OS22B-01 INVITED 10:20h
The National Tsunami Hazard Mitigation Program
The National Tsunami Hazard Mitigation Program (NTHMP) is a state/Federal partnership that was created to reduce the impacts of tsunamis to U.S. Coastal areas. It is a coordinated effort between the states of Alaska, California, Hawaii, Oregon, and Washington and four Federal agencies: the National Oceanic and Atmospheric Administration (NOAA), the Federal Emergency Management Agency (FEMA), the U.S. Geological Survey (USGS), and the National Science Foundation (NSF). NOAA has led the effort to forge a solid partnership between the states and the Federal agencies because of it's responsibility to provide tsunami warning services to the nation. The successful partnership has established a mitigation program in each state that is developing tsunami resilient coastal communities. Inundation maps are now available for many of the coastal communities of Alaska, California, Hawaii, Oregon, and Washington. These maps are used to develop evacuation plans and, in the case of Oregon, for land use management. The NTHMP mapping technology is now being applied to FEMA's Flood Insurance Rate Maps (FIRMs). The NTHMP has successfully upgraded the warning capability in NOAA so that earthquakes can be detected within 5 minutes and tsunamis can be detected in the open ocean in real time. Deep ocean reporting of tsunamis has already averted one unnecessary evacuation of Hawaii and demonstrated that real-time tsunami forecasting is now possible. NSF's new Network for Earthquake Engineering (NEES) program has agreed to work with the NTHMP to focus tsunami research on national needs. An overview of the NTHMP will be given including a discussion of accomplishments and a progress report on NEES and FIRM activities.
http://www.pmel.noaa.gov/tsunami-hazard/
OS22B-02 INVITED 10:35h
Mapping Coastal Flood Zones for the National Flood Insurance Program
The National Flood Insurance Program (NFIP) was created by Congress in 1968, and significantly amended in 1973 to reduce loss of life and property caused by flooding, reduce disaster relief costs caused by flooding and make Federally backed flood insurance available to property owners. These goals were to be achieved by requiring building to be built to resist flood damages, guide construction away from flood hazards, and transferring the cost of flood losses from taxpayers to policyholders. Areas subject to flood hazards were defined as those areas that have a probability greater than 1 percent of being inundated in any given year. Currently over 19,000 communities participate in the NFIP, many of them coastal communities subject to flooding from tides, storm surge, waves, or tsunamis. The mapping of coastal hazard areas began in the early 1970's and has been evolving ever since. At first only high tides and storm surge were considered in determining the hazardous areas. Then, after significant wave caused storm damage to structures outside of the mapped hazard areas wave hazards were also considered. For many years FEMA has had Guidelines and Specifications for mapping coastal hazards for the East Coast and the Gulf Coast. In September of 2003 a study was begun to develop similar Guidelines and Specifications for the Pacific Coast. Draft Guidelines and Specifications will be delivered to FEMA by September 30, 2004. During the study tsunamis were identified as a potential source of a 1 percent flood event on the West Coast. To better understand the analytical results, and develop adequate techniques to estimate the magnitude of a tsunami with a 1 percent probability of being equaled or exceeded in any year, a pilot study has begun at Seaside Oregon. Both the onshore velocity and the resulting wave runup are critical functions for FEMA to understand and potentially map. The pilot study is a cooperative venture between NOAA and USGS that is partially funded by both agencies and by FEMA. The results of the pilot study will help FEMA determine when tsunamis should be considered in mapping coastal hazards, how to predict their impact, how they should be mapped and possibly the construction standards for zones mapped as having a 1 percent or greater chance of suffering a tsunami.
OS22B-03 INVITED 10:50h
Probabilistic Tsunami Hazard Analysis in Japan
Several projects for estimating tsunami hazard for the Japanese coasts are ongoing. I will introduce four of them, two by the government and the others by professional society and a research institute. The Earthquake Research Committee of the Headquarters of Earthquake Research Promotion of the Government (http://www.jishin.go.jp/main/index-e.html) has evaluated the probabilities of large subduction-zone earthquakes, as well as inland earthquakes on active faults, in and around Japan. In the Tokachi-oki region along Kuril Trench, for example, the probability of an M~8 earthquake occurrence by 2033 was estimated as ~ 60 %. Six months after this evaluation has publicized, the September 26, 2003 Tokachi-oki earthquake (M 8.0) took place and generated up to 4 m of tsunami on the Pacific coast of Hokkaido. Similarly high probabilities (> 40%) of M~8 event have been estimated for parts of Nankai Trough, Japan Trench and Kuril Trench. Central Disaster Management Council of the Cabinet Office has estimated earthquake and tsunami hazards for Nankai Trough (Tokai, Nankai and Tonankai earthquakes). They established scenario sources of tsunami, estimated the coastal tsunami heights (on 50 m grids), and assessed the damage on human, structures, and lifelines. The results are used for countermeasures and disaster prevention by the coastal communities, local and national governments. One of the outcomes is large-scale coastal hazard maps. Similar projects for Japan and Kuril Trenches has recently started. Tsunami Subcommittee of Japan Society of Civil Engineers (Annaka et al., this meeting) is making probabilistic tsunami hazard curves (tsunami heights as a function of its exceeding probability) for specific coastal sites. The hazard analysis is based on tsunami numerical simulations from earthquake faults. Logic-tree approach is adopted to evaluate the fault parameters. Geological Survey of Japan is making large-scale (1:25,000) tsunami inundation maps, as well as smaller-scale (1:500,000) map for coastal tsunami heights, based on the studies of tsunami deposits (Nanayama et al., Nature, 2003) combined with tsunami numerical simulation for possible sources. The maps will be released as CD-ROM.
OS22B-04 11:05h
Probabilistic Tsunami Hazard Assessment: the Seaside, Oregon Pilot Study
A pilot study of Seaside, Oregon is underway, to develop methodologies for probabilistic tsunami hazard assessments that can be incorporated into Flood Insurance Rate Maps (FIRMs) developed by FEMA's National Flood Insurance Program (NFIP). Current NFIP guidelines for tsunami hazard assessment rely on the science, technology and methodologies developed in the 1970s; although generally regarded as groundbreaking and state-of-the-art for its time, this approach is now superseded by modern methods that reflect substantial advances in tsunami research achieved in the last two decades. In particular, post-1990 technical advances include: improvements in tsunami source specification; improved tsunami inundation models; better computational grids by virtue of improved bathymetric and topographic databases; a larger database of long-term paleoseismic and paleotsunami records and short-term, historical earthquake and tsunami records that can be exploited to develop improved probabilistic methodologies; better understanding of earthquake recurrence and probability models. The NOAA-led U.S. National Tsunami Hazard Mitigation Program (NTHMP), in partnership with FEMA, USGS, NSF and Emergency Management and Geotechnical agencies of the five Pacific States, incorporates these advances into site-specific tsunami hazard assessments for coastal communities in Alaska, California, Hawaii, Oregon and Washington. NTHMP hazard assessment efforts currently focus on developing deterministic, "credible worst-case" scenarios that provide valuable guidance for hazard mitigation and emergency management. The NFIP focus, on the other hand, is on actuarial needs that require probabilistic hazard assessments such as those that characterize 100- and 500-year flooding events. There are clearly overlaps in NFIP and NTHMP objectives. NTHMP worst-case scenario assessments that include an estimated probability of occurrence could benefit the NFIP; NFIP probabilistic assessments of 100- and 500-yr events could benefit the NTHMP. The joint NFIP/NTHMP pilot study at Seaside, Oregon is organized into three closely related components: Probabilistic, Modeling, and Impact studies. Probabilistic studies (Geist, et al., this session) are led by the USGS and include the specification of near- and far-field seismic tsunami sources and their associated probabilities. Modeling studies (Titov, et al., this session) are led by NOAA and include the development and testing of a Seaside tsunami inundation model and an associated database of computed wave height and flow velocity fields. Impact studies (Synolakis, et al., this session) are led by USC and include the computation and analyses of indices for the categorization of hazard zones. The results of each component study will be integrated to produce a Seaside tsunami hazard map. This presentation will provide a brief overview of the project and an update on progress, while the above-referenced companion presentations will provide details on the methods used and the preliminary results obtained by each project component.
OS22B-05 11:20h
Size and Age Characteristics for West Coast Tsunamigenic Landslides
Multibeam bathymetric imagery is now available for a number of well-defined submarine landslide deposits along the west coast of the United States. Several of these landslides are known to have caused damaging tsunamis and others are of sufficient size to have generated tsunamis when they occurred, assuming that their motion was rapid. These failures are located off Palos Verdes Peninsula, California, and within Santa Barbara Channel, California, Commencement Bay, Washington, Resurrection Bay, Alaska, and Port Valdez, Alaska. For two of these failures, ages were determined by identifying acoustic reflectors in the vicinity of the failed masses that either clearly postdate or predate the landslide events. The ages of the reflectors are determined by tracing them to the locations of nearby ODP borings or to piston cores dated using radiocarbon methods. Three of the landslides produced tsunamis during historic time (post 1750 AD) so the ages are well constrained. High-resolution subbottom reflection profiles also allow us to estimate the dimensions of the failed masses. Although the examples selected clearly do not represent all scales of tsunamigenic west coast landslides, this information is useful in providing input to statistically based models of landslide-induced tsunamis.
OS22B-06 11:35h
Probabilistic Risk Analysis of Run-up and Inundation in Hawaii due to Distant Tsunamis
Risk assessment of natural hazards usually includes two aspects, namely, the probability of the natural hazard occurrence and the degree of damage caused by the natural hazard. Our current study is focused on the first aspect, i.e., the development and evaluation of a methodology that can predict the probability of coastal inundation due to distant tsunamis in the Pacific Basin. The calculation of the probability of tsunami inundation could be a simple statistical problem if a sufficiently long record of field data on inundation was available. Unfortunately, such field data are very limited in the Pacific Basin due to the reason that field measurement of inundation requires the physical presence of surveyors on site. In some areas, no field measurements were ever conducted in the past. Fortunately, there are more complete and reliable historical data on earthquakes in the Pacific Basin partly because earthquakes can be measured remotely. There are also numerical simulation models such as the Cornell COMCOT model that can predict tsunami generation by an earthquake, propagation in the open ocean, and inundation onto a coastal land. Our objective is to develop a methodology that can link the probability of earthquakes in the Pacific Basin with the inundation probability in a coastal area. The probabilistic methodology applied here involves the following steps: first, the Pacific Rim is divided into blocks of potential earthquake sources based on the past earthquake record and fault information. Then the COMCOT model is used to predict the inundation at a distant coastal area due to a tsunami generated by an earthquake of a particular magnitude in each source block. This simulation generates a response relationship between the coastal inundation and an earthquake of a particular magnitude and location. Since the earthquake statistics is known for each block, by summing the probability of all earthquakes in the Pacific Rim, the probability of the inundation in a coastal area can be determined through the response relationship. Although the idea of the statistical methodology applied here is not new, this study is the first to apply it to study the probability of inundation caused by earthquake-generated distant tsunamis in the Pacific Basin. As a case study, the methodology is applied to predict the tsunami inundation risk in Hilo Bay in Hawaii. Since relatively more field data on tsunami inundation are available for Hilo Bay, this case study can help to evaluate the applicability of the methodology for predicting tsunami inundation risk in the Pacific Basin. Detailed results will be presented at the AGU meeting.
OS22B-07 11:50h
Progress Towards a Probabilistic Tsunami Hazard map for New Zealand
New Zealand sits on the Pacific `ring-of-fire' and has a long shoreline exposed to the open ocean. As such it is at a moderate to high risk from tsunami, a conclusion which is confirmed by records of historical tsunami and geological evidence of paleo-tsunami. In recent years there has been a considerable impetus for development of coastal areas. For these reasons maps of tsunami inundation hazard, akin to maps of seismic hazard, would be very useful for planning purposes. However the task of producing such maps is made difficult by the diversity of potential tsunami sources and the short span of historical records in New Zealand. As a first step we have decided to concentrate on the most frequently recurring source of hazardous tsunami for New Zealand, namely large subduction-zone earthquakes along the South-American coast. In this presentation we demonstrate our initial attempts to model the spatial and seismic properties of the South-American subduction zone, and our method for relating simulated earthquake events to tsunami hazard at the New Zealand coastline using the MOST tsunami propagation model [1]. Results from this initial modelling will be described and compared against qualitative expectations of hazard based on historical events and the morphology of the seabed. The techniques for producing tsunami hazard maps are at an early stage of development and we will discuss the limitations of our work so far and our plans for future improvements. [1] Titov, V. V. and Gonzalez, F. L. (1997). Implementation and testing of Method of Splitting Tsunami (MOST) model. NOAA Technical Memorandum ERL PMEL-112.
OS22B-08 12:05h
Quantifying tsunami risk at the Pisco, Peru LNG terminal project
We examine and quantify the tsunami risk near Pisco, Peru, where a major Liquefied Natural Gas facility is in project at Playa Loberia. We re-assess the historical record of tsunami damage along the coast of Central and Southern Peru, from 9 deg. S (Chimbote) to 19 deg. S (Arica), building seismic models of the events involved, and conducting numerical simulations of the run-up at Pisco that such models predict. We then evaluate possible return periods for the main seismic events under consideration, from a combination of historical datasets and plate tectonics arguments. We classify tsunami hazard according to the amplitude of their run-up on the coast: decimetric tsunamis (0.1 to 1 m) do not carry a specific hazard over and beyond that presented by storm waves. Metric tsunamis (a few meters) can inflict severe damage to coastal and harbor communities, and result in inundation distances of up to 1 or 2 km. Finally, dekametric tsunamis (10 m and above) are catastrophic events leading to the total destruction. We estimate that a scenario of metric run-up, which could substantially damage port facilities and lead to a number of fatalities, may have a repeat time at Pisco of about 60 years. A catastrophic tsunami of dekametric amplitude, capable of totally destroying harbor infrastructures, may have a repeat time of about 110 years. This result is also consistent with the "back-of-the-envelope" observation that the city was destroyed four times over the past 400 years. The last such tsunami took place 136 years ago.