3
Data Dissonance in Disasters

Demographic data are some of the most common data collected worldwide. The collection of demographic data, however, does not guarantee their availability before, during, or after disasters. More importantly, such data are generally not sufficient to assist the affected population. New crises need new data analyses. People who are responding to disasters complain that they are often operating in a data vacuum. These complaints vary in levels of frustration, but they are heard with respect to every kind of disaster and in all countries, whether rich or poor. Without appropriate demographic data, responders have difficulty setting short-term priorities, allocating scarce resources efficiently, or establishing strategic plans for longer-term recovery efforts. This chapter addresses the dissonance created by the existence of detailed local demographic data and the data vacuum that appears in the midst of most disasters.

This dissonance is illustrated through examples from three types of natural disasters (earthquakes, hurricanes, and tsunamis) in four different parts of the world (Turkey, Pakistan, India, and the United States) with affected populations in urban and rural settings. Although the examples are all natural disasters, the observations made from examination of the four case studies can also apply to populations-at-risk data in any setting.

This chapter discusses some of the important reasons why the existing data are underutilized by decision makers and, most significantly, how existing local area data would have to change to become more useful to decision makers who provide humanitarian assistance. The chapter also places the data dissonance in the context of broader information management, training, and technology considerations. Decreasing the dissonance



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 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises 3 Data Dissonance in Disasters Demographic data are some of the most common data collected worldwide. The collection of demographic data, however, does not guarantee their availability before, during, or after disasters. More importantly, such data are generally not sufficient to assist the affected population. New crises need new data analyses. People who are responding to disasters complain that they are often operating in a data vacuum. These complaints vary in levels of frustration, but they are heard with respect to every kind of disaster and in all countries, whether rich or poor. Without appropriate demographic data, responders have difficulty setting short-term priorities, allocating scarce resources efficiently, or establishing strategic plans for longer-term recovery efforts. This chapter addresses the dissonance created by the existence of detailed local demographic data and the data vacuum that appears in the midst of most disasters. This dissonance is illustrated through examples from three types of natural disasters (earthquakes, hurricanes, and tsunamis) in four different parts of the world (Turkey, Pakistan, India, and the United States) with affected populations in urban and rural settings. Although the examples are all natural disasters, the observations made from examination of the four case studies can also apply to populations-at-risk data in any setting. This chapter discusses some of the important reasons why the existing data are underutilized by decision makers and, most significantly, how existing local area data would have to change to become more useful to decision makers who provide humanitarian assistance. The chapter also places the data dissonance in the context of broader information management, training, and technology considerations. Decreasing the dissonance

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises in the underutilization of existing data requires better organization and management of the data and does not necessarily invoke significant additional direct costs. The difficulties in effecting better organization stem from a number of causes, both internal and external to the organization. These difficulties in obtaining and using data within existing organizational structures are highlighted in this chapter and discussed in detail in Chapter 4. THE LOCAL CONTEXT The examples presented in this chapter span diverse natural disaster settings and demonstrate that disasters can strike both rich and poor countries in a variety of inland and coastline terrains and can occur either instantaneously (sudden onset) or with ample notice. These disasters represent recent events for which the pre-impact planning and data were judged to be better than most and the committee had some knowledge based on the members’ own field work in affected areas. The common theme is that in all four scenarios, the existing population and spatial data were underutilized, making the disaster response less effective and ultimately worsening the plight of victims. The Izmit, Turkey Earthquake, 1999 Turkey is in a seismically active region affected by the northward collision of the Arabian plate with the Eurasian plate along the Anatolian fault system, which is more than 900 kilometers long (Figure 3.1). Eight major earthquakes have occurred along this fault during the past century resulting in more than 88,000 fatalities (CRED, 2006). On August 17, 1999, a 7.4 magnitude earthquake, the latest in this earthquake series, struck the Marmara region southeast of Istanbul (Figure 3.1) and resulted in more than 17,000 fatalities and more than half a million people left homeless (Scawthorn, 2000). Turkey is a middle-income country with a per-capita income of $7,680 in 2005 (PRB, 2005). At the time of the earthquake, Turkey’s population was approximately 73 million. Although the most recent population census had been taken in 1990, a housing census was taken as recently as 1997, two years prior to the Marmara earthquake. The next population census was not scheduled until 2000. In the immediate period after the earthquake struck, geospatial data and technologies were employed, and the resulting images and maps were made available to responders, and more generally, on the Internet. For example, Landsat 5 images of the devastated area were widely available two days after the earthquake, and the U.S. Department of State also produced five maps using high-resolution satellite images for in-country

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises FIGURE 3.1 Location of major earthquakes in Turkey along the North Anatolian Fault prior to the 1999 Izmit earthquake. SOURCE: Adapted from USGS, http://quake.wr.usgs.gov/research/geology/turkey/images/turkey_loc.gif.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises use. However, the maps and images were not combined with existing demographic data for the country and thus did not contain any information on the plight of the residents affected by the earthquake or relevant, local planimetric data such as street names; for the city of Adapazari, near the epicenter, the only local street map available to responders was a tourist map. Five weeks after the earthquake, no detailed disaster assessments were available when a joint Turkey-U.S. team went into the field (Eguchi et al., 2000a,b). At that time, physical damage assessments were made on the ground using geographic positioning and geographic information systems (GIS, including global positioning systems [GPS] data); although a delayed input to the recovery situation, these technologies made it possible to do in hours what used to take days to do. The new data were able to provide information about what roads were open and closed and the extent of the collapse of densely populated buildings; this kind of information was used to make some estimates of the number of people who had been affected by the event. The fact that the epicenter was located near Istanbul, which was not extensively damaged, facilitated accessibility to the area and thus made the relief efforts more effective than they might otherwise have been, given the lack of correspondence between remotely sensed satellite images and maps, local planimetric data, and national and subnational population data. The proximity of the Marmara earthquake to Istanbul also encouraged authorities in Turkey to develop better emergency preparedness plans for the city, which is responsible for 60 percent of Turkey’s gross national product and is home to 12 million people. Plans are currently under way to engage the public in disaster mitigation and preparedness and to invoke GIS and GPS technology and existing population data for this purpose (Ulgen, 2006). For the rest of the nation, however, population data are not available at subnational levels such as a city block or similar enumeration level. While personal locators and handheld GPS devices, or potentially, cell phones, for individual responders or the general public may be part of the future in some countries or specific communities, current technologies and population survey methods are still challenged by the issue of collecting demographic data when people are forced to migrate in response to a disaster. This issue is also explored in the section below on Hurricane Katrina and has been the topic of an earlier National Research Council report (NRC, 2001). South Asian (Kashmiri) Earthquake, 2005 The South Asian earthquake (also referred to as the Kashmir or Pakistan earthquake) of 2005 occurred in an inaccessible part of one of the poorest countries in the world (Figure 3.2). The epicenter of the 7.6

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises FIGURE 3.2 Status map of affected population and infrastructure on November 3, 2005. SOURCE: OCHA Humanitarian Information Centre Pakistan (HICP) and available on Relief Web, http://www.reliefweb.int/rw/RWB.NSF/db900LargeMaps/JOPA-6J5D2H?OpenDocument&emid=EQ-2005-000174-PAK&rc=3/.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises magnitude earthquake was near Muzaffarabad. The earthquake resulted in an estimated 90,000 deaths and more than 80,000 injuries. Approximately 3 million people were made homeless, and about 575 health facilities were partially or fully damaged (Durrani et al., 2005). The location of the event in the disputed territory of Kashmir, now controlled by Pakistan, and the timing of the October 8, 2005, event just before the onset of winter in the Himalayas, were significant factors in rescue, relief, and recovery efforts. These circumstances exacerbated the demographic and geospatial data dissonance that occurred at the time of the crisis. Pakistan has a population of 165 million, which is 66 percent rural. Its gross domestic product per capita is less than one-third that of Turkey (CIA, 2007). Pakistan’s most recent census had been taken in March 1998. Extensive mapping of the communities affected by the Kashmir earthquake was undertaken by a number of agencies and organizations. The maps provided detailed information, satellite imagery, and radar data showing the changing landscape as a result of the earthquake and massive landslides and the damage to buildings, roads, bridges, and other infrastructure. For example, the U.S. Geological Survey (USGS; http://earthquake.usgs.gov/eqcenter/eqinthenews/2005/usdyae/), the Earthquake Engineering Research Institute (EERI; http://www.eeri.org/lfe/clearinghouse/kashmir/observ1.php), the Centre for the Observation and Modeling of Earthquakes and Tectonics (COMET; http://comet.nerc.ac.uk/news_kashmir.html), and the Mid-America Earthquake (MAE) Center (http://mae.ce.uiuc.edu/Publications/cdseries/Cd%20series%20files/05-04/Report05-04.pdf), among many others, collected, analyzed, and distributed a significant amount of data, generated maps, and have made accessible a number of reports focusing on the physical impact of the earthquake (Figure 3.2). In some of the most severely affected regions, more than 95 percent of the buildings were destroyed, while lifelines (water, electricity) and the communication infrastructure were essentially rendered useless. The magnitude and extent of the Kashmir earthquake encouraged significant local and international response to this disaster. Initially, disaster relief aid was extremely slow to reach disaster victims, particularly in the most affected regions (e.g., Muzaffarabad, Balakot) and in remote mountain villages. A number of these remote communities were inaccessible for weeks after the earthquake and remained cut off from relief supplies even by helicopter or airplane. The rough terrain and the severity of the disaster combined with the lack of adequate disaster planning and response initiatives at the local level to hinder disaster relief strategies in the initial days following the event. As is typical following a disaster of such proportions, a massive convergence of materials and supplies from international sources resulted in an excessive amount of goods. This aid did not reflect the immediate needs of

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises the local population, was not responsive to local norms and circumstances, and was not delivered to the places most in need. The fact that the census information was dated made its use problematic; the census data had not been supplemented by newer sources of information that might have been derived from household surveys or a Demographic and Health Survey (DHS), nor were the census data available in electronic form for responders. The distribution of disaster relief aid following the Pakistan earthquake was more a function of who was at the right place at the right time, than of any systematic assessment and placement of vehicles for distribution of the necessary aid (Bilsborrow, 2006). Information on the affected population, including people’s demographic and economic characteristics, and their locations—data that should be available from a prior census or household survey—was difficult to obtain and thus could not be utilized in a meaningful way in the distribution of disaster relief aid or other humanitarian assistance once data from the post-earthquake community mapping initiatives were available. South Asian Tsunami, 2004 The South Asian tsunami was produced by a large, undersea 9.0 magnitude earthquake near the Andaman Islands on December 26, 2004. Within 15 minutes, the large wave began affecting the Thai and Indonesian coasts and radiated eastward from the epicenter. A short time thereafter, it began to impact other countries bordering the Indian Ocean to the north and west including Sri Lanka and India. Seven hours after the earthquake, the resulting tsunami had traveled across the Indian Ocean to Somalia (Figure 3.3). The human death toll remains uncertain years later, but official estimates suggest that approximately 250,000 perished (UNEP, 2006) and millions of people were displaced from their homes and livelihoods. Overall damage to the region was estimated at $10 billion (UNEP, 2006). Indonesia, India, and Sri Lanka suffered the highest fatalities. All three countries had taken population and housing censuses every 10 years since 1961, except Sri Lanka, which had a 20-year lapse from 1981 to 2001. Indonesia had a census in 2000, and India and Sri Lanka both had censuses in 2001. Therefore, up-to-date, local area demographic data were available in these countries when the tsunami struck. In India, the national statistical office (NSO) had also digitized population maps based on the last census. This information existed in the NSO and the state government offices of Tamil Nadu. The data, however, were not available to the first, second, and third responders to the disaster at the local level (Subramanian, 2006). Considerable aerial photography was collected of affected areas (Figure 3.4), but few ancillary data were available to responders on the number of

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises FIGURE 3.3 Impact area of the 2004 South Asian tsunami. SOURCE: International Coordination Group for the Tsunami Warning System in the Pacific (http://ioc3.unesco.org/itic/). Image available at http://www.unep.org/tsunami/images/ image001.png.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises FIGURE 3.4 The top satellite picture of Kalutara, Sri Lanka, was taken about an hour after the first tsunami wave hit on December 26, 2004. Water is rushing back out to sea after inundating the land. The lower picture shows what the same area looks like under standard daily conditions. SOURCE: Courtesy of Windows to the Universe, http://www.windows.ucar.edu. Image available at http://www.windows.ucar.edu/earth/images/tsunami_NASA_ EO_sm.jpg, 324 × 426 pixels- 160k, image may be scaled down and subject to copyright.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises people likely to have been in the area at the time, including both residents and tourists. While other data sources, such as administrative records, vital statistics, or perhaps tax records, could have been used, these data were largely unavailable for local coastal villages, so information about the pre-disaster population was in many cases derived from interviews with survivors. Lacking digitized census maps to determine how many people had been affected, relief teams in the field relied on hand-drawn maps of local villages such as Kameswaram (Figure 3.5). Ultimately, the numbers of people who were killed or affected were reconstructed through interviews of survivors, but these numbers lacked the support that could have been provided from pre-existing demographic data. Furthermore, data on the age and gender of the victims were not recorded, so it was difficult to project the longer-term impact on demographic patterns in the affected regions. Responders traveling through India and Sri Lanka in the immediate aftermath of the tsunami stated (Rodríguez et al., 2006, pp. 170-171) that they: observed a variety of irregularities or inequities, particularly related to the distribution of disaster relief aid. We also received reports of challenges in the provision of relief and recovery services. For example, in some instances, NGOs duplicated efforts or provided assistance not suited to the locale or to the varying population sizes. Further, while in some communities there seemed to be an abundance of aid, in other communities, particularly remote ones, the distribution of aid seemed to be quite slow and limited. In Sri Lanka, the ongoing conflict between the government and the Liberation Tigers of Tamil Eelam (better known as the Tamil Tigers) generated a variety of concerns regarding how aid was distributed which made understanding the difference between political and disaster response issues complicated. Hurricane Katrina 2005 In the previous three cases of disasters, the countries were relatively poor, the governments had few resources for population data collection and analysis, and the disasters occurred without warning. None of the three conditions existed in the case of Hurricane Katrina when it struck the United States. Unlike the previous three examples, Hurricane Katrina struck the world’s richest country, which had a wealth of data appropriate to tackle the disaster. The hurricane developed over a period of days, with hourly warnings on many public broadcast systems. Disaster preparation plans had been developed long in advance for just such a circumstance and were in place when the hurricane struck New Orleans and other parts of Louisiana, Alabama, and Mississippi. The amount of data and plans avail-

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises FIGURE 3.5 Hand-drawn maps of Kameswaram village, derived from information collected from villagers by aid responders (Subramanian, 2006). The map on the left shows the village prior to the tsunami, and the map on the right shows the water inundation of the village on the day of the tsunami. Despite an enormous contribution of international relief aid to areas affected by the earthquake and tsunami, and the advanced technology available to some responders, immediate relief efforts and assessments of the condition and numbers of the affected population relied on hand-drawn maps in this region. This map contains not only subnational, but village-specific, population data relevant to assistance providers and demonstrates that adequate aid does not necessarily require advanced technology—rather advanced technology requires appropriate and timely input of data. SOURCE: Courtesy of Abbiah Subramanian, Madras Christian College, India.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises BOX 3.4 Nongovernmental Organizations with Collections of Regional or Global Aggregations of Population Estimates ACAP specializes in demographic research and training to maximize the use of African census microdata for academic and policy-oriented research to benefit African governments and research (http://www.acap.upenn.edu). CIAT is a nonprofit organization that conducts social and environmental research oriented toward reducing hunger and poverty and preserving natural resources in developing countries (http://www.ciat.cgiar.org). CIESIN specializes in online data and information management, spatial data integration and training, and interdisciplinary research related to human interactions in the environment (http://www.ciesin.org). IPUMS collects and provides in database format U.S. census microdata and census data from around the world, for social and economic research at no cost to the user (http://www.ipums.umn.edu). example of such a data gateway is housed at the Socioeconomic and Data Applications Center (SEDAC; http://sedac.ciesin.columbia.edu) at CIESIN, Columbia University, which serves simply as an example of the architecture; this particular gateway currently serves gridded demographic data but not population data associated with finely resolved subnational units. Data Availability Simply sharing existing population data is admirable but may be insufficient if disaster management is to be effective. Other data—on population characteristics (e.g., demographic, health, socioeconomic) as well as the location of roads, facilities (e.g., hospitals, clinics, schools), and elevation and slope—may be critical. With the exception of elevation and slope, none of these data are currently available to the extent of population data. Roads are perhaps the most glaring omission of a data set for which there is a clear, unmet need and for which the private investment (e.g., commercial road maps) might make the construction of a public global good particularly challenging. A recent initiative supported by academia, the private sector, and UN agencies is developing a GIS data model for humanitarian action (Box 3.5). Data collected and acquired in the relief and recovery process may be useful for longer-term development activities. However, data acquired under conditions of disaster response may be the most perishable and decentralized and subject to the greatest variation in ownership, consistency, and quality. Nevertheless, these data may present important details about

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises TABLE 3.1 Spatial Resolution Associated with Available Population Data: Examples from the Gridded Populationof the World (GPW) Data Collection   Country Resolution (km) Population per Unit (in thousands) Area (km2) Administrative Area Type Number of Administrative Units Population (in thousands) (UN, 2000) 1 South Africa <1 1 1,217,645 Enumeration Area 83125 43,309 2 Guam 2 1 546 Block Group 203 155 3 Slovenia 2 0 20,224 Settlement 5989 1,988 4 Malta 2 6 315 Locality 67 390 5 Macao 3 142 19 Peninsula/Island 3 444 6 Maldives 3 13 189 Atoll 21 291 7 Malawi 3 1 94,958 Enumeration Area 9219 11,308 8 Mauritius 3 6 1,993 Municipal Ward/Village Council Area 186 1,161 9 Netherland Antilles 3 2 818 Geozone 71 215 10 United StatesVirgin Islands 3 3 374 Blocks 32 121 11 Czech Republic 4 2 78,616 Obec 6258 10,272 12 Switzerland 4 2 38,975 Commune 2912 7,170

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises   Country Resolution (km) Population per Unit (in thousands) Area (km2) Administrative Area Type Number of Administrative Units Population (in thousands) (UN, 2000) 223 Papua New Guinea 152 241 464,043 Province 20 4,809 224 Botswana 156 71 559,502 Census District 23 1,541 225 Serbia and Montenegro 159 2,658 101,561 Settlement 4 10,552 226 Sudan 171 358 2,492,385 Muhafazat 85 31,095 227 Algeria 219 634 2,302,498 Wilaya 48 30,291 228 Libyan Arab Jamahiriya 254 223 1,611,363 Mohafada 25 5,290 229 Angola 264 1 1,251,924 Municipio 18 13,134 230 Mongolia 265 108 1,546,294 Aimag 22 2,533 231 Chad 298 527 1,243,139 Sous-Prefecture 14 7,885 232 Saudi Arabia 386 1,604 1,938,837 Emirate 13 20,346 NOTE: “Resolution” is calculated as the square root of the land area divided by the number of administrative units. Only the finest and coarsest resolutions are shown. “Population per unit” is the average population per administrative unit (2000 population estimate divided by the number of units).

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises BOX 3.5 GIS Data Model for Humanitarian Action The purpose of the Humanitarian Data Model (HDM) is to promote GIS usage and data sharing among humanitarian organizations. The model would provide a “ready-to-use” template or framework from which various agencies and organizations could build their internal system. The current model consists of both geodatabase design and application framework. Those currently participating in the ongoing development of the model are the International Charter on Space and Major Disasters; UN OCHA Humanitarian Information Centers; UNOSAT; the RESPOND Consortium; UN Geospatial Information Working Group (UNGIWG); the U.S. Department of State’s Humanitarian Information Unit; the Open GIS Consortium (OGC); and the Institute for Crisis, Disaster, and Risk Management (ICDRM) at George Washington University. Proposed Applications of the Model: Risk assessment Capacity and vulnerability analysis Loss estimation Emergency relief management Infrastructure mapping Base Universal Layers: Critical infrastructure Settlements Land use Simple geography Political boundaries The draft HDM will consist of (1) a conceptual database design document, (2) an analysis diagram, (3) Unified Modeling Language (UML) documentation and Geography Markup Language (GML) schema, and (4) sample database and map documents. After it has been developed and stabilized through peer review, the HDM will then be ready for publication on ArcOnline and the ICDRM web site. The developers would then begin construction of a short book that documents the data model design. Published by ESRI Press, the book would provide a concise description of the thematic groups and classes in the model, and would serve as a reference book for teams working on projects. The paper discussing the development of the model referenced the Homeland Security, Defense-Intelligence, and Transportation Data Models are potential references for the structure of the HDM. SOURCE: http://www.humanitariangis.com/?q=node&from=10.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises migration flow and individual, household, and community characteristics, all of which might aid longer-term recovery and development efforts. However, agencies that collect and share data for relief aid may differ from those that collect data for longer-term recovery and development. Data Documentation Metadata for humanitarian applications are often not available or are poorly documented. Government agencies should be held responsible for including metadata using one of the standard available formats, as long as doing so does not hamper the publication of data during the crisis. This requirement could be incorporated with a larger mission of cost-effective investment of national resources to improve the level of geographic detail. Data should be subnational and should include more spatial units, boundaries, and sufficient detail for future updates and replication either by the original data collector or by another entity. These requirements could be explored in practice through disaster exercises or dress rehearsals where national-level actors interact in a simulated crisis environment to see where communication and data flows can break down. Data Sharing Needs During the Response Phase Rapid access to relevant information is crucial for the immediate response to the event and to mitigate its consequences. Various data about the affected areas are needed for this purpose, including building plans and footprints, local streets, utility lines, and critical facilities such as hospitals. Emergency responders must to be able to search, retrieve, assemble, and use existing geospatial information quickly (Goodchild, 2003a). Data sharing among agencies or units would greatly facilitate emergency response, but such exchanges are constrained by operating cultures (see Chapter 4) and information technology issues of interoperability. Interoperability is the ability of systems to exchange information, based on shared understanding of meaning (semantic interoperability) and mutually agreed formats (syntactic interoperability) (Goodchild et al., 1999). The data needed may be difficult to assemble or integrate because of incompatible formats and inaccuracies. Geospatial data standards that facilitate data sharing and interoperability need to be developed and adopted long before disasters occur. In addition, there are several important considerations when establishing the database or information system for emergency response (Goodchild, 2003b). First, the system should not only incorporate important geographic information but also have spatial analytical and modeling capabilities to facilitate effective response to disasters—for example, the capability to dis-

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises play, identify, and analyze critical spatial patterns or relationships among event locations, shelters, transportation routes, and the population at risk. The system should also permit interactive and dynamic visualization of the temporal progression of both the disaster situation and the evacuation of the affected population from the disaster site (see Box 3.2, for example). Since the risks at a disaster site can change swiftly and unexpectedly, decision support in real time for both emergency response personnel and the affected population is an essential function of such a system. This means that the system should be able to collect and disseminate information about the current condition of the disaster site in real time. The system should also allow responders to model or simulate possible trajectories of change in the disaster conditions and to formulate alternative decision scenarios. Furthermore, the database infrastructure should have a distributed architecture and allow for wireless and mobile deployment. As emergency crews work at a site, mobility and self-reliance (without requirement for hard-wired connections, even if available) are critical for information and decision support. The ability to communicate decisions and desirable actions effectively among all affected persons and emergency personnel is, therefore, essential for any geospatial data sharing infrastructure (Messick, 2006). To remain operational during a disaster situation, such infrastructure could be built upon a highly flexible and distributed system architecture, where the geographic database and decision support functionalities remain accessible to emergency personnel through multiple channels including wireless and mobile communications technologies. Most importantly, strong coordination mechanisms should be created at the field level between all actors involved in population data collection. For this to happen, population data management could be considered a separate “cluster,” similar to water, sanitation, and shelter. Although population data are cross-cutting, creating a population data cluster in each emergency would ensure proper resources and coordination opportunities. Training for Effective Data Use Of course, better data management and technology will require better training for the people who will be expected to use the data more efficiently. Some people will have to be trained to integrate demographic data with GIS technology, but many more people will have to be trained to analyze the demographic data, in both tabular and spatial form. This training does not necessarily have to be sophisticated, but it does have to be appropriate to emergency situations in general. Training is not free, but it need not be expensive if it is integrated into the job requirements of those who will be expected to use data during a disaster. Demographic data training is essential for basic disaster management competency and is also important

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises for promoting an organizational culture that is more efficient, flexible, and collaborative during disasters and beyond. Not every person involved in disaster relief needs data training; in fact, the selection of a subset of people, such as geographers and demographers, for training will be more cost efficient. One of the obstacles to the full employment of spatial demographic data during disasters, despite the clear need to do so, is the pressing human resource issue. Responding to disasters and humanitarian crises requires shared geographic and demographic thinking and training. At present, relatively few units, especially in developing countries, have sufficient trained expertise in both demography and geospatial tools and technologies. Improvements in the training and commitment of the national statistical office and other staff for each country are essential, to include both demographic projection methodology in local areas and the use of appropriate spatial administrative units in map form. There are a number of mechanisms for building such capacity, the first of which is recognizing the importance of the skill sets required for disaster preparedness and response. The second is formalized training. Such training programs could be part of overall capacity building and funded by bilateral aid programs, such as the U.S. Agency for International Development, or through broader country capacity-building programs, such as those conducted by the U.S. Census Bureau with support from the World Bank or the United Nations. The United States experience suggests that improvement in the capacity to prepare for and respond to disasters saves lives and reduces economic costs when an event occurs. SUMMARY The selected vignettes of disasters in countries with different levels of resources all demonstrate the same problem. A pervasive gap exists between the data that are available for use in disaster situations and the underutilization of those data when they are most needed. Each disaster seems to receive a new approach without any systematic understanding of what information is needed and the form it should take. Although each disaster has its own unique characteristics and each country has its own unique data collection and dissemination systems, the uniqueness of the disaster and of the national, local, and urban institutions does not obviate the more fundamental requirements for demographic data before and after disasters strike. The sense of each disaster being unique (in terms of both the event characteristics and the impact on particular places) has hindered the ability to generalize across disasters and countries in terms of common information requirements, especially among response communities.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises Remedies for data dissonance in disasters rely less on new money than on the better organization and coordination of the data and resources that already exist. National, local, and urban governments should coordinate the release of specific data sets that are vital to disaster management and planning based on what has been called the “good Samaritan” principle (Iwata and Chen, 2005). Protocols for data sharing drawn up before disasters could facilitate sharing during and after disasters. Protocols for data formatting could also be developed before disasters. Local and urban area data sharing before disasters would also facilitate disaster response. A template of illustrative local area population and other data could be developed for different kinds of disasters for different kinds of responders over time. At least two efforts to do this are taking place, the Humanitarian Data Model (see Box 3.5) and the Structured Humanitarian Assistance Reporting Effort (SHARE) effort by OCHA. Both efforts stress common formats for reporting data. Focus of both efforts on short-term response and on what is needed for long-term recovery efforts would be useful. To improve the effectiveness of response, the integration of spatial data and demographic data would be most useful prior to the disaster, not afterward. The development of MEDS to provide the baseline for preparedness and response is one avenue for reducing the demographic dissonance that plagues so many responders to disasters. RECOMMENDATIONS Based on the preceding discussion, the committee makes the following recommendations: Develop a template of minimum acceptable population and other geospatial data sets that are required by disaster responders. The data sets should be updated frequently (at least mid-decade if not more frequently) and include digital census enumeration units and other census maps in digital form. [Report Recommendation 4] The standard of open-access census data and sharing (as practiced, for example, by Brazil, South Africa, and the United States) should serve as a model for other agencies and for countries that currently do not operate in an open geospatial environment. This access includes spatial data such as digital boundary files or subnational units of countries of the world. Governments should release specific data sets that are vital to disaster planning and response. Furthermore, international standards should be developed for the release of subnational population data to maintain confidentiality. Countries

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises financially unable to comply with confidentiality standards should be offered incentives to do so. [Report Recommendation 5] Establish a centralized system of access, such as distributed archives and data centers for publicly available subnational data, including data from surveys. The archive would function as such a repository for shared local data and would have the primary responsibility for re-dissemination of data to the appropriate response communities during a disaster. The archives should build upon existing data resources. [Report Recommendation 6] REFERENCES Anselin, L., 2004. GeoDa 0.9.5-i Release Notes. Spatial Analysis Laboratory (SAL), Department of Agricultural and Consumer Economics, University of Illinois, Urbana-Champaign. Bagiire, V., 2006. Bridges.org; presentation to the committee at a workshop held on March 13-14 at the National Academies Keck Center, Washington, D.C. Presentation available through the National Academies Public Access Records Office. Bilsborrow, R., 2006. University of North Carolina; presentation to the committee on April 20 at the National Academies Keck Center, Washington, D.C. Presentation available through the National Academies Public Access Records Office. Buchanan-Smith, M., and S. Davies, 1995. Famine Early Warning and Response: The Missing Link. London: Intermediate Technology Publications. CIA (Central Intelligence Agency), 2006. The World Factbook. Available online at www.cia. gov/cia/publications/factbook/ [accessed October 3, 2006]. CIA, 2007. The World Factbook. Available online at https://www.cia.gov/cia/publications/factbook/index.html [accessed March 19, 2007]. CRED (Center for Research on the Epidemiology of Disasters), 2006. “Turkey” country profile. EM-DAT: The OFDA/CRED International Disaster Database. Available online at http://www.em-dat.net/disasters/countryprofiles.php [accessed October 3, 2006]. Doocy, S., 2007. Identify ways in which sub-national demographic and geographic data and tools could be used to help decision makers provide useful information to populations at risk. In National Research Council, Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises. Washington, D.C.: The National Academies Press. Doocy, S., A. Rofi, G. Burnham, and C. Robinson, 2006. Tsunami mortality in Aceh Province, Indonesia. Presented at the Population Association of America: Annual Meeting 2006, Los Angeles, March 30-April 1. Durrani, A.J., A.S. Elnashai, Y.M.A. Hashash, S.J. Kim, and A. Masud, 2005. The Kashmir Earthquake of October 8, 2005: A Quicklook Report. Mid-America Earthquake Center, CD Release 05-04, December. Eguchi, R.T., C. Huyck, B. Houshmand, B. Mansouri, M. Shinozuka, F. Yamazaki, and M. Matsuoka, 2000a. The Marmara Earthquake: A View from Space. Section 10: The Marmara, Turkey Earthquake of August 17, 1999: Reconnaissance Report, Technical Report MCEER-00-0001, Multidisciplinary Center for Earthquake Engineering Research, University at Buffalo, New York. Eguchi, R.T., C. Huyck, B. Houshmand, B. Mansouri, M. Shinozuka, F. Yamazaki, M. Matsuoka, and S. Ülgen, 2000b. The Marmara, Turkey earthquake: Using advanced technology to conduct earthquake reconnaissance. In Research Progress and Accomplishments

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises Report 1999-2000: A Selection of Papers Chronicling Technical Achievements of the Multidisciplinary Center for Earthquake Engineering Research (Buffalo, New York). Available online at http://mceer.buffalo.edu/publications/resaccom/00-SP01/ [accessedOctober 3, 2006]. Entwisle, B., and P. Stern (eds.), 2005. Population, Land Use and Environment: Research Directions. Washington D.C.: The National Academies Press. FEMA (Federal Emergency Management Agency), 2004. Hurricane Pam Exercise Concludes. Press release R6-04-093, July 23. Available online at http://www.fema.gov/news/newsrelease.fema?id=13051 [accessed October 3, 2006]. GNOCDC (Greater New Orleans Community Data Center), 2006. Rapid Population Estimate Survey Report, January 28-29. Available online at www.gnocdc.org/reports/NOLAPopEstimate.pdf [accessed October 3, 2006]. Goodchild, M.F., 2003a. Geospatial data in emergencies. In S.L. Cutter, D.B. Richardson, and T.J. Wilbanks, eds., The Geographical Dimensions of Terrorism. New York: Routledge, pp. 99-104. Goodchild, M.F., 2003b. Data modeling for emergencies. In S.L. Cutter, D.B. Richardson, and T.J. Wilbanks, eds., The Geographical Dimensions of Terrorism. New York: Routledge, pp. 105-109. Goodchild, M.F., M.J. Egenhofer, C. Kottman, and R. Fegeas, 1999. Interoperating Geographic Information Systems. New York: Springer. Isaaks, E.H. and R.M. Srivastava, 1989. An Introduction to Applied Geostatistics. New York: Oxford University Press. Iwata, S., and R.S. Chen, 2005. Science and the digital divide. Science 310(5747):405. Kaiser, R., P.B. Speigel, A.K. Henderson, and M.L. Gerber, 2003. The application of geographic information systems and global positioning systems in humanitarian emergencies: Lessons learned, programme implications and future research. Disasters 27(2): 127-140. Laska, S., 2004. What if Hurricane Ivan had not missed New Orleans? Natural Hazards Observer 29(2); available online at http://www.colorado.edu/hazards/o/nov04/nov04c.html. Liverman, D., E. Moran, R.R. Rindfuss, and P.C. Stern (eds.), 1998. People and Pixels: Linking Remote Sensing and Social Science. Washington, D.C.: National Academy Press. MacDonald, R., 2005. How women were affected by the tsunami: A perspective from Oxfam. PLoS Med 2(6):e178. MacEachren, A.M., and G. Cai, 2006. Supporting group work in crisis management: Visually mediated human-GIS-human dialogue. Environment and Planning B: Planning and Design 33(3):435-456. Mather, M., K.L. Rivers, and L.A. Jacobson, 2005. The American Community Survey. Population Reference Bureau. Population Bulletin 60(3):5-24. Maxwell, D., and B. Watkins, 2003. Humanitarian information systems and emergencies in the Greater Horn of Africa: Logical components and logical linkages. Disasters 27(1):72-90. Messick, S., 2006. Veterans For America; presentation to the committee at a workshop held on March 13-14, at the National Academies Keck Center, Washington, D.C. Presentation available through the National Academies Public Access Records Office. Montana, L., 2006. ORC Macro; presentation to the committee at a workshop held on March 13-14 at the National Academies Keck Center, Washington, D.C. Presentation available through the National Academies Public Access Records Office. NRC (National Research Council), 2001. Forced Migration and Mortality. Washington, D.C.: National Academy Press. NRC, 2007. Successful Response Starts with a Map: Improving Geospatial Support for Disaster Management. Washington, D.C.: The National Academies Press.

OCR for page 72
Tools and Methods for Estimating Populations at Risk from Natural Disasters and Complex Humanitarian Crises Noji, E.K., 2005. Estimating population size in emergencies. Bulletin of the World Health Organization 83(3):164. PRB (Population Reference Bureau), 2005. 2005 World Population Data Sheet of the Population Reference Bureau. Available online at http://www.prb.org/pdf05/05WorldDataSheet_Eng.pdf [accessed October 3, 2006]. Peckman, R., 2004. GIS, visualization and spatial analysis for major hazards in Europe. Poster Session at Seventh AGILE Conference on Geographic Information Science, Heraklion, Greece, April 29-May 1. Rodríguez,H., T. Wachtendorf,J.Kendra, and J. Trainor, 2006. A snapshot of the 2004 Indian Ocean tsunami: Societal impacts and consequences. Disaster Prevention and Management 15(1):163-177. Scawthorn, C., 2000. Introduction, The Marmara Turkey Earthquake of August 17, 1999: Reconnaissance Report. MCEER Multidisciplinary Center for Earthquake Engineering Research, University at Buffalo. Available online at http://mceer.buffalo.edu/publications/Reconnaissance/00-0001/default.asp?sH2=-1&oH0=-1&oH1=-1&oH3=-1&oH4=-1 [accessed October 3, 2006]. Subramanian, A., 2006. Madras Christian College, India; presentation to the committee at a workshop held on March 13-14 at the National Academies Keck Center, Washington, D.C. Presentation available through the National Academies Public Access Records Office. Tufte, E.R., 1983. The Visual Display of Quantitative Information. Cheshire, Connecticut: Graphics Press. Ulgen S., 2006. United Nations; presentation to the committee at a workshop held on March 13-14 at the National Academies Keck Center, Washington, D.C. Presentation available through the National Academies Public Access Records Office. UNEP (United Nations Environment Programme), 2006. After the Tsunami: Rapid Environmental Assessment. Available online at http://www.unep.org/tsunami/reports/TSUNAMI_report_complete.pdf [accessed October 3, 2006]. UNHCR (United Nations High Commissioner for Refugees), 1990. Handbook for Emergencies. Geneva, p. 370.