. 15, 2006, an 8.3 magnitude earthquake rocked the Kuril Islands, located between Japan and the Kamchatka Peninsula of Russia. That quake generated a small tsunami. Waves from the tsunami propagated approximately 4,000 nautical miles across the Pacific Ocean in about nine hours-- a speed of about 445 nautical miles per hour when this commercial buoy first detected them. Throughout that event, the tsunami buoy system showed excellent correlation with data collected by a NOAA DART buoy located 28 nautical miles north of it. Subsequent analysis revealed that the STB matched DART operational capabilities and performed flawlessly. The buoy proved its capabilities again on Jan. 13, 2007, when an 8.1 magnitude earthquake occurred in the same region, and the STB detected the seismic event. As a result of the successes of this entire project, SAIC recently applied for and received a license from NOAA to build DART systems.
S53A-1030
EC-NEAREST project: First Tsunami Underwater Observatory in the Gulf of Cadiz
On August 25 at 10:15 p.m. G.M.T. the first prototype of an underwater tsunami observatory was successfully deployed at 3207 m water depth in the Gulf of Cadiz. The deployment,150 km South of Portugal, was planned within the NEAREST project (Integrated observations from NEAR shore sources of Tsunamis: towards an early warning system – EC, cont. 37110) The observatory is based on the GEOSTAR platform, previously developed by a consortium led by INGV. The ocean bottom equipment includes a broad-band seismometer, hydrophone, gravitymeter and pressure gauge, allowing for cross-checking of the signals. Pre-processing of seismic and pressure data is done underwater. A relay surface buoy, equipped with additional meteorological sensors, is in acoustic link with the seafloor platform and performs satellite real-time transmission to the shore stations, where the integration of the marine with land data from operating networks will improve the reliability of the tsunami detection strategy. Additional sensors can be installed, in a framework of a multi-parameter seafloor observatory, able to characterise earthquake/ tsunami processes and to extend its capacity to a larger set of monitoring strategies. The deployment, performed by the R/V Urania, was supervised by a team of scientists and engineers from the partners in the project. The site of the seafloor observatory was selected according to the available geological and bathymetric data, to maximise proximity to the potential sources; its present position is 36º 21.875'N;09º28.885'W; 3207 m w.d. Data are currently being acquired and processed on an experimental basis and integrated with information from land monitoring networks of: Portugal, Spain and Morocco.This observatory is an important milestone towards a European tsunami warning system - UNESCO IOC resolution of December 2005 to develop the "North East Atlantic and Mediterranean Tsunami Warning System".
S53A-1031
An innovative tsunami detector operating in tsunami generation environment
On August 25th 2007 a tsunami detector installed onboard the multi-parameter observatory GEOSTAR was successfully deployed at 3200 b. s. l. in the Gulf of Cadiz, Portugal. This activity is within the NEAREST EC Project (http://nearest.bo.ismar.cnr.it/ ). Among other deliverables, the NEAREST project will produce and test the basic parts of an operational prototype of a near field tsunami warning system. This system includes an onshore warning centre, based on the geophysical monitoring networks which are already operating, and a tsunami detector deployed on board GEOSTAR at the sea bottom. On land the warning centre is in charge of collecting, integrating, and evaluating data recorded at sea. At the sea bottom data is recorded and processed by an advanced type of tsunami detector which includes: a pressure sensor, a seismometer and two accelerometers. The detector communicates acoustically with a surface buoy in two-way mode. The buoy is equipped with meteo station, GPS and tiltmeter and is connected to a shore station via satellite link. The prototype is designed to operate in tsunami generation areas for detection-warning purpose as well as for scientific measurements. The tsunami detector sends a near real time automatic alert message when a seismic or pressure threshold are exceeded. Pressure signals are processed by the tsunami detection algorithm and the water pressure perturbation caused by the seafloor motion is taken into account. The algorithm is designed to detect small tsunami waves, less than one centimetre, in a very noisy environment. Our objective is to combine a novel approach to the tsunami warning problem, with a study of the coupling between the water column perturbations and sea floor motion, together with the long term monitoring of geophysical, geochemical and oceanographic parameters. http://nearest.bo.ismar.cnr.it/
S53A-1032
Use for Hydroacoustic (T-phases) for Tsunami Warning
We have identified a method to exploit T-phases recorded at the Comprehensive Test Ban Treaty's International Monitoring System in-water hydroacoustic arrays for tsunami warning by examining the T-phase spectral slope and high-frequency duration. Over a dataset consisting of several hundred M>=5.5 events recorded at the hydroacoustic stations, long-duration and shallow spectral slope T-phases originating from subduction-zones region are associated with tsunamigenic earthquakes. These observations can be explained based on the following concepts: the observed T-phase results from propagation within the earth convolved with propagation within the water column nd a earth-water conversion factor. Propagation within the water column at 5-100 Hz is simple as there is minimal spectral deformation. Within the earth, the propagation is more complicated, with scattering, multiple phases, and anelastic attenuation. As the primary factor impacting the spectral shape is the anelastic attenuation, and assuming Q is constant with frequency, that means that the observed T-phase spectrum is dependent only on the value of Q and the number of wave cycles. Thus, the presence of a shallow slope means that either Q is high or the distance is small. In subduction zones, the energy travels through the low-Q accretionary wedge; thus implying that a shallow slope in subduction zone environment means the rupture is very close to the water interface. Similarly, the T-phase duration can be used to estimate the source duration. At higher frequencies, the less direct seismic energy will be attenuated prior to entering the water column. Therefore, the duration of the higher frequency energy gives the source duration. We had also hoped to use the direction of the T-phase to determine source finiteness. However, the analysis indicated that the broadband T-phase directionality is influenced by many factors including the source finiteness; we found that smaller, deep events show a broad range of back-azimuths, presumably the result of a longer solid-earth T-phase path. In addition, if the propagation to the hydroacoustic station aligns with the fault orientation, then source finiteness estimates cannot be obtained. It is possible that directionality from higher frequency T-phase energy will prove to be a robust indicator.
S53A-1033
Array Analysis of T-Waves Generated by Large Tsunamigenic and Non-tsunamigenic Earthquakes
Hydroacoustic stations of the International Monitoring System are now operating in the Atlantic, Pacific and Indian Oceans. Each station consists of one or two three-element arrays with a bandwidth of 100 Hz and precise signal direction finding capability on the order of 0.25 degrees. These arrays routinely record T-waves generated by sub- sea earthquakes and provide the capability to localize the points of sound generation on the ocean floor. The utility of these arrays for understanding T-wave generation was demonstrated with data from the Great Sumatran Earthquake of December 26, 2004, where T-waves of 15-minutes in time and nearly 1000-km in source extent could be localized along the Sumatran trench. This correlates well with seismic determinations of the fault dynamics and implies the capability to perform real-time estimates of source complexity, which impacts tsunami warning. Since then, we have applied array-processing techniques to the analysis of T-waves from dozens of sub-sea earthquakes in the Atlantic, Pacific and Indian Oceans. Here we report on a comparison of results from the larger events in our dataset (magnitudes > 6.5 Mw), some of which have generated recordable or reportable tsunamis. These include: March 28, 2005 Northern Sumatra, Mw 8.6; May 16, 2006 Northern Sumatra, Mw 6.8; March 30, 2007 Kamchatka, Mw 6.5; and August 15, 2007 Coastal Peru, Mw 8.0. T-waves from all of these events can be localized to an extended source area, which is often more than 100-km long, but the relationship of this source area to tsunamigenisis is complicated. This source area may or may not be directly indicative of the fault zone extent, since T-wave generation results from a complex interaction seismic waves with ocean bottom topography. This interaction must be accounted for to relate acoustic source area to seismic faulting. One way to accomplish this is by using smaller events (aftershocks) as Green's functions for the main shock. Array analysis of T-waves from aftershocks often show acoustic excitation patterns that mimic those of the main shock. Addition factors related to tsunamigensis include spectral content and source duration.
S53A-1034
Mwpd: Rapid Determination of Earthquake Magnitude and Tsunamigenic Potential from P Waveforms
We present a duration-amplitude procedure for rapid determination of an earthquake moment magnitude, Mwpd, from P-wave recordings at teleseismic distances. The Mwpd magnitude can be obtained within 20 minutes or less after the event origin time since the required data is available in near-real time. The procedure determines apparent source durations, T0, from high-frequency, P-wave records, and estimates moments through integration of broadband displacement waveforms over the interval tP to tP+T0, where tP is the P arrival time. We apply the duration-amplitude methodology to a number of recent, large earthquakes (Global Centroid-Moment Tensor magnitude, MwCMT, 6.6 to 9.3) with diverse source types. The results show that a correction to the moment estimates for interplate thrust and possibly tsunami earthquakes is necessary to best match MwCMT. With this correction, Mwpd matches MwCMT typically within ±0.2 magnitude units, with a standard deviation of σ=0.11, outperforming other approaches to rapid magnitude determination. In addition, Mwpd does not exhibit saturation for the largest events, or, equivalently, ΔM=Mwpd-MwCMT does not become more negative with increasing MwCMT. The explicit use of the source duration for integration of displacement seismograms, the moment correction, and other characteristics of the duration-amplitude methodology make it an extension of the widely used, Mwp, rapid-magnitude procedure. The obtained duration-amplitude moments and durations allow rapid estimation of the energy-to-moment ratio Θ used for identification of tsunami earthquakes. The need for a moment correction for interplate thrust and possibly tsunami earthquakes may have important implications for the source physics of these events. http://www.alomax.net/posters/duration-amplitude/
S53A-1035
Rapid estimation of earthquake size using the broadband P-wave magnitude mB
The size of an earthquake is the most fundamental characteristic besides its location. In order to adequately respond to an earthquake, especially when human life is at stake, an accurate size estimate must be available rapidly. Quick quantification of earthquake magnitude, however, continues to present a challenge particularly for large events. Especially following the disastrous December 26, 2004, Sumatra-Andaman event, a number of new methods have been proposed to overcome this difficulty. As early as 1956, Gutenberg and Richter introduced the body wave magnitude mB = log10(A/T)max + Q(Δ,z). They determined the displacement amplitude A and dominant period T from intermediate- period displacement seismograms. The calibration function Q was derived accordingly. Later the same formula and calibration function was adopted for the narrow-band, WWSSN short-period magnitude mb, resulting in systematically underestimated magnitudes for events larger than about 5.5. We demonstrate that by using the full broadband P-waveform, the original but long-forgotten mB provides excellent magnitude estimates up to at least magnitude 8. It is probably the simplest of all P-wave based magnitudes, since modern broadband instruments record ground velocity directly and (A/T)max may simply be replaced by Vmax/2π. Based on broadband recordings of more than 1000 large earthquakes since 1990, we derived a new mB calibration function, spanning the distance range from 5 to 100 degrees. By using this calibration function, first mB estimates may become available after as little as 2 minutes following the origin time. The trivially simple computation makes mB a competitor for MWP. The latter involves double integration of velocity seismograms and is known to suffer from stability problems. Furthermore, it is highly susceptible to noise. In contrast, mB is guaranteed to be numerically stable and noise is of less concern. This is an important consideration especially in automated setups where processing speed is crucial and time-consuming manual interaction has to be avoided.
S53A-1036
New Theory for Tsunami Propagation and Estimation of Tsunami Source Parameters
In numerical studies based on the shallow water equations for tsunami propagation, vertical accelerations and velocities within the sea water are neglected, so a tsunami is usually supposed to be produced by an initial free surface displacement in the initially still sea. In the present work, new theory for tsunami propagation across the deep sea is discussed, that accounts for the vertical accelerations and velocities. The theory is based on the solutions for the water surface displacement obtained in [Mindlin I.M. Integrodifferential equations in dynamics of a heavy layered liquid. Moscow: Nauka*Fizmatlit, 1996 (Russian)]. The solutions are valid when horizontal dimensions of the initially disturbed area in the sea surface are much larger than the vertical displacement of the surface, which applies to the earthquake tsunamis. It is shown that any tsunami is a combination of specific basic waves found analytically (not superposition: the waves are nonlinear), and consequently, the tsunami source (i.e., the initially disturbed body of water) can be described by the numerable set of the parameters involved in the combination. Thus the problem of theoretical reconstruction of a tsunami source is reduced to the problem of estimation of the parameters. The tsunami source can be modelled approximately with the use of a finite number of the parameters. Two-parametric model is discussed thoroughly. A method is developed for estimation of the model's parameters using the arrival times of the tsunami at certain locations, the maximum wave-heights obtained from tide gauge records at the locations, and the distances between the earthquake's epicentre and each of the locations. In order to evaluate the practical use of the theory, four tsunamis of different magnitude occurred in Japan are considered. For each of the tsunamis, the tsunami energy (E below), the duration of the tsunami source formation T, the maximum water elevation in the wave originating area H, mean radius of the area R, and the average magnitude of the sea surface displacement at the margin of the wave originating area h are estimated using tide gauges records. The results are compared (and, in the author's opinion, are in line) with the estimates known in the literature. Compared to the methods employed in the literature, there is no need to use bathymetry (and, consequently, refraction diagrams) for the estimations. The present paper follows closely earlier works [Mindlin I.M., 1996; Mindlin I.M. J. Appl. Math. Phys. (ZAMP), 2004, vol.55, pp. 781-799] and adds to their theoretical results. Example. The Hiuganada earthquake of 1968, April, 1, 9h 42m JST. A tsunami of moderate size arrived at the coast of the south-western part of Shikoku and the eastern part of Kyushu, Japan. Tsunami parameters listed above are estimated with the theory being discussed for two models of tsunami generation: (a) by initial free surface displacement (the case for numerical studies): E=1.91· 1012J, R=22km, h=17.2cm; and (b) by a sudden change in the velocity field of initially still water: E=8.78· 1012J, R=20.4km, h=9.2cm. These values are in line with known estimates [Soloviev S.L., Go Ch.N. Catalogue of tsunami in the West of Pacific Ocean. Moscow, 1974]: E=1.3· 1013J (attributed to Hatori), E=(1.4 - 2.2)· 1012J (attributed to Aida), R=21.2km, h=20cm [Hatory T., Bull. Earthq. Res. Inst., Tokyo Univ., 1969, vol. 47, pp. 55-63]. Also, estimates are obtained for the values that could not be found based on shallow water wave theory: (a) H=3.43m and (b) H=1.38m, T=16.4sec.
S53A-1037
Far-field tsunami amplitude estimations from numerical simulations and empirical laws
Tsunami amplitudes in the far field can be simply estimated from the seismic moment M sub 0 of the source event, which primarily controls tsunami excitation [Talandier and Okal, 1989]. This formula used only a distance correction (1) combining geometrical spreading on the spherical Earth with a mild amount of dispersion. In the context of tsunami warning in French Polynesia, this relation has been in operational use for close to 20 years, using data in several harbors: Papeete (Tahiti), Taiohae (Marquesas) and Rikitea (Gambier Island). However, it is possible to refine it by considering the influence of additional parameters of the source, such as focal depth, source directivity, focusing effect and site amplification. In the present study, tsunamis due to 10 large earthquakes (1946 Aleutian, 1964 Alaska, 1960 and 1995 Chile, 2006 Tonga, etc. plus 3 virtual generic events) have been modeled for 57 virtual sensors distributed in the deep sea throughout the Pacific Basin and its boundaries. Computed amplitudes are compared with the values predicted by the 1989 formula, in order to discuss: (2) the constant of proportionality between tsunami amplitude and seismic moment; and (3) the possibility of defining an empirical correction for source directivity, expressed as a function of receiver azimuth relative to fault strike. Furthermore, tsunami amplitudes recorded in Papeete harbour and Nuku Hiva bay, have been compared to our numerical simulations, in order to define: (4) a constant of site amplification; and (5) a focusing constant, depending on source-receiver geometry, and expressing refraction due to variable bathymetry. The influence of focal depth is also examined. The final formula obtained, integrates the above five terms.
S53A-1038
Tsunami Simulations With Unstructured Grids in Support of a Tsunami Early Warning System for the Indian Ocean
A Finite Element Model for simulations of Tsunami waves (TsunAWI) has been developed. It is part of the German Indonesian Tsunami Early Warning System (GITEWS). Model results will be the main source for the prediction of arrival times and expected wave heights. The unstructured triangular grid has a relatively low resolution in the ocean interior (about 10 km) coastal regions however are very well resolved (up to 80 m). This flexibility allows for a good representation of the wave propagation in the deep ocean as well as inundation processes without the need of nesting different grids. Numerical experiments simulating the Indian Ocean Tsunami generated by the earthquake of Dec. 26 in 2004 have been conducted. The role of the model bathymetry and topography (based on the GEBCO dataset as well as data from the SRTM satellite mission and ship cruises) has been investigated. The inundation obtained in the simulations were compared to field measurements as well as to satellite images of Banda Aceh region. Furthermore the results were compared to simulations of the same event by the finite difference model TUNAMI- N3 with three nested grids and resolutions ranging from 900 m in the coarsest grid to 90 m in the finest nested grid. It turned out that the two models coincide fairly well with respect to wave propagation and inundation. However in both approaches a good knowledge of topography and bathymetry especially in the near shore range turned out to be crucial for realistic results.
S53A-1039
Deep-ocean tsunami modelling based on the teleseismic source model of the 2006 Kuril earthquake
The importance of tsunami modelling has increased since the 2004 Sumatra-Andaman earthquake and the Indian Ocean tsunami that followed it. A magnitude 8.3 earthquake occurred in the Kuril subduction zone on 15 November, 2006 and resulted in teletsunamis widely recorded by bottom pressure recorders deployed in the northern Pacific Ocean. Because these recordings were unaffected by shallow complicated bathymetry near the coast, this provides a unique opportunity to investigate whether seismic rupture models can be inferred from teleseismic waves with sufficient accuracy to be used to forecast teletsunamis. In this study, we estimated the rupture model of the 2006 Kuril earthquake by inverting the teleseimic waves and used that to model the tsunami source. The tsunami propagation was then calculated by solving the linear long-wave equations. We finally found that the simulated 2006 Kuril tsunamis compared very well to the ocean bottom recordings when simultaneously using P and surface waves in the teleseismic inversion.
S53A-1040
Role of Splay Faulting and Dispersion in Tsunami Waveforms
Tsunami theory based on standard earthquake source models and shallow-water gravity wave methodology suggests that the 2004 Indian Ocean tsunami should have arrived in Sri Lanka as a large wave followed by a depression. However, eyewitness accounts of the tsunami arrival indicate that there were multiple wave arrivals separated by a recession (Liu et al., 2005). The tsunami waveform was also measured by the Jason-1 satellite two hours after the event, which is approximately the same time as the first arrival in Sri Lanka. The altimetry data agrees with those accounts and shows a double peaked lead wave (Smith et al., 2005). This may be the result of two areas of vertical uplift in the source region caused by slip on both the plate interface and on a steeper splay fault branching from this detachment. Only a small amount of slip is needed on the steep splay to vertically displace the seafloor enough to create a double peaked wave. There have been attempts to distinguish between coseismic slip on the detachment and slip on a splay or a combination of both of these surfaces. GPS inversions by Banerjee et al., attempted to distinguish between these possibilities, but one model of slip distribution was not preferred. The tsunami waveform and arrival time should vary with these models, but inversions using tide gauge data from the 1944 Tonankai event were similarly unable to prefer a model based on splay slip, detachment slip or a combination of the two (Baba et al., 2006). We suggest that the satellite altimetry data can be used as a constraint on splay fault activation. We model the seafloor deformation based on a variety of splay fault locations and slip distributions to determine the criteria for a double peaked arrival at Sri Lanka. The Jason-1 track data indicates that the waves are dispersive (Kulikov, 2005), so we compare the results of linear wave propagation using shallow water theory (c = \sqrt{gh}) versus including wave dispersion (c = \sqrt{g \tanh(kh) / k}), albeit for a model with uniform sea depth h; here k is wave number. The distance between and the relative magnitude of the peaks is determined by the fault position and amount of moment transferred to the splay. Aftershock distribution following the 2004 Sumatra Andaman earthquake hints that splay faults were activated during the event at 70 km and 110 km inward of the trench (Araki et al., 2005). A splay fault location of 70 km inboard does not result in a double peaked wave matching the observed wave, but a fault located 110 km inboard results in a modeled spacing of 106 km between peaks which correlates well with the observed ~115 km from the Jason-1 data. When the effects of dispersion are accounted for in the wave propagation, the shape of the wave changes with the distance from the source. An initially sharp rise in the wave front becomes broader due to the lag of the high frequency components, and the waveform can evolve into two peaks of approximately the same wavelength and magnitude. Such a wave structure is similar to that observed by the satellite. We conclude that dispersion is important to the shape of the propagating wave, if not to its arrival time, and that the source process likely involved a significant component of splay faulting.
S53A-1041
Observed and Simulated High Amplitude Tsunami Offshore of Northern Banda Aceh During the 2004 Sumatra Earthquake
Several isolated high waves were observed about 0.5 to 20 km away from the west coast of northern Sumatra based from the interviews undertaken with fishermen who were on their fishing vessels when the December 2004 Sumatra earthquake occurred. The fishermen were on their wooden boats powered by motor engine manned by one to 20 crew members. Most of them felt the earthquake shocks and some of the crews immediately recognized the shaking phenomena as an earthquake based on their previous experiences. About half of interviewees had difficulties in controlling their boats during the shaking, which lasted about 10 to 15 min. About 5 to 15 minute after the shaking their boats were struck by the 10-20m high waves. Moreover, their boats, located more than several km off the coast of northern Sumatra, were struck repeatedly by more or less isolated high waves. Based from these interviews it is ascertained that tsunami can also pose danger even offshore when the tsunami amplitudes are high and water depths are shallower than about 50m. These observations also provided new evidence that the 2004 tsunami is already high enough to damage or overturn ships offshore. We simulated tsunami waves using a source model of the 2004 earthquake by Fujii and Satake (2007). Since the seafloor topography is not clear for both continental shelf and slope in northern Sumatra, the depth of 150m at 30 km off the coast varying gently towards the coast is presumed. This depth is assumed based on the type of fish caught in the area that are likely to live near the rocky seafloor at depths around 50m. For the numerical analysis, the non-linear long wave equation and the non-linear dispersive long wave equation were used for depths deeper and shallower than 150m, respectively. As a result, at depths around 20m, dispersive short- period solitary waves with height of 6-8m were generated that are similar to the waves that struck the fishing vessels. However, the tsunami waves that the fishermen encountered were as high as 10-20m. This means that the incoming tsunami in the shallower depths was higher than the wave heights derived from Fujii and Satakefs model.
S53A-1042
Detailed Modeling of the 2004 Tsunami Flooding in the Banda Aceh and Lhok Nga Districts (Sumatra, Indonesia)
Within the French-Indonesian TSUNARISQUE programme (http://www.tsunarisque.cnrs.fr) a great amount of observational data has been gathered in order to quantify the impact of the 2004 tsunami on the northern region of Sumatra, near Banda Aceh and Lhok Nga. Tsunami studies and numerical modeling in particular have rarely been confronted to such a database. Now this consortium also provides a detailed bathymetric and topographic mapping of the area, which allows us to refine our previous hydrodynamic modeling of the tsunami flooding. Our models succeed in reproducing the overall trend of the measured wave heights. We focus on the timing of the wave arrivals on the various locations where temporal observations are available, such as the clocks broken when struck by the tsunami, or as the collision of the wave merging across Banda Aceh and the incoming wave at Lhok Nga. We discuss the delayed arrival time at Lhok Nga and the early estimation at Banda Aceh obtained by the previous models. Since slip heterogeneities and rupture kinematics can infer various tsunami impacts on the area, we test several source mechanisms by letting vary the coseismic slip and fault geometry along the thrust.
S53A-1043
Source of Run-up Tsunami Deposits Based on Foraminiferal Tests Transportation and their Hydrodynamic Verification
Foraminiferal tests are often yielded in tsunami deposits and provide important information on the source of sediment supply. Water depth of sediment source areas and transport distance from the source areas estimated from the foraminiferal tests may be useful criteria to identify the tsunami deposits. Tractive force by tsunami waves is in inverse proportion to the water depth at depositional sites. Horizontal distance of sediment transport by tsunami waves is proportion to the period and amplitude of the waves. Then, threshold amplitude and period of tsunami waves to explain the sediment transport depend mainly on the weight of sediment grains and water depth of depositional sites. Some washover and near-shore deposits yield the foraminiferal tests transported long distance from deep marine bottom. These deposits are likely to result from tsunamis with large wave period and amplitude. Foraminiferal tests reported from some tsunami deposits indicate that they were originated from 100m or deeper sea floor and transported a long distance, up to several kilo meters, to the coasts. However, theoretical amplitude of tsunami waves to explain the transport of the foraminiferal tests is unusually large. Change in the action of the tsunami water mass affected by local topography, such as submarine canyons, is convincing cause of the contradictions in sediment transportation estimated from foraminiferal tests and from theories of tsunami propagation.
S53A-1044
Tsunami Casualty Model
More than 4500 deaths by tsunamis were recorded in the decade of 1990. For example, the 1992 Flores Tsunami in Indonesia took away at least 1712 lives, and more than 2182 people were victimized by the 1998 Papua New Guinea Tsunami. Such staggering death toll has been totally overshadowed by the 2004 Indian Ocean Tsunami that claimed more than 220,000 lives. Unlike hurricanes that are often evaluated by economic losses, death count is the primary measure for tsunami hazard. It is partly because tsunamis kill more people owing to its short lead- time for warning. Although exact death tallies are not available for most of the tsunami events, there exist gender and age discriminations in tsunami casualties. Significant gender difference in the victims of the 2004 Indian Ocean Tsunami was attributed to womenfs social norms and role behavior, as well as cultural bias toward womenfs inability to swim. Here we develop a rational casualty model based on humansf limit to withstand the tsunami flows. The application to simple tsunami runup cases demonstrates that biological and physiological disadvantages also make a significant difference in casualty rate. It further demonstrates that the gender and age discriminations in casualties become most pronounced when tsunami is marginally strong and the difference tends to diminish as tsunami strength increases.
S53A-1045
Mortality distributions in time and space during tsunamis: implications for the performance of tsunami warning systems and the importance of Education for Self Warning and Voluntary Evacuation (ESWAVE)
Tsunami warning systems take advantage of the time windows of opportunity between occurrence of the source event and the impact of the tsunami on vulnerable coastlines. These time windows are taken up by instrumental detection of the event or the tsunami, distribution of warning messages, and the time needed for effective mitigation by evacuation of the expected inundation zone. Time distributions of mortality during a variety of well documented historical tsunamis show that with one exception (the 1946 Aleutians tsunami) at least 80 percent of tsunami mortality occurs within the first hour after the source event, even in the case of transoceanic tsunamis. This usually leaves less than 30 minutes for the warning systems to function if they are to prevent these 80 percent of deaths in tsunamis. However, in most cases the 80 percent of deaths also occur within zones of strong felt seismic shaking from tsunamigenic earthquakes, or within zones where other warning signs are evident to the vulnerable populations. Even outside these zones premonitory signs of tsunamis like leading negative or small positive waves commonly provide sufficient warning, in daylight at least, for effective evacuations by aware or educated populations. Instrument – based tsunami warning systems need to be significantly more rapid than at present to be as effective as Education for Self – Warning and Voluntary Evacuation (ESWAVE), and both need to be coupled with coastal land use planning to prevent mass mortalities in sites where evacuation is impossible within the tsunami propagation time.