Nat Hazards (2011) 59:597–616 DOI 10.1007/s11069-011-9778-8 ORIGINAL PAPER
GIScience research challenges for emergency management in Southeast Asia Chen-Chieh Feng • Yi-Chen Wang
Received: 8 July 2009 / Accepted: 3 March 2011 / Published online: 18 March 2011 Ó Springer Science+Business Media B.V. 2011
Abstract The study examines the current application of geospatial information technologies and highlights the challenges of and constraints on GIScience research for emergency management that are particularly pertinent to Southeast Asia. Based on the generic GIScience research priorities noted by the University Consortium for Geographic Information Science, four topics most relevant to Southeast Asia are examined and discussed, including issues relating to use of spatial data, advancement and adoption of technology, dynamic representation of geographic processes, and public participation in emergency management. In-depth consideration of these can contribute to more appropriate and effective uses of geospatial information technology in Southeast Asia for emergency management purpose. Keywords
Emergency management GIScience Natural disaster Southeast Asia
1 Introduction The economic losses sustained by disaster worldwide since the 1950s has increased about sixfold, averaging some US$67 billion per year in the 1990s. The average monetary loss per year has increased to US$81 billion between 2000 and 2007 and reached a staggering US$181 billion in 2008 due mainly to the Sichuan earthquake in China (UNISDR 2009). Also, since 1990, natural disasters have been responsible for 86 percent of all disasterrelated deaths, of which three quarters occurred in Asia (Guha-Sapir et al. 2004). Natural disasters affect all subregions of Asia, but over the period of 2000–2009, Southeast Asian countries have experienced the most fatalities from the geophysical disaster of the 2004 Indonesia earthquake and the meteorological disaster of the 2008 Myanmar cyclone according to the International Disaster Database (EM-DAT). Wildfire, tsunami, and seasonal flooding have also cost the region tremendously (Fig. 1). Moreover, population C.-C. Feng (&) Y.-C. Wang Department of Geography, National University of Singapore, AS2, 1 Arts Link, Singapore 117570, Singapore e-mail:
[email protected]
123
598
Nat Hazards (2011) 59:597–616
Fig. 1 Top ten most costly natural disasters in Southeast Asia from 1990 to 2007 with date and type labeled in the map. The circles indicating damage costs are placed close to where these disasters took place. Note that the 2004 tsunami caused significant losses in four countries (from north: Myanmar, Thailand, Indonesia, and Malaysia). The dates in the figure are the beginning dates of the disasters. Data source: Compiled from EM-DAT, The OFDA/CRED International Disaster Database, www.em-dat.net—Universite´ Catholique de Louvain, Brussels, Belgium (last accessed 5 Nov 2010)
growth (e.g., an increase of 26% from 1990 to 2005) and the concomitant infrastructural development (e.g., increases of 61.2 and 149% in road density per 1,000 km2 from 1990 to 2005 for Malaysia and Laos) are putting substantial stress on the region’s unstable and fragile physical environment (UNESCAP 2009). This, coupled with the higher chances of climatic extremes due to global environmental changes, has resulted in the increased magnitude and frequency of floods, fire, and droughts in the region. Therefore, the management of emergency situations related to the natural disasters is increasingly seen as an institutional and research priority in Southeast Asia. Geographic information plays an essential role in this context because many critical issues that arise in these emergency situations have a strong spatial component. Geospatial information technologies have thus become fundamental tools in emergency management (Cova 1999; Radke et al. 2000). Here, the term geospatial information technologies is used to include a suite of geo-information processing and visualization tools such as geographic information systems (GIS), remote sensing, and global positioning systems (GPS) that allow the processing of spatial data for making decisions about some portion of the earth. The term geographic information science (GIScience) is used to refer to the science underlying the principles of the geospatial information technologies (Mark 2003). Prior studies have outlined the GIScience research challenges in emergency management, citing examples from various natural and human-induced hazards, the US 2001 World Trade Center disaster and the 2005 Hurricane Katrina (Cutter 2003; DeCapua and Bhaduri 2007; NRC 2007a). A study on this topic from the perspective of Southeast Asia, however, is still required for at least two reasons. First, the complex interaction between the physical and human systems has made the region especially vulnerable to natural disasters (Pereira 2001). The landforms of the region display great physical variations, consisting of large
123
Nat Hazards (2011) 59:597–616
599
river basins, a number of peninsulas, and thousands of islands forming the world’s two largest archipelagos. This landform assemblage has resulted from a combination of plate tectonics, Pleistocene environmental change, Holocene geomorphic processes, and anthropogenic modifications. Unlike the rest of the world that is shape by different combinations of these processes, in Southeast Asia all four processes are important (Gupta 2005). The physical structure of the region has influenced markedly the population distribution. In mainland Southeast Asia, main population centers are in the deltaic areas formed by large rivers such as the Mekong. Conversely, in insular Southeast Asia, main population centers are in areas of fertile volcanic materials often within the reach of potentially dangerous volcanoes. For example, the average population density on the flanks of the active Merapi volcano on island Java, Indonesia, is about 1,140 inhabitants per km2 (Thouret and Lavigne 2005). The main population centers are thus susceptible to extreme events, such as volcanic activities, earthquakes, and the resulting tsunamis. These natural disasters are not contained by country boundaries. Timely decisions in emergency management thus require the collaboration and coordination of scientists, responders, and residents from multiple countries with various standards and policies. It is desirable to know where and how geospatial information technologies have been used and whether the region has the same GIScience research challenges as other areas that have been examined previously. Second, the growing interests from public sectors and non-governmental organizations of Southeast Asia in geospatial information technologies have led to spatial database development, especially in online geo-portals and geospatial warehouses, such as the Integrated Land Information Service by the Singapore Land Authority. The region’s capacity of using geospatial information technologies in emergency management is also growing, such as the training of local people in Aceh and Nias to use GIS and GPS by the Reconstruction and Rehabilitation Agency in Indonesia (locally known as Badan Rahabititasi dan Rekonstrucki, the BRR) for post-tsunami reconstructions (Abdulharisa et al. 2006). Such provisions have resulted in many GIScience research issues in emergency management that need to be addressed. The goals of this paper are to examine the current application of geospatial information technologies and to identify the challenges of and constraints on GIScience research for emergency management that are particularly pertinent to coping with natural disasters in Southeast Asian contexts. The paper contains three parts. First, a brief overview of the current applications of geospatial information technologies for emergency management in Southeast Asia is provided. Second, selected research priorities from the University Consortium for Geographic Information Science (UCGIS) are employed to investigate challenges of GIScience research in emergency management. Third, the identified challenges and issues for Southeast Asia are discussed and compared with those in the USA. The paper is not intended to provide a complete discussion on the issues raised but to highlight particular areas that require immediate research investment from the GIScience research community. In so doing, it contributes to a broadened awareness of more appropriate and effective uses of geospatial information technology in Southeast Asia for emergency management.
2 Overview of applying geospatial information technologies for emergency management in Southeast Asia Emergency management is defined as the discipline and profession of applying science, technology, planning, and management to deal with extreme events that can injure or kill
123
600
Nat Hazards (2011) 59:597–616
large numbers of people, cause extensive damage to property, and disrupt community life (Waugh and Tierney 2007). In this paper, we focus on emergency caused by natural disasters. Many projects and studies of Southeast Asia have used geospatial information technologies in emergency management for natural disasters. Reviewing a portion of these will not only serve to point out where GIScience has contributed to emergency management in Southeast Asia, but also help identify the GIScience research challenges in emergency management in the next section. Managing disaster-induced emergency typically involves four phases: mitigation and preparedness activities to reduce the impact of potential and forecasted disasters, and response and recovery actions undertaken during or immediately after the disaster as well as longer-term efforts to restore the social and economic infrastructure and clean up, to the extent possible, the environment of affected areas (FEMA 1996). Countries of Southeast Asia have used geospatial information technologies in all four phases of emergency management. For example, during the mitigation and preparedness phases, the integration of various in situ data and a fire danger rating system for monitoring forest fire and transboundary haze provides planners information on the traveling paths of haze in Indonesia and Malaysia (de Groot et al. 2007). The use of MODIS data and AMSR-E microwave measurements has enabled flood forecasting and flood control planning (Brakenridge and Anderson 2006), and GPS networks have been used to capture information on land deformation caused by tectonic activities (Vigny et al. 2005). The modeling of tsunami propagation and inundation (Titov et al. 2005) presents an additional example of how geospatial information technologies can guide mitigation initiatives for coastal communities. Mobilizing rescue and relief operations such as coordinating rescue teams and dispatching resources to the right locations in the initial response phase are greatly facilitated by the spatial-enabled Incident Command Systems that have been incorporated by Southeast Asian countries into their emergency response systems (ADPC 2007). Examples of geospatial information technologies being used to identify hazard impacted areas and magnitude also include using RADARSAT to provide near real-time data in flooded areas of the Mekong River (Bonn et al. 2005), using synthetic aperture radar to monitor the earthquake damage in Java, Indonesia (Matsuoka and Yamazaki 2006), and using distributed tidal gauge data to assess the areal extent impacted by the 2004 tsunami (Borrero et al. 2006). In addition, geospatial information technologies have been used in the recovery phase to manage information about the tsunami-altered environment as well as delineate new hazard areas before selecting suitable sites for post-tsunami reconstruction in both Thailand (Tanasescu et al. 2006) and Indonesia (Abdulharisa et al. 2006).
3 Examining the GIScience research challenges Examples from Sect. 2 demonstrate how geospatial information technologies do and have played an important role in managing natural disasters in Southeast Asia. They also provide the basis on which related GIScience research challenges in the region can be examined. To this end, we consider the generic GIScience research priorities recommended by UCGIS (McMaster and Usery 2004), the research issues noted in prior studies (Radke et al. 2000; Cutter 2003), and the GIScience research agenda provided on the UCGIS website (http://www.ucgis.org/). Although these sources of information are strongly US based, their frameworks are comprehensive and sufficiently generic, making them the ideal basis on which GIScience research challenges in Southeast Asia can be
123
Nat Hazards (2011) 59:597–616
601
examined. Note that the UCGIS 2002 research agenda has a short-term research priority specifically on emergency data acquisition and analysis, and some issues, such as interoperability, have been considered as individual challenges in other studies (e.g., Cutter 2003). Because the purpose of the paper is not to provide a complete list of these challenges and our viewpoint is based on GIScience research, some of these issues are integrated into four major challenges, which will be elaborated below: use of spatial data; advancement and adoption of technology; dynamic representation of geographic processes; and public participation in emergency management. 3.1 Use of spatial data Spatial data, such as the whereabouts of residents, are indispensable in emergency management. The main GIScience research challenge in Southeast Asia concerns spatial data provision and usage because of prevailing perceptions of data ownership. Four issues critical to the use of spatial data in emergency management in the region include: (1) the accessibility of digital spatial data, (2) data quality, (3) geospatial clearinghouses, and (4) semantic interoperability. 3.1.1 Accessibility of digital spatial data The types of data required for various emergency management tasks are enormous. A committee of experts in consultation with individuals and agencies in the emergency management field has suggested the need for data on spatial and temporal distributions of population, resources, critical infrastructures, risks, hazards, service provided, and areas affected by disaster events (NRC 2007a). Developing an exhaustive list of the data needs is virtually impossible because it depends on the emergency management phase and the disaster type. It also requires tremendous effort to secure consents to release necessary information from various agencies and people. On the other hand, the foundational and framework data sets (NRC 1995) are essential for georeferencing, mapping, and extracting information from these data in every phase of emergency management and for most disaster types. Given the critical role of the two data sets for GIScience research in emergency management, we examined the accessibility of foundational and framework data sets for Indonesia, Malaysia, Singapore, and Thailand (Table 1). We examine these four countries for two reasons. First, they have been developing digital spatial data longer than other countries in the region, thereby maintaining better digital spatial data availability than other Southeast Asian countries, a necessary criterion for examining digital spatial data accessibility. Second, these four countries were judged to maintain stable contents on official websites, i.e., information on these websites is more reliable than that on other countries in the region. Note that the accessibility of the two data sets for the USA is also included in Table 1 to facilitate the comparison between the four countries and a highly developed one. The result in Table 1 suggests that none of the four countries has a complete set of the spatial database in the two data sets. The unchecked cells in Table 1 may not only indicate that data are inaccessible, but also mean that data exist in analog formats and huge dataprocessing effort is required before they can be analyzed using geospatial information technologies. This would suggest that countries in the region seldom embrace the notion that the value of data comes from its use. Legal concerns such as data custodianship, copyright and licensing, and national security have further complicated the issue of data accessibility (Abdulharisa et al. 2005; Hossain and Katiyar 2006). Even if the data are
123
602
Nat Hazards (2011) 59:597–616
Table 1 Availability of foundation and framework GIS data layers advertised on the government websites of Indonesia, Malaysia, Singapore, and Thailand Singapore
Malaysia
Thailand
Indonesia
USA
H1
H4
H9
H11
H12
1
5
Foundation data set Geodetic control Benchmark Land survey system
H
H
H13
*
Orthoimagery Orthophoto
H6
H10
H11
H14
H6
H9
H11
H14
11
H15
Elevation DEM
H
Bathymetry Framework data set Hydrography River Lake
H2
H6
H10
H11
H16
2
6
10
11
H16
H
H
H
H
10
H16
H
Dam Transportation Road Navigation chart
H2
H6
3
5
H
H
H10 10
H
Airport
H11
H17
H
11
H18
H
11
H19
Cadastral information Parcel information Land use/cover
H1
H7
2
8
H
H
H9 10
H
H20 11
H
H22
10
H22
H
Law enforcement
H
Response district
H21
10
Boundary Country Province/state Municipality
H2
H6
H9
H11
H17
**
H
6
9
H
11
H17
H
6
H
11
H17
**
H
9
H
A column for the USA is added for comparison. A ‘‘H’’ in the cell indicates that readily available digital data suitable for spatial analysis can be found on the corresponding government website * Not applicable for Thailand because it was the only non-European colony in Southeast Asia ** Not applicable for Singapore as it is a city state
Data availability and spatial coverage vary in different states
Source: 1 http://www.inlis.gov.sg; 2 http://www.sla.gov.sg/htm/ser/ser0503.htm; 3 http://www.mpa.gov.sg/ sites/port_and_shipping/port/charts_tidal_information_and_hydrography/charts/singapore_electronic_navig ational_chart%28enc%29.page; 4 http://www.jupem.gov.my/SPPMG/dv/geodaticSearch.aspx?p=g; 5 http:// www.water.gov.my/; 6 http://www.jupem.gov.my/SPPMG/dv/MappingProd.aspx?p=m; 7 http://www.jupem.gov. my/SPPMG/dv/cadastralSearch.aspx?p=c; 8 http://www.mygeoportal.gov.my/; 9 http://www.gistda.or.th; 10 http:// www.rtsd.mi.th/service/; 11http://www.bakosurtanal.go.id/?m=153; 12 http://www.ngs.noaa.gov/; 13 http://www.blm. gov/nils/GeoComm/home_services.html; 14 http://seamless.usgs.gov; 15 http://www.ngdc.noaa.gov/mgg/bathymetry/; 16 http://nhd.usgs.gov/; 17 http://www.census.gov/geo/www/tiger/tgrshp2009/tgrshp2009.html; 18 http://www.nautical harts.noaa.gov/mcd/enc/index.htm; 19 http://www.bts.gov/publications/north_american_transportation_atlas_data/; 20 http://www.fgdc.gov/library/whitepapers-reports/annual%20reports/2009/web-version/AppendixC.html; 21 http:// edcftp.cr.usgs.gov/pub/data/LULC/100K/; 22 http://gisinventory.net/data_layers.html (last accessed 5 Nov 2010)
123
Nat Hazards (2011) 59:597–616
603
published, they are often prohibitively expensive for studying physical processes that affect multiple countries. Evidently, efforts are needed to widen spatial data accessibility to emergency scientists and researchers.
3.1.2 Data quality The fitness of use of existing spatial data poses another challenge for GIScience research in emergency management in the region as data may not be collected at sufficient spatial or temporal resolution. This is a problem not unique to but more severe in Southeast Asia than in the USA because data are often collected without considerations of their potential applicability. For example, to accurately model the run-up distribution and extreme run-up height of tsunamis in the near shore areas, gridded bathymetric data with a horizontal resolution of 10–50 m or better are needed (Tivov and Gonzales 1997). Yet, only 25% of the countries hit by the 2004 tsunami have bathymetry data that meet this requirement (UNESCO 2006). Maps of various hazard types covering the region or certain individual countries are available, usually at a cartographic scale smaller than 1/24,000 (e.g., Petersen et al. (2007), UNOCHA (2007), and Pailoplee et al. (2008)). While these maps are useful for risk analyses at the regional or country level, they are unsuitable for the analysis at the community or settlement level (OAS 1991). Countries such as Thailand and Malaysia are developing flood-hazard maps at larger cartographic scales, but the accuracy of the maps varies (Amnatsan 2009; Anun 2009). Population data are another data set with quality issues for emergency management. In addition to inaccurate numbers that can be found in any population databases such as those in the Thailand census database (CIESIN and CIAT 2005), not every country in the region provides population data with a spatial resolution down to the finest administrative level possible, limiting the capability to reduce uncertainties in estimating the population at risk (NRC 2007a). Along the temporal dimension, population data commonly have two constraints impeding GIScience research in emergency management: unable to capture changes within the decadal scale of population census and unable to account for day-time population distributions (NRC 2007b). Countries in this region are also restricted by the same constraints. The temporal inaccuracy of the population data is further compounded by the region’s relatively high rate of irregular cross-border migration (Avanta and Arifin 2004). The LandScan Global Population Project provides partial solution to the problem by offering day-time population estimates that cover Southeast Asia (Budhendra et al. 2002), though the spatial resolution of the data, approximately at 1 9 1 km, may be insufficient for estimating affected population in smaller areas. Apart from these issues on temporal resolution, long-term historical data for estimating the return period of major hazards are generally of varying reliability and unavailable in digital form (van Westen et al. 2006), especially in the developing countries in the region. Problems due to poor data quality are amplified when various data are integrated for emergency management, most commonly gaps in the data sets, mismatches between datums, and outdated information (Syafi’i 2006). In addition, bathymetric data and digital elevation data seldom match along the coastal areas, making tsunami prediction difficult, which is a particularly serious drawback for a country like Indonesia consisting of thousands of islands. Given that a seamless database that integrates these two data sets is essential to improve inundation modeling (Titov et al. 2005), it is critical to resolve these data quality problems, especially those caused by data integration.
123
604
Nat Hazards (2011) 59:597–616
3.1.3 Geospatial clearinghouses The ability of GIS to translate information into action depends heavily on a geospatial clearinghouse that supports data discovery and transfer. Countries in Southeast Asia are in various stages of developing their national geospatial clearinghouses. Our investigation of various data search supports in the geospatial clearinghouses of the four analyzed countries suggests that Malaysia and Thailand have the richest support in terms of the search functions for geospatial data (Table 2). Their geospatial clearinghouses also provide online map services that enable data search by pointing at the exact location on a map. One common problem found in these clearinghouses, however, is the lack of a search-andretrieve function that allows immediate data access after the suitable data are identified. Efforts are thus needed to enhance supports for this function in order to facilitate data discovery and transfer. Most search functions in Table 2 are based on the metadata associated with the geospatial data in a clearinghouse, as searches are done mainly through comparing keywords specified by users and those in the metadata. While a higher number of search functions tend to lead to better search outcomes, these search functions are still limited by the metadata standard adopted in these clearinghouses, normally the ISO/TC211 19115 Geographic Information/Geoinformatics Standard. This standard is suited for describing spatial data, but understandably insufficient for supporting spatial data exchange between GIS and environmental models (Crosier et al. 2003), which is important for predicting impacts of natural disasters such as fire spreading or flooding (Ren and Xie 2004; Zerger and Wealands 2004). In addition, existing metadata standard typically allows optional metadata elements. This may cause inconsistency in how metadata are interpreted, thereby Table 2 Geospatial clearinghouse search functions and metadata availability for the four analyzed countries in Southeast Asia Singapore1
Malaysia2
Thailand3
Indonesia4
USA5
Keywords
H
H
H
H
H
Themes
H
H
H
Organizations
H
Search function H H
H
H
H
Product types
H
H
Time
H
H
H
Map (specific location)
H
H
H
H
Geo names
H
H
H
H
Zip codes/postal codes
H
Coordinates
H
H
H
H
H
H
H
H
H
H
H
H
H H
Spatial scales Clearinghouse nodes
H
Search and retrieve Metadata availability
A column for the USA is added for comparison. The search functions evaluated here are based on the common search functions on a spatial data infrastructure suggested by Crompvoets et al. (2004) Source: 1 http://www.onemap.sg/; 2 https://www.mygeoportal.gov.my; 3 http://thaigos.gistda.or.th:8080/ geonetwork/srv/th/main.home; 4 http://clearinghouse.bakosurtanal.go.id:8080/clearinghouse/srv/en/main. home; 5 http://geodata.gov (last accessed 5 Nov 2010)
123
Nat Hazards (2011) 59:597–616
605
confusing users who rely on the metadata for data search and retrieval. Furthermore, in order to maximize the search efficiency in the context of international collaborative efforts, metadata should ideally be entered in English, a requirement that poses challenges for counties in which English is not a working language. In Thailand, for example, the metadata content of the clearinghouse is mostly provided in Thai (CICC 2006), which could prevent regional or international researchers from finding suitable data. In addition to the geoportals at the country level, the region can benefit significantly from a regional geospatial clearinghouse as disasters tend to transcend country boundaries. Also, a regional geospatial clearinghouse avoids piecing together data in an ad hoc fashion using a combination of public and private provides and assets (Cutter 2003). The need of such a regional clearinghouse is addressed by various organizations such as the Pacific Disaster Center (http://www.pdc.org/mde/explorer.jsp) and two working groups of the Permanent Committee on GIS Infrastructure for Asia and the Pacific (PCGIAP) that aim at developing a regional geodetic infrastructure and a seamless spatial data set. Four constraints, however, have limited the success of such development (PCGIAP 2007). First, the type and the quality of geodetic data possessed by countries in Southeast Asia are still unclear. Second, the sharing of data sets of country border regions to facilitate data integration needs to be furthered. Third, individual country policies for releasing various data sets hinder the development of a complete regional database. And fourth, the adaption of geospatial data sharing standard to facilitate data exchange is still plagued by technical difficulties, further complicated not only by the constant revision of these standards, but also problems in keeping up with the latest standards for some countries (CICC 2006). 3.1.4 Semantic interoperability Semantic interoperability of geospatial data refers to the ability to exchange data from various sources and have the meaning encoded in the data accurately interpreted by the target system. It is typically enabled through the use of ontologies, which make precise the meaning of the data in computer readable forms (Klien et al. 2006; Xu and Zlantanova 2007), and of algorithms that determine semantic similarity between the data requests and the data available in the databases (Schwering 2008). Achieving semantic interoperability is essential in emergency management because most of the management decisions rely on integrating data from various sources. Despite the continual development of spatial data ontologies and the advancement in the semantic similarity algorithms, achieving semantic interoperability of the spatial databases in Southeast Asia is still challenging due to cognitive heterogeneity, i.e., one entity in reality being conceptualized as different feature types (Bishr 1997). Semantic non-interoperability as a result of cognitive heterogeneity within each country’s spatial database or between the spatial databases of multiple countries in Southeast Asia is significant. There are at least four reasons. First, an entity in reality may be represented by different sets of properties in two databases. In the Malaysian Standard for Geographic Information/Geomatics—Feature and Attribute Code (MS1759) (http://www.mygeoportal.gov.my/en/31.aspx), a ‘‘rail station’’ can be categorized as an ‘‘institutional building’’—characterized by properties such as address, building type, and number of floors, or as a feature in the ‘‘land transportation’’ category—characterized by properties such as station status (e.g., in operation or abandoned), rail track category (e.g., single rail or standard gauge rail), and station type (e.g., station or halt). Rail station data defined as institutional building will be needed when planning temporary shelters, while rail station data defined as a type of land transportation
123
606
Nat Hazards (2011) 59:597–616
would be needed for planning evacuation; neither definition alone generates data to support both emergency response needs. Second, data types with overlapping meanings can compromise emergency management by making the data selection process error prone. For example, although ‘‘residential complex’’ and ‘‘housing estate’’ in MS1759 both represent areas with high evacuation priority during an emergency, recording them as two distinct data types could lead to decision makers selecting data from only one data type if the commonality in meaning between the two data types cannot be established. Third, huge problems can arise if an entity carries two meanings due to the use of different reference datum. For example, the coastline data on Indonesia’s nautical charts are based on the average high tide levels, while the coastline data collected by the Indonesian National Coordinating Agency for Survey and Mapping are based on the water level prevailing when aerial photographs were taken. This results in obvious discrepancies in coastline length and land area and difficulty in integrating coastline data from the two differently defined data sources; it also affects the integration of land and marine data in tsunami inundation modeling. Fourth, emergency response can be compromised if an entity is conceptualized differently due to varying data collection goals. An example is found for ‘‘watershed’’, an important entity for flood monitoring in Southeast Asia. In Malaysia where the main datacollection goal of the Department of Irrigation and Drainage is to support river management, the term ‘‘watershed’’ is used interchangeably with the term ‘‘river basin’’ (T. S. Lee, personal communication, 2008). In Thailand, however, the main purpose for collecting watershed data is to identify the inherent capacity of landscape units (Tangtham 1992), which means that watershed data sets carry a highly specific meaning defined by slope, elevation, landform types, geology conditions, and soil types. The use of the same term watershed for these two data sets, however, can easily lead the users to believe that both data sets store the same type of data, resulting in erroneous prediction results and less accurate, if not inaccurate, management decisions. Achieving semantic interoperability of geographic data, which is presumed in any effective use of GIS in emergency management, thus requires fundamental GIScience research into the exact meanings of the entities in each country’s spatial databases in the region and the encoding of entity semantics into ontologies. Research on the degree of correspondence between entity types needs to be conducted to achieve semantic interoperability in geographic data for critical emergency management purposes within and between different countries. 3.2 Advancement and adoption of technology Disasters put stringent requirements on technology used to support emergency management. They can easily damage the telecommunication infrastructure and disrupt vital communication between a command center and local relief personnel. It is thus critical to maintain operations of geospatial technologies during an emergency event and ensure the timely delivery of accurate and up-to-date information, particularly during the response phase. To cope with these requirements, researchers and practitioners have emphasized the need to advance geospatial technologies for supporting emergency management. Five research areas have been suggested: (1) developing distributing spatial databases across the Internet, a security measure ensuring the accessibility of GIS during disasters (Tsou and Sun 2007); (2) adopting new information technology to automatically conflate data, create useful maps, and enable geo-collaborative activities (Zhang et al. 2007;
123
Nat Hazards (2011) 59:597–616
607
Fuhrmann et al. 2008); (3) designing light-weight mobile devices suitable to be operated under stringent conditions during an emergency (Radke et al. 2000); (4) delivering geospatial information easily understood by local responders (Cutter 2003); and (5) developing certain networking and wireless technologies that function well with limited infrastructural support (Mundie 2005). Although research area (5) is not specific to the geospatial domain, maintaining well-functioned communication among emergency command centers, local responders, and residents during an emergency is fundamental to the successful achievement of research areas (1) to (4). In Southeast Asia, there has been considerable progress along the suggested research areas (1), (2), and (5) (Table 3). For example, the region is now covered by a newly developed multimodal alert system, the Indonesia Tsunami Early Warning System (InaTEWS). The region also has quicker information delivery via the Real Simple Syndication (RSS) from Relief Web; easier map-generating mechanisms based on the digital earth platform (e.g., Google Earth) and the Geographically Encoded Objects for RSS Feeds (GeoRSS); and more Internet-based web services from the Global Disaster Alert and Coordination System (GDACS). Research has also been conducted to provide a peer-topeer communication system that remains operational after a disaster has occurred (Majid and Ahmad 2007) (Table 3). Table 3 Examples of new technology adoption and advancement in emergency management Example
Research area
Technologies adopted
Main function
Indonesian Tsunami (1) Early Warning System (InaTEWS)1
Short message service (SMS), Delivers tsunami warning; generates fax, radio, and web GIS maps of earthquake locations
Sentinel Asia2
(1)
Web GIS and digital earth platform
Provides remote sensing images and maps in AcrobatÒ portable document format (PDF); provides text descriptions for disasters
Relief Web3
(1), (2)
Web, email, and real simple syndication (RSS)
Provides text descriptions for disasters; provides tailored maps in PDF
Global Risk Data Platform4
(1), (2)
Web services, web, and Geographically Encoded Objects for RSS Feeds (GeoRSS)
Offers remote access to raw geospatial data for further analysis; offers data download through traditional access means (e.g., HTTP access); provides geo-tagged data for quick and easy map production
The Global Disaster Alert and Coordinate System (GDACS)5
(1), (2)
SMS, email, RSS, Common Alert Protocol (CAP), web, and digital earth platform
Sends location-aware alerts to the subscribers; provides up-to-date or historical information about disasters
An ad hoc framework for disaster or emergency communication
Maintains communication among victims of a disaster or an emergency event during infrastructure failure
Cross-layer (5) framework for postdisaster communication6
The numerical numbers in the ‘‘research area’’ column correspond to the research areas suggested in Sect. 3.2 Source: 1 http://balittanah.litbang.deptan.go.id/dokumentasi/prosiding/post%20tsunami/1harjadi.pdf; 2 https:// sentinel.tksc.jaxa.jp/; 3 http://www.reliefweb.int/rw/dbc.nsf/doc100?OpenForm; 4 http://preview.grid.unep.ch/ index.php?preview=map&lang=eng; 5 http://www.gdacs.org (last accessed 5 Nov 2010, July 2009); 6 Majid and Ahmad (2007)
123
608
Nat Hazards (2011) 59:597–616
However, efforts on the suggested research areas (3) designing light-weight mobile devices suitable to be operated under stringent conditions and (4) delivering geospatial information easily understood by local responders are still scant in Southeast Asia. The scarcity of reported efforts on designing light-weight mobile GIS technology for emergency management suggests that research in this area is almost non-existence in Southeast Asia, although an attempt was reported in the 2009 conference on Geoinformation Technology for Natural Disaster Management and Rehabilitation in Thailand. Training courses on using mobile GIS devices for emergency management were also hosted by the Asian Disaster Preparedness Center in 2008 in Bangkok. Nevertheless, the focus has been on field data collection for the mitigation and recovery phases, missing the response phase. It is unknown if the mobile GIS developed elsewhere, such as Lee (2008) for forest fire management in South Korea, can be adapted to the regional infrastructure specificity for communication between different countries and be integrated into the rescue operations in the region. As for delivering geospatial information easily understood by local responders, a proper design of GIS is desirable that can efficiently merge information from mobile and wireless units and provide real-time information understandable to non-technical emergency managers and responders. This may lessen the time that emergency responders have to spend on evaluating unorganized messages and requests sent from affected areas before directing and dispatching appropriate assistance to the local level because natural disasters normally have short response time. Hence, efforts are needed to investigate mobile GIS designs that are suitable for the response phase of a disaster and can be adapted to the regional specificity; to examine ways of representing analytical output that is understandable to the non-technical emergency managers and responders; and to develop methodologies to distinguish useful information for different groups of emergency response personnel to achieve intelligent information delivery. 3.3 Dynamic representation of geographic processes One long-standing research challenge in GIScience is the representation of geographic processes (McMaster and Usery 2004). Being able to capture the dynamic nature and interactions of the physical and social processes that account for an emergency remains a practical necessity because it allows the exploration of different emergency management scenarios and a better understanding of people’s vulnerability to a disaster, thereby developing better mitigation and response strategies. Traditional GIS cannot represent spatial dynamics beyond using layers of time-stamped data. These systems do not encode information on how processes occur and how they change over space and time (Rietsma and Albrecht 2005), making it impossible to support temporal queries. Newer data models based on object orientation provide generic tools to solve the problem (Worboys and Duckham 2004). Emergency management, however, requires the customization of these tools with entities, as defined by the processes causing the emergency, and with relations between entities, which describe explicitly the interactions between entities, processes, and entities and processes. This leads to the challenge of how our understanding of physical and social processes and their interactions can be encoded in GIS to support emergency management tasks. Landslide is used as an example of a frequent hazard in the mountain regions of tropical Southeast Asia. The physical processes accounting for landslide occurrences (e.g., land use change, land mass behavior, fluid dynamics, and rainfall) have been well studied (Alexander 2008). Interactions of these processes result in various landslide types, such as
123
Nat Hazards (2011) 59:597–616
609
mud slide or translational slide. The identification of these types represents our knowledge of the regional environment that is highly dynamic, shaped by the interaction of physical processes, soil and bedrock characteristics, and a threshold value over which a landslide occurs. This knowledge can be used to inform a dynamic representation model within which the composition of entities (e.g., land masses) and processes (e.g., change in soil saturation) form the basis for encoding information for developing landslide warning systems. Our examination on landslide studies involving the use of GIS in the region shows that there are at least two missing components in the context of emergency management. First, most of the studies do not go beyond correlating field observations with time-stamped data layers from geospatial information technologies to identify factors causing landslides or map landslide susceptibility (e.g., Lee and Sambath 2006; Yumuang 2006; Lee and Pradhan 2007). It is thus important to formalize these research findings for the development of dynamic representations of landslides that will be beneficial in the response phase of a landslide emergency. Second, prior studies do not characterize social processes in terms of evolving states of vulnerability. The lack of this component is rather unsatisfactory because it is the people’s vulnerability and changing capacities to deal with disasters over time that matter. In a study of social vulnerability of a coastal Vietnam district, Adger (1999) showed, among other factors, how this was affected by the increased regional autonomy, de-collectivization, and changes in land allocation systems over time and across different communes. In Indonesia and Thailand, the shift toward a more decentralized administration in the later half of the 1990s engendered fire management strategies that potentially decreased local vulnerability to fire (Murdiyarso and Lebel 2007). In Iloilo, Philippines, Heijmans (2004) reported how farmers of different financial means mixed and changed their cropping strategies and crop types to lower their vulnerability to drought. Clearly, social vulnerability to natural disasters is dynamic and depends on a combination of economic and institutional contexts (Wisner 1978). However, while studies such as these could inform representational frameworks that encode key variables influencing vulnerability, there is very little work done on representational frameworks in the region. The research challenge for this region is thus the design and development of a better representation framework incorporating an understanding of the physical and social processes influencing the nature and severity of natural disasters, and the possible interactions between them. Such frameworks would allow us not only to answer basic spatiotemporal queries, but also to perform spatiotemporal reasoning tasks. Most importantly, these frameworks will enable emergency managers to examine various scenarios efficiently and accurately.
3.4 Public participation GIS and emergency management Public participation plays an essential role in emergency management because the acceptance of the management plans by local people ensures the sustainability of the management plans and thus the success of the management strategies. Geospatial information technologies are important tools in encouraging public participation, which is essential in sustaining local acceptance of and compliance in emergency management strategies (Pierce 2003). Maps are often the best way to present information for policy making and opinion solicitation. The combination of public participation and GIS (hereafter referred to as PPGIS) permits better portrayal of the problems at hand and enhanced
123
610
Nat Hazards (2011) 59:597–616
communication between stakeholders, thereby aiding the formulation of better management plans (Sieber 2006). Public participation and GIS approach have been applied in Southeast Asian countries by various international agencies, such as the International Federation of the Red Cross and Red Crescent Society and Oxfam, for bridging the gaps between technocratic and localcentered disaster management strategies. Our examination of these applications identifies four issues that require GIScience research attention. The first concerns the representation of spatial knowledge solicited from local stakeholders in GIS. Traditionally, representation of local knowledge in GIS is done by forming a data layer of points, lines, or polygons that contains some value-based information. Advances in GIS can now capture qualitative information, such as personal narratives and video recordings. Nevertheless, data collected ultimately require human interpretation, often by experts who do not have local knowledge. The challenge here is how to apply the experts’ modeling skills to represent local knowledge in GIS without distorting the existing information or introducing any ‘‘external’’ knowledge. The second issue relates to the effectiveness of applying PPGIS in emergency management. One goal of applying PPGIS is to allow the local voices to be heard and incorporated into emergency management plans. Most work conducted in Southeast Asia reports local people’s coping strategies, how they perceive disasters, and the factors affecting their vulnerability (de Dios 2002). Little is found on how these results are incorporated into emergency management plans or, indeed, on the degree of such plans do lower local people’s vulnerability to natural disasters. Studies on the best practices of PPGIS in emergency management in Southeast Asia are thus desirable. The third issue concerns the sustainability of a PPGIS project. During the recovery phase, attention and aid from international communities are often shifted to a disaster in another area. For example, the 28 December 2004 Sumatran earthquake claimed at least one thousand lives and destroyed hundreds of buildings in Nias, Indonesia. During our participation in the recovery phase there, we discovered that although GIS was proposed to facilitate the reconstruction, the necessary funding support from international organizations was planned for only 1 year, and the coordination by BRR in Indonesia was only shortterm. Local efforts were expected to shoulder long-term reconstruction tasks. Last, efforts are needed to distinguish knowledge that is transferable across different parts of Southeast Asia from that which is not. Such work is important for developing best practice guidelines for applying PPGIS for emergency management in the region. The PPGIS approach is almost always dependent on national and local laws, culture, politics, or the history and perceptions of affected communities. In the study of volcanic eruptions in the Philippines, Heijmans (2001) found that the local residents’ views on their own safety differed from that of the central government’s. The Philippine government’s management plan held that local residents should retreat to emergency shelters whenever major seismic activities are detected. Local residents, however, considered themselves in danger only when the volcanic cloud and ashes from the eruption are visible. Furthermore, retreating to an emergency shelter is not perceived as safer than staying at home because of the lack of water supply and the crowdedness of shelters. While such findings provide extremely useful information for fine-tuning specific and place-based emergency management plans, the level of specificity almost always precludes their applicability elsewhere in the region to provide best practice guidelines for the whole region. In the meantime, the transferable research insights and results might be clarified so that they can be used as guidelines for other regions.
123
Nat Hazards (2011) 59:597–616
611
4 Comparing challenges in Southeast Asia and in the USA Our examination in Sect. 3 shows that Southeast Asia has similar generic GIScience research challenges in the context of emergency management to those in the USA. Conversely, many issues and constraints highlighted under the four main research challenges are more severe in Southeast Asia than those in the USA because the region is heterogeneous, characterized by multiple data sharing policies, various spoken languages, and diverse cultural backgrounds. Two issues related to spatial data appear to be more challenging. First, unlike the USA where spatial data are in general accessible to the general public (cf. Tables 1, 2), the unwillingness to open up spatial databases impedes spatial data access in Southeast Asia. The cause of this unwillingness may be that the data custodians are uncertain about the consequences of granting access of their data to researchers. There is also a misconception that the value of the data is measured by the revenue from data sale. The notion that the value of data comes from its use (Onsurd and Rushton 1995) is still not well perceived in the region. To remediate these problems, efforts are needed to evaluate the effects of opening up spatial data, specifically on the benefits of supporting research in emergency management and to advocate the idea that the value of the data can be better measured by the amount and frequency of their usage. Second, the development of an integrated regional database that houses spatial data for emergency management is much more difficult than the development of a seamless database in the USA. The degree of semantic non-interoperability is higher in the region than that of the USA because Southeast Asia has a higher level of cognitive heterogeneity. Compared to the abundance of research on resolving semantic non-interoperability in the USA (Rodriguez et al. 1999; Wiegand and Garcia 2007) and in Europe (Visser 2004; Kavouras et al. 2005), the investigation on this topic is virtually non-existent in the region. Little is known about the severity of this issue in the region and how it may affect crossborder emergency management. Alternatively, the region is fast in adopting new technologies for emergency management, as evident in the adoption of international standards such as the ISO/TC211 19115 and Common Alert Protocol (CAP), of clearinghouse models of the Federal Geosaptial Data Committee (Table 2), and of communication technologies such as SMS and GeoRSS (Table 3). However, the adoption process is often incomplete, hampering GIScience research in emergency management in Southeast Asia. One obvious example is the geospatial clearinghouses (Table 2), many of which miss the crucial search-and-retrieve function. The need to customize these technologies to the users in the context of emergency management in the region is also often overlooked. Finally, the uneven advancements in information and communication technology—with Singapore ranked 15 while Myanmar and Cambodian ranked 191 and 121, respectively, among the 154 countries surveyed in 2007 (ITU 2009)—hold back many benefits that might have been brought forth by these technologies. The demand for better dynamic representations of geographic processes is the same for regions where natural disasters are common. This issue deserves to be further underscored for Southeast Asia because in the region, engineering approaches are often given priority over other approaches for emergency management (Karim 2004). The negligence of social processes and local knowledge as an integral part of the emergency management practices, similar to what Opadeyi et al. (2003) reported in Caribbean, results in that emergency management tools produce over-simplified estimates for vulnerability. This may account for the lack of ready-to-use geospatial tools capable of representing social processes.
123
612
Nat Hazards (2011) 59:597–616
5 Conclusion During the past decades, natural disasters in the Southeast Asia region have caused significant casualties and economic loss. This article highlights four interrelated GIScience research challenges in emergency management that are most pertinent to the Southeast Asia region, including the use of spatial data, advancement and adoption of technology, dynamic representation of geographic processes, and PPGIS. Our examination suggests that these four challenges for GIScience research in emergency management are paramount in Southeast Asia because the use of GIS in the region has focused more on developing infrastructures than on handling issues that are more delicate and contextual, such as facilitating the emergency management operations and considering the needs of the locals. The access to spatial data remains a constraint on research efforts in Southeast Asia even though national spatial data infrastructures are being developed by the respective government agencies. The slowness in moving toward openness in data access is exhibited in various forms—from not publishing the data to charging fees that may be cost prohibitive to researchers. The implementation of geospatial metadata remains problematic mainly because the metadata standard adopted is not sufficiently rich to support all emergency management tasks. Achieving semantic interoperability is especially challenging for the region because of the cognitive heterogeneity in encoding various spatial entities. The need to integrate data across the region is exacerbated by the aforementioned problems, as the inconsistencies caused by different national systems have to be resolved. To allow people to respond to a disaster more efficiently and to shorten the recovery duration, research is needed to advance mobile GIS with special attention to the user interface of the GIS software. To enable emergency managers to examine various scenarios and perform spatiotemporal reasoning, improvement in dynamic representation is needed for this region. The improvement should focus not only on the physical aspect of disasters, such as representing how landslides can occur, but also on the social aspect, such as how vulnerability can change over time. This leads to the use of PPGIS approach to fill the gaps formed by emergency management plans that are technocrat centered or that lack consideration of people’s views on the ground. The application of PPGIS in this region requires research efforts on several fronts, of which correct interpretation of data appears to be the most important. Acknowledgments Funding supporting from the National University of Singapore funding number R-109-000-070-101/123 and the Staff Research Support Scheme are appreciated. We wish to thank Yikang Feng for GIS mapping and Xinyi Liang for Thai translation.
References Abdulharisa R, van Loenenb B, Zevenbergenc J (2005) Legal aspects of access to geo-information within Indonesian spatial data infrastructure. In: Proceedings of the ISPRS workshop on service and application of spatial data infrastructure, Hangzhou, China, pp 147–158 Abdulharisa R, Hakim DM, Riqqi A, Zlatonova S (2006) Assessment of geo-information utilization at the tsunami affected areas in Aceh and Nias, Indonesia. In: Fendel E, Rumor M (ed) Proceedings of urban data management society (UDMS) ‘06. Urban Data Management Society, Aalborg, Demark, pp 6.98–6.97 Adger WN (1999) Social vulnerability to climate change and extremes in coastal Vietnam. World Dev 27:249–269
123
Nat Hazards (2011) 59:597–616
613
ADPC (Asian Disaster Preparedness Center) (2007) Incident command system for building faster responses to disasters response in the Indian Ocean countries. http://www.adpc.net/v2007/IKM/EVENTS% 20AND%20NEWS/ADPC/2007/Oct/Abstract_ICS.pdf. Accessed 5 Nov 2010 Alexander DE (2008) A brief survey of GIS in mass-movement studies, with reflections on theory and methods. Geomorphology 94:261–267 Amnatsan S (2009) Progress report: flood hazard mapping in Thailand. http://www.pwri.go.jp/icharm/ training/2009seminar/progressreport2009_thai2.pdf. Accessed 5 Nov 2010 Anun P (2009) Progress report on flood hazard mapping activities in Malaysia. http://www.pwri.go.jp/ icharm/training/2009seminar/progressreport2009_malaysia.pdf. Accessed 5 Nov 2010 Avanta A, Arifin EN (2004) Should Southeast Asia borders be opened? International migration in Southeast Asia. In: Avanta A, Arifin EN (ed) Institute of Southeast Asian Studies, Singapore, pp 1–27 Bishr Y (1997) Overcoming the semantic and other barriers to GIS interoperability. Int J Geogr Inf Sci 12:299–314 Bonn F, Aube´ G, Mu¨ller-Poitevien C, Be´nie´ G (2005) Emergency preparedness system for the Lower Mekong River Basin: a conceptual approach using earth observation and geomatics. In: van Oosterom P, Zlatanova S, Fendel EM (eds) Geo-information for disaster management. Springer, Berlin, pp 269–280 Borrero JC, Sieh K, Chlieh M, Synolakis CE (2006) Tsunami inundation modeling for western Sumatra. Proc Natl Acad Sci 103:19673–19677 Brakenridge R, Anderson E (2006) MODIS-based flood detection, mapping and measurement: the potential for operational hydrological applications. In: Marsalek J, Stancalie G, Balint G (eds) Transboundary floods: reducing risks through flood management, IV. Earth and environmental sciences, vol 72. Springer, The Netherlands, pp 1–12 Budhendra B, Bright E, Coleman P, Dobson J (2002) LandScan: locating people is what matters. Geoinformatics 5:34–37 CICC (Center for the International Cooperation of Computerization) (2006) Report of the conference for the Asian IT standardization—GIS and geospatial data interoperability. Tokyo, Japan CIESIN and CIAT (Center for International Earth Science Information Network Columbia University and Centro Internacional de Agricultura Tropical) (2005) Gridded population of the world version 3 (GPWv3): population grids. Palisades, NY. http://sedac.ciesin.columbia.edu/gpw. Accessed 5 Nov 2010 Cova TJ (1999) GIS in emergency management. In: Longley PA, Goodchild MF, Maguire DJ, Rhind DV (eds) Geographical information systems: principles, techniques, management and applications. Wiley, New York, pp 845–858 Crompvoets J, Bregt A, Rajabifard A, Williamson I (2004) Assessing the worldwide developments of national spatial data clearinghouses. Int J Geogr Inf Sci 18:665–689 Crosier SJ, Goodchild MF, Hill LL, Smith TR (2003) Developing an infrastructure for sharing environmental models. Environ Plan B Plan Des 30:487–501 Cutter SL (2003) GIScience, disasters, and emergency. Trans GIS 7:439–445 de Dios HB (2002) Participatory capacities and vulnerabilities assessment: finding the link between disasters and development. Quezon City, Philippines de Groot WJ, Field RD, Brady MA, Roswintiarti O, Mohamad M (2007) Development of the Indonesian and Malaysian fire danger rating systems. Mitig Adapt Strateg Glob Change 12:165–180 DeCapua C, Bhaduri B (2007) Applications of geospatial technology in international disasters and during Hurricane Katrina. Mississippi State University Coastal Research and Extension Center, Gulfport, MS. http://www.gri.msstate.edu/research/katrinalessons/Documents/GeoSp_Tech_Applications.pdf. Accessed 5 Nov 2010 FEMA (Federal Emergency Management Agency) (1996) State and local guide for all-hazard emergency operations planning. Federal Emergency Management Agency. http://www.fema.gov/pdf/plan/ 0-prelim.pdf. Accessed 5 Nov 2010 Fuhrmann S, MacEachren A, Cai G (2008) Geoinformation technologies to support collaborative emergency management. In: Chen H, Brandt L, Gregget V, Traunmu¨ller R, Dawes S, Hovy E, Macintosh A, Larson CA (eds) Digital government—E-government research, case studies, and implementation. Springer, New York, pp 395–420 Guha-Sapir D, Hargitt D, Hoyois P (2004) Thirty years of natural disasters 1974–2003: the numbers. Presses Universitaires de Louvain, Belgium Gupta A (2005) Landforms of Southeast Asia. In: Gupta A (ed) The physical geography of Southeast Asia. Oxford University Press, Oxford, pp 38–94 Heijmans A (2001) Vulnerability: a matter of perception. International Conference on Vulnerability in Disaster Theory and Practice, London, UK
123
614
Nat Hazards (2011) 59:597–616
Heijmans A (2004) From vulnerability to empowerment. In: Bankoff G, Frerks G, Hilhorst D (eds) Mapping vulnerability—disaster, development, and people. Earthscan, London, pp 115–127 Hossain F, Katiyar N (2006) Improving flood forecasting in international river basins. Eos Trans Am Geophys Union 87:49–60 ITU (International Telecommunication Union) (2009) Measuring the information society—the ICT development index 2009. http://www.itu.int/net/pressoffice/backgrounders/general/pdf/5.pdf. Accessed 5 Nov 2010 Karim N (2004) Hazard mitigation in South and Southeast Asia. In: Stoltman JP, Lidstone J, Dechano LM (eds) International perspectives on natural disasters: occurrence, mitigation, and consequences. Springer, Dordrecht, pp 211–230 Kavouras M, Kokla M, Tomai E (2005) Comparing categories among geographic ontologies. Comput Geosci 31:145–154 Klien E, Lutz M, Kuhn W (2006) Ontology-based discovery of geographic information services—an application in disaster management. Comput Environ Urban 30:102–123 Lee B (2008) Forest fire situation and main study activities in South Korea. http://wwwoa.ees.hokudai.ac. jp/people/yamazaki/G8/S4-all.pdf. Accessed 5 Nov 2010 Lee S, Pradhan B (2007) Landslide hazard mapping at Selangor, Malaysia using frequency ratio and logistic regression models. Landslides 4:33–41 Lee S, Sambath T (2006) Landslide susceptibility mapping in the Damrei Romel area, Cambodia using frequency ratio and logistic regression models. Environ Geol 50:847–855 Majid S, Ahmad K (2007) Cross-layer framework for post-disaster communications. 14th IEEE international conference on telecommunications (ICT) and Malaysia international conference on communications (MICC) 2007, IEEE, Penang, Malaysia, pp 725–728 Mark DM (2003) Geographic information science: defining the field. In: Duckham M, Goodchild MF, Worboys MF (eds) Foundations of geographic information science. Taylor & Francis, London, pp 3–18 Matsuoka M, Yamazaki F (2006) Use of SAR imagery for monitoring areas damaged due to the 2006 mid Java, Indonesia earthquake. 4th International workshop on remote sensing for post-disaster response. Cambridge, UK. http://www.arct.cam.ac.uk/curbe/matsuoka06.pdf. Accessed 5 Nov 2010 McMaster R, Usery EL (2004) A research agenda for geographic information science. Taylor & Francis, London Mundie M (2005) Technology transforming disaster response. ICT solutions for disaster recovery management and global warning. Phuket, Thailand. http://www.microsoft.com/presspass/exec/craig/0218PhuketICT.mspx. Accessed 5 Nov 2010 Murdiyarso D, Lebel L (2007) Local to global perspectives on forest and land fires in Southeast Asia. Mitig Adapt Strateg Glob Change 12:3–11 NRC (National Research Council) (1995) A data foundation for the national spatial data infrastructure. The National Academies Press, Washington, DC NRC (National Research Council) (2007a) Successful response starts with a map: improving geospatial support for disaster management. The National Academies Press, Washington, DC NRC (National Research Council) (2007b) Tools and method for estimating populations at risk, from natural disasters and complex humanitarian crisis. The National Academies Press, Washington, DC OAS (Organization of American States) (1991) Primer on natural hazard management in integrated regional development planning. Washington, DC Onsurd HJ, Rushton G (1995) Sharing geographic information systems. Center for Urban Policy Research, New Brunswick Opadeyi J, Ali S, Chin E (2003) Status of hazard maps vulnerability assessments and digital maps in the Caribbean: final report. Saint Michael, Barbados Pailoplee S, Sugiyama Y, Charusiri P (2008) Probabilistic seismic hazard analysis in Thailand and adjacent areas using regional seismic source zones. In: Proceedings of the international symposia on geoscience resources and environments of Asian terranes (GREAT 2008), 4th IGCP 516 and 5th APSEG, Bangkok, Thailand, pp 405–409 PCGIAP (Permanent Committee on GIS Infrastructure for Asia and the Pacific) (2007) Report—13th meeting of the PCGIAP. http://219.238.166.217/pcgiap/PCGIAP13/final%20report/full%20report.pdf. Accessed 5 Nov 2010 Pereira JJ (2001) Environmental geoscience in Southeast Asia: current trends and future challenges. Episodes 24:115–117 Petersen M, Harmsen S, Mueller C, Haller K, Dewey J, Luco N, Crone A, Lidke D, Rukstale K (2007) Documentation for the South East Asia seismic hazard maps. US Geological Survey, Reston Pierce L (2003) Disaster management and community planning, and public participation: how to achieve sustainable hazard mitigation. Nat Hazards 28:211–228
123
Nat Hazards (2011) 59:597–616
615
Radke J, Cova T, Sheridan MF, Troy A, Lan M, Johnson R (2000) Application challenges for geographic information science: implications for research, education, and policy for emergency preparedness and response. Urban Reg Inf Syst As J 12:15–30 Ren AZ, Xie XY (2004) The simulation of post-earthquake fire-prone area based on GIS. J Fire Sci 22:421–439 Rietsma F, Albrecht J (2005) Modeling with the semantic web in the geosciences. IEEE Intell Syst 20:86–88 Rodriguez AM, Egenhofer MJ, Rugg RD (1999) Assessing semantic similarities among geospatial feature class definitions. In: Vckovski A, Brassel K, Shchek H-J (eds) Interoperable geographic information systems, Interop ‘99. Springer, Berlin, pp 189–202 Schwering A (2008) Approaches to semantic similarity measure for geo-spatial data: a survey. Trans GIS 12:5–29 Sieber R (2006) Public participation geographic information systems: a literature review and framework. Ann As Am Geogr 96:497–507 Syafi’i MA (2006) The integration of land and marine spatial data set as part of Indonesian spatial data infrastructure development. Seventeenth United Nations regional cartographic conference for Asia and the Pacifics, Bangkok, Thailand. United Nations Economic and Social Council. http://unstats.un.org/ unsd/geoinfo/E-CONF97-6-IP-46.pdf. Accessed 5 Nov 2010 Tanasescu V, Gugliotta A, Domingue J, Leticia G-V, Davies R, Rowlatt M, Richardson M, Stincˇic´ S (2006) A semantic web GIS based emergency management system. In: Cruz I, Decker S, Allemanget D, Preist C, Schwabe D, Mika P, Uschold M, Aroyo LM (eds) The semantic web—ISWC 2006. Springer, Heidelberg, pp 959–966 Tangtham N (1992) Principles and application of watershed classification in Thailand. Asian Institute of Technology, Bangkok Thouret J, Lavigne F (2005) Hazards and risks at Gunung Merapi, Central Java: a case study. In: Gupta A (ed) The physical geography of Southeast Asia. Oxford University Press, Oxford, pp 275–299 Titov VV, Gonzales FI, Bernard EN, Eble MC, Mofjeld HO, Newman JC, Venturato AJ (2005) Real-time tsunami forecasting: challenges and solutions. Nat Hazards 35:41–58 Tivov VV, Gonzales FI (1997) Implementation and testing of the method of splitting tsunami (MOST) model. National Oceanic and Atmospheric Administration. www.pmel.noaa.gov/pubs/PDF/tito1927/ tito1927.pdf. Accessed 5 Nov 2010 Tsou M-H, Sun C-H (2007) Mobile GIServices applications in disaster management. In: Drummond J, Billen R, Joao E, Forrest D (eds) Dynamic and mobile GIS: investigating change in space and time. CRC Press, London, pp 213–236 UNESCO (United Nations Educational Scientific and Cultural Organization) (2006) Indian Ocean tsunami warning and mitigation system IOTWS—Implementation plan. Third session of the intergovernmental coordination group for the Indian Ocean tsunami warning and mitigation system (ICG/IOTWS-III). http://ioc3.unesco.org/icg-iii/documents/IOC_TS_71web.pdf. Accessed 5 Nov 2010 UNESCAP (United Nation Economic and Social Commission for Asia and the Pacific) (2009) Annual core indicators online database. http://www.unescap.org/stat/data/swweb/DataExplorer.aspx. Accessed 5 Nov 2010 UNISDR (United Nationa International Strategy for Disaster Reduction) (2009) CRED disaster figures Geneva. http://www.unisdr.org/eng/media-room/facts-sheets/2008-disasters-in-numbers-ISDR-CRED. pdf. Accessed 5 Nov 2010 UNOCHA (United Nation Office for the Coordination of Humanitarian Affairs) (2007) Hazard maps. http:// ochaonline.un.org/roap/MapCentre/HazardMaps/tabid/3725/language/en-US/Default.aspx. Accessed 5 Nov 2010 van Westen CJ, van Asch TWJ, Soesters R (2006) Landslide hazard and risk zonation—why is it still so difficult? Bull Eng Geol Environ 65:167–184 Vigny C, Simons WJF, Abu S, Bamphenyu R, Satirapod C, Choosakul N, Subarya C, Socquet A, Omar K, Abidin HZ, Ambrosius BAC (2005) Insight into the 2004 Sumatra-Andaman earthquake from GPS measurements in Southeast Asia. Nature 436:201–206 Visser U (2004) Terminological representation and reasoning, semantic translation. In: Carbonelle JG, Siekmann J (eds) Intelligent information integration for the semantic web. Springer, Berlin, pp 141–150 Waugh WL, Tierney K (eds) (2007) Emergency management: principles and practice for local government, 2nd edn. ICMA Press, Washington, DC Wiegand N, Garcia C (2007) A task-based ontology approach to automate geospatial data retrieval. Trans GIS 11:355–376 Wisner B (1978) An appeal for a significantly comparative method in disaster research. Disaster 2:80–82 Worboys MF, Duckham M (2004) GIS: a computing perspective, 2nd edn. CRC Press, Boca Raton
123
616
Nat Hazards (2011) 59:597–616
Xu W, Zlantanova S (2007) Ontologies for disaster management response. In: Li J, Zlatanova S, Fabbri A (eds) Geomatics solutions for disaster management. Springer, Berlin, pp 185–200 Yumuang S (2006) 2001 Debris flow and debris flood in Nam Ko area, Phetchabun province, central Thailand. Environ Geol 51:545–564 Zerger A, Wealands S (2004) Beyond modelling: linking models with GIS for flood risk management. Nat Hazards 23:191–208 Zhang A, Qi Q, Jiang L (2007) GeoRSS based emergency response information sharing and visualization. Third international conference on semantics, knowledge, and grid (SKG2007), Xi’an, China, pp 596–597
123