CHAPTER TWO
Aligning Priorities with Societal Risks from Tsunamis

SUMMARY

This chapter reviews progress toward understanding the nation’s tsunami risk, which is the first step in a comprehensive tsunami program. The knowledge of the hazards tsunamis pose is evolving. The nation is just beginning to understand the populations and social assets that tsunamis threaten, the readiness of individuals and communities to evacuate, and the losses of life and property they may cause. Although much has been learned about the nation’s tsunami risk, the nation remains far from understanding enough of its tsunami risk to set priorities and allocate resources for tsunami mitigation efforts based on risk.

The chapter evaluates progress and identifies opportunities for the National Oceanic and Atmospheric Administration (NOAA) and the National Tsunami Hazard Mitigation Program (NTHMP) to advance the goal of hazard and evacuation mapping and a comprehensive national risk assessment. Among other things, hazard modeling serves as the basis to produce evacuation maps, which are a critical tool in educating and preparing the public. The committee concludes that it is unclear whether current evacuation maps are sufficient for enabling effective evacuations or preparing the public due to the absence of uniform quality standards, evaluative metrics, or guidelines on what constitutes effective mapping approaches.

The advice, directed mainly at NTHMP partners and listed here in summary form, includes:

  • Completion of an initial, national tsunami risk assessment in the near term to inform program prioritization.

  • Periodic reappraisal of tsunami sources and modeling codes, achieved in part through workshops and peer review.

  • Greater consistency, across state boundaries, in the methods, criteria, and judgments employed in modeling of tsunami inundation, achieved in part through collaboration among federal and state partners and through external review of inundation maps.

  • Evaluations of the effectiveness of hazard maps, leading to standards that increase the overall quality and consistency of these maps.

  • Recurring inventories of the number and kinds of people in tsunami hazard zones, with special attention to high-risk groups including children, the infirm, and tourists.

  • Identification of areas where successful evacuation from a tsunami would require buildings or engineered berms.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 37
CHAPTER TWO Aligning Priorities with Societal Risks from Tsunamis SUMMARY This chapter reviews progress toward understanding the nation’s tsunami risk, which is the first step in a comprehensive tsunami program. The knowledge of the hazards tsunamis pose is evolving. The nation is just beginning to understand the populations and social assets that tsunamis threaten, the readiness of individuals and communities to evacuate, and the losses of life and property they may cause. Although much has been learned about the nation’s tsunami risk, the nation remains far from understanding enough of its tsunami risk to set priorities and allocate resources for tsunami mitigation efforts based on risk. The chapter evaluates progress and identifies opportunities for the National Oceanic and Atmospheric Administration (NOAA) and the National Tsunami Hazard Mitigation Pro- gram (NTHMP) to advance the goal of hazard and evacuation mapping and a comprehensive national risk assessment. Among other things, hazard modeling serves as the basis to produce evacuation maps, which are a critical tool in educating and preparing the public. The commit- tee concludes that it is unclear whether current evacuation maps are sufficient for enabling effective evacuations or preparing the public due to the absence of uniform quality standards, evaluative metrics, or guidelines on what constitutes effective mapping approaches. The advice, directed mainly at NTHMP partners and listed here in summary form, includes: • Completion of an initial, national tsunami risk assessment in the near term to inform program prioritization. • Periodic reappraisal of tsunami sources and modeling codes, achieved in part through workshops and peer review. • Greater consistency, across state boundaries, in the methods, criteria, and judgments employed in modeling of tsunami inundation, achieved in part through collaboration among federal and state partners and through external review of inundation maps. • Evaluations of the effectiveness of hazard maps, leading to standards that increase the overall quality and consistency of these maps. • Recurring inventories of the number and kinds of people in tsunami hazard zones, with special attention to high-risk groups including children, the infirm, and tourists. • Identification of areas where successful evacuation from a tsunami would require buildings or engineered berms. 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S INTRODUCTION Understanding risk is a prerequisite to effectively preparing and warning endangered people of potential and imminent tsunamis. To that effect, officials must understand in advance (1) where and to what extent tsunami inundation may occur and the amount of time before waves arrive, (2) the characteristics of the population and communities in tsunami-prone areas, and (3) how prepared individuals and communities are for evacuation. Risk, as used in this report, is a concept used to give meaning to things, forces, or circum- stances that pose danger to people or what they value (see also Box 2.1). It takes into account BOX 2.1 Definitions Risk is a concept used to give meaning to things, forces, or circumstances that pose danger to people or what they value. Risk descriptions are typically stated in terms of the likelihood of harm or loss of a vulnerable thing or process (e.g., health of human beings or an ecosystem, personal property, quality of life, ability to carry on an economic activity) due to a physical event (i.e., hazard) (National Research Council, 1996a). Some researchers have used the term risk to quantify the likelihood of future tsunamis, while others have defined it as a product of the probability of tsunami-attributable social damage (e.g., buildings, lives, businesses) and the magnitude of that damage. Research shows that managers, policy makers, and members of the public rarely define risk as an objective calcula- tion; instead, perceptions vary according to differences in awareness, experiences, and social context (Fischhoff et al., 1984; Weichselgartner, 2001). Hazard is the physical characteristics of an event (e.g., tsunami: speed of onset, impact forces, currents, inundation area) that can pose a threat to people and the things they value. Vulnerability is the personal or situational conditions that increase the susceptibility of people or resources to harm from the hazard. Inundation refers to the process of coastal flooding due to tsunamis or storm waves regardless of the impact to human activities. Run-up height is the vertical elevation of the most landward penetration of the tsunami wave with respect to the initial sea level (figure opposite page). Run-up is a vertical distance, while inunda- tion is a horizontal distance. Inundation models determine the areas likely to be flooded by a tsunami and involve numerical computations of tsunami evolution for specific tsunami scenario or consider an ensemble of tsunami scenarios that might affect the map area. Hazard maps depict inundation areas on base maps that typically include contours, imagery, buildings, roads, and/or critical infrastructure and take into account local geologic knowledge. Evacuation maps depict areas that need to be evacuated in the event of a tsunami and to show evacuation routes to safe havens. Evacuation maps are based on the same inundation zones in hazard 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis the likelihood and physical characteristics of tsunamis (i.e., the hazard), the personal or situ- ational conditions that increase the susceptibility of people or resources to harm from the tsunami (i.e., vulnerability), and associated uncertainties. This chapter first introduces tsunami risk assessment by identifying roles it may play in reducing the losses of life and property to tsunamis and by summarizing broad approaches to risk assessment. The rest of the chapter reviews progress in assessing tsunami hazard and tsunami vulnerability. Aspects of perceptions, knowledge, and preparedness levels that influ- ence individual resilience are discussed in the following chapter. TSUNAMI FORM IN THE OPEN OCEAN NEARSHORE TSUNAMI FORM dI H R CONTINENTAL INITIAL SHELF SHORELINE CONTINENTAL SLOPE ABYSSAL PLAIN H = tsunami wave height R = run-up height d I = inundation depth The depth in the continental shelf is typically less than 200 m. The average depth in the abyssal plain in the Pacific is about 4,000 m. The definition of tsunami run-up height. SOURCE: Committee member. Figure (Box 2.1).eps maps but these zones are typically enlarged to allow easier identification of landmarks, such as major intersections or highways. Evacuation modeling refers to the process of simulating how people evacuate any given area and is used to quantitatively evaluate whether high ground can be reached in time. 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S COMPREHENSIVE RISK ASSESSMENT Tsunami risk assessment is fundamental to the nation’s tsunami programs, because it can help support and guide risk reduction efforts, including tsunami education, preparedness plan- ning, and warning system development. In particular, it can help allocate resources according to greatest risk and thus ensure that efforts are prioritized to protect the most people. The following three examples illustrate the benefits of such an assessment: • Priorities among states. A comprehensive tsunami risk assessment could improve resource allocations among states, territories, and commonwealths. In 2006, the Gov- ernment Accountability Office (GAO) observed that “with the likely expansion of the NTHMP from 5 state participants to potentially 28 state and territorial participants in 2006, it will be difficult for NOAA to ensure that the most threatened states receive the resources they need to continue and to complete key mitigation activities without an updated, risk-based strategic plan” (Government Accountability Office, 2006). To this date, NTHMP resources are being allocated on the basis of estimated hazards, without regard for vulnerability (National Tsunami Hazard Mitigation Program, 2009a). • Priorities among program elements. A comprehensive tsunami risk assessment coupled with cost-benefit analyses could also help NOAA weigh the value of warning and forecasting on the one hand against that of education and community preparedness on the other. The GAO’s 2006 report noted that NOAA’s initial strengthening efforts emphasize detection and warning for distant tsunamis, while the greater risk to most locations in the United States— according to NOAA data as well as the National Science and Technology Council’s (NSTC) December 2005 report on tsunami risk reduction—[is] likely to be posed by local tsunamis. For example, the deployment of Deep-ocean Assessment and Reporting of Tsunamis (DART) stations and warning center en- hancements will not reduce the local tsunami risk as directly as other strategies such as educating vulnerable populations to immediately head for high ground when the earth shakes near the coast (Government Accountability Office, 2006). • Program development. Understanding tsunami risk helps averts surprises. In the 1980s, the Cascadia subduction zone became known as a source of catastrophic tsunamis, first from geophysical clues that it might produce such waves and then through geo- logical signs that it had (Atwater et al., 2005). These discoveries about the tsunami haz- ard were the first indication of the need for Washington, Oregon, and California to take steps to prepare to build tsunami-resilient communities before the next great Casca- dia tsunami strikes. These discoveries also helped create the NTHMP itself. When the NTHMP originated in the mid-1990s, it was founded in part on then-recent discoveries about tsunami risk. Oregon’s concern about Cascadia tsunami hazards played a central role in the NTHMP’s establishment, according to committee members’ interviews with founding members of the NTHMP steering group. 0

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis Tsunami risk assessments are challenging due to (1) the paucity of information about the frequency, sources, and characteristics of past tsunamis, (2) the complex and interdependent nature of coastal communities in larger economic or sociopolitical systems, (3) the poorly defined mix of near-field and far-field tsunami hazards that coastal communities face, and (4) the uncertainty of potential impacts of future events. Although societal risk from tsunamis is challenging to assess, it is critical information for the development and prioritization or risk- reduction efforts, such as education, preparedness planning, warning-system development, mitigation planning, and response strategies. An effective risk characterization should be undertaken with local decision makers in mind and should be directed toward informing specific choices of public officials and affected individuals (National Research Council, 1996a). When developing the analysis for decision tools (e.g., inundation maps, vulnerability assessments, evacuation maps), scientists need to en- gage and deliberate with decision makers to assess their information needs and create useful decision tools (e.g., Fischhoff et al., 1978; National Research Council, 1996a, 2007; Gregory and Wellman, 2001; Bostrom et al., 2008; Renn, 2008). Risk analysis involves the systematic collection and interpretation of quantitative or qualitative data to better understand hazards and vulner- able communities (National Research Council, 1996a). Risk deliberation is any communicative process in which technical experts, public officials, and affected parties collectively consider risk issues to ensure that decision-relevant knowledge and diverse perspectives are included in any risk reduction process (National Research Council, 1996a). There is no single method to properly characterize all aspects of risk from tsunamis; differ- ent techniques are needed to address different aspects (e.g., demographic sensitivity, structural fragility, financial exposure) and their potential risk-reducing adjustments (e.g., education pro- grams, structural mitigation, insurance). However, best practices suggest including assessments of possible exposure sources and pathways (including geospatially specific susceptibilities to tsunamis), potential consequences, the effects of feasible risk reduction options, and the prob- abilities and uncertainties of exposures and ensuing effects. Probabilistic assessment can be a tool to provide a basis for cost-benefit analysis and design considerations for tsunami mitiga- tion efforts (e.g., design criteria of an evacuation structure) and a transparent basis for prioritiz- ing resources. Geographic information system (GIS)-based analyses of socioeconomic exposure to tsunamis are useful for identifying demographic sensitivities within a community that could impact evacuations. Assessments of adaptive capacity and resilience can identify educational needs and pre-event preparedness levels. The level of sophistication, accuracy, resolution, and format required for assessing societal risk to tsunamis will depend on the intended use of the information. For example, education efforts designed to raise hazard awareness, such as teaching people to recognize natural cues, may only require oral histories from tsunami survivors (Dudley, 1999) or coarse delineations of run-ups from past events (Theilen-Willige, 2006) to indicate tsunami-prone areas (see Box 2.1 for definitions). By contrast, evacuation planning usually requires computer modeling of various tsunami evacuation scenarios (Katada et al., 2006; Yeh et al., 2009). Urban planning or detailed assessments of economic impacts require even higher resolution calculations (Borrero et al., 2005). 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S So far, no comprehensive risk assessments have been undertaken that could guide the set- ting of priorities at the state or national level. Existing risk assessments are uneven and are typi- cally isolated efforts for specific sites (e.g., González et al., 2006). The only national assessment to date is an evaluation of tsunami hazards, based on the written historical records available at the NOAA/National Geophysical Data Center (NGDC) (Dunbar and Weaver, 2008) with the rationale being that “the state of geologic knowledge [of tsunami sources] does not permit the calculation of meaningful probabilities of occurrence.” Although hazard researchers may be uncomfortable with the lack of information about occurrence, probabilistic risk assessment can be an indicator for the relevance and import of existing evidence, despite the uncertainties. Probabilistic risk assessment for nuclear power plants and for earthquakes show how risk assessments can be constructed for rare, high- consequence events in order to inform planning efforts and designs (e.g., McGuire and Becker, 2004; McGuire, 2008). Where data are sparse, expert judgment can be used to qualify the avail- able data appropriately. For example, tsunami planning has been so far based on worst-case scenarios, with inconsistent choices made about specific scenarios and little understanding of consequences or expected losses. Where extremely rare, high-consequence events are the worst case, communities that are much more likely to see smaller yet damaging tsunamis may not be prioritized for funding. Such tradeoffs deserve to be explicitly considered and can be incorporated into quantitative and qualitative risk assessments using expert judgment in a deliberative process (National Research Council, 1996a). Conclusion: The United States lacks a national tsunami risk assessment that characterizes the hazards posed by tsunamis, inventories the populations and social assets threatened by tsunamis, measures the preparedness and ability of individuals and communities for successful tsunami evacuations, and forecasts expected losses. This information is needed to help spur and prioritize investments in preparedness, education, detection, and warning efforts and for developing long-term strategic planning at the local, state, and federal level. Recommendation: NOAA and its NTHMP partners, in collaboration with researchers in social and physical sciences, should complete an initial national assessment of tsunami risk in the near term to guide prioritization of program elements. The national tsunami risk assessment should (1) incorporate the best possible relevant science (social and behavioral science, geography, economics, engineering, oceanography, and geophysics) and (2) include broad stakeholder and scientific participation to ensure that efforts are responsive to the needs of at-risk communities and decision makers. 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis TSUNAMI HAZARD ASSESSMENT Tsunami hazard assessments focus on characterizing and visualizing the physical charac- teristics of future tsunamis (e.g., speed of onset, impact forces, currents, inundation area) that can pose a threat to people and the things they value. Understanding where tsunami inunda- tion is likely and how much time at-risk individuals have to evacuate frame the discussion of societal risk to tsunamis and are the foundations upon which education, preparedness plans, response plans, and mitigation strategies are developed. Tsunami hazard assessments typi- cally entail three elements: (1) inundation models to determine the areas likely to be flooded, (2) hazard maps that portray inundation-model outputs on community base maps (e.g., roads, elevation, structures), and (3) evacuation maps that depict areas that may need to be evacu- ated in the event of tsunamis. The purpose of this section is to describe each of these elements, progress in the development and implementation of each element, and areas for improvement. Before discussing each individually, it is important to distinguish the differences between the three elements. For example, although hazard and evacuation maps are both used for re- ducing tsunami risk, the two types of tsunami maps are developed in different ways and for dif- ferent purposes (Figure 2.1). Inundation model outputs simply denote the physical character- istics of tsunami generation, propagation, and inundation areas and do not recognize political boundaries. Tsunami hazard maps portray inundation model outputs with some modification to reflect local knowledge of land conditions, are organized by communities, and include basic societal assets (e.g., roads, major structures). Tsunami evacuation maps incorporate the same inundation areas as hazard maps, but typically denote larger zones to accommodate local risk tolerance and to allow for easier identification of landmarks (e.g., at-risk individuals can identify major roads easier than a specific elevation contour). Evacuation maps also differ from hazard maps in that they are educational tools designed to be easily understood by non-scientists and typically identify evacuation shelters and assembly sites (both natural and manmade), sug- gested routes, locations of warning loudspeakers, and transportation infrastructure to facilitate evacuations (e.g., streets, bridges). Inundation models and hazard maps are developed by sci- entists to identify and communicate maximum inundation areas and flow conditions, whereas evacuation maps are designed by scientists in collaboration with local, state, and federal emer- gency management and public safety agencies to identify areas of public safety concern. Inundation Modeling Credible tsunami inundation modeling requires three elements: (1) an understanding of the tsunami source that generates the tsunami, usually through estimation of seafloor displacement; (2) accurate and precise bathymetric and topographic data to understand the surface over which the waves propagate; and (3) a robust hydrodynamic computational model to simulate tsunami evolution. Each of these three elements are treated below in turn and offer two sets of conclusions and recommendations—one on ways of reducing uncertainties about tsunami sources and the other on hydrodynamic modeling. 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S (A) (B) Figure 2.1a.eps bitmap FIGURE 2.1 Tsunami maps for Cannon Beach, Oregon, including (A) a tsunami hazard map developed for a Cascadia subduction zone earthquake scenario and (B) an evacuation map that includes evacuation zones for a Cascadia-related tsunami (in yellow) and a far-field tsunami (in orange). SOURCE: http:www. oregongeology.org/sub/default.htm; image courtesy of DOGAMI. Tsunami Sources The societal value of inundation models depends largely on estimates of tsunami sources (Synolakis et al., 1997). If a modeler underestimates a tsunami source, a real tsunami may inundate places the modeling had deemed safe and lives could be unnecessarily lost as a result. If instead the modeler overestimates the tsunami source, risk reduction efforts may be cost-prohibitive and more people and businesses may be unnecessarily affected in future evacuations. Large uncertainties remain concerning the sources of tsunamis that could inundate U.S. shorelines. These sources include subduction zones of the Pacific Rim and Caribbean, under- water landslides off the Atlantic and Gulf Coasts and off southern California, and volcanoes in 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis Alaska (Figure 2.2, Appendix A). The next six paragraphs illustrate unknowns that are spelled out more fully as questions in Appendix A. Far-field heights of tsunamis from circum-Pacific sources. For hazard and evacuation mapping, simulations of far-field tsunami heights on U.S. Pacific shores require estimates about tsunami sources at subduction zones on the Ring of Fire (Figure 2.2). The simulations are sensitive to earthquake size because far-field tsunami height increases substantially with earthquake magnitude (Abe, 1979). For example, to make worst-case simulations for Pearl Harbor, a NOAA group used 18 tsunami sources, together spanning all subduction zones on the Pacific Rim ( Tang et al., 2006). For each source, the group used an earthquake of magnitude 9.3 (McCaffrey, 2008). For some of the subduction zones, the assumed magnitude 9.3 differs from the largest known earthquake magnitude of the past, which falls short of 8.0 or 8.5 (blue or brown, respec- tively, in Figure 2.2). This difference in assumed source magnitude results in the correspond- ing worst-case tsunami on distant U.S. shores to be larger by a factor of about 10 from the tsunamis of the past. Far-field hazards from Alaskan sources. Earthquake-generated tsunamis from sources along the Aleutian-Alaskan subduction zone pose far-field tsunami hazards in Washington, Oregon, California, and Hawaii. If a hazard map is to give each of these sources appropriate weight, it is necessary to estimate how large the earthquakes can get and how often they happen (Geist and Parsons, 2006; González et al., 2006). For tsunamis from most of the Aleutian-Alaskan subduction zone, currently there is almost no basis for such weighting except for geophysical estimates of how much plate motion gets spent on earthquakes of various sizes (González et al., 2006; Wesson et al., 2008). Little is known about how often the tsunamis actually recur ex- cept for the source of the oceanwide 1964 Alaskan tsunami (Carver and Plafker, 2008). Even for the 1964 source, it is unclear whether the next large tsunami is expected to recur sooner than average because the 1964 earthquake ended a recurrence interval, which was 300 years longer than the previous 600-year average. Near-field hazards from Cascadia tsunamis. Although documented geological histories of great Cascadia earthquakes extend thousands of years into the past, persistent uncertainties about them have yielded wide-ranging estimates of tsunami hazards. A probabilistic analysis of tsunami hazards in Seaside, Oregon, showed the hazard to be sensitive to variability in earth- quake size and recurrence (González et al., 2006). A subsequent hazard map for nearby Cannon Beach doubled the maximum tsunami height relative to previous estimates (Priest et al., 2009). Caribbean hazards. Islands of the Caribbean are threatened by tsunamis mainly from local tectonic sources and the subsequent potential for submarine landslide. One tsunami source, near Puerto Rico, may even threaten coastlines as far away as Massachusetts. Local tsunami sources caused loss of life in the Virgin Islands in 1867 and in western Puerto Rico in 1918 (O’Loughlin and Lander, 2003). The 1867 tsunami began during an earthquake (Reid and Taber, 1920; McCann, 1985), as did the 1918 tsunami (Reid and Taber, 1919). However, the 1918 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S FIGURE 2.2 Illustrations of various tsunami sources. (A) Global sketch of subduction zones and landslides known or inferred to pose tsunami hazards. The blue, brown, and yellow lines do not necessarily represent Figure 2.2a,b,c.eps maximum earthquake size, as discussed in the text and further illustrated in Appendix A. The depiction of bitmap landslides emphasizes those regarded as posing a tsunami threat to the United States and its territories. SOURCE: Committee member. (B) The offshore area of Puerto Rico. SOURCE: Image courtesy of Uri ten Brink. (C) The offshore area near Los Angeles. SOURCE: Normark et al., 2004; with permission from Elsevier. 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis tsunami may owe most of its size to a submarine landslide in the earthquake’s focal region (López-Venegas et al., 2008). The 1867 and 1918 disasters are probably just the tip of an iceberg; they represent but a sample, during the geological instant of the past 150 years, of Caribbean tsunami sources that can be inferred from the region’s active tectonics (McCann, 1985; Grindlay et al., 2005; Mercado-Irizarry and Liu, 2006) and from its abundance of steep submarine slopes (ten Brink et al., 2004, 2006). Probably the biggest open question about these many Caribbean sources is the tsunami potential of the highly oblique subduction zone marked by the Puerto Rico Trench. This hypothetical tsunami source faces the low-lying metropolis of San Juan (pop. 0.5 million) and, farther afield, may threaten the Atlantic seaboard from the Carolinas to Massachusetts (Geist and Parsons, 2009). In addition to having all these tsunami sources of its own, the Caribbean bore the brunt of the documented far-field effects of the 1755 Lisbon tsunami (Barkan et al., 2009; Muir-Wood and Mignan, 2009). The tsunami did not appear to have a significant effect on San Juan, based on the absence of documentation in the extensive Spanish-language records from that part of the 18th century (McCann et al., undated). Near-field hazards from slides off U.S. coasts. Submarine slides abound off the Atlantic coast, par- ticularly in the Caribbean (above) and off New England and the Middle Atlantic states (Twichell et al., 2009). Submarine slides are also present beneath the Gulf of Mexico (Trabant et al., 2001) and off southern California (Lee, 2009). The probabilistic tsunami hazard the slides pose is poorly known. It may be low because most of the sliding appears to have occurred soon after the last glaciation, at a time when sediment supply and sea levels greatly differed from today’s (Lee, 2009). Much remains to be learned about slide size, speed, and duration (Locat et al., 2009), all of which affect a slide’s efficiency in setting off a tsunami (Geist et al., 2009). Tsunami sources that have escaped notice. That such sources remain undiscovered can be inferred from the recent identification of tsunami threats that had previously gone unrecognized—from great earthquakes on the Cascadia subduction zone (Atwater et al., 2005), faults and landslides beneath Puget Sound (Bucknam et al., 1992), outsize subduction earthquakes off northeast Japan (Nanayama et al., 2003), and landslides off Norway (Haflidason et al., 2005), Puerto Rico (ten Brink et al., 2006), the U.S. Atlantic coast (ten Brink, 2009), and southern California (Lee, 2009). Determining worst-case source scenarios. Decisions about worst-case tsunami sources for the purpose of inundation modeling (see below) vary among NTHMP members. Inundation model- ing in Alaska uses historical events (e.g., the 1964 Great Alaskan Tsunami) as well as a set of hypothetical tsunami scenarios unique for each local community for the tsunami sources. Inundation modeling in Hawaii is also based on historical distant tsunamis (1946 Aleutian, 1952 Kamchatka, 1957 Alaska, 1960 Chile, and 1964 Alaska tsunamis). Inundation modeling in California is based on 6 to 15 local and distant sources (depending on map location) that result in a single maximum tsunami inundation scenario. The primary subduction-type fault threat for northern California is the Cascadia scenario, and other potentially important tsunami sources include distant tsunamis (e.g., earthquakes near Alaska or Japan) and submarine landslides. 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S earthquake faults in Puget Sound (e.g., the Seattle Fault, Tacoma Fault). California has recently completed new tsunami inundation mapping and is currently producing tsunami evacuation maps for all populated areas of the coastline. In other states and territories (Alaska, Puerto Rico, Guam), tsunami evacuation maps are available only for certain communities, and few are avail- able online. No tsunami evacuation maps currently exist for communities in the Gulf of Mexico or the eastern U.S. coastline. As outlined in its strategic plan (National Tsunami Hazard Mitiga- tion Program, 2009a), the NTHMP plans to inventory all tsunami evacuation maps by 2010 and to thereafter increase the number of maps annually by 10 percent. The majority of evacuation maps across the United States are available as digital PDFs or printed brochures (Table 2.4). Public access to digital copies of evacuation maps is not consis- tent or intuitive across the states. For example, local evacuation maps are available online at the state geology department in Oregon and at a university-based seismic network organization in Puerto Rico, while in most other states, they are provided by the state emergency management or civil defense agency (Table 2.4). Hawaii was the first state to create a dynamic online map- TABLE 2.4 Availability and Format of Tsunami Evacuation Maps in Several U.S. States and Commonwealths State/ Online Location for Commonwealth Tsunami Evacuation Maps Format of Available Maps • laska Department of Natural Resources • azard maps for selected Alaska A H communities (pdf ) • alifornia Emergency Management • rochures (available online as PDF) California C B • nline mapping application— Agency O “My Hazards” • uam Office of Civil Defense • ext-based descriptions of Guam G T evacuation plans • awaii State Civil Defense • Stand-alone evacuation maps (PDF) Hawaii H • nline mapping application O (Google-enabled) • regon Department of Geology and • rochures (available online as PDF) Oregon O B Mineral Industries • regon Coastal Atlas O • uerto Rico Seismic Network, based at the • rochures (available online as PDF) Puerto Rico P B University of Puerto Rico-Mayaguez • ashington State Emergency • rochure (available online as PDF) Washington W B Management Division • ashington Dept. of Natural Resources W SOURCE: Committee member. 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis ping application (based on Google Maps) that allows individuals to find specific addresses and see tsunami evacuation zones relative to more identifiable landmarks (Appendix D). Because it is critical for individuals to understand the extent of tsunami evacuation zones in their commu- nities, the committee commends the NTHMP for acknowledging a deficiency in map availabil- ity and citing the need for guidelines for the approval and distribution of maps in its 2009-2013 draft strategic plan. Efforts to centralize the availability of evacuation maps, or at the very least to develop a NTHMP portal that guides individuals to disparate state archives, will enable indi- viduals to more easily find evacuation materials and prepare for future events. In addition to inconsistencies in map access, tsunami evacuation maps are inconsistent with regard to formats, colors, and noted landmarks (e.g., bridges, assembly areas, hospitals, hotels to help tourists, instructions on whether to evacuate by foot or car). The formats of currently avail- able tsunami evacuation maps vary and include community-based maps that emphasize land- marks (see Appendix D; Alaska), single tsunami evacuation zones (see Appendix D; Washington, California, Oregon, Puerto Rico), and multiple tsunami evacuation zones to differentiate between local and distant tsunamis (see Appendix D; Cannon Beach, Oregon; Figure 2.1). Recent updates to the Cannon Beach map that show two tsunami evacuation zones (one for a distant tsunami and another for a near-field event; see Figure 2.2) may be the most scientifically justified, but no studies have been conducted to determine whether at-risk individuals in this community under- stand these differences and would know whether to evacuate certain areas given live drills of the two scenarios. In general, there is no rigorous evaluation of how people respond to or interpret maps. Evacuation maps in Oregon, Washington, California, and Puerto Rico use yellow to denote tsunami evacuation zones, while in Hawaii the static maps found online denote the hazard zones as gray and the maps in phone books in red. Variations also exist among tsunami evacuation maps with regard to accompanying text on the map products that explains how to use the map and to prepare for future tsunamis (Table 2.5). To date, state mapping efforts have largely relied on their own state advisory groups to guide evacuation map development. Because tsunami scenarios that form the basis for evacu- ation maps will vary among states (e.g., distant versus local events or worst-case versus most likely), map content will always need to be tailored to the special facilities and populations in tsunami-prone areas. However, the preparation and presentation of this information (e.g., symbols, colors) in a consistent way across the United States helps create a consistent voice in public education and is encouraged (see Chapter 3 for additional details). For example, a resident of the Oregon coast who works in a neighboring coastal town in Washington or is vacationing in Hawaii should be able to recognize and understand tsunami evacuation maps with little interpretation. State agencies have not evaluated the effectiveness of various map formats (either based on surveys and interviews or by testing their utility during evacuation drills) in promoting individuals to take protective action, and there are no NTHMP guidelines for evacuation map preparation. Although consistency in evacuation map preparation does not currently exist, the commit- tee commends the NTHMP for noting the need for guidelines on evacuation map preparation (including assistance to non-English speaking communities and criteria for defining evacuation routes and sites) in its 2009-2013 strategic plan (National Tsunami Hazard Mitigation Program, 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S TABLE 2.5 Availability of Explanatory Text with Tsunami Evacuation Maps in Various NTHMP States and Commonwealths Hawaiia On-line Text Included in Static Mapping Puerto Mapb Ricob Tsunami Evacuation Maps Application Oregon California Washington Tsunami explanation X X X X Instructions to evacuate X X X X tsunami zone Distant vs. local X X X X tsunamis distinction Risk-reduction strategies X X X X (e.g., family plan, emergency kit) Natural cues described X X X X Evacuation signage X X X described Vehicle- vs. foot-based X X X X evacuation discussed Where to evacuate to X X X X (e.g., assembly areas) 1 Directions for more X X X information Checklist for immediate X X X X action after warning Post-tsunami X X X actions to take Mention of special- X X needs populations a Three of the four links for more information at county civil defense departments did not work. The fourth link (to Hawaii County Civil Defense) was active, but led to the static PDFs of evacuation maps for the island. b For the static evacuation maps in Hawaii and in Puerto Rico, no accompanying documentation was included in the map or could be found. SOURCE: Committee member. 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis 2009a). The need for best practices or standards for evacuation map preparation was also noted in the 2007 NTHMP five-year review. These guidelines or best practices need to be based on an evaluation of how people process the maps and what conventions are most effective. Conclusion: Evacuation maps are critical tools for understanding and communicating population vulnerability to tsunamis. Most at-risk communities in the Pacific states (Hawaii, Alaska, Washington, Oregon, and California) and Puerto Rico have produced evacuation maps with a single line indicating a worst-case scenario, except for Cannon Beach, Oregon, where the map includes evacuations zones for far- and near-field tsunamis. Approaches to evacuation map production vary greatly among NTHMP members (e.g., format, choice of scenarios); therefore, at-risk populations are expected to interpret different state- developed representations of tsunami risk. It is unclear whether current evacuation maps are sufficient for enabling effective evacuations or preparing the public due to the absence of uniform quality standards, evaluative metrics, or guidelines on what constitutes effective mapping approaches. Recommendation: The NTHMP Mapping and Modeling Subcommittee should develop guidelines on evacuation map production that foster consistency in format and quality across the nation and that are based on sound cartographic principles, although map content must be tailored to the relevant facilities, populations, and characteristics of the local communities. To improve public access to evacuation maps, the NTHMP should develop a national, online repository for tsunami evacuation maps and host a consistent online mapping application for all tsunami evacuation zones across the United States. The NTHMP should annually update the inventory of evacuation maps relative to the number of at-risk communities. VULNERABILITY ASSESSMENTS Societal vulnerability to tsunamis refers to the physical, social, economic, and environ- mental conditions or processes that increase the potential for individuals or communities to incur losses or damages from future tsunamis (International Strategy for Disaster Reduction, 2004). Common elements of vulnerability within the natural hazard literature include exposure, sensitivity, and resilience (Dow, 1992; Hewitt, 1997; Cutter, 2003; Turner et al., 2003). Exposure refers to hazard proximity, sensitivity refers to differential degrees of potential harm given simi- lar exposure (e.g., different building types), and resilience addresses the coping and adaptive capacities of an individual or community during and after an extreme event. Understanding societal vulnerability to tsunamis provides emergency managers with the required information to protect their communities and to determine whether individuals have the capacity to take protective actions. Although this information is considered critical to reducing tsunami risk, until recently rela- tively little has been written about societal vulnerability to tsunamis compared to the amount 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S of literature devoted to the physical characteristics of tsunamis (Keating, 2006; National Research Council, 2007). One possible reason is that vulnerability is not an integral part of the NTHMP and is not addressed by its subcommittees (Warning Coordination, Mapping and Modeling, Mitigation and Education). The NTHMP Mitigation and Education Subcommittee notes the importance of understanding vulnerability to tsunamis in the development of miti- gation and education strategies but has not dedicated resources to doing such assessments or providing guidelines on how to do so. Efforts to describe societal vulnerability to tsunamis have come from other federal agencies (e.g., the USGS) and academic institutions. As the tsunami research community increases its efforts into assessments of population exposure and other elements of vulnerability assessment, it would benefit from leveraging ongoing efforts for other hazards in the various agencies. Individual and community vulnerability to tsunami hazards are dynamic processes that require monitoring due to changing coastal populations, risk perceptions, and use of tsunami- prone areas. Methods of characterizing vulnerability vary depending on the intended use of the results (e.g., evacuation planning, land-use planning, infrastructure siting, and mitigation projects). Due to the committee’s focus on national preparedness to tsunamis, we limit our discussion of vulnerability in this chapter to issues that relate to an individual’s ability to evacu- ate tsunami-prone areas, including (1) population exposure and sensitivity to tsunamis and (2) evacuation potential for at-risk individuals in tsunami-prone areas. The purpose of this sec- tion is to briefly describe each element, progress in the nation’s understanding of vulnerability, and areas for improvement. Population Exposure and Sensitivity Tsunamis pose risks only if they have the potential to impact humans or things they value. Therefore, a first step in understanding vulnerability is to inventory the number and types of individuals in tsunami hazard zones. Population exposure can be estimated for small geo- graphic areas (e.g., a single coastal community) via building inventories in tsunami-prone areas (Morgan, 1984; Papathoma et al., 2003; Wood and Good, 2004; Dall’Osso et al., 2006) or commu- nity workshops that leverage local knowledge (Wood et al., 2002). For large geographic areas (e.g., counties, states), decadal population data gathered by national census agencies (e.g., U.S. Census Bureau) or business databases gathered by private companies can be integrated with tsunami hazard data using GIS tools to determine the number of individuals in tsunami-prone areas. Other regional approaches to estimating population exposure to tsunamis include global population models (Balk et al., 2005), and landcover data (Wood, 2009). In addition to determining the number of individuals in tsunami-prone areas, it is impor- tant for emergency managers to assess their demographic characteristics, as these can amplify an individual’s potential for losses and affect their ability to receive and understand warning messages (Mileti and Sorenson, 1990; Miller et al., 1999; Morrow, 1999; Cutter, 2003). People in tsunami-prone areas will vary in their hazard awareness, risk perception and tolerance, and ability to prepare or respond to an extreme event. If officials are to effectively motivate people 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis to take protective measures in response to a tsunami warning, then they need to understand who they are trying to motivate and their capacity to respond. Assessing the types of people in tsunami-prone areas helps officials determine placement of warning signage and technology (e.g., sirens), tailor the format and delivery of education efforts to reach different populations (see Chapter 3 for additional discussion on education), and identify those who may need special assistance during an evacuation (e.g., elderly populations). Characterizing the demographic attributes of individuals in tsunami-prone areas does not imply that all individuals of a certain demographic group will exhibit identical behavior during or after a tsunami since multiple demographic characteristics of an individual or neighborhood interact and likely amplify each other. Variations in local cultures and situations, as well as in individual and community resilience, will influence the extent of these demographic sensitivi- ties. Individuals have multiple demographic characteristics (e.g., age, gender, economic status) and the interaction of factors may heighten or reduce their sensitivity to tsunamis (Wood et al., 2010). The importance of demographic sensitivities will also be influenced by characteristics of the hazard; for example, a high fraction of elderly people in a tsunami-prone area will be a larger issue if the likely warning time before inundation is 30 minutes (e.g., near-field tsunami) compared to several hours (e.g., far-field tsunami). With these caveats in mind, the following demographic groups may have higher sensitivity to tsunami hazards: • the very young and the very old (Balaban, 2006; McGuire et al., 2007); • households of racial and ethnic minorities becuase of historical societal inequalities (Laska and Morrow, 2006) and to potential exclusion from disaster preparedness efforts (Morrow, 1999); • renters, who are less likely than homeowners to prepare for catastrophic events and have more limited exposure to hazard information (Morrow, 1999; Burby et al., 2003); • individuals with pre-existing socioeconomic issues (e.g., homeless, living in poverty, low literacy levels, inability to speak the primary language of an area) that may inhibit their ability to prepare for future events (Wisner et al., 2004); • individuals at hospitals, psychiatric facilities, adult residential care centers, daycare centers, schools, and correctional facilities that may have difficulty (e.g., hospital patient) or be incapable (e.g., correctional-facility inmate) of evacuating on their own and will require external assistance to evacuate; and • employees or tourists who may have low or no exposure to awareness efforts or evacuation drills (Wood and Good, 2004; Johnston et al., 2007). The committee found that by 2010, reports attempting to inventory the number and types of people in tsunami hazard zones in coastal communities have been completed for the tsunami-prone areas of Hawaii (Wood et al., 2007), Oregon (Wood, 2007), and the open-ocean coast of Washington (Wood and Soulard, 2008). In each of these reports, census block data from the 2000 U.S. Census and national business data were merged with tsunami hazard data to identify, by community, the number and types of residents, employees, and facilities that attract tourists and house special-needs populations in tsunami-prone areas. As a first ap- 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S proximation of economic exposure, these reports also include inventories of the amount and percentage of tax-parcel values, employee distributions, and business sales volume in tsunami hazard zones of each community and county. In reviewing the tsunami and risk/vulnerability assessment sections of several FEMA-approved state mitigation plans, the committee found little information related to the diversity of populations in tsunami-prone areas of the various coastal states. Some state plans estimate the number of individuals in tsunami hazard zones of each county, some simply list the communities with populations with tsunami risk, and some only discuss general tsunami-related population issues. Many plans, however, did include detailed information on financial, structural, and critical facility exposure, suggesting that plans are written to help document potential long-term economic impacts and not preparedness issues to save lives. Conclusion: There is no national assessment of population exposure and sensitivity to tsunamis, including the number and types of individuals in tsunami hazard zones. The NTHMP’s current subcommittee structure (Mapping and Modeling, Warning Coordination, and Mitigation and Education) does not coordinate mapping the exposure of at-risk individuals and communities and therefore does not provide leadership on this topic. The absence of this information and the leadership to attain it impacts national preparedness to tsunamis in several ways. First, the NTHMP will be unable to reach its stated goal of a national tsunami risk assessment (see earlier conclusion in this chapter on this topic). Second, efforts to develop realistic evacuation plans are compromised in communities threatened by far-field tsunamis if the magnitude of the at-risk population is not known. Third, communities threatened by near-field tsunamis cannot develop public awareness and education efforts that are tailored to local conditions and needs (discussed in the following chapter) without an understanding of the types of at-risk individuals. Recommendation: The NTHMP should collaborate with state and federal agencies (e.g., the USGS, the Census Bureau) to periodically inventory the number and types of people in tsunami hazard zones at intervals no less frequently than that of the U.S. Census, with special attention to children, the infirm, tourists, and other groups whose heightened sensitivity to tsunamis could constrain their ability to prepare for and evacuate from future tsunamis. The NTHMP should expand Mapping and Modeling Subcommittee efforts to explicitly include community vulnerability. The NTHMP should establish a Science Advisory Committee to help develop guidelines on consistent approaches for identifying and mapping populations in tsunami-prone areas. The NTHMP should also provide guidelines on how to use this information to tailor evacuation planning and education efforts. Evacuation Potential In addition to population exposure and sensitivity, an individual’s capacity to learn from past disasters, implement risk reduction measures, adapt during an event, and persevere after an 

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis event is another significant factor in understanding societal vulnerability to tsunamis (National Science and Technology Council, 2005). These resilience factors influence the ability of an individual to prepare for future tsunamis and to take self-protective measures when a tsunami occurs. This sub-section on evacuation potential focuses on assessing the ability of individuals to evacuate tsunami-prone areas based on physical characteristics of the tsunami hazard zone (e.g., distance to higher ground, integrity of egress routes, island with no high ground), while the role of perceptions, knowledge, and preparedness levels in increasing resilience is discussed in the following chapter on education, preparedness, and evacuation coordination. Evacuation modeling estimates the amount of time necessary for people to reach safe havens from various locations in tsunami hazard zones (Post et al., 2008), which is especially critical for U.S. coastal communities that are threatened by tsunamis reaching shores in an hour or less. Such modeling efforts have been used to study the influence of congestion due to crowds and road bottlenecks on the ability of individuals to evacuate tsunami hazard zones (e.g., Lammel et al., 2008); the effectiveness of official routes in managing the typical number of people who will need to evacuate (Ismail et al., 2008); the likelihood of casualties (Koshimura et al., 2006); or the need for vertical evacuation structures (e.g., buildings, engineered berms) in places where time is not available to reach naturally occurring higher ground (Yeh et al., 2005; Federal Emergency Management Agency, 2008). Such analysis can answer fundamental questions that might ultimately determine the survival of people: Under what circumstances is it most effective to evacuate on foot instead of car due to congestion or earthquake-damaged infrastructure? Under what circumstances is it best to use vertical evacuation (e.g., tall build- ings, engineered berms) because ground-based evacuations may not be practical in communi- ties that must evacuate thousands of people from a large tsunami-prone area? Although there is basic research focused on tsunami evacuation modeling, the committee found little applied evacuation modeling research (e.g., Yeh et al., 2009) to examine specific U.S. coastal communities that may only have minutes to an hour to evacuate thousands of indi- viduals from tsunami-prone areas (e.g., Seaside, Oregon; Ocean Shores, Washington). In areas where ground-based evacuations may not be feasible (due to short times before inundation and substantial distances to higher ground) and where there are no existing structures or fea- tures capable of serving as a vertical refuge, workshops are being held in coastal communities (e.g., Cannon Beach, Oregon, in September, 2009; Long Beach, Washington, in January 2010) to further discuss the opportunities and constraints of vertical evacuation structures for tsunamis (e.g., buildings, engineered berms). Although engineering guidelines have been published and officials in some coastal communities are expressing interest in new structures (Federal Emer- gency Management Agency, 2008), the committee found no case studies that delve into the social and economic aspects of vertical evacuation structures in at-risk communities. In addi- tion, because there is very limited information about current velocities as the tsunami interacts with the built environment (and currently no measurements are being taken), it is difficult to estimate the forces involved with a tsunami flow field and to assess what structures might re- main intact during a tsunami. Before communities commit significant time and funds to these structures, communities need a careful determination of the feasibility and issues related to near-field tsunami evacuations. 

OCR for page 37
T S U N A M I WA R N I N G S A N D P R E PA R E D N E S S To date, no structures have been built in the United States specifically to serve as vertical evacuation sites for tsunamis. In communities with high-rise buildings in low-lying areas (e.g., Honolulu, Los Angeles), vertical evacuation into existing buildings to avoid inundation from far- field tsunamis is promoted by local emergency managers. In communities threatened by near- field tsunamis, officials should examine the structural integrity of buildings that may be used for vertical evacuation. Earthquakes that precede the tsunami may make certain building types (e.g., unreinforced masonry) unsafe for entry or for tsunami refuge and the subsequent tsunami waves may overtop or destroy wood-based buildings that survive the initial earthquake. Strong ground motions, ground failure, and land subsidence from earthquakes that pre- cede near-field tsunamis may also damage key egress routes, bridges, and critical facilities in coastal communities, thereby putting additional constraints on an individual’s ability to evacu- ate a tsunami-prone area. Communities in Alaska, the Pacific Northwest, and Puerto Rico are likely to experience several minutes of strong ground motions with tsunami inundation arriv- ing only minutes later. An initial large earthquake would likely result in damage to critical infra- structure in the evacuation zone (e.g., roads, bridges) and create barriers for individuals trying to evacuate from an imminent tsunami (e.g., toppled power lines, building debris in roads, etc.). Initial observations of the Chilean earthquake in February 2010 indicate this was the case in many coastal communities. In addition, critical facilities, such as emergency manage- ment offices, police stations, and fire stations, could be destroyed by the original earthquake or blocked by earthquake-related debris, possibly leaving emergency responders unable to man- age local evacuations. Radio and television stations and the towers that transmit their signals could also be damaged, thereby limiting the dissemination of warning or all-clear messages. The number of critical facilities in tsunami-prone areas has been documented in several studies (e.g., Charland and Priest, 1995; Lewis, 2007; Wood, 2007) and hazard mitigation plans (e.g., State of Hawaii Multi-Hazard Mitigation Plan). However, these efforts are simple inventories that do not delve into evacuation and response consequences of earthquake-damaged infrastructure and facilities or whether there are redundant facilities and access routes. Conclusion: Although many communities in the United States are threatened by a tsunami that originates from a source at close or intermediate distance, few evacuation studies have been conducted to evaluate the ability of at-risk individuals to reach higher ground before tsunami waves arrive. A related problem is that there have been no studies to assess the potential impact of local earthquakes that generate near-field tsunamis on egress routes, supporting infrastructure (e.g., bridges), or facilities considered critical in response efforts. Without such information, emergency managers are not able to identify where targeted outreach is needed and where potential vertical evacuation structures (e.g., buildings, engineered berms) may be warranted. Recommendation: For all communities with close or intermediate proximity (i.e., arrival times ranging from minutes to about an hour) to a potential tsunami source, the NTHMP should conduct evacuation modeling studies to assess the likelihood of successful horizontal evacuations. These studies should include the potential impacts of preceding 0

OCR for page 37
Aligning Priorities with Societal Risks from Tsunamis earthquakes on key egress routes and consideration of any special-needs populations in tsunami-prone areas. In communities where the time required for at-risk individuals to reach higher ground is likely greater than predicted tsunami wave arrival times, the NTHMP should conduct feasibility and effectiveness studies of various vertical evacuation strategies (e.g., buildings, engineered berms) that include engineering considerations and social and economic constraints of at-risk communities. 

OCR for page 37