Igi Global Proof

IJAGR Editorial Board Editor-in-Chief: Donald Patrick Albert, Sam Houston State U., USA

Associate Editors: Jonathan Comer, Oklahoma State U., USA Thomas Crawford, East Carolina U., USA G. Rebecca Dobbs, U. of North Carolina - Chapel Hill, USA Sonya Glavac, U. of New England, Australia Carol Hanchette, U. of Louisville, USA Tony Hernandez, Ryerson U., Canada Jay Lee, Kent State U., USA Shuaib Lwasa, Makerere U., Uganda John Strait, Sam Houston State U., USA Fahui Wang, Louisiana State U., USA David Wong, George Mason U., USA

IGI Editorial: Heather A. Probst, Senior Editorial Director Jamie M. Wilson, Director of Journal Publications Chris Hrobak, Journal Production Manager Christen Croley, Journal Production Assistant

International Editorial Review Board:

Bhuiyan M. Alam, The U. of Toledo, USA David Martin, U. of Southhampton, UK Badri Basnet, The U. of Southern Queensland, Australia Luke Marzen, Auburn U., USA Rick Bunch, U. of North Carolina - Greensboro, USA Darrel McDonald, Stephen F. Austin State U., USA Ed Cloutis, U. of Winnipeg, Canada Ian Meiklejohn, Rhodes U., South Africa Kelley Crews, U. of Texas at Austin, USA Joseph Messina, Michigan State U., USA Michael DeMers, New Mexico State U., USA William A. Morris, McMaster U., Canada Steven Fleming, United States Military Academy, USA Petri Pellikka, U. of Helsinki, Finland Doug Gamble, U. of IGINorth Carolina GLOBAL - Wilmington, USA François PROOF Pinet, Cemagref - Clermont Ferrand, France Gang Gong, Sam Houston State U., USA Wei Song, U. of Louisville, USA Carlos Granell, European Commission, Italy Brad Watkins, U. of Central Oklahoma, USA William Graves, U. of North Carolina - Charlotte, USA Dion Wiseman, Brandon U., Canada Bin Jiang, U. of Gävle, Sweden Zengwang Xu, Brown U., USA C. Peter Keller, U. of Victoria, Canada Xinyue Ye, Bowling Green State U., USA C. Gichana Manyara, Radford U., USA

IGIP IGI Publishing www.igi-global.com CALL FOR ARTICLES International Journal of Applied Geospatial Research

An official publication of the Information Resources Management Association

The Editor-in-Chief of the International Journal of Applied Geospatial Research (IJAGR) invites authors to submit manuscripts for consideration in this scholarly journal.

Mission The International Journal of Applied Geospatial Research (IJAGR) publishes research that exemplifies the usage of geographic information science and technology (GIS&T) to explore and resolve geographical issues from various application domains within the social and/or physical sciences. IJAGR is designed to provide planners and policy analysts, practitioners, academicians, and others using GIS&T useful studies that might support decision-making activities.

Topics of Interest: • Biogeography • Business and marketing geography • Climatology • Economic geography • Geography of crime • Geomorphology • Historical geographyIGI GLOBAL PROOFISSN 1947-9654 • Medical geography eISSN 1947-9662 Published quarterly • Military geography • Natural hazards • Political geography • Population geography • Soil geography • Tourism geography • Transportation geography • Other geographic subfields All submissions should be e-mailed to: Donald P. Albert, Editor-in-Chief [email protected]

Ideas for Special Theme Issues may be submitted to the Editor-in-Chief.

Please recommend this publication to your librarian. For a convenient easy-to-use library recommendation form, please visit: http://www.igi-global.com/ijagr. International Journal of Applied Geospatial Research

April-June 2012, Vol. 3, No. 2

Table of Contents Editorial Preface i Free and Open Source Software: Development and Utilization Donald P. Albert, Sam Houston State University, USA Research Articles

1 To Emphasize Openness Ken Hartness, Sam Houston State University, USA

6 Open Source Based Deployment of Environmental Data into Geospatial Information Infrastructures José Gil, Institute of New Imaging Technologies, Universitat Jaume I, Spain Laura Díaz, Institute of New Imaging Technologies, Universitat Jaume I, Spain Carlos Granell, Institute for Environment and Sustainability, Italy Joaquín Huerta, Institute of New Imaging Technologies, Universitat Jaume I, Spain

24 A Review of Geospatial Information Technology for Natural Disaster Management in Developing Countries Sam Herold, University of Ottawa, Canada Michael C. Sawada, University of Ottawa, Canada Reports IGI GLOBAL PROOF

63 Embracing Geographic Analysis Beyond Geography: Harvard’s Center for Geographic Analysis Enters 5th Year Weihe (Wendy) Guan, Harvard University, USA Peter K. Bol, Harvard University, USA

72 A Reflection on the PhD Program in Spatially Integrated Social Science at the University of Toledo Bhuiyan Monwar Alam,, The University of Toledo, USA Jeanette Eckert, The University of Toledo, USA Peter S. Lindquist, The University of Toledo, USA i

Editorial Preface Free and Open Source Software: Development and Utilization

Donald P. Albert, Sam Houston State University, USA

This issue of the International Journal of Ap- by Wendy Guan and Peter Bol. The CGA devel- plied Geospatial Research is dedicated to free oped WorldMap, a geoportal which conforms and open software (FOSS) including its develop- to Open Geospatial Consortium standards, and ment and utilization. Ken Hartness begins with therefore fits well into the scope of this IJAGR an historical overview and current assessment FOSS theme. of open sourceIGI software withGLOBAL a piece titled “To PROOF Emphasize Openness.” Next José Gil, Laura Díaz, Carlos Granell1 and Joaquín Huerta from Donald P. Albert the Institute of New Imaging Technologies at Editor-in-Chief the Universitat Jaume I (Spain) discuss the de- IJAGR ployment of environmental data into geospatial information structures. Sam Herold and Michael Sawada offer an extensive review of free and ENDNOTE open source software (FOSS) for geospatial information technology (GIT). Their review of FOSS is framed within context of natural 1 Dr. Carlos Granell is currently affiliated with disaster management in developing countries. the European Commission, Joint Research Completing this issue is a status report on Har- Centre, Institute for Environment and vard’s Center for Geographic Analysis (CGA) Sustainability, Spatial Data Infrastructures Unit, Ispra, Italy.

Donald Patrick Albert is a professor of geography in the Department of Geography and Geol- ogy at Sam Houston State University (Texas, USA). He earned geography degrees from Salem State College (BS), Appalachian State University (MA), and the University of North Carolina at Chapel Hill (PhD). International Journal of Applied Geospatial Research, 3(2), 1-5, April-June 2012 1

To Emphasize Openness

Ken Hartness, Sam Houston State University, USA

ABSTRACT

Although open source software has existed, in a sense, throughout the history of computing, it has only more recently become recognized as a valid means of producing professional-quality software. Although primarily conceived as a zero-cost alternative to commercial software, open source software also supports customization and verification as a result of the software being available to all users in human-readable form. The availability of free software supports both researchers with limited budgets and those who seek to confirm the findings of researchers or use similar methods in related research.

Keywords: Commercial Software, Free Software Foundation, Open Science, Open Source, Openness

TO EMPHASIZEIGI OPENNESS GLOBALlike IBM, PROOF are now embracing the open source design methodology and releasing professional- In the golden age of commercial software, it quality software as “open source.” Schools and is not unusual to pay between a hundred and individuals on tight budgets are able to legally a thousand dollars for popular software. More copy productive software tools at no charge. specialized software can be quite a bit more. Although the phrase “open source” is often These prices do not entitle the customer to used as synonymous with the word “free,” the customization of the product to their needs and phrase, itself, actually conveys the idea that any current operating procedures. Custom software user of a program should be able to examine developed just for one customer is even more the inner workings of the software and modify expensive. On the other hand, computer owners it to better suit the user’s needs. This quality of can download a version of the Linux operating openness is valuable to auditors and scientists system, legally, for no charge. OpenOffice.org who seek to validate the results produced by a is a suite of tools similar to those in commer- piece of software. Organizations and individuals cially available office packages, including a needing customized software can use related word processor and spreadsheet tool. While the open source software as a starting point instead concept of open source software is not new, the of paying programmers to create the software more recent success and dependability of open from scratch. Although the phrase “open source” source products has forced companies to take only requires that the software be available in a them seriously. Companies that previously were form that can be read, customized, and extended jealously protective of their software secrets, by computer programmers, globally open software is often available free of charge or for a nominal fee to cover creation and distribution DOI: 10.4018/jagr.2012040101 of a DVD-ROM.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 2 International Journal of Applied Geospatial Research, 3(2), 1-5, April-June 2012

The open source community makes a professional to modify an existing open source distinction between “free” software and other program rather than develop a complete custom forms of “open source” software (Scacchi, application from scratch should greatly reduce 2007). The Free Software Foundation consid- the cost of development. Programmers can add ers software to be free if its source code, the a graphical visualization package to a statisti- human-readable description of its nature and cal package designed to generate text reports. capabilities, is freely available to all users and is One could modify the GPS software for a cell only used to create free software. Note that their phone to take a picture and save the current GNU General Public License (GPL) does not location and picture for immediate or even- object to the software being sold commercially tual transmission to a central database server. or otherwise incorporated into a commercial A team of researchers working on a shared product. However, any product incorporating document at different locations might have an GPL software must, itself, fall under the GPL; open source word processor modified to keep as such, the creators of the commercial product them up-to-date on contributions made by other must make the source code available to any and team members. Rather than facing the difficult all users and cannot tell those users what they task of determining how to create a statistical can and cannot do with the software with one package, GPS logger, or word processor, one exception; they cannot violate the GPL and only has to discover how to interface with the restrict the freedom of others (Free Software existing software. This is sometimes easier said Foundation, 2007). Therefore, it is perfectly than done, but clarity is important in a software acceptable for a company to sell free software development culture where the participants may for $1,000 and equally acceptable for one of rarely get together in a meeting room. In some its customers to post the source code on their cases, the software comes already designed to web page. Naturally, this tends to limit the ex- support additional “plug-ins,” greatly easing pense of a commercial product under the GPL, the process of customization. although manyIGI people areGLOBAL willing to pay for PROOF convenient installation and support. Other open source licenses allow the open source software EXAMPLES to be combined with commercially-developed, Open source software has existed in one form or proprietary software; although the open source another for as long as there have been comput- software remains available to its users, the other ers. Many people, however, view the creation elements of the product can remain secret. The of the Linux operating system and its rise as a Free Software Foundation is part of a social rival for more commercial operating systems movement that embraces the free exchange as the beginning of an open source movement of ideas and insists upon this free exchange (Moody, 2001). Richard Stallings started the in GPL applications. Open source is a soft- GNU project in the hope of one day creating ware design methodology where, potentially, an open source version of the UNIX® operat- any user can make a contribution to the final ing system. He was able to write many of the product. Developers working under a license tools familiar to users of UNIX®, but he and his other than the GPL may occasionally choose to supporters were looking at a lengthy develop- compromise and combine open source software ment effort to create an operating system kernel with proprietary software in an effort to create that would allow their GNU operating system a quality product. “More simply, free software to stand on its own. Linus Torvalds, in 1991, is always OSS [open source software], but OSS received Andrew S. Tanenbaum’s source code is not always free” (Scacchi, 2007, p. 459). for MINIX (Tanenbaum, 1997), a relatively Although open source software is usually simple operating system that he created for available at little or no cost, even if it were not, teaching the internal operations of operating the fact that an organization can ask a software

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 1-5, April-June 2012 3 systems. MINIX inspired Torvalds to create a OpenFTS (Bartunov et al., 2010) supports professional-quality operating system kernel. full text search of documents. It builds an index He made the source code freely available and for the documents and allows the user to obtain invited others to make contributions. As the a list of possibly relevant documents that don’t Linus Torvalds kernel became ready for public otherwise meet the search criteria by calculat- use before the GNU kernel, the GNU project has ing a proximity value indicating how close one incorporated it into the GNU/Linux operating document is to another. system. The Linux kernel has had thousands PostGIS (http://postgis.refractions.net/) of contributors, but oversight from Torvalds adds support for spatial locations for the de- and the commitment of programmers who are velopment of databases that support geographic also users of the operating system has created information systems. PostGIS is currently being a quality product. Combined with tools from used in a number of projects, including Sistema the GNU project and increasingly sophisticated de Información Territorial Estatal en Línea for open source applications, Linux-based operat- sharing positional data related to health, security, ing systems have expanded their user base from mobility, etc., among Mexican governmental hobbyists to become a valid choice for corpora- agencies and EcoTrust, an organization that tions concerned about security and reliability. seeks to increase ecological awareness with The common computer user has become supporting data concerning the Pacific Rim more likely to consider a Linux-based operat- Bio-region. ing system as open source versions of familiar software products have been created. One of these is OpenOffice.org, a suite of tools created OPEN SCIENCE as a single integrated package (Oracle, 2009). Many papers on scientific research are col- The suite includes tools for word processing, lected into journals or conference proceedings spreadsheets, simple databases, and drawing where the full text of the material is only avail- charts or manipulatingIGI images. GLOBAL PROOF able to those who subscribe to the journal or Powerful tools for creating web serv- purchase the proceedings. A publisher must ers and database servers have been released make a living and cover printing costs (or the as open source. These include the Apache overhead of electronic storage, indexing and HTTP Server (Apache Software Foundation, web server maintenance, in the case of elec- 2009) for connecting your computer and its tronic journals). However, the free exchange web pages to the Internet, Samba (2010) for of information possible over the Internet has making operating systems like GNU/Linux prompted some researchers to consider whether work with sophisticated file sharing and re- another means of dissemination could be used. lated networking protocols used by Microsoft The use of computers for storing and analyzing Windows®, MySQL (Oracle, 2010), a popular data suggests that researchers do not have to database management system for managing the be content with reporting a brief summary of information utilized by dynamic web pages, their results. Gezelter (2009) expresses concern and PostgreSQL (http://www.postgresql.org/ that verifiability and falsifiability need to be an about/), an object-relational database manage- important part of scientific research; journal ment system that supports many features usually articles, alone, often do not convey enough only found in powerful commercial database methodology to support these qualities. The management systems. Thanks to the ability to phrase “open science” refers to a free exchange customize PostgreSQL’s handling of custom of ideas, methodologies, and data. Where data data objects, it has been selected as a founda- is accessible by software means or produced tion for a number of other open source projects, by software, researchers should freely share including OpenFTS and PostGIS. the data and/or means of production.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 4 International Journal of Applied Geospatial Research, 3(2), 1-5, April-June 2012

Large collections of freely accessible data available tools to support collaboration. And, are starting to form (Schroeder, 2010b). The maybe, our computers won’t crash as often. web site data.gov provides data sets collected by the executive branch of the United States government. An application called GeoViewer ACKNOWLEDGMENT allows GIS data from a variety of sources to UNIX is a registered trademark of the Open determine if it meets a researcher’s needs. Group in the United States and other countries. Mesirov (2010) proposes a “reproducible Windows is a registered trademark of Microsoft research environment,” a package that contains Corporation in the United States and other data and software, or links to long-term copies countries. accessible over the Internet, and a document preparation program that allows for easy incorporation of different analyses of the data into a document. The package would automati- REFERENCES cally track the user responsible for collection, Apache Software Foundation. (2009). Welcome! The analysis, and reporting of different aspects of Apache HTTP server project. Retrieved September the data. Of course, the creators of proprietary 16, 2010, from http://httpd.apache.org software generally don’t want researchers freely Bartunov, O., Demetriou, N., Sigaev, T., & Wick- copying and sharing their software. strom, D. (2010). OpenFTS Primer. Retrieved Sep- Schroeder (2010a) believes that open tember 16, 2010, from http://openfts.sourceforge. source software will come to be the software net/primer.html of choice for scientific research. In addition to Free Software Foundation. (2007). The GNU General the importance of an open sharing of knowledge Public License – GNU Project. Retrieved September and a need for verifiability in the scientific 14, 2010, from http://www.gnu.org/licenses/gpl.html community, he makes the point that scientific IGI GLOBALGezelter, PROOF D. (2009). Being scientific: Fasifiability, software must evolve to incorporate new meth- verifiability, empirical tests, and reproducibility. odologies. Open source software projects are Retrieved September 24, 2010, from http://www. already designed for agile development to meet openscience.org/blog/ changing requirements of its users; typically, Mesirov, J. (2010). Accessible reproducible research. an open source software project includes some Science, ▪▪▪, 22. of the very users whose needs are changing on its development team. Moody, G. (2001). Rebel code: Inside Linux and the Open Source revolution. Cambridge, UK: Perseus. Oracle. (2009). Why: Why OpenOffice.org. Retrieved CONCLUSION September 16, 2010, from http://why.openoffice.org

Open source software has a lot to offer. Be- Oracle. (2010). MySQL: The world’s most popular open source database. Retrieved September 16, cause it is typically available at little or no 2010, from http://www.mysql.com cost, third world countries are legally using productivity software and students are able to Samba. (2010). Opening Windows to a wider world. use professional quality products on their own Retrieved September 16, 2010, from http://www. samba.org laptop that might, otherwise, be only available in a lab, assuming the lab has the budget Scacchi, W. (2007). Free/open source software de- to purchase the software. Open data and open velopment. In Proceedings of the 6th Joint Meeting of the European Software Engineering Conference source software will enhance the exchange of and the ACM SIGSOFT Symposium on the Founda- information and vastly improve the ability of tions of Software Engineering. New York, NY: ACM. researchers to compare and incorporate each others’ work. Work will be enhanced by freely

Schroeder, W. (2010a). Why open source will rule Schroeder, W. (2010b). Why open source will rule scientific computing [Web log post]. Retrieved scientific computing (Part 2) [Web log post]. Re- August 23, 2010, from http://www.kitware.com/ trieved August 23, 2010, from http://www.kitware. blog/home/post/3 com/blog/home/post/6 Tanenbaum, A. (1997). Operating systems: Design and implementation (2nd ed.). Upper Saddle River, NJ: Prentice Hall.

Ken Hartness has a PhD from the University of North Texas (2004) and a Master of Science from Texas A&M University. All college degrees are in computer science. He has taught for ten years in higher education, not including four years as a teaching fellow while pursuing graduate degrees. Dr. Hartness primarily teaches Artificial Intelligence, programming in Java or C++ COBOL, and programming and advanced information systems processing.

IGI GLOBAL PROOF

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 6 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

Open Source Based Deployment of Environmental Data into Geospatial Information Infrastructures

José Gil, Institute of New Imaging Technologies, Universitat Jaume I, Spain Laura Díaz, Institute of New Imaging Technologies, Universitat Jaume I, Spain Carlos Granell, Institute for Environment and Sustainability, Italy Joaquín Huerta, Institute of New Imaging Technologies, Universitat Jaume I, Spain

ABSTRACTIGI GLOBAL PROOF Today, scientists use local and closed geospatial solutions to run their models and store their results. This may limit their ability to share their models, and results with other interested colleagues. This scenario is changing with the advent of new factors such as the rapid growth and rise of open source projects, or new paradigms promoted by government organizations to manage environmental data, such as Infrastructure for Spatial Information in the European Community (INSPIRE) directive, or the massive use of Web 2.0 techniques where users are looking for applications with a high degree of collaboration, interactiveness, and multimedia effects. Many authors address the versatility of Spatial Data Infrastructures where resources are shared and accessed via standard service according to complex specifications. In this context, the authors point out the need to merge the traditional building and maintenance of these infrastructures, driven by official providers, with these more participative methodologies where users can participate in creating and integrating information. It seems necessary to develop new geospatial tools which integrate these new trends. This paper proposes a unified solution offering to the scientific field an open development framework, based on standards and philosophies focused on new technologies and scientific needs.

Keywords: Capacity Building, Data Integration, Data Sharing, Geospatial Services Deployment, INSPIRE, Interoperability, Open Source (OS), Rich Internet Applications (RIA), Spatial Data Infrastructure (SDI)

INTRODUCTION or near-surface of our planet (Goodchild, 2008). The Earth is a multi-dimensional system made One of the challenges we are facing today is to of complex interactions highly interconnected better understand the processes that occur over and continuously evolving at many spatial and temporal scales (GEO Secretariat, 2007). This means that to understand such interactions, DOI: 10.4018/jagr.2012040102

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012 7 scientists and environmental experts need to promote interoperability, standards specifica- collect and integrate different data sets referred tions, and seamless integration of services and to physical and biological aspects of the Earth information to improve this accessibility to geo- (Giuliani et al., 2011). In addition, these data spatial resources (see next sections). However, sets are often georeferenced, that is, they de- the participation of users in these infrastructures scribe physical phenomena tied to a concrete has been delayed due to the complexity of cur- geographical location in terms of points, lines rent methodologies to participate and integrate and sets of polygons, and we refer to them as new resources; the necessity of skills about how geospatial data. to incorporate real time data with historical, the Scientists and environmental experts need use of standard services specification to deploy to manage vast amounts of these geospatial resources and promote interoperable resources, data to turn them into meaningful information. the integration of data from different domains, The applicability of complex environmental or simply looking for these data sources have processes and models has been the focus of become tedious tasks in current developments intense research in this domain (Poch, Co- (Yang & Raskin, 2009). This makes scientists mas, Rodríguez-Roda, Sánchez-Marrè, & continue using traditional methods like local Cortés, 2004; Minsker et al., 2006; Reichert processing. & Schuwirth, 2010). For instance, models for The aim of this article is to present an open reduction of energy consumption, computation source-based framework that enables both: of atmospheric emissions, and simulation of standard access and processing of data, and also forest fire require multidisciplinary teams of the deployment of new content into geospatial geospatial technologists and environmental information infrastructures. The processing experts. Such models, however, have been and access of environmental data sets and the traditionally performed using multiple, iso- deployment and sharing of processing results lated desktop geospatial tools: a specialist in will be addressed in this paper. All this, in the Digital Elevation ModelsIGI might GLOBAL perform slope simplest PROOFway possible, through an immersive computations while other colleague might ap- and user-friendly interface, based entirely on ply remote sensing techniques over satellite geospatial open source technologies. imaginary to produce a snow cover product. The rest of this article is structured as fol- In this simple example, various GIS software lows. The following three sections outline the packages could be involved in the activities of basic concepts underlying the topics tackled a single environmental model. However, this in this article: architectural styles for informa- paradigm of closed solutions, which operates tion infrastructures, concrete realizations of with local data sets, provides a way of hiding geospatial information infrastructures, and the shared results and collaborative research, and concept of open source applied to geospatial more often, hindering further analysis tasks. technologies. Given the pieces of the puzzle, Better decisions and concrete actions at we present a toolbox that tries to connect all local and global scale to underpin sustainable pieces to facilitate the sharing, deployment and development can be taken only based on pre- publication of new geospatial and environmen- cise knowledge of our environment (Lannotta, tal data sets into existing geospatial information 2007). However, geospatial datasets, and other infrastructures. We discuss the applicability of resources are often isolated, hidden to potential our framework through a use case in the forest interesting stakeholders that could benefit from fire domain. The paper concludes by looking to them. Therefore, the ease of access and retrieval futures challenges in automatic deployment of of geospatial and environmental data is a key user-generated content and identifies a number aspect for information infrastructures (Yang, of issues that are in need of further research. Raskin, Goodchild, & Gahegan, 2010). There are consensus-based initiatives which aim is to

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 8 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

DESIGNING INFORMATION At the time of implementation SOA-based INFRASTRUCTURES services must make use of concrete languages and protocols. Here is where web service tech- Geospatial and environmental research is a nology gains importance because it increasingly field that involves multidisciplinary teams of is becoming the choice to implement SOA-based scientists and technologists working together. applications. Web services (Papazoglou, 2008) The application of environmental models may are, by definition, loosely coupled independent span from local to global scope, what undeniably units and are well described (interface descrip- requires suitable information infrastructures to tion contains functional properties), thereby share and do research on the web (Hey & Tre- promoting one of the goals of SOA: enabling fethen, 2005). This complex scenario demands interoperability or the ability of services to the use of new architectural styles which oppose interact with minimal knowledge of the under- centralized, isolated solutions, and instead sup- lying structure of other services (Goodchild, port distributed processing capabilities and re- Egenhofer, Fegeas, & Kottman, 1999). Interop- mote communications, necessary ingredients to erability is achieved (or optimized) by using successful collaborative and multidisciplinary standard interfaces. Web service technology research. Service-Oriented Architectures (SOA) includes various standards such as Web Service and Resource-Oriented Architectures (ROA) are Description Language (WSDL) for the descrip- currently the architectural styles adopted in the tion of service interfaces, Universal Description, development of collaborative, distributed Web Discovery and Integration registry (UDDI) systems and applications. for their advertisement and discovery, and SOA is an architectural style to design ap- Simple Object Application Protocol (SOAP) plications based on a collection of best practices, that enables communication among services principles, interfaces, and patterns related to the (Curbera, Duftler, Khalaf, Nagy, Mukhi, & central concept of service (Bell, 2008). In SOA, Weerawarana, 2002). services play IGIa key role andGLOBAL become the basic PROOFAs opposed to SOA, focused on distributed computing unit to support development and capabilities and services, ROA is an archi- composition of larger, more complex services, tectural style devoted to manage distributed, which in turn can be used to create flexible, heterogeneous domain resources. In SOA client ad hoc and dynamic applications. The main applications interacts with a distributed Web design principle behind SOA is that a service applications through delegation, that is, by is a standards-based, loosely-coupled unit specifying the desired capability to a service composed of a service interface and a service component instead of directly acting on the re- implementation. The service interface describes sources themselves (Mazzetti, Nativi, & Caron, the functional capabilities of a service. The 2009). In ROA, through, client applications service implementation codifies what a service interact directly with the exposed resources. should execute. This principle provides a clean The main constraints behind ROA-based separation of concerns especially between applications are the set of architectural prin- service interfaces (what services offer to the ciples known as Representational State Transfer public community) and internal implementa- (REST) (Fielding, 2000), namely: tions (how services work). Essentially SOA introduces a new philosophy for building • Resources should be identified properly distributed applications, where services can be using global identifiers (URI). However, discovered, aggregated, published, reused, and REST proposes to use URIs not only for invoked at the interface level, independently identification but also for physical access of the specific technology used internally to to the actual resource representation. One implement each service. important feature is that each resource must

be addressable via a HTTP (Hypertext GEOSPATIAL INFORMATION Transfer Protocol) URI. INFRASTRUCTURES • Uniform interface through the use of HTTP as the unique application-level protocol. Several administrations are recognizing the role HTTP has a small, fixes of operational of geospatial data in e-commerce, sustainable methods with specific purpose and mean- development and government activities, and ing. For instance, the GET method is for the ability to be efficiently coordinated and retrieving representations of target re- managed for the interest of public in general sources, the POST method for creating new (Rajabifard, Binns, Masser, & Williamson, resources, PUT for updating resources, and 2006). In consequence, the notion of Spatial finally, the DELETE method to eliminate Data Infrastructures (SDI) refers to the spe- a given resource. cialization of information infrastructures for • Resources are manipulated through the geospatial sciences, allowing discovering, representations since clients and servers access, publishing, sharing, maintaining, and exchange self-descriptive messages each integrating geospatial data (Nebert, 2004). another. Many initiatives at different administra- • Interaction stateless since servers only tive levels have been recently flourishing to record and manage the state of their re- promote the creation of SDIs (Masser, 2005). sources they exposes. Client sessions are These initiatives coordinate actions and poli- not maintained in the server. cies that encourage awareness of institutional • Hypermedia as the engine of application agreements, common standards and effective state, that is the application state is built mechanisms for data access, harmonization, following hyperlinks according to the standardization, and the development of navigation paradigm so that clients can interoperable geospatial data and technolo- discover related resources from a given one. gies to support decision making for multiple IGI GLOBALpurposes. PROOF One of these initiatives to eliminate As in the case of SOA-based services, jurisdictional, cultural or domain boundaries is ROA-based applications can be implemented the Infrastructure for Spatial Information in the in many different ways, such as RESTful web European Community (INSPIRE) Directive of services (Richardson & Ruby, 2007). Designing the European Commission (European Parlia- RESTful web services following the REST con- ment and Council, 2007). INSPIRE Directive straints described above has gained popularity establishes a legal framework for the creation within the Earth science, mobile and location and implementation of the European Commu- based services community (Mazzetti, Nativi, nity SDI, forcing organizations to publish their & Caron, 2009). geospatial resources (focusing on the environ- The selection of one or another architectural mental domain) under standard data services style depends often on the context of the target and by promoting the adaptation of the SOA application. Both approaches should not be dis- paradigm and geospatial web service standards. joint but complementary. In this paper, we take In essence, geospatial web services al- the complementary strategy for the proposed low users to access, manage, and process framework implementation that makes use of geospatial data in a service-oriented manner both architectural patterns to communicate with (Granell, Gould, & Esbri, 2008). The demand remote geospatial services. Accessing to com- for interoperability has boosted the develop- mon geospatial services, as we describe in the ment of standards and tools to facilitate data following, are built upon a SOA approach, while transformation and integration, mostly in terms recent open source implementations already of standard service interfaces specified by Open expose RESTful interfaces to manipulate and Geospatial Consortium (OGC1). The Web Map deploy remote geospatial resources. Service (WMS) (OGC, 2006), the Web Feature

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 10 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

Service (WFS) (OGC, 2005), the Web Cover- Bocher, 2009), these can be a serious solution age Service (WCS) (OGC, 2008), the Sensor to address the lack of mechanisms for democ- Observation Services (SOS) (OGC, 2007a) and ratizing the capacity building of geospatial the Web Processing Services (WPS) (OGC, information infrastructures. 2007b) are some prominent examples of OGC A key reference in geospatial technologies’ interfaces for geospatial services with a clear state of the art is the world conference for Free applicability to environmental applications and Open Source Software for Geospatial (FOS- (Granell, Díaz, & Gould, 2010). For instance, S4G2). One can find the latest innovative tools a WMS could be used to provide a map of the and developments in the geospatial landscape, protected areas of a given region, a SOS may such as OpenGeo Suite or GeoNode, which are serve detailed meteorological observations from examples of open source technology related to near weather stations, and a WCS may offer those proposed in this paper. OpenGeo Suite satellite imagery of the area of interests and its (Pumphrey, 2009) is an open source project that surroundings. The central building-blocks for brings together different technologies (PostGIS, data, as well as service discovery, are provided GeoServer, GeoWebCache, GeoExt, OpenLay- by the Catalogue Services for the Web (CSW) ers) to develop an architecture based on OGC (OGC, 2007c). The CSW provides one access standards. The main objective is that users can point to users that publish and search environ- easily build a platform that provides maps and mental data. data via web applications, mobile and desktop In the proposed framework implementa- clients. GeoNode provides a new SDI solution tion, we will use a broad list of OGC services by adding a new application discovery layer (WMS, WFS, WCS, WPS, and CSW) to let so that scientists can easily discover the data experts to perform not only retrieval but also services. The result is a platform that facilitates deployment of geospatial data sets. the creation, sharing and collaborative use of geospatial data (Benthall & Gill, 2010). IGI GLOBAL PROOFMerging our work with other domains not GEOSPATIAL OPEN strictly in touch with geospatial domain, we must SOURCE TECHNOLOGIES review the take off of many web applications that following the Web 2.0 philosophy have The use of geospatial applications for environ- created an environment that brings users to mental studies has experienced a significant participate very actively and add new content boom in the last years (Minsker et al., 2006; to be accessed and shared by many users. This Pezzoli, Marciano, & Robertus, 2006; Morisette way of content sharing brings new forms of et al., 2009). Its development has led to progress study. Not only with the incorporation of the in making decisions in real time (Ramaprivan, social factor but with a new relationship with the 2008), it has created new relationships between user. Applications are required to have higher different areas of knowledge (Ball et al., 2008) degree of interaction and attractive interfaces and has developed new processes to understand that facilitate complex tasks. Solutions such the behavior of natural phenomena (Díaz, as RIA (Rich Internet Applications) can be a Granell, Gould, & Olaya, 2008). key element in future developments. In this For all these efforts to continue to improve, paper we describe how we propose to merge they need to be associated with technological these distinct disciplines. Having an underlying advances that gradually eliminate barriers that geospatial information infrastructure based on scientists found in their research. This requires standard services and components we add on advanced tools (processing, analysis and visu- top a framework composed by mechanisms and alization) and sophisticated information infra- Web 2.0 interfaces to assist users in not only structures that provide access to large volumes accessing but also deploying and sharing new of data. Given the wide array of open source geospatial resources in this SDI. technologies for geospatial data (Steiniger &

FRAMEWORK FOR technologies may lead to underuse such infor- DEPLOYMENT OF mation infrastructures for efficient analysis and ENVIRONMENTAL DATA decision-making tasks. This means scientists who produce interesting results through models In the previous section we have seen numer- and processes often use information for access, ous technological trends that can potentially not to publish their own data sets. This implies in become the basis of a set of solutions in multiple low rates of user participation in the process of scenarios. On one hand, common policies and injecting environmental datasets into informa- agreements in terms of the notion of SDI, un- tion infrastructures for collaborative research. derstood as geospatial data stores collected by For this reason, we present an open source government agencies and private institutions, framework based on INSPIRE architecture where data are published via open, standards- that has been extended with some modules based services, can be easily incorporated into for not only accessing but also deployment of scientific applications (Nerbert, 2004). The environmental data into geospatial information INSPIRE directive, which creates a collabora- infrastructures (like SDI). The main character- tive SDI, accessible, designed to facilitate data istics are the following: sharing and standardization of specifications (European Parliament and Council, 2007), re- • It allows the access and retrieval of envi- inforced by standardization organizations, such ronmental data by means of interoperable as the Open Geospatial Consortium (OGC) and OGC-based services), International Organization for Standardization • It allows the integration and processing of (ISO). On the other hand, the advent of recent new information by means of deployment Web 2.0 technologies to create interoperable mechanisms geospatial web portals (Yang, Li, Xiao, Raskin, • It allows the generation and publication & Rambacus, 2007;IGI Wilkins-Diehr, GLOBAL Gannon, of metadataPROOF descriptions in OGC-based Klimeck, Oster, & Pamidighantam, 2008) and catalogs for further search. that allows users to publish and access data in an easy way. The rise of open-source service- Figure 1 illustrates the conceptual archi- oriented geospatial environments and tools, tecture of the proposed framework. It basically provide a variety of solutions and knowledge follows the INSPIRE technical architecture, that that can be easily extrapolated or incorporated is, three-layer SOA: the Application layer (top) directly into new scientific applications (Pierce, contains client side services, application logic Fox, Choi, Guo, Gao, & Ma, 2009). and presentation modules; the Service layer Why then the diverse geospatial technolo- (middle) with server side services that use a gies, open source tools, and geospatial informa- standard interface for communication; and the tion infrastructures have not received the same Data layer (bottom). Grey boxes represents level interests in environmental applications the components contained within the Service as in others domains? After all, SDI and open Framework. In the following we outline the source geospatial technologies presents many main functionality of each layer. Next section benefits for environmental scientists and ex- will describe in detail each component in grey perts, since the intrinsic multidisciplinary aspect in Figure 1 in environmental research can be viewed as a source of expertise for collaborative applica- Application Layer tions between different domains (Westlund, 2010). Our contribution on this layer is to add the From the user’s perspective of environ- Service Deployer and Publisher that allows mental and geospatial sciences, the increasing user not only access resources but also deploy complexity of current methodologies, tools and new resources in the SDI and a user interface

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 12 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

Figure 1. Multi-layered architecture for the proposed framework

IGI GLOBAL PROOF that following web 2.0 principles offers a entry Service Layer point to the underlying SDI that allows users to participate in the building and maintenance The Service layer is responsible for mediating of the SDI. between the Application and the Data layer. Its Following the Model-View-Controller development focuses on the paradigm INSPIRE/ (MVC) pattern (Freeman, Sierra, & Bates, 2004) SDI/SOA to minimize problems of accessibility to build the client side, we find the Application and interoperability, so that wherever possible layer is divided into three levels according to requests for data on this layer are performed their proximity to the user. At the level closest using OGC standards. However, on certain to the user, we have the visual part of the ap- occasions such as the insertion of new data, plication, multimedia effects and interaction we need to use a RESTful approach. In this with the user. At the intermediate level, we layer reside the instances of the OGC services find the application logic associated with the in which we deploy the new content that users controller of the actions produced by the user. will provide using the framework interface. Ser- At the lower level we have client side’s services vice Framework is preconfigured to work with that allow communication with the geospatial existing instances of geospatial web services. services deployed at the Service layer, through OCG standards and RESTful technology. Data Layer In this part we find the final elements for data storage, such as databases or simply memory

Figure 2. Interfaces used to expose the components functionality

disk regions, where the Service layer can store mented components to be delivered as open their data. It is out of the scope of this paper to source projects as well. describe further this layer since it will be something managed by the OGC standard service Application Layer implementations that we use to deploy new data content and that reside in the service layer. The Application layer as we can see in both Fig- ures 2 and 3 is composed of several sub-layers IGI GLOBALand components. PROOF At the top of the Application OPEN SOURCE BASED Layer, we find the components that interact COMPONENTS directly with the user. Next we describe each of these components regarding the interfaces As we have seen throughout the paper, there and the technology used. are many open source projects in the geospatial domain and in particular many projects that User Interface, Map implement open source components according Viewer and Searcher to OGC standard specifications. We describe next the service framework components regard- Interface. The main purpose of the User Inter- ing the technology used to implement them. face is to facilitate the user interaction with For the Service Framework implementation we the application. The easiness and comfort have reuse some of these existing open source of use and the level of interaction are some components and also we have self-develop factors that determine the success or failure. some components to be offered through an The User Interface integrates the other two open source license. components that interact directly with the The rest of this section describes our solu- user: the Map Viewer and the Searcher. tion from two points of view: interfaces and The former component offers a common technologies used. Figure 2 shows the interfaces map-based interface for data visualization implemented by each component in the Service and exploration. The latter, the Searcher Framework. Figure 3 shows the technology component, follows an OpenSearch3 that has been used to implement each of these interface. This simple specification pro- components. Some of these technologies are poses a standard search interface, used in existing open source projects and self imple- most modern browsers to manage custom

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 14 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

Figure 3. Technology used to implement the components

searches. In addition, a geospatial extension In our work OpenLayers has been the chosen (OpenSearch-Geo) has been proposed that one. OpenLayers is a JavaScript client that of- adds geographic filtering capabilities and fers interesting features like very good perfor- support for geographic response formats, mance, the ability to consume many different obtaining then an optimal approach to geographic formats, and client side support search for geospatial data. of standards such as OGC Web Map Service Technology. As we shown in Figure 3, the inter- (WMS) and Web Feature Service (WFS). The face is implementedIGI usingGLOBAL RIA technology. communication PROOF between the interface and the It proposes a new evolution of the client map viewer component is done through calls server paradigm that allows executing a to specific methods (using the JavaFX Applet- portion of the application on the user’s StageExtension class) that are interpreted by local system. This reduces the communi- the JavaScript runtime engine in the browser. cation with the server and increases the Furthermore, OpenLayers has been the user’s autonomy (Weaver, Gao, Chin, & container extended with a new self developed Iverson, 2009). control called OpenSearch-geo control which has been integrated in OpenLayers for search- From among the wide range of valid options ing purposes (Fonts, Huerta, Díaz, & Granell, like FLEX, SilverLight or JavaFX, we have 2010). This new control in the client translates selected JavaFX because of its fully integration user queries to the standard OpenSearch proto- with Java which allows us to easily integrate col, using geospatial parameters and supported the components reused and developed in our formats, and offers user the retrieved documents. framework. JavaFX is structured according As in the previous case, the JavaFX Applet- to the classical concept of scene graph for the StageExtension class is the preferred interface. implementation of graphical applications, with a primary scene node on which iteratively hang Controller the rest of nodes. These nodes represent mainly Interface. Following the MVC pattern, the behavior animation, transition effects, visual logic of the application is located in the content and user interaction (Weaver, Gao, Controller component. This responds to Chin, & Iverson, 2009). events produced by the user, and invokes In the open source community there exist requests to the model by delegation to the several map viewers components available.

lower components of the Application Layer Currently versions 0.4 and 1.0.0 are sup- (Service Connector, Service Deployer and ported. This component is programmed Service Publisher). also under open source license. Technology. This component is developed using the Singleton pattern (Freeman, Service Deployer Sierra, & Bates, 2004) so that any instance requested on it to access the same content Until now the described components offered and execution. The interactions between accessing capabilities to resources deployed the view (user interface) and the controller in the infrastructure. The aim of the Service are performed via a new feature in JavaFX Deployer component is to assist the user to de- that synchronizes variables between the ploy resources as services in the infrastructure. user interface and the controller. Examples These resources can be either data resources or are the list of local resources loaded (lo- processes. In this paper we focus on the deploy- calFileResources) as a view update and ment of data resources. the method loadLocalFile() as an input to the controller (and therefore execution of Interface. The Data Wrapper component al- actions on the model). lows users to upload geospatial data to be deployed and delivered as standard OGC Service Connector data services, such as WMS, WFS and WCS. The Data Wrapper interface contains Interface. The Service Connector component in two methods: WrapFile and WrapData to the application layer contains components deploy data available in a File or included that implement the client side interfaces as a parameter in the request. of the standard services deployed in the Service layer.IGI For instance GLOBAL it contains the To deployPROOF data as an OGC Service, this component Web Processing Service Appli- component should connect to OGC Service cation Programming Interface (WPSAPI) instances available in the infrastructure. Some that implements the OGC WPS client of these OGC standards (such as WFS) include interface to be able to access and run distrib- transactional operations in their interface uted processing services. Another module specification. In this case, the resources can developed in the Service connector is the be deployed by implementing the transactional CSWAPI (Catalog Service for Web API). interface in the client side. This is the ideal Technology. Currently, Service Connector case, because the component implementation modules like WPSAPI and CSWAPI is independent of service implementations. In are included in a self–developed Java other cases like deploying data for visualiza- library. For instance WPSAPI offers a tion purposes as WMS, the specification does programmatic way to send and receive not consider transactional operations; therefore getCapabilities request using Java objects. this component has to deal with the concrete It implements objects to send and receive services implementation and their own proto- Process description requests to get the col and interface. Currently, in our prototype description of processes in order to see the implementation this is the case of the Data parameters and the way a process within a Wrapper component when deploying data in WPS has to be invoked. Finally, it permits WMS implementations using the Geoserver sending an execute request to invoke one RESTful API4. of the available processes and it is able to retrieve and parse the response in order Technology. The Data Wrapper component is a to extract the process results. The object self developed Java library in form of API model is scalable to support new versions.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 16 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

exposing the described interface. In order Technology. The Service Publisher is again a to deploy the data, we have first associated self-developed Java library delivered with this component with an instance of OGC open source license. The Service Publisher Data Service that implements WMS, WFS uses the Service Connector to perform the and WCS interfaces. This OGC Services two main steps of its workflow: first, uses implementation is an open source project the Service Connector to send GetCapabili- called Geoserver, which offers a RESTful ties request to the services in which we have API to deploy content. Geoserver is then deployed the resources and get a minimum the responsible for publishing data through set of metadata elements (available in the OGC interfaces, transforming the resources Capabilities document) and second, it uses processed in WMS for viewing and WFS the Service Connector CSWAPI to connect and WCS for download purposes. to the catalog(OGC CSW) and deploy this metadata using the transactional interface, Service Publisher which provides methods to insert, delete, and update records in the catalog. Service Publisher component is in charge of deploying data and process descriptions to Service Layer an open catalog, so they can be found in the infrastructure. As we based our infrastructure on INSPIRE technical guidelines, the service layer must Interface. Like the Data Wrapper, the Service provide services for discovery, viewing and Publisher needs to implement an interface downloading of data available in the infra- to deploy content in a specific Service structure. type. In this case the content is not data but metadataIGI and the service GLOBAL type is the Dis- Interface PROOF. The INSPIRE technical architec- covery Service Type, which recommended ture shows that the different service types interface according to INSPIRE is the OGC defined by the Directive are placed in the CSW specification. In this case CSW offers service layer resulting in the so-called Ser- a transactional operation in its interface to vice Network. Each type defines common deploy new metadata in the service, thus capabilities offered by a group of services. opposite to the Data Wrapper the interface Regarding this functionality, the INSPIRE implemented by Service Publisher is a Service Types we consider are the Discov- standard interface, the CSW-T. For these ery (INSPIRE Network Services Drafting transactional requests, we build and send Team, 2009b), View (INSPIRE Network a HTTP POST request where the body is a Services Drafting Team, 2008) and Down- standard metadata element included in the load (INSPIRE Network Services Drafting standardized XML structure according to Team, 2009a) service types. In addition we the OGC service. also consider the Processing type to add processing capabilities to the SDI (Díaz, On the other hand once deployed the meta- Granell, & Gould, 2008). data, it will be available for searching purposes through the CSW interface. Since the CS-W Since OGC standards are proven to be implementation of choice offers other inter- mature, they have been chosen to implement the faces, the deployed metadata will be available interfaces of the services available in the Service not only through the CSW but also through the layer as the INSPIRE directive recommends Open Search interface. in its implementation rules (see parenthesis content: such as OGC CSW (to implement Dis- covery Service type), OGC WMS (View Service

Types), OGC WFS, OGC WCS (Download we choose 52North WPS framework because it Service types) and OGC WPS (pointed out to be is an open source Java framework that enables useful to implement distributed environmental easily design, implementation and deployment applications (Granell, Díaz, & Gould, 2010). of WPS-based services.

Technology. There are many implementations of OGC specifications in the open source USE CASE community. GeoServer is the project of To evaluate our proposal and illustrate with a choice to deploy data for visualization and practical example the behavior of the described download purposes, that is, through WMS, components we present a particular use case WFS and WCS specifications. It offers focused on the evaluation of burned regions these interfaces and in its latest version on protected environmental areas in Valencia GeoServer RESTful API allows the Data (Spain). The analysis of fire impact requires a Wrapper component to deploy resources real time interaction between different datasets, and configure the services. processes and users. For these reasons it represents a perfect case study to look at the overall Geonetwork is another open source project functionality of the framework. broadly extended thanks to its implementation of A technician must assess the impact of the the OGC CSW specification. This is the project some summer fires in the protected areas of Va- of choice to deploy the discovery service type in lencia, she should find and access the protected our framework. Regarding WPS specification, areas dataset, the burned areas dataset (for a

Figure 4. SequenceIGI diagram GLOBAL PROOF

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 18 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

Figure 5. Local data deployment

IGI GLOBAL PROOF

certain year) and perform the intersection pro- standard OGC data service. Thus, the user can cess. Traditionally the result of the assessment upload it in the application. Figure 5 shows a can be used to write a report or make a decision; screenshot of how this step can be performed here we only describe how this technician can in the user interface. deploy it easily in the underlying SDI. Figure 4 The second step shown in the sequence shows a sequence diagram with the operations diagram (Figure 4) is the execution of a dis- and actors involved in each step of this scenario. tributed process to perform the intersection and The first step is to find the initial data sets, the generation of the assessment results (Figure in our case the protected areas of Valencia and 6). The Service Framework interface allows the burned areas in 2005. For the first data the user to connect and execute to existing WPS entry (protected areas), we use a local data services, rather than using heavy desktop ap- provider using a standard OGC data service. plications. In this step we must select a WPS These data can be found using the OpenSearch- server and a specific process, as in our case the geo client component. Intersection (a process that will return the If the data is not accessible through an OGC protected areas that burned in 2005). All com- service and it is only available locally like the munication with the service is through the input data related to the burned areas in 2005, Service Connector component using the WPS Service Framework provides the functional- API. In the diagram of Figure 4, we can see ity to upload local datasets to be deployed as how internally several WPS interface calls are

Figure 6. Process inputs view and results

IGI GLOBAL PROOF

performed. First a describeProcess request to standard services. After the data deployment, get the process description regarding input data the Service Publisher component creates a and afterwards (after the user identifies the minimum set of metadata elements to be pub- correct input data, in this case references to the lished in an instance of CS/W so that the data- WFS where we have deployed the data), the set can be found by other users through common execute request to run the process. discovery mechanisms in SDI. The last step consists of deploying the processing results to explore the results (Figure 6). Users can thus view it and analyze it through CONCLUSION the integrated map viewer component based in We have presented a framework that follows OpenLayers. Again as we described with the a hybrid SDI building methodology. We pro- upload of the burned areas the most interesting vide an feature of the Service Framework is that the INSPIRE-based architecture, whose Service Framework assists the user to deploy components assist users not only in accessing the processing results for visualization and but also in deploying resources as INSPIRE- download purposes into standard services, OGC compliant services. In this way we provide WMS and OGC WFS respectively. The Service an entry point for users to participate in SDI Framework uses the GeoServer RESTful API building and maintenance. As a result, users and creates a new data store accessible through

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 20 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012 can massively deploy resources improving their REFERENCES availability in a standard manner. To deploy resources automatically as Ball, W. P., Brady, D. C., Brooks, M. T., Burns, R., standard services, the Service Framework uses Cuker, B. E., & Di Toro, D. M. (2008). Prototype a concrete implementation of the OGC WMS, system for multidisciplinary shared cyberinfrastructure: Chesapeake Bay environmental observatory. WFS, WCS specification. The Service Frame- Journal of Hydrologic Engineering, 13(10), 960–970. work uses Geoserver’s own protocol to deploy doi:10.1061/(ASCE)1084-0699(2008)13:10(960) resources. We still could migrate the solution to other implementations, but some program- Bell, M. (2008). Service-Oriented Modeling (SOA): Service analysis, design, and architecture. Hoboken, ming should be done. To avoid this, it would NJ: John Wiley & Sons. be useful to deploy resources using standard transactional operations. This would require to Benthall, S., & Gill, S. (2010). SDI Best Practices with GeoNode. Retrieved June 1, 2010, from FOSS4G have the OGC WMS and WCS specifications website: http://2010.foss4g.org/tutorial10.php including a transactional interface like WFS and WPS is considering in its current and future Curbera, F., Duftler, M., Khalaf, R., Nagy, W., Mukhi, versions. This would detach the solution from N., & Weerawarana, S. (2002). Unravelling the Web Services Web: An Introduction to SOAP, WSDL, the vendor implementation. In this sense OGC and UDDI. IEEE Internet Computing, 6(2), 86–93. standards have shown to be mature enough to doi:10.1109/4236.991449 provide specifications to create interoperable Díaz, L., Granell, C., & Gould, M. (2008). Case web services to deploy and share resources study: Geospatial processing services for web based and provide the functionality needed to create hydrological applications . In Sample, J. T., Shaw, K., distributed applications on SDI. Tu, S., & Abdelguerfi, M. (Eds.), Geospatial services The use of open source components has the and applications for the Internet (pp. 31–47). New advantage of giving us full control over the code York, NY: Springer. to modify everything. A disadvantage, however, Díaz, L., Granell, C., Gould, M., & Olaya, V. (2008). is that they haveIGI caused GLOBALsome inconveniences An openPROOF service network for geospatial data process- because most of them use libraries that are not ing. In Academic Proceedings of the Free and Open fully developed in the context of this project. Source Software for Geospatial Conference, Cape Town, South Africa (pp. 410-419). Documentation scarcity is always a big issue to face. Therefore, there is a need to invest European Parliament and Council. (2007). INSPIRE an important amount of time and effort in the EU Directive. Retrieved January 28, 2009, from development to have satisfying results. http://eur-lex.europa.eu/LexUriServ/site/en/oj/200 7/1_108/1_10820070425en00010014.pdf Fielding, R. (2000). Architectural styles and the ACKNOWLEDGMENTS design of network-based software architectures (Unpublished doctoral dissertation). University of This work has been partially supported by California, Irvine, CA. the European project EuroGEOSS and by the Fonts, O., Huerta, J., Díaz, L., & Granell, C. (2010). CENIT “España Virtual” project through the OpenSearch-geo: The simple standard for geographic Instituto Geográ□co Nacional (IGN). We would web search engines. In Proceedings of the IV Jornadas like to thank Jorge Suarez from the Conselleria de SIG Libre, Girona, Spain. Retrieved September de Medi Ambient, Aigua, Urbanisme i Habitatge 1, 2010, from http://www.sigte.udg.edu/jornadas- siglibre/uploads/Articles/a29.pdf (Environmental, Water, Urban planning and living regional government of Valencia), who Freeman, E., Sierra, K., & Bates, B. (2004). Head first kindly granted the forest fire data for experi- design patterns. Sebastopol, CA: O’Reilly Media. mentation purpose.

Giuliani, G., Ray, N., Schwarzer, S., De Bono, A., Minsker, B., Myers, J., Marikos, M., Wentling, T., Peduzzi, P., & Dao, H. (2011). Sharing environmen- Downey, S., Liu, Y., et al. (2006). NSCA environ- tal data through GEOSS. International Journal of mental cyberinfrastructure demonstration project: Applied Geospatial Research, 2(1). doi:10.4018/ Creating cyber environments for environmental jagr.2011010101 engineering and hydrological science communities. In Proceedings of the ACM/IEEE Conference on Goodchild, M. F. (2008). The use cases of digital Supercomputing, Tampa, FL. New York, NY: ACM. earth. International Journal of Digital Earth, 1(1), 31–42. doi:10.1080/17538940701782528 Morisette, J. T., Richardson, A. D., Knapp, A. K., Fisher, J. I., & Abatzoglou, J. (2009). Tracking the Goodchild, M. F., Egenhofer, M., Fegeas, R., & rhythm of the seasons in the face of global change: Kottman, C. (1999). Interoperating geographic in- Phenological research in the 21st century. Frontiers formation systems. Boston, MA: Kluwer Academic. in Ecology and the Environment, 7(5), 253–260. Granell, C., Díaz, L., & Gould, M. (2010). Service- doi:10.1890/070217 oriented applications for environmental models: Nerbert, D. (2004). Developing spatial data infra- Reusable geospatial services. Environmental Mod- structure: The SDI cook book, version 2.0. Victoria, elling & Software, 25(2), 182–198. doi:10.1016/j. Australia: GSDI. envsoft.2009.08.005 Open Geospatial Consortium Inc. (OGC). (2005). Granell, C., Gould, M., & Esbrí, M. A. (2008). Web feature service implementation specification Geospatial web service chaining . In Karimi, H. 1.1.0. Redlands, CA: Author. A. (Ed.), Handbook of research on geoinformatics (pp. 189–195). Hershey, PA: Information Science Open Geospatial Consortium Inc. (OGC). (2006). Reference. Web map service implementation specification 1.3.0. Redlands, CA: Author. Hey, T., & Trefethen, A. E. (2005). Cyberinfraestruc- ture for e-Science. Science, 308(5723), 817–821. Open Geospatial Consortium Inc. (OGC). (2007a). doi:10.1126/science.1110410 Sensor observation service (Version 1.0.0). Redlands, CA: Author. INSPIRE Network Services Drafting Team. (2008). Draft technical guidanceIGI to implement GLOBAL INSPIRE view Open Geospatial PROOF Consortium Inc. (OGC). (2007b). services. Retrieved from http://inspire.jrc.ec.europa. Web processing service (Version 1.0.0). Redlands, eu/reports/ImplementingRules/network/Draft_Tech- CA: Author. nical_Guidance_View_Services_v1.0.pdf Open Geospatial Consortium Inc. (OGC). (2007c). INSPIRE Network Services Drafting Team. (2009a). OpenGIS® catalogue service implementation Draft technical guidance for INSPIRE download specification (Version 2.0.2). Redlands, CA: Author. services. Retrieved from http://inspire.jrc.ec.europa. eu/reports/ImplementingRules/network/Draft_Tech- Open Geospatial Consortium Inc. (OGC). (2008). nical_Guidance_Download_Services_v1.0.pdf Web coverage service implementation specification 1.1.2. Redlands, CA: Author. INSPIRE Network Services Drafting Team. (2009b). Technical guidance for INSPIRE discovery ser- Papazoglou, M. P. (2008). Web services: Principles vices. Retrieved from http://inspire.jrc.ec.europa. & technology. Essex, UK: Person Education. eu/documents/Network_Services/Technical%20 Pezzoli, K., Marciano, R., & Robertus, J. (2006). Guidance%20Discovery%20Services%20v2.0.pdf Regionalizing integrated watershed management: Lannotta, B. (2007). GEOSS: A global view. Aero- A strategic vision. In Proceedings of the 7th Annual space America, 45(8), 36–40. International Conference on Digital Government Research, San Diego, CA (pp. 444-445). New York, Masser, I. (2005). GIS worlds: Creating spatial data NY: ACM. infrastructures. Redlands, CA: ESRI Press. Pierce, M. E., Fox, G. C., Choi, J. Y., Guo, Z., Gao, X., Mazzetti, P., Nativi, S., & Caron, J. (2009). & Ma, Y. (2009). Using Web 2.0 for scientific applica- RESTful implementation of geospatial services tions and scientific communities. Concurrency and for Earth and Space science applications. Inter- Computation, 21(5), 583–603. doi:10.1002/cpe.1365 national Journal of Digital Earth, 2(1), 40–61. doi:10.1080/17538940902866153

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 22 International Journal of Applied Geospatial Research, 3(2), 6-23, April-June 2012

Poch, M., Comas, J., Rodríguez-Roda, I., Sànchez- Westlund, S. (2010). The use of geospatial tech- Marrè, M., & Cortés, U. (2004). Designing and nologies in disaster management. International building real environmental decision support sys- Journal of Applied Geospatial Research, 1(3), 17–30. tems. Environmental Modelling & Software, 19(9), doi:10.4018/jagr.2010070102 857–873. doi:10.1016/j.envsoft.2003.03.007 Wilkins-Diehr, N., Gannon, D., Klimeck, G., Oster, Pumphrey, M. (2009). OpenGeo Suite. Retrieved June S., & Pamidighantam, S. (2008). TeraGrid science 1, 2010, from FOSS4G website: http://wiki.osgeo. gateways and their impact on science. Computer, org/wiki/FOSS4G_2009_Demonstration_Theatre 41(11), 32–42. doi:10.1109/MC.2008.470 Rajabifard, A., Binns, A., Masser, I., & William- Yang, C., & Raskin, R. (2009). Introduction son, I. (2006). The role of sub-national govern- to distributed geographic information process- ment and the private sector in future spatial data ing research. International Journal of Geo- infrastructures. International Journal of Geo- graphical Information Science, 23(5), 553–560. graphical Information Science, 20(7), 727–741. doi:10.1080/13658810902733682 doi:10.1080/13658810500432224 Yang, C., Raskin, R., Goodchild, M., & Gahegan, Ramaprivan, H. K. (2008). NASA’s Earth science M. (2010). Geospatial Cyberinfrastructure: Past, data systems: A ‘bit of history’ and observations. In present and future. Computers, Environment and Proceedings of the Cyberinfrastructure for Environ- Urban Systems, 34, 264–277. doi:10.1016/j.com- mental Observations, Analysis, and Forecasting: A penvurbsys.2010.04.001 Cyberinformatics Forum. Yang, P., Li, W., Xiao, D., Raskin, R., & Rambacus, Reichert, P., & Schuwirth, N. (2010). A generic M. (2007). Earth information exchange - Sharing framework for deriving process stoichiometry in earth science information through interoperable environmental models. Environmental Modelling approach and cyberinfrastructure. In Proceedings of & Software, 25(10), 1241–1251. doi:10.1016/j. the SPIE Conference on Geoinformatics: Geospatial envsoft.2010.03.002 Information Science, Nanjing Richardson, L., & Ruby, S. (2007). RESTful Web Services. Sebastopol,IGI CA: O’ReillyGLOBAL Media. PROOF Secretariat, G. E. O. (2007). The full picture: GEO. ENDNOTES Geneva, Switzerland: Tudor Rose. 1 All specifications for service interfaces defined Steiniger, S., & Bocher, E. (2009). An overview by OGC are publicly available at http://www. on current free and open source desktop GIS opengeospatial.org developments. International Journal of Geo- 2 Workshops, tutorials and presentations for graphical Information Science, 23(10), 1345–1370. every conference edition are publicly available doi:10.1080/13658810802634956 at http://foss4g.org 3 Weaver, J. L., Gao, W., Chin, S., & Iverson, D. (2009). The OpenSearch specification and its exten- Pro JavaFX Platform: Script, Desktop and Mobile sions can be found at http://www.opensearch. RIA with Java Technology. New York, NY: APress. org 4 http://docs.geoserver.org/stable/en/user/ extensions/rest/index.html

José Gil received his MSc in computer science from the Jaume I University in 2004. Currently he is a researcher at the Interactive Visualization Group at Jaume I University and he is pursuing his PhD in computer science from Jaume I University. His research areas include real-time visualization, computer simulations and Java technology.

Laura Díaz graduated in Computer Engineering (2000) (Universitat de València). She obtained a Master in Intelligent Systems (2008) and received her doctorate in Geospatial Science (2010) at the Universitat Jaume I of Castellón where she is currently a postdoctoral researcher. Her research lines are approaches for improving geoservices descriptions and deployment, distributed geoprocessing and service interoperability. She has participate in R&D projects in the Institute of Robotics of the Universitat de València and in GIS companies such as Geodan (The Netherlands) and Iver TI (Spain).

Carlos Granell, PhD (Universitat Jaume I, Spain), is a postdoctoral researcher interested in the study of Spatial Data Infrastructures (SDI), interoperability, geoprocessing services, workflow, and the composition and reuse of geospatial services applied to the environmental domain. He has taken part in several public funded research projects, both at Spanish and European levels. He has carried out research visits at SINTEF (Norway), ITC (the Netherlands), University of Nottingham (UK). Dr. Granell is currently affiliated with the European Commission, Joint Re- search Centre, Institute for Environment and Sustainability, Spatial Data Infrastructures Unit, Ispra, Italy.

Joaquín Huerta is an associate professor in the Department of Information Systems at UJI, where he teaches GIS and Internet Technologies. He holds a PhD in Computer Science from Jaume I University.IGI He is currently GLOBAL leading two EU projects PROOF at UJI: EUROGEOSS (FP7) and eSDI-NET+ (FP6) and a Spanish project “España Virtual” funded by Cenit Programme. His current research interests are Geospatial Technologies and Computer Graphics. He is Director of the Master in Geospatial Technologies funded by Erasmus Mundus program. In addition to academic activities Dr. Huerta is founding board member of an Internet service provider and, thus, possesses considerable business experience as well as experience in systems integration.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 24 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

A Review of Geospatial Information Technology for Natural Disaster Management in Developing Countries

Sam Herold, University of Ottawa, Canada Michael C. Sawada, University of Ottawa, Canada

ABSTRACT

Disasters are deadly and destructive events, particularly in developing countries where economic, social, political and cultural factors increase natural hazard vulnerability. The recent devastation of the Haiti earthquake on January 12th, 2010 was a prime example of the human toll a natural disaster can take in developing regions of the world.IGI There GLOBALis an imminent need to improve naturalPROOF disaster management capacity in developing countries to reduce disaster impacts. Given that disasters are spatial phenomenon, the application of geospatial information technology (GIT) is essential to the natural disaster management process. However, in developing countries there are numerous barriers to the effective use of GIT, especially at the local level, including limited financial and human resources and a lack of critical spatial data required to support GIT use to improve disaster management related decision making processes. The results of a thorough literature review suggests that currently available free and open source GIT (FOS GIT) offers great potential to overcome some of these barriers. Thus, disaster management practitioners in developing countries could harness this potential in an attempt to reduce hazard vulnerability and improve disaster management capacity. The use of FOS GIT significantly reduces software costs and can help build local level GIT knowledge/technical skills that are required for successful GIT implementation.

Keywords: Disaster, FOSS, Geographic Information System, GIS, GIT, Hazard, Open Source Software, Remote Sensing, Software

INTRODUCTION that time (IFRC, 2001). Ultimately contributing to this trend are environmental degradation, Although the United Nations designated the rapid urbanization and social marginalization 1990’s as the International Decade for Natural (McEntire, 1999), particularly in developing Disaster Reduction (IDNDR), there was a global countries. The increasing number of disasters failure to reduce natural disaster impacts during suggests that vulnerability to natural hazards is also rising and so equates to changing the geography of risk. By way of elaboration, more DOI: 10.4018/jagr.2012040103

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 25 people are living in low-lying coastal zones, significantly enhanced through the effective seismically hazardous areas and concentrated use of GIT (Goodchild, 2006). Even though urban environments (Amendola et al., 2008; the natural processes (e.g., floods, earthquakes, Briceño, 2004; Burton et al., 1993; El-Masri landslides, etc.) that generate disasters might be & Tipple, 2002). Vulnerable populations will fundamentally different, the techniques to assess be at increased risk, for example, as the geog- and mitigate risk, evaluate preparedness, and raphy and magnitude of hydrometerological assist response have much in common and can hazards that are historically associated with share and benefit from advances in geographic some of the greatest disasters (Kondratyev et information science (GIScience) (e.g., data ac- al., 2002) change with global climate (Smith, quisition and integration; issues of data owner- 2004; IPCC, 2007). Defining the geography of ship, access, and liability; and interoperability) risk is of a major concern in general and in par- (Radke et al., 2000). We propose that currently ticular in developing countries, where disasters available free and open source software (FOSS) jeopardize important social development goals can fulfill many GIT requirements needed to such as addressing poverty, ensuring adequate improve disaster management capacity at the food, water, and sanitation, and protecting the local level. While FOSS can create additional environment. Because natural disasters have challenges compared to commercial solutions the greatest overall impact in developing coun- (Camara & Onsrud, 2004), with a clear under- tries (Alexander, 1995; Bui et al., 2000; IFRC, standing of the barriers and benefits of FOSS 2001), this is where geospatial information from a developing world perspective, FOSS is technologies (GIT) have the greatest potential a capable and effective alternative. to mitigate causalities. In the first section of this review, we begin The purpose of this paper is to examine the by laying a set of brief contextual explanations use of GIT for natural disaster management, and definitions of natural hazards and disasters with an emphasis on how these technologies, while emphasizing their spatial components, in particular free andIGI open sourceGLOBAL GIS (FOS and then PROOFdescribe some of the factors that dif- GIT) can be effectively utilized at the local ferentiate developed and developing countries level in developing countries. Although natural from a natural disaster vulnerability standpoint. disasters cannot entirely be prevented, disaster Included in this section is a more detailed losses (including human, environmental and examination of the most commonly cited GIT infrastructure/personal property) can be mini- implementation barriers faced by developing mized with effective disaster management – the countries. This is followed by a brief review of process of mitigation, preparation, response and the phases that comprise the disaster manage- recovery. The field of disaster management has ment cycle. In the next section, which forms greatly benefited from recent advancements in the bulk of this paper, we review the extensive computers and related information technologies. literature that describes and explores the many Geospatial information technologies (GIT), uses of GIT in the field of natural disaster man- including geographic information systems agement. We then examine GIT-based FOSS, (GIS), remote sensing (RS), global position- highlighting its potential as well as limita- ing systems (GPS) and Internet GIS (IGIS) tions in terms of its ability to fulfill disaster are currently being employed in a variety of management related requirements. Finally, we ways to support all phases of disaster man- discuss and describe our vision of how FOSS agement. Since each phase is geographically can greatly improve the ability of local level related to where people, places, and things disaster managers to implement GIT, and thus are spatially located (Gunes & Kovel, 2000), improve overall disaster management capacity the entire disaster management process can be and reduce vulnerability to natural hazards.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 26 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

BACKGROUND events of rapid-onset is increasing (Oloruntoba, 2005). In particular, hydrometeorological re- Natural Hazards and lated natural disasters (e.g., floods, landslides/ Disasters: An Overview avalanches, forest/scrub fires, wind storms and waves/surges) have more than doubled since To provide context for the pursuant issues, a 1996 and caused over 90 percent of deaths from brief overview of the relation between a natural natural disasters during the 1990’s (IFRC, 2001). hazard and a natural disaster is in order. Natural Herein, our main focus is on the role of GIT for hazards vary greatly in terms of frequency, dura- mitigation of rapid-onset natural hazards that tion, scale, impact, etc., and these differences cause the risk of natural disaster. partly determine the spatial data and technol- Large-scale disasters represent a complex ogy needed to effectively mitigate, prepare for, and multidisciplinary problem for local disaster respond to and recover from potential disasters managers and related organizations, as well as that may result from their occurrence. International humanitarian/aid organizations. Simply put, natural hazards are un- While many natural disasters are characterized predictable acts of nature, characterized by by short reaction/response times, overwhelming extremes in physical processes (Zerger & damage to property and infrastructure, and a Smith, 2003). Examples of natural hazards strain on the resources of the affected com- include earthquakes, tsunamis, hurricanes, munity, those less frequent large-scale natural typhoons, droughts, wildfires, tropical storms disasters are much more deadly. Among the and floods. The fundamental determinants of largest-scale globally predominant natural natural hazards are location, timing, magnitude disasters with 50,000 victims or more, three and frequency (Alexander, 2000). The spatial hazard types can be singled out: earthquakes, scale and duration of natural hazards can vary tropical cyclones (with coastal inundation), and greatly, which is important from a GIT perspec- river floods (Kondratyev et al., 2002). These tive, and in particularIGI from GLOBAL the perspective of types PROOF of natural hazards have caused the worst data requirements. Landslides, for example, calamities both in the 20th century and the entire have a local impact, whereas major floods can history of humanity, namely: the 1970 flood and affect a large region. Earthquakes occur with cyclone in Bangladesh (300,000 victims), the little warning and last only a few seconds to 1976 earthquake in China (242,000 victims), minutes, while a drought may build up over a and the 1931 flood in China (140,000 victims). period of months over large regions and last even The 2004 earthquake that occurred off the coast longer. Thus, a distinction can be made between of Sumatra, Indonesia, and resulting tsunami ‘rapid-onset’ natural hazards such as floods that devastated many countries surrounding and earthquakes, or slower ‘creeping crises’ the Indian Ocean also ranks among the dead- hazards like drought or disease (De Paratesi, liest natural disasters in history, with 283,000 1989). However, Coppock (1995) points out reported fatalities (Lay et al., 2005). that those slowly developing hazards have more in common with natural resource management, Disaster Vulnerability: Developed at least from a mitigation perspective vs. Developing Countries McEntire (2001) describes natural disasters as the disruptive and/or deadly and Although natural hazards pose a considerable destructive outcome of triggering agents when threat to all countries, historically, developing they interact with, and are exacerbated by, vari- countries have been disproportionately affected ous forms of vulnerability. Simply put, when a (Briceño, 2004). About 95% of deaths caused by hazard intersects the zone of human use there natural hazards occur in developing countries is the risk of disaster. The number of people af- (Bui et al., 2000), and the loss of GNP due to fected by disasters resulting from natural hazard disasters is 20 times that of developed nations

(Alexander, 1995). These alarming statistics the scale of a disaster, but why are people in cannot be attributed to a greater frequency of developing countries so vulnerable? natural hazards in the developing world, but can The vulnerability of a geographic area is be partly explained by differences in natural haz- determined by its natural and human-made ard vulnerability. The concept of vulnerability environmental conditions, climatic patterns, is complex – it depends on a number of param- and its political, social and economic ability to eters – and varies depending on the research withstand and respond to natural hazard events orientation and perspective (for example, see (Jayaraman et al., 1997). Hazards tend to be Cutter, 1996; Cutter et al., 2000, 2003; McEn- more destructive in developing countries where tire, 2001; Morrow, 1999; Weichselgartner, political, social and economic instability lead to 2001). Thus, there is no single accepted way poor organizational infrastructure and no adap- to assess natural hazard vulnerability (Simpson tive capacity as compared to most developed & Human, 2008). However, regardless of how countries. Henderson (2004), in examining the it is conceptualized or assessed, history shows pervasive risk of natural disaster faced by de- that developing countries are more vulnerable veloping countries, emphasizes that conditions to natural hazards than developed countries, of poverty, poor housing, lack of information and as a result experience a greater number of about disaster risk, poor telecommunications, natural disasters. and inadequate physical infrastructure fre- Although location and proximity to natural quently exacerbate natural disasters. Insufficient hazards is certainly an important factor, many disaster management resources, the absence researchers propose that the scale of disaster of enforced laws and the shortage of trained impact is more a function of human vulnerability experts all increase the difficulty in coping rather than of the physical magnitude of the with natural disasters (Guinau et al., 2005). hazard (Hewitt, 1995; Quarantelli, 1998; Smith, For example, the high cost and skills required 2001). In this sense, natural disasters can more for the application of GIS technology has hin- accurately be seen asIGI social phenomenon, GLOBAL where dered its PROOFutilization for disaster management in the overall damage due to natural hazards is the developing countries (Rudyanto et al., 2001). result of both natural events that act as ‘trig- The lack of ability to effectively utilize relevant gers’, and a series of societal factors (McEntire, technology is a societal factor that contributes 2001; Weichselgartner, 2001). Conceptualizing to increasing disaster vulnerability. vulnerability in terms of societal factors as op- Developed countries have a greater aware- posed to biophysical factors helps to explain the ness and understanding of the importance of wide range of human impact that can result from natural disaster management. Bui et al. (2000) events of comparable magnitude. For example, point out that they invest more in mitigation and Montoya and Masser (2005) point out that the prevention, and have more resources available to 1988 Spitak earthquake in Armenia (former enforce legislation that may help reduce human USSR) and the 1989 Loma Prieta earthquake vulnerability. In developed countries, insurance in California were of similar magnitude and absorbs more than half of the economic losses affected populations of comparable size; how- from natural disasters, in contrast, less than 2% ever, the Armenian event killed 25,000 people of the losses are insured in developing countries whereas the California earthquake killed 63. (Freeman & Pflug, 2003). However, as a general Hurricanes Hugo in 1989 and Andrew in 1992 trend, in both developed and developing nations, caused less than 50 deaths each in the U.S. On disaster management becomes a more press- the other hand, cyclones in Bangladesh killed ing concern only after the disaster has struck over half a million people in 1970 and another (Currion et al., 2007). Perhaps the devastating 140,000 in 1991 (Bui et al., 2000). Clearly, the impact of recent large-scale disasters, such as extent of human (or societal) vulnerability is the Indian Ocean tsunami (2004), Hurricane a major factor that contributes to determining Katrina (2005), and the Haiti earthquake (2010)

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 28 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 will encourage disaster managers in both the of well-identified barriers that often limit the developed and developing world to focus more use of such technology, thus increasing vul- on mitigation and preparation, rather than re- nerability to natural disaster. These common sponse and recovery. barriers include: There are many factors that contribute to determining disaster vulnerability, and it is not 1. A lack of financial resources our intention to discuss them all, but rather to demonstrate that the concept is pervasive within Generally, using geospatial technology is the disaster management literature, particularly costly: requiring a substantial investment in from a developing countries perspective. In ad- computer software/hardware, spatial data, and dition, since many of the factors that determine human resources (education and training). In vulnerability vary spatially, vulnerability is well many cases disaster management responsibili- suited to be assessed using GIT. For example, ties and duties are decentralized to local gov- McEntire (2001) considers vulnerability to be ernments without being accompanied by the the dependent component of disaster that is necessary funds (Montoya & Masser, 2005). determined by the degree of risk, susceptibility, Considering that local governments have to resistance and resilience. Risk is created because provide many essential services (e.g., health, of proximity to potential hazards, which varies education and infrastructure services such as spatially; resistance refers to the ability of build- water, electricity and sewage, to name a few), ings and the infrastructure to resist the strain they are often not predisposed to purchasing off- exerted by natural hazards, which also varies the-shelf commercial GIS software due to other from one area to another depending on building demands for funding (Renyi & Nan, 2002). It is design, building codes and the materials used, difficult to justify spending limited budgets on among other factors. Susceptibility, which is technology to enhance disaster mitigation and a product of IGI social, political, GLOBAL economic and preparation PROOF when basic needs have priority. In cultural forces that determine the proneness of addition, software has no benefits without the groups and individuals to be adversely affected ability to purchase or collect data. Thus, limited by disasters, can also be analyzed in spatial finances are a major barrier and if GIT is go- terms. For example, characteristics such as ing to play a role in local level natural disaster age, gender, race and socioeconomic status are management financially feasible solutions such generally accepted to be influential components as FOSS can help. of social vulnerability (Cutter et al., 2003), and these characteristics have a high degree 2. A lack of local expertise/knowledge (hu- of spatial variability, particularly within large man/technical resources) urban areas. Such characteristics are typically captured in a census, and this information can be analyzed and mapped at varying scales using Effectively use of GIT requires consider- GIT, such as GIS. able technical knowledge and skills, obtained Considering that assessing and understand- through education and training, as well as prac- ing hazard vulnerability is a key component of tical experience. Mohamed and Plante (2002) disaster management, and that vulnerability emphasize the lack of local expertise and capac- varies spatially, it is well suited to be analyzed ity to operate and maintain a GIS in developing using GIT. Therefore, GIT must play a key role countries and general awareness is strongly in any comprehensive and effective disaster lacking (Murgia et al., 2002). In contrast, the management strategy, yet there are dramatic well established GIT communities built up over differences in GIT use between developed and time in developed countries are an important developing countries. While each developing resource for individual practitioners to acquire country is unique, there exists a common set new knowledge, methods or techniques, discuss

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 29 ideas and/or obtain support. In some develop- empowered individuals and groups rather than ing countries, practitioners may be physically organisational structures. separated by hundreds of kilometres and be without Internet or telephone (Britton, 2000). 4. A lack of spatial data As such, local disaster management organizations would have trouble in trying to find local Accurate and comprehensive spatial data personnel who could effectively manage the play a critical role in all phases of disaster various GIT components/aspects of a holistic management, and are required for effective GIT natural disaster management strategy. When use, yet in many developing countries reliable GIT has been implemented at the local level spatial data are a scarce resource (ESRI, 2006; in developing countries, it is often the result Dewan, Islam, Kumamoto, & Nishigaki, 2007). of externally sponsored/funded technology Murgia et al. (2002) suggest that part of the development programs, rather than through the problem are weak national data providers such gradual building of local GIT capacity (Britton, as mapping agencies, census organizations and 2000). Local workforce development and capac- cadastres, and that many projects that have been ity building projects are essential for sustained initiated are stand-alone and lack continuity and GIT implementation and maintenance that does consistent funding or are not embedded in stable not rely on external support or funding. institutions or are politically misused. In addition, governments in developing countries tend 3. Institutional/political instability to have a very conservative approach in terms of data and information management (Shrestha, Disaster management is dependent on the 1994). For example, among the most common functional and effective operation of institu- obstacles to be found in Asia are official restric- tions, whether formal or informal, and at the tions on geospatial data for security reasons as local level where itIGI matters most GLOBAL (Pande, 2006). well as thePROOF rigidity and compartmentalization The effectiveness of GIT and the ability to of government bureaucracies, which consider convert information into action also depends certain types of information as their property on the existence of supporting organizations, (Brodnig & Mayer-Schönberger, 2000). In which are generally lacking in developing Costa Rica, for example, access to census tract countries (Coppock, 1995). Political barriers level data is restricted (Montoya & Masser, have also been cited as obstacles to local level 2005). Additional factors that contribute to GIT use (Montoya & Masser, 2005). High the lack of spatial data include: (1) the remote- resolution census data, for example, contains ness of many areas; (2) the lack of technical a wealth of information that could be utilized capacities to collect and assess bio-physical in a variety of GIT-based natural disaster and socio-economic data; and (3) competing management related scenarios but this data is priorities in the fledging economies (Brodnig often unavailable in developing countries for & Mayer-Schönberger, 2000; Shrestha, 1994). various reasons. Governments in developing Montoya and Masser (2005) emphasize the need countries can be unstable, and political terms to identify or develop cost-effective data collec- can be short. According to Murgia et al. (2002) tion methods for producing spatially referenced politicians have their own circle of supporters information in developing countries. The role who will be appointed in public institutions, of spatial data in all areas of natural disaster and thus political instability is expanded to in- management cannot be underestimated; in fact, stitutional instability. Considering institutional Mansourian et al. (2004) propose the develop- instability, Ramasubramanian (1999) suggests ment of a spatial data infrastructure (SDI) as a that GIS implementation in developing coun- framework to facilitate disaster management. tries would be more successful in the hands of Aside from the barriers to GIT implementation just described, Ramasubramanian (1999)

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 30 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

Figure 1. Disaster management cycle (Adapted from Alexander, 2000)

also identifies cultural and language barriers Disaster Management in terms of GIS implementation in developing and the Role of GIT countries (forIGI more on GLOBAL specific GIS imple- PROOF mentation issues in developing countries see Disaster management can be understood as Ramasubramanian, 1999). Given the aforemen- a cycle (Figure 1) which includes an effort tioned barriers, disaster managers in developing to mitigate against, prepare for, respond to, countries are unable to utilize GIT in the same and recover from a disaster (Montoya, 2002). ways as developed countries, which in effect, The core components of this cycle are also further increases their vulnerability to natural commonly referred to within the emergency disasters. However, recent growth in the domain management literature (e.g., Cutter, 2003), of free and open source software (FOSS), and and often in conjunction with natural disasters in particular GIT-based FOSS, provides new (Cova, 1999). Thus, emergency and disaster opportunities that were previously unavailable. management are closely linked; after all, a For example, the cost alone of commercial GIT disaster clearly constitutes an emergency, but software (such as GIS) can prevent it from being in this paper it will be referred to as the disaster used at the local level in developing countries management cycle. (Ranyi & Nan, 2002); with FOSS, the issue of Mitigation refers to efforts that aim to cost is eliminated. For this reason, and others eliminate or reduce the risk to humans and/or that will be described later in this paper, FOSS property caused by natural or man-made hazards is increasingly being recognized within the (e.g., risk assessment, insurance, engineering disaster management literature as a good fit standards, land use management, public educa- for developing countries. tion, etc.). Preparedness refers to activities

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 31 necessary to the extent that mitigation measures effective disaster management strategy. Figure have not, or cannot, prevent disasters. This 2 incorporates GIT within the disaster manage- involves developing operational capabilities ment cycle to emphasize the central role it plays for responding to a sudden disastrous situation. during all phases. In this phase governments, organizations, and Disaster management is complex, and individuals develop plans to save lives and involves the participation and collaboration of minimize potential disaster damage. This in- many institutions/organizations/agencies oper- cludes emergency planning, training exercises, ating at international, national, regional and implementing hazard warning systems, evacu- local levels. Such institutions/organizations/ ation procedures and stock-piling of critical agencies use a range of GIT depending on their supplies. In addition, preparedness measures specific information requirements, financial also seek to enhance disaster response opera- resources and technical capabilities. While a tions. Response refers to those actions taken strong national disaster management initiative just prior to, during, and immediately after a is important, Henderson (2004) emphasizes that disaster that save lives, reduce property damage disaster management capabilities should be or improve recovery. The most important aspect decentralized to the regional or local level of this phase involves providing emergency given the wide variations in demographic, assistance for victims (e.g., search and rescue, socioeconomic, cultural and infrastructural emergency shelters, medical care and food/ conditions within regions and local areas of a water). Disaster responders also seek to stabilize nation. Regional and local authorities involved the situation and reduce the probability of in disaster management should ideally have the secondary damage. Recovery (often used in capacity to effectively utilize GIT, yet even in conjunction with the word ‘relief’) includes developed countries limited financial resourc- those activities that (1) restore vital life support es can prevent such utilization within this ad- systems and (2) return the area/population to a ministrative level (Laben, 2002). Theoretically, pre-disaster state. IGIThe former GLOBAL can be seen as this approach PROOF is increasingly being recognized part of the short-term recovery plan while the to be more effective than the centralized ap- latter as part of the long-term, which may con- proach, but as Montoya amd Masser (2005) tinue for a number of years after a disaster. note, local authorities are often not provided Reconstruction is closely linked with mitigation, with the necessary financial resources to de- and is undertaken in ways that aim to reduce velop and implement effective policies and vulnerability and improve preparedness, thus plans, which should incorporate GIT. Laben disaster management occurs in a cyclical fash- (2002) emphasizes that it is extremely important ion as depicted in Figure 1. to have GIT alternatives available that can be Geospatial information technology (GIT), used by all levels of the emergency and disas- and its ability to acquire, interpret, analyze, map ter management communities. and disseminate information, are essential in all areas of natural disaster management. Because GIS is a spatial decision support tool it is in- GEOSPATIAL INFORMATION valuable when effectively used in a field like TECHNOLOGY AND disaster management that deals with critical spa- DISASTER MANAGEMENT tial decisions (Cova, 1999). For instance, GIT Geospatial Information Technology is used provides the basis for estimating and mapping for many disaster management functions, in- risk, planning evacuation routes, determining cluding hazard and risk assessment (Dewan, suitable areas for shelters, identifying disaster Kabir, Islam, Kumamoto, & Nishigaki, 2007; victims, and assigning resources during recov- Ramli et al., 2005), vulnerability assessment ery (Goodchild, 2006). It follows then, that GIT (Cutter et al., 2000; Kienberger & Steinbruch, must play a key role in any comprehensive and

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 32 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

Figure 2. A disaster management cycle incorporating GIT (adapted from Hussain et al., 2005)

2005; Weichselgartner, 2001), vehicle dispatch to help answer essential questions and make and supply routingIGI (Dong, GLOBAL 2005), damage as- informed, PROOF timely and appropriate decisions sessment (Chen et al., 2005; Rivereau, 1995; that can help save lives. Perhaps the greatest Zhang et al., 2002) and resource mobilization strength of GIS are their ability to integrate a (Goodchild, 2006), among many other essential wide range of data types, including geographic, tasks. In this section we examine the use of GIS, social, economic, and political data into a single remote sensing and Internet GIS in the area of system (Dash, 1997). natural disaster management. However, to utilize GIS not only requires the software, but also hardware, data and trained Geographic Information personnel. Thus, a GIS capability involves not Systems (GIS) only the software itself, but the necessary spatial and descriptive (attribute) data, computer A GIS is an “organized collection of computer hardware and personnel who can effectively hardware, software, geographic data, and per- utilize it (Dash, 1997; Montoya, 2002). As a sonnel designed to efficiently capture, store, result, the use of GIS for disaster management up-date, manipulate, analyze, and display all varies widely between individual countries, and forms of geographically referenced informa- within disaster related organizations and institution” (ESRI, 1993, pp. 1-2). A GIS is a key tions at different levels of government (Laben, component of any effective and comprehensive 2002). These differences in GIS capability are disaster management strategy, and are used to clearly reflected in the wide range of literature display, integrate, map, analyse, and model that examines and describes the use of GIS for data and information derived from satellites, natural disaster management. This section of the and other spatial data sources (Kumar et al., paper will review the GIS and disaster manage- 1999). GIS function primarily as a support tool ment literature in order to identify examples of

Table 1. Crucial spatial data for disaster management (adapted from Gunes & Kovel, 2000)

Data/Information Type Description Disaster forecast Information concerning the extent of a particular hazard or disaster Vulnerability analysis Information on critical facilities (hospitals, schools, shelters, police and fire facilities, dams, trauma centers, industrial facilities, etc.); Information regarding human vulnerability (age, gender, socioeconomic status, etc.) Damage assessment Data/imagery of the actual impact of a hazard Resource inventory Location information regarding supplies, equipment, vehicles, or other material resources Infrastructure Shows transportation networks (roads, railroads, bridges, traffic control points, and evacuation routes) as well as complete utility grids (electric, gas, water, and sewer) Mass care/shelter status Monitors the movement of people to and from government or voluntary agency shelters by providing information on capacity, availability, supplies, and suitability to victims’ needs

GIS capabilities (software, data, hardware, and of the implementing institution/organization. level of expertise required) that are feasible for As an example, Table 1 describes some of the implementation in developing countries. These various types of spatial data/information ap- are all important considerations that will be plicable for disaster management that can be emphasised and discussed further. utilized in a GIS environment. Since there are distinct differences between Over the years, developments in spatial using a GIS for pre- or post-disaster manage- data collection and use have led to the concept ment activities, thisIGI section hasGLOBAL been divided in of spatial PROOF data infrastructure (SDI). Musinquzi half, one-half that deals with pre-disaster GIS et al. (2004) suggest that an ideal SDI includes use and another that examines post-disaster use. “technologies, policies, standards, and human However, before these sections are presented resources to acquire, process, store, distribute, it is first necessary to discuss the importance and improve utilization of geo-spatial informa- of spatial data, since it is the primary input for tion” (p. 790). Generally speaking, the success GIS and is required for their use. of disaster management largely depends on the availability, dissemination and effective use of The Importance of Spatial Data information, much of which is spatial (Ven- katachary et al., 2002). Accordingly, then, there The success in utilizing GIS depends largely on is a relationship between GIS, SDI and the the availability of spatially-referenced data; the extent to which spatial information is utilized quality of which is determined by locational pre- for disaster management related purposes. cision, the characteristics of the attribute data, Most developed countries are at various and by the extent to which standards are adopted stages of developing national SDIs, and their that allow for data transfer (Coppock, 1995). The success can be linked to high levels of technol- importance of good quality spatial data for use ogy, availability of funds, trained personnel and with GIS to assist disaster management cannot political support and stability (Musinguzi et al., be underestimated. There are several types of 2004). Many of these factors that contribute spatial information that are useful for disaster to successful national SDI development are management related decision making – the type, lacking in most developing countries, and as scale and complexity of such data depend on the a result, there are vast differences in spatial type, scale and stage of the disaster (De La Ville data quality and quantity between developed et al., 2002), as well as on the GIT capabilities and developing countries. As well, without a

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 34 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 national SDI there is no effective mechanism line, its discussion is reserved for subsequent for coordinating data collection efforts or for sections, and spatial data in the present context sharing spatial data among agencies or depart- refers to vector data only. ments. This often results in duplication of data Scale can also limit the usefulness of and wastage of human/financial resources. online spatial data. Currently, data available Limitations in accessing and sharing spatial for developing countries is most often at the data are particularly problematic for disaster national scale and derived from small-scale managers, whose spatial data needs typically cut map sources. Such data is inappropriate for re- across departmental/agency boundaries (Rego, gional or local level use for a variety of reasons, 2001). However, some developing countries including locational inaccuracy and a lack of are making progress, including Sri Lanka, for attribute information. Even if the scale of the example, who just recently (2008) initiated a data is appropriate, some data may require ad- national SDI (http://www.survey-dept.slt.lk/). ditional time investments to clean, correct, or Their policy statement makes specific refer- add the relevant attribute data before they can ence to disaster management, and how a SDI be effectively utilized. In addition, problems is required to reduce existing inefficiencies. can arise when trying to combine data from While there are distinct differences between disparate sources due to variations in coordinate developed and developing countries in terms of systems and map projections. Engler and Hall the quality and quantity of currently available (2007) emphasize that, in many cases, it may spatial data, efforts are being made to improve be better to create or purchase the required data the current situation. Nebert (2004) points out elsewhere, as there may be too much work to that “many national, regional, and international be done or uncertainty associated with using programs and projects are working to improve data freely available from the Internet. access to available spatial data, promote its Spatial data are essential in all areas of reuse, and ensure that additional investment in disaster management, yet good quality spatial spatial informationIGI collection GLOBAL and management data PROOF may be hard to come by, especially at the results in an ever-growing, readily available local level in developing countries. To improve and useable pool of spatial information” (p. 6). and increase the use of GIS for disaster man- Although such initiatives may not be part of a agement will require a substantial effort in data formal SDI policy, they undoubtedly help the collection, one that can be facilitated with an global GIS community and others who may effective SDI policy. Rego (2001) stresses that use spatial data. the development of spatial databases should be The Internet has also drastically increased built bottom up from the lowest administrative access to spatial data, as numerous websites unit in country (e.g., the sub-district or district). now offer free access to a wide range of data The district databases would then feed into the (e.g., political/administrative boundaries, roads, state/provincial database and then into the na- hydrology, digital elevation, land cover, etc.). tional database. A bottom-up approach to spatial However, despite the rapid growth in available database production is especially relevant for spatial data there has been little attention paid local-level GIT use in developing countries. to aspects of its quality, including currentness, lineage, locational accuracy, completeness, Pre-Disaster and overall usefulness (Engler & Hall, 2007). This is unfortunate, because the quality of A GIS has much to offer in the pre-disaster man- spatial data is particularly important when it is agement phases of mitigation and preparedness. used for disaster management, when lives are To effectively mitigate and prepare for a disaster potentially at risk. Although satellite imagery requires not only detailed knowledge/informa- is an essential type of spatial data for disaster tion about the expected frequency, character, and management, and is abundantly available on- magnitude of hazardous events in an area, but

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 35 also the vulnerability of the people, buildings, and eight social indicators (e.g., age, race/ethnic- infrastructure and economic activities in a poten- ity, income levels, gender, building quality, etc.) tially dangerous area (Rego, 2001; Van Westen their study demonstrates how a GIS can be used & Hofstee, 2001). This information forms the to integrate both biophysical and social factors cornerstone of preparedness planning and helps that contribute to hazard vulnerability. Social determine appropriate mitigation strategies. A characteristics of the population are available GIS allows for the synthesis and analysis of from most national censuses, and census data such data/information to help determine risk is often used in a GIS to map human vulner- levels, assess vulnerability, model scenarios, ability to various natural hazards. The research plan evacuation routes, determine resource methodology and conceptualization of hazard requirements, and create a variety of useful vulnerability provides a template for others to information products to aid decision-making. follow, and helps fill a void in the literature on Some specific examples of GIS use in the pre- spatial analytical approaches to vulnerability disaster phases are particularly relevant for assessment. If vulnerability to natural hazards developing countries and so we elaborate upon can be identified, appropriate steps can then be these as cases of effective GIT implementation taken to reduce the social and economic impacts in pre-disaster management. of potential disasters. However, the authors Van Westen and Hofstee (2001) present acknowledge that implementing this approach a case study that used GIS to create a spatial at the local level may pose challenges in terms database of buildings, land parcels, roads and of availability of funds for training and data other infrastructure by digitizing features from acquisition. Furthermore, results are based on aerial photographs. Subsequently, a field team a data intensive methodology, and while this of investigators collected specific attribute data approach is possible in developed countries, it for each digitized land parcel and building, us- is unlikely feasible in most developing counties ing a pre-determined checklist. For example, due to a lack of detailed spatial data, in addition land use of the parcelIGI (residential, GLOBAL industrial, to other GISPROOF implementation issues previously commercial, etc.), building material, building described. However, it seems likely that this age, number of floors and whether previous methodology could be simplified and still pro- hazard damage had occurred were among the duce useful hazard vulnerability information. attribute data collected. The resulting data- Guinau et al. (2005) used a GIS to create base, when combined with historical data on a hazard susceptibility map to help mitigate previous disasters, such as flood depth, was landslide risk. The methodology applied fo- used to generate a variety of vulnerability and cused solely on the biophysical factors that risk maps. The database and maps that can be cause landslides. The authors digitized present generated serve as a basis for future develop- and past landslides from aerial photographs to ment and planning that takes into account both create a landslide inventory, which was then biophysical and human vulnerabilities. This overlaid on terrain data, such as lithology, study was conducted for the city of Turrialba, slope, soil characteristics and land use. Analy- Costa Rica, which has a population just over sis of terrain conditions in areas affected by 30,000. Data collection and analytical methods landslides made it possible to determine zones are relatively simple to repeat, making this an with similar characteristics, and through further excellent example of the type of GIS use that analysis the delineation of low, medium and is practical for local level disaster managers in high susceptibility zones. Perhaps most im- developing countries. portant about this study is that it demonstrates Cutter et al. (2000) present a county level a relatively simple GIS-based methodology to “hazards-of-place” (Cutter, 1996) based GIS assess landslide vulnerability that is feasible method for assessing hazard vulnerability in spa- for developing countries. tial terms. Using twelve environmental threats

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 36 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

Dewan, Islam, et al. (2007) integrated GIS imagery or digitized from map sources, includ- and remote sensing techniques to analyze the ing land use, land parcels, roads and drainage. To flood hazard and risk levels in Dhaka, Ban- derive land cover, for example, a combination gladesh. A major impediment was unavailable of techniques was used, including unsupervised digital geospatial data, and as a result, a number (to identify spectral clustering) and supervised of the required data layers had to be created. classification based on local knowledge, and Flood-affected frequency and flood depth were the use of aerial photographs for verification. estimated from multi-date Synthetic Aperture Major roads were digitized on-screen in the GIS Radar (SAR) data (from RADARSAT), based from the satellite imagery, and the remaining on previous flood events. Land-cover was data were generated using a variety of low-tech generated using a combination of approaches methods, including visual image interpretation and data sources, including on-screen digitiza- (and digitization) and field work. Using basic tion of features from high resolution remotely GIS functions, such as overlay and buffering, sensed imagery, recently produced topographic a final map was produced that identified sites maps, field surveys and the use of hand-held possessing appropriate criteria for low-income GPS units. A geomorphic map was also de- housing: non-built, greater than 25 hectares, veloped using a 1999 LANDSAT TM image within 1 km of a road, and least prone to flood- in conjunction with available paper maps and ing. The map can be effectively utilized at large field observations. Elevation data in the form and medium scales (e.g., 1:20,000 – 1:50,000), of a DEM was obtained from the Institute of and the analysis methods and techniques are Water Modeling (IWM), Bangladesh. All the considered feasible for cities in developing vector data layers were then converted to raster countries, in terms of available data and required at identical resolutions. Using a relatively simple skill sets. A better understanding of suitable procedure that involved assigning a weighted locations for potential housing sites could help score to each data type (to represent varying guide development in ways that reduce natural significance) andIGI the use GLOBAL of GIS overlay func- hazard PROOF vulnerability. tionality, the authors were able to create maps Knowledge of the spatial characteristics depicting flood hazard potential and flood risk of hazards and vulnerability before a disaster zones. The data collection methods, analytical strikes enables disaster management authorities techniques and general approach used in this and emergency workers to identify areas that are study demonstrate innovative and adaptive use likely to be most affected, and thus efficiently of GIT in the context of a city-scale hazard as- focus their resources. Such knowledge is critical sessment for developing countries. to the development of effective mitigation and Uncontrolled and informal housing devel- preparation strategies that will ultimately help opment is a recurring problem in cities in de- reduce the devastating impacts that can result veloping countries (Thomson & Hardin, 2000), from natural disasters. and such development practices contribute to increasing natural hazard vulnerability. To ad- Post-Disaster dress this problem, Thomson and Hardin (2000) used GIS and satellite imagery (LANDSAT Disaster responders, emergency personnel, aid TM) to identify potential low income housing workers and anyone involved in the response sites in the eastern portion of the Bangkok Met- and relief effort need timely and up-to-date ropolitan Area, where flood risk is a concern. information, such as the extent of damage and Location, infrastructure, land cover, and natural location of potential victims, the location of criti- environment factors were used to assess the cal facilities (e.g., shelters, hospitals, air strips, characteristics appropriate for public housing etc.), available resources (e.g., food, water, sites. Due to a shortage of spatial data, all GIS blankets, medical supplies, etc.), infrastructure coverage required were derived from satellite conditions (e.g., damaged roads/bridges/utility

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 37 lines, etc.), and evacuation or supply drop off and to study the factors (e.g., slope, geology, points. Much of this information is spatial and land cover, etc.) that contributed to causing is thus well suited to be compiled and ana- the various types of mass movements. All the lyzed in a GIS and then disseminated as maps. information derived from this study was used However, using GIS post-disaster is different by the corresponding government agencies as and poses greater challenges than using it pre- the basis for the preparation of reconstruction disaster, since time becomes the critical factor plans for the affected areas. (Goodchild, 2006). Immediately following a Following the Indian Ocean tsunami disaster, information must be quickly collected, of 2004, and after evoking the International analyzed and assembled into useful informa- Charter on Space and Major Disasters to action products that can assist response efforts. quire pre- and post-disaster high resolution Thus, to the extent possible, GIS data should satellite imagery, Magsud et al. (2005) used be collected and/or analysed in advance of an a GIS for rapid damage assessment of build- event, rather than having to put it all together ings in Galle, Sri Lanka. The GIS was used during the aftermath (ESRI, 2006). Bottlenecks primarily to combine different data types and in information flow and dissemination can to support a visual analysis of building dam- mean the difference between life and death. age. QuickBird multi-spectral imagery and a Therefore, having a database of the most critical 1:5000 scale vector layer of buildings (obtained and useful spatial data before a disaster strikes from the Survey Department) were overlaid is essential. It allows for quick updates of the in a GIS to accurately identify construction data to reflect ground conditions and reduces that existed prior to the disaster. In addition, a the time it takes to produce critical information ground survey of 81 buildings was undertaken products required by disaster responders and that identified the level of damage following related emergency response personnel. the disaster (completely destroyed, partially Following the initial response period, and damaged mainly inside, partially collapsed with once the situation IGIhas stabilized, GLOBAL a GIS can be roof intact, PROOF and slightly damaged). GPS was used to analyze disaster impacts and help plan used in conjunction with ground photographs the rehabilitation process in a way that reduces to accurately record the location and building potential vulnerabilities. Development of this damage, respectively. This allowed for a com- fashion is termed ‘invulnerable development’ parison between the damage level in the photo- (McEntire, 2001). The following few examples graphs and satellite imagery. Results indicated demonstrate the value and necessity of GIS that heavily damaged buildings can easily be during the post-disaster management phases. identified, but partial damage, particularly if the Impact analysis is one fundamental use of roof was still intact, was difficult to determine GIS in the post-disaster phase. Impact analysis from the satellite imagery alone. The mapped can assist response efforts by identifying those results clearly show the location of destroyed areas most in need, and can help guide recon- buildings, which is valuable information that struction efforts in a way that will minimize could be used to prioritize response operations. the potential for future disasters; for example, With an adequate amount of technically trained through improved land use planning that takes personnel, the authors’ propose that a near real into account local hazard vulnerabilities. In time damage assessment could be possible. The conjunction with IKONOS panchromatic im- analytical methods and data inputs used in this ages, De La Ville et al. (2002) used GIS to study are relatively simple, yet effective, and evaluate the distribution of landslide erosion are therefore considered appropriate/suitable scars and their effects on several urban areas for disaster managers in developing countries situated among six mountain catchments in Ven- given that suitable data availability and delivery ezuela. The GIS was used to analyze and map methods are in place. the distribution of scars and deposition zones,

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 38 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

In general, GIS are well suited to provide recovery) has repeatedly been emphasized (e.g., valuable information to assist disaster response Becking, 2004; Chen et al., 2005; Jayaraman et operations. For example, GIS has been used to al., 1997; Mansor et al., 2004; Rivereau, 1995), determine the extent of a disaster and estimate and is reflected in over 400 scientific articles damage (Ranyi & Nan, 2002), organize resource between 1972 and 1998 (Showalter, 2001). For inventories and their geographic distribution instance, remotely sensed imagery can answer (Hussain et al., 2005), monitor shelter/refugee questions such as: what did/does the area look camp status and the state of transportation like pre-/post-disaster? RS data can show the infrastructure (Gunes & Koval, 2000) and in- land cover and topographic features in an area tegrate disparate spatial data sources that may and can illustrate infrastructure and population be required to guide response (Amdahl, 2001). density (Becking, 2004). Showalter (2001) As such, GIS can help with search and rescue, provides an in-depth review of the use of RS providing medical services, debris removal, in hazard and disaster research, and concludes sheltering, and infrastructure repairs. However, that the technique is primarily used to detect, a large number of spatial data layers are required identify, map, survey and monitor existing for planning and coordinating such operations, hazards and/or their effects; secondary goals and without them the value and usefulness of a of RS focus on damage assessment, improved GIS decreases substantially, possibly to a point planning, or the provision of data for disaster where it may no longer be required. Therefore, management functions. Simonovic (2002), in to some extent, the usefulness of GIS follow- examining the repeat frequency, spatial resolu- ing a disaster hinges upon the existence of the tion, and types of sensors on-board, provides required framework datasets like roads, critical insight into the suitability of specific satellites facilities, and population density. to different natural disasters. Multispectral Finally, Zerger and Smith (2003) empha- scanners (optical sensors) and radar collection size that the suitability of GIS for planning systems are probably the two most widely rec- verses real-timeIGI applications GLOBAL is quite different. ognized PROOF RS capabilities that are used in support Results from a test scenario indicated that the of disaster management (GDIN, 1997). utility of GIS for real-time decision making is The applicability of remote sensing for di- questionable owing to a number of practical saster management is perhaps best exemplified and implementation impediments, including in the case of flooding, and many researchers a lack of training and the need for temporal address its use for this disaster type. Satellite resolution rather than spatial detail. Using GIS imagery can be used to assess the extent of past for pre-disaster management functions, such as flood events (Dewan, Islam et al., 2007) and vulnerability assessment or evacuation route aid in the development of flood hazard potential planning, poses fewer challenges than trying maps. Zhang et al. (2002) and Jayaraman et al. to use the technology post-disaster, when time (1997) highlight the potential of RS technology is critical and ground conditions may be con- to drastically assist flood response and relief stantly changing. Nevertheless, assuming there operations by providing inundation mapping is adequate data and personnel to effectively and damage assessment. Disasters resulting utilize GIS, few would question its ability to from floods are a logical choice for RS analysis assist disaster response and recovery operations. because: (1) floods generally cover large areas, and thus occur at spatial scales much larger than Remote Sensing the spatial resolution of most satellite imagery, and (2) water has a unique spectral reflectance The potential of remote sensing (RS) to pro- which makes it clearly discernable from other vide critical earth observation information for ground features (Showalter, 2001). In contrast, disaster management (e.g., hazard assessment, earthquakes, for example, may cause significant disaster mitigation, preparedness, response and damage to buildings and infrastructure, but

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 39 without high resolution imagery or change found that Landsat 7 ETM+ data (30m resolu- detection capabilities it can be difficult or tion) could be used to identify and separate impossible to identify. out settlements, but that the detection of linear Remotely sensed digital elevation data features such as roads, railways and waterways (often termed DEM (digital elevation model) was unsatisfying. Integrating the panchromatic or DTM (digital terrain model)) are a digital band (15 m resolution) into the analysis failed representation of surface topography and are to significantly improve results. Thus, while frequently used in the study of natural hazards. Landsat imagery can be used to identify settle- DEMs are a critical data input for assessing ments it cannot be used as a source to accurately landslide susceptibility (Guinau et al., 2007), derive all spatial features on the ground. delineating flood risk potential (Dewan, Kabir Repeat frequency is also an important et al., 2007), flood hazard mapping (Dewan, consideration, as time becomes a critical factor Islam et al., 2007; Sanyal & Lu, 2003) and for in post-disaster use of RS imagery, some satel- a variety of coastal hazard (e.g., tsunami, storm lites may not be available within an appropri- surges, etc.) and disaster assessment purposes ate/suitable time frame (San Miguel-Ayanz et (for example, see Chen et al. 2005). DEMs al., 2000). For example, repeat frequency can are also commonly used to derive new data vary from 12 hours (NOAA) to 35 days (ERS) that is required for specific types of disaster (Simonovic, 2002). In addition, Cutter (2003) management related analysis or visualization. points out that the pre- and post-processing time The types of new datasets that can be generated for remote sensing images may negate their use include, but are not limited to: slope, aspect, in immediate response activities. Even if they contour lines, flow direction, flow accumula- are able to pass over a disaster affected region, tion, watersheds, solar insolation, viewsheds, cloud cover and the time of day (day vs. night) hillshade visualizations, and many others. can constrain the use of optical sensors for im- The most important factors that determine the age acquisition (Coppock, 1995). Therefore, the suitability of a DEMIGI for any particularGLOBAL disaster type of sensorPROOF onboard a particular satellite as management related application are the spatial well as the atmospheric conditions will partly resolution and vertical accuracy. determine the suitability of satellites for a given The potential of remotely sensed data to disaster related requirement. Even partial cloud assist disaster management is very clear, how- cover can be a significant obstacle that limits ever, there are some limitations and potential the use of optical imagery. In cloudy condi- obstacles, including image resolution, repeat tions satellites employing radar systems are frequency (temporal resolution) and the suit- required, but they too have their limitations ability of particular sensors. Perhaps the most (Simonovic, 2002) and are at times not suitable significant obstacle facing the use of satellite for acquiring the type of information that may imagery for a number of disaster management be required. Shadows can also be a limiting requirements is low pixel resolution. In large factor in both optical and radar imagery. Mag- urban centers, satellite image resolution can be sud et al. (2005) found that shadows prevented less than 1 meter, but in more remote areas and the visual identification of damaged buildings many parts of the developing world, only 15-30 using high resolution QuickBird imagery fol- meter Landsat or ASTER imagery are publicly lowing the Indian Ocean tsunami of 2004. The available (Nourbakhsh et al., 2006). Low im- same authors also note that intact roof tops age resolution can limit the range of potential can be deceiving, as initially certain buildings uses of RS data, including using it as a source appeared to be damage free, but upon ground to create digital datasets of basic, yet funda- truthing the building was essentially hollowed mental information for disaster management. out, with the roof top and support beams being For example, Rüdenauer and Schmitz (2005) the only remnant. Accordingly, caution must be used supervised classification techniques and used when drawing initial conclusions based on

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 40 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 visual imagery analysis, and ground truthing the US WorldView and QuickBird platforms, should be considered before any significant Canada’s RADARSAT-2 and the ESA’s ERS-2 actions are taken. and Envisat sensors. Moreover, the delivery of Recognizing the importance of high these datasets was facilitated by Google (http:// resolution satellite imagery in disaster manage- www.google.com/relief/haitiearthquake/ ment, many major space agencies and satellite geoeye.html) and GeoEye Inc. (www.geoeye. operators signed the International Charter on com) to leverage Google’s web-presence and Space and Major Disasters by the year 2001. bandwidth to provide 100s of GBs of high This multilateral agreement stems from the fact resolution GeoEye 1 and IKONOS images. In that no single operator or satellite can match addition, many Google pages like the Google the data-related challenges of natural disaster Earth Library (http://www.gelib.com/haiti- management. The International Charter aims earthquake.htm) offered critical infrastructure at providing a unified system of space data data for direct viewing in Google Earth format- acquisition and delivery to those affected by ted KML. The wide availability of post-disaster natural or man-made disasters (Space Agen- KML datasets allowed for rapid assessments by cies, 2005). The charter has been successfully non-GIS experts, thereby, and it is too soon to evoked on numerous occasions by countries all tell, increasing the effectiveness of relief efforts. around the world – from Ecuador to India to Overall, the geospatial response to the Haitian Russia – providing indispensable high resolu- disaster was quite rapid when compared to the tion imagery for no fee. For example, in the 2004 Indian Ocean tsunami, largely due to the aftermath of the Indian Ocean tsunami several need for rapid geospatial assessments during agencies and private companies provided RS disasters. By way of illustration, in December imagery for response and relief work (ESRI, 2006, the United Nations General Assembly 2006). DigitalGlobe provided 60-centimeter established the United Nations Platform for QuickBird images, which were the highest Space-based Information for Disaster Manage- resolution commercialIGI satelliteGLOBAL images avail- ment PROOF and Emergency Response - UN-SPIDER able. Although developing countries cannot (http://www.un-spider.org/haiti). A visit to UN- send satellites into orbit, opportunities do exist SPIDER provides a plethora of metadata and for acquiring high resolution remotely sensed links to disaster mapping for Haiti. Numerous imagery for disaster management purposes. other private organizations like ESRI Inc. (http:// However, Magsud et al. (2005) point out that www.esri.com/haiti/) created a central earth- once satellite data are received by the coor- quake response in support of Haiti, providing dinating office it is then up to the authorized maps and mapping support for the relief phase. user to effectively use the data – there are few By contrast, the first web-based mapping and mechanisms in place to help countries in need of delivery platform for the 2004 Indian Ocean support to use the data. In developing countries tsunami was the Tsunami Disaster Mapping this can pose a problem because potential data Portal launched on January 10th, 2005, and users are not fully aware of the capabilities of was led by DM Solutions Group, collaborators GIS and remote sensing, and they require ad- including researchers and scientists from the ditional assistance in order to put the data to Laboratory for Applied Geomatics and GIS good use. Moreover, the International Charter Science (LAGGISS), University of Ottawa, does not provide free imagery for pre-disaster Asian Institute of Technology (AIT), Pathum GIT operations. Thani, Thailand, Chulalongkorn University, The Haitian earthquake of January 2010 Bangkok, Thailand, and Media Center, Osaka was an example of the International Charter City University, Osaka, Japan. By comparison, on Space and Major Disasters being invoked there were well over 45 map-based websites or and imagery being acquired from satellites data delivery sites 14 days after the devastation including Japan’s ALOS, CNES’s Spot-5, in Haiti. Our point here is that there has been a

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 41 significant amount of incremental knowledge termath of a disaster. Organizations such as the gained about the critical nature of geospatial data Global Disaster Information Network (GDIN) needs in a time of disaster and this recognition provide evidence of the importance and value has led to a greater degree of cooperation in of disaster-related information, as well as the terms of data availability, timeliness and mode of need to be able to obtain and share it effectively. delivery for different practitioner communities. The integration of GIS and the Internet In sum, RS technology acts as a tool to first began in the early 1990’s (Plewe, 1997), gather spatial information that can support many and has grown rapidly since that time. This disaster management functions, especially when growth is partly attributable to advancements combined with other spatial data (points, lines in computer and information technology and and polygons) in a GIS. Satellite imagery is the building of spatial data infrastructure (SDI) critical for natural disaster response in remote worldwide (Yang et al., 2005). It is therefore and inaccessible regions, or regions where pertinent in this review to briefly define and primary (or basemap) geospatial data are non- describe the capabilities of Internet GIS, discuss existent or difficult to collect. In such regions, the implications of this relatively new technol- satellite imagery often offers the best means of ogy on the field of disaster management, and obtaining the necessary information (Kerle & examine its potential from a developing coun- Oppenheimer, 2002). This is particularly true in tries perspective. many developing countries which tend to lack In this emerging field of Internet GIS, there digital spatial data infrastructures (Brodnig & is no general agreement on the terms used to Mayer-Schönberger, 2000; ESRI, 2006). describe GIS-based programs on the Internet. They are commonly referred to as Internet GIS, Internet GIS web-based GIS, distributed GIS and On-line GIS, among others. Peng and Tsou (2003) help Internet GIS (IGIS), the integration of GIS and to better differentiate between these terms and the Internet, has quicklyIGI evolved GLOBAL over the last classify thePROOF various types of applications. The decade and has achieved significant recognition term ‘Internet GIS’ (IGIS) will be used hence- within the disaster and emergency manage- forth herein and is defined as network-based ment community for two main reasons. First, geographic information services that can utilize GIS provides a way to centralize and visually wired or wireless Internet protocols to access display critical information relevant to a di- geographic information, spatial analysis tools saster, since most of the data requirements are and GIS Web services (Peng & Tsou, 2003). It of a spatial nature and can be located on map is worth noting here that although the term GIS (Johnson, 2000). Second, the Internet provides is included with “Internet”, IGIS focus mainly an ideal medium for multiple users with a range on displaying geographic information (in map of backgrounds and skill sets to access spatial form) as well as data dissemination tasks but tend information and mapping capabilities. Radke et to lack comprehensive GIS capabilities common al. (2000) emphasize that data acquisition and to most desktop software. Kraak (2004) points integration may be the single largest contribu- out that most IGIS applications currently in use tion area needed for emergency and disaster are limited to (interactive) mapping (with zoom, response – IGIS can help address both these pan, measure distances, identify spatial features, issues. During the response phase, access to etc.), although some offer basic GIS functions pertinent spatial data/information is among the such as address matching, proximity searches, most essential requirements (Amdahl, 2001; and route planning like Google Maps for ex- Jayaraman et al., 1997). Disaster managers, lo- ample. However, more recent advancements cal authorities, aid workers and the public need are allowing for the development of Internet up-to-date data/information to enable quick and distributed GIServices with the capabilities to effective decision making in the immediate af- interact with multiple and heterogeneous sys-

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 42 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 tems and servers that support more advanced and difficult to update and distribute to all par- GIS functions (Dragićević, 2004). Such capa- ties involved. bilities are reflected in the rapidly developing Experience has shown that a top-down ap- standards for web-based GIS services (http:// proach to data sharing is not entirely effective www.ogc.org). (Radke et al., 2000), as disaster responders must The characteristics and capabilities of gain access to a number of department manag- IGIS offer considerable potential in the field ers and organizations, their unique maps and of disaster management, particularly during the data. Knowing which data is where and how response phase when access to spatial informa- to effectively access it can be a significant bar- tion is a key requirement (Radke et al., 2000). rier. Furthermore, this type of approach often Disaster impacts can significantly alter the results in duplication of efforts. For example, landscape (natural and built environments), and following the Indian Ocean tsunami, many often vary across organizational, sociological, agencies created damage maps at almost the political, and geographic boundaries. This is same time (ESRI, 2006). An alternative option especially true following large-scale disasters. is to utilize current IGIS architectures to create Information about the disaster zone needs to be a disaster information system with wide acces- disseminated to many stake-holders involved sibility, which allows multiple users to access in the response, at local, regional, national and relevant spatial data/maps provided by a variety international levels, both public and private of organizations through distributed servers/ (GDIN, 1997). Much of this information is spa- sources. An IGIS-based mapping system that is tial, such as the location of the most devastated intuitive, user-friendly and easy to access could towns or villages, the location of supply drop-off help provide geographic awareness to disaster points and the extent of in-tact transportation responders, many of whom have limited or no networks. IGIS are well suited to fulfill these GIS experience but could certainly benefit from information requirements since (1) they allow access to spatial information and maps. Andre for the integrationIGI of various GLOBAL spatial datasets and PROOFSmith (2003) note that disaster respond- and (2) they can be accessed from any location ers often require simple cartographic products with an Internet connection. (for example, the location of impassable roads The often ad-hock nature of disaster identified on a road network map) as opposed to response involves people and organizations products derived from advanced spatial analysis. (local to international) that require spatial Experience to date indicates that IGIS information relevant to their own logistical has played a key role in the collection and operations. Comfort (2000) reports that dur- dissemination of spatial information during ing disaster response there are often dynamic the disaster response and recovery of recent and spontaneous actions taken by responding natural disasters, including the Indian Ocean organizations and individual groups of people. tsunami (2004) and Hurricane Katrina (2005). These types of efforts are encouraged, but could However, the success/failure of IGIS in that be more effective and better coordinated with regard must be interpreted within the context an improved spatial awareness of the disaster of the very young IGIS field. With that caveat zone, achieved through the use of appropriate in mind, following these natural disasters IGIS spatial information and maps accessed using sites were established that integrated a large IGIS-based websites. Becking (2004) states number of spatial datasets, including coastlines, that obtaining geographic awareness is one of satellite imagery, damage maps, transportation the most important steps in effectively under- networks, population centers and other data standing a disastrous situation, and being able intended to provide a visual overview of the re- to make appropriate decisions. Traditionally, gions impacted by the disasters. Figure 3 shows geographic information has been distributed images of the main interface of two such sites. using paper maps, which are costly to produce

Figure 3. Examples of two IGIS-based natural disaster map viewers: a) South East Asia and Indian Ocean Tsunami Response Map Viewer (produced by DM Solutions Group, http://www. dmsolutions.ca/showcase/), and b) Hurricane Katrina Disaster Viewer (produced by ESRI Inc., Redlands California, USA)

The user friendly map based information al., 2006), which reduces the potential for ac- systems created using IGIS can increase access cessing information using IGIS sites. to relevant spatial information required during In general, the ability to develop an IGIS for disaster response. However, their design, de- disaster response requires excellent GIT skills velopment and implementationIGI GLOBAL require sub- and a large PROOF amount of spatial data, which may stantial GIS/technical knowledge and signifi- exist at differing scales and in varying formats. cant human and financial resources. As Currently, the GIT requirements and skills as- demonstrated earlier, in many developing sociated with creating IGIS are beyond those of countries each of these requirements represents disaster managers in developing countries, and a challenge, and could prevent them from cre- we suggest that trying to implement IGIS may ating IGIS for disaster response. In fact, the not be the best use of their limited resources. two known IGIS sites created to assist with the Indian Ocean tsunami response were not created by any developing country that was af- GEOSPATIAL INFORMATION fected by the disaster; one was created by a TECHNOLOGY SOFTWARE Canadian software development company (DM We have reviewed many uses and barriers of Solutions Group – http://www.dmsolutions.ca) geospatial information technology (GIT) to and the other by the Pacific Disaster Center, natural disaster management. However, there which is based out of Hawaii. Both have sig- has been no discussion regarding specific as- nificant GIT resources and capabilities, and pects of GIT software, including the difference were able to establish these IGIS sites in a between commercial and free and open source relatively quick period of time. This is an im- (FOS) software (FOSS), and the implications of portant point, as it is unlikely that GIT practi- these differences for adoption within developing tioners in developing countries have the exper- countries. Turning our attention to that issue, tise required to effectively develop an IGIS site we now discuss the main differences between in a reasonable time frame. In addition, Internet commercial/proprietary software and FOSS, connections in many parts of the developing and explore the significance of these differences world can be slow and patchy (Nourbakhsh et

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 44 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 from a developing country perspective. We also freedom, not necessarily zero price, although briefly discuss the FOS GIT software domain. almost all are available at no cost. Free software Although there are some well documented implies that the user has the freedom to run, drawbacks to using FOSS, and challenges to view, copy, modify, and distribute a piece of overcome, the potential benefits that can be software, irrespective of financial limitations. gained far outweigh any downside, and thus Thus, users can improve the source code, by FOSS should be considered a viable model either enhancing its existing functionality or by (Fitzgerald, 2004; Hoe, 2006). adding new functions. In contrast, proprietary software indicates ownership, and has its source Free and Open Source code closed – it cannot be viewed, modified Software (FOSS) or redistributed, as is stated in the EULA (End User License Agreement) that you must agree While some consider “free” and “open source” to prior to installation. Free, as in price, is software to be something close but not identical, perhaps the most significant characteristic of they are similar enough that for the purposes FOSS that differentiates it from commercial of this paper they are considered the same. We software, which is inherently sold for profit. can lump them together because for the gen- Another characteristic of FOSS is that it can eral software user the difference is negligible, be accessed (downloaded) from anywhere with although philosophical and legal differences an Internet connection; this is typically not true do exist (Rajani et al., 2003; Steiniger & Bo- of commercial/proprietary software. Table 2 cher, 2009). Although there are a number of outlines some of the key differences between criteria that a software package must meet to proprietary and FOS software products, in terms be considered FOSS, three essential features of advantages and disadvantages. With certain capture the essence of the semiofficial “Open types of software, including GIT software, there Source Definition” (http://www.opensource. are also considerable differences in terms of org/docs/osd):IGI GLOBALfunctionality PROOF and ease of use. The FOSS movement is continually gain- • The source code must be distributed with ing momentum, but has already garnered sig- the software or otherwise made available nificant attention in some areas. For example, for no more than the cost of distribution. Apache dominates the web server market, as • Anyone may redistribute the software for over 65% of all active web sites use it (Weber, free, without royalties or licensing fees to 2004). Ramsey (2007) points out that its success the author. can be linked to a powerful user community • Anyone can modify the software or derive that is committed to maintaining the Apache other software from it, and then redistribute platform. Corporate mainstays such as IBM the modified software under the same terms and HP, government agencies and academic (Weber, 2004). contributors are all part of the Apache community. Other FOSS projects that have become Licensing agreements such as the General very popular include office suites such as Ope- Public License (GPL) define the rights users nOffice, database systems like MySQL, the have over the software product (http://www. Mozilla web browser, and the Linux operating gnu.org/licenses/gpl.html). Wu and Lin (2002) system. As of February 2009, there were over as well as Cook and Horobin (2006) provide 230,000 registered projects and over 2,000,000 further information about the various FOSS registered users at Sourceforge.net, the world’s licensing models. largest open source software development What makes FOSS different than pro- website. In fact, most users can find an applica- prietary or commercial software is that the tion that will exactly meet their specific needs source code is “free.” In this context free means (Wu & Lin, 2001). According to Wheeler (2007),

Table 2. Key differences between proprietary and FOS software

Proprietary Software Free and Open Source Software Advantages warranty of developing company on no licence fees product (holds for every computer) unrestricted use (e.g., no limits for the number of components should work together installations) well documented software no update enforcement regular release times for new versions. support of open standards regular service packs support usually available from several providers customization at the API level customisation at source code level platform independent Disadvantages software price and maintenance fees installation know-how necessary in many cases training costs training costs maintenance fees tied to specific interoperability issues between FOSS licensed companies, software options quality (but self-correcting if actively used) and time period no responsible authority customised development can be dif- support can lack for some packages ficult due to available resources of vendors support only as long as software company exists Some limitations on out-of-box functionality where vendor partner’s are necessary for upgraded functionality reliance - retraining costs when software versions change (e.g., ESRI’s ArcView3.2 to ArcGIS 8.x) or data models change at vendor’s whim or IGIdevelopment GLOBAL cycle PROOF who provides quantitative data, FOSS reliabil- deploy FOSS freely offer advice to one another, ity, performance, scalability, security and total sharing insights and lessons learned (Fitzgerald, cost of ownership are at least as good as or 2004). As a result, solutions for many typical better than its proprietary competition, and problems can often be found at no cost. When under certain circumstances are superior. This they cannot, there is a growing FOSS support evidence suggests that FOSS can, and already and custom development industry that can be is, rivaling certain commercial software do- utilized in such circumstances. mains. Another concern about FOSS relates to its Although a strong case can be formed for long-term survival. Fitzgerald (2004) reports the FOSS model, it also has some often cited dis- that studies of Freshmeat.net and Sourceforge. advantages compared to commercial software. net (two popular FOSS development websites) For starters, software support and technical as- revealed that most projects have only one or sistance is a major issue facing the adoption of two developers, and that follow up studies FOSS. However, FOSS proponents are quick to reported no change in version number or size point out that support may be found within the of code base for many listed projects several FOSS community, in the form of user groups and months later. This sort of vague analysis of the archives of past queries and answers available FOSS domain can be deceiving, since there on the Internet (Ramli et al., 2005). Installation are so many FOSS projects it is not surprising and user documentation is also common among that most have only one or two developers, and the more mature FOSS products (Steiniger & that new versions are not coming out regularly, Bocher, 2009). Furthermore, organizations that especially if the particular package is sufficient

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 46 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 for its purpose. That is the nature of the FOSS the MapServer Foundation (Schutzberg, 2005). model, as it encourages individuals or small The Foundation is expected to provide a stable teams to develop and share software. It is up to infrastructure for the now extended MapServer potential users of FOSS to examine individual family’s code base and its growing community. products, consider potential advantages and The FOS GIT software community is disadvantages, and choose what best fits their steadily growing, and since 2006 has been needs. Câmara and Onsrud (2004) examined spearheaded by the Open Source Geospatial FOSS GIT and identified many differences Foundation (OSGeo) (http://www.osgeo.org). that exist – in terms of support, maturity and Their mission is to support and promote the functionality – between products led by a single collaborative development of open geospatial individual, products produced by small teams technologies and data (OSGeo, 2008). The and corporate led products. They conclude OSGeo hosts an increasing number of software that corporate led projects tend to be of better projects, publishes the OSGeo journal, founded quality, at least from these three standpoints. an education and curriculum committee, and presents the annual Free and Open Source Soft- Free and Open Source Geospatial ware for Geospatial (FOSS4G) international Information Technology Software conference (Steiniger & Bocher, 2009). The comprehensive list of links to FOS Turning our attention now to the specific do- GIT related software projects/products available main of FOS GIT software, there is reason for at opensourcegis.org and freegis.org provides optimism. As opposed to questionable long-term evidence of an active development community. survival, Steiniger and Bocher (2009) empha- Some notable GIT programs, which offer a range size the increasing interest and development of functionality from simple data/map viewing of FOS GIT products. They point out in their to more advanced spatial analysis and Inter- overview of FOS GIS (e.g., gvSIG, Quantum net GIS capabilities, include: Quantum GIS, GIS, SAGA, uDig,IGI GRASS, GLOBAL etc.) that: DIVA-GIS, PROOF OpenEV, uDig, gvSIG, GRASS (Geographic Resources Analysis Support Sys- a. Four out of ten desktop projects examined tem), MapServer and OSSIM (Open Source receive governmental funding support; Software Image Map). Other programs focus on b. There is an increase in the download rate more specific tasks, such as data management, of FOS GIS software; and format processing, geostatistical analysis and c. There is an increasing number of use cases data visualization. Currently, the freegis.org of FOS GIS website contains information and links to over 300 FOS GIT projects. Ramsey (2007) provides Furthermore, Ramsey (2007) points out an excellent review of some of the more mature that “existing products are now entering a projects within the FOS GIT software domain, phase of rapid refinement and enhancement… categorized by development/implementation (FOS) software can provide a feature-complete languages, such as C, Java and .Net. For a more alternative to proprietary software in most in-depth overview of current FOS desktop GIS, system designs.” Some commercial software in terms of organizations, software groups and manufacturers are even starting to back FOS functionality, see Steiniger and Bocher (2009). GIT initiatives, which is encouraging especially from a user support and longevity perspective. In FOSS and Developing Countries late 2005 the software industry giant Autodesk, in association with the MapServer community From the perspective of developing countries and DM Solutions Group, announced that it the FOSS model is a particularly good fit, for would support and promote open source web reasons that include: cost, freedom, accessibil- mapping (a form of IGIS) through the creation of ity, customizability, compatibility, capacity

Table 3. Approximate costs of selected commercial GIS/RS software (the wide price ranges reflect possible variations in license type, software options [software extensions], purchasing agreements, etc.)

Application Price range (USD) Software High end GIS/RS $10, 000 to $50, 000 ArcGIS (ArcINFO) Low end GIS/RS $1, 000 to $3, 000 Genasys ENVI ERDAS Imagine ER-Mapper MapInfo ArcGIS (ArcView) ArcView extensions ERDAS Imagine Essential development and reducing the overall so-called software in the case of developing countries, ‘digital divide’ (Hoe, 2006). The lower cost of and could also help reduce their reliance, and FOSS is definitely the most significant factor inadvertent support and encouragement, of the that makes it attractive to developing countries illegal software market. (Rajani et al., 2003), with the absence of licens- The freedom to access and study the source ing fees seen as a major benefit (Hoe, 2006). code is another fundamental advantage of using The software that comprises the commercial FOSS. As a result, the choice to use FOSS is GIT domain is particularly expensive. Table not only a software choice, but also a means of 3 provides a few examples of some popular acquiring knowledge about the software itself. commercial GIS and remote sensing software In developing countries this is important for applications and theirIGI general GLOBAL price ranges. capacity PROOFbuilding of the local population, and Even the most basic and required propri- can help them better understand and deploy etary software necessary for many computing new technologies successfully (Hoe, 2006). needs, such as Windows XP and MS Office, Mohamed and Plante (2002) emphasize that can be too expensive to purchase in a develop- local workforce development and capacity ing country (Roets et al., 2007). With that in building are critical for system maintenance and mind, it is not surprising that the highest software operation over the long run. In addition, using piracy rates occur in developing countries such FOSS can contribute to the development of the as Vietnam, China and Indonesia, with piracy local information and communications tech- rates of 97, 94 and 89 percent, respectively nologies (ICT) industry (of which GIT is a part (Rajani et al., 2003). In such countries is not of), as support, development and maintenance unusual to have new computers come with a contracts can be provided by local businesses range of pre-installed pirated proprietary soft- that offer FOSS services (Cook & Horobin, ware (Weerawarana & Weeratunge, 2004). This 2006). This will, in turn, provide more jobs in evidence demonstrates that users in developing the FOSS industry, attract trained professionals, countries most often don’t, and perhaps more and help stimulate the local economy. importantly, cannot afford to pay for computer The freedom to customize the source code software. Although piracy is very common, it is also important from a developing country actually devalues the economic benefits of perspective. Krakowski (2006) points out that FOSS by falsely reducing the price of propri- proprietary software developed within the West- etary software (Weerawarana & Weeratunge, ern world is designed to fit well with Western 2004). The zero cost of FOSS should be con- culture. As a result, it may not be well-suited for sidered a major advantage over proprietary use in other regions of the world, where local

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 48 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 customs and practices may be quite different. data storage device such as a USB or external With FOSS, applications can be tailored to meet hard drive. This makes computer software the specific needs of the users, and take into ac- accessible to potential users who would have count cultural variations and social practices that absolutely no ability to purchase or otherwise deviate from industrialized societies. Similarly, obtain such software. Second, the freedom to the majority of people in developing countries do install FOSS on as many computers as is desired not understand English, yet proprietary software is another fundamental benefit not offered by is often only available in English (Hoe, 2006; proprietary software. If a single use license is Nonogaki et al., 2004). Clearly, interface and purchased from a proprietary company only documentation language represents a major one computer can legally run the software, and obstacle to users who do not understand Eng- multi-use licenses are expensive and considered lish; however, FOSS can be adapted to meet unaffordable for most software users in devel- local language needs and in doing so helps to oping countries, even governments (Roets et tackle the digital divide. For example, Ubuntu al., 2007). This limits the accessibility of the is an African adjusted Linux distribution that software, preventing it from being installed attempts to resolve language problems (Kra- and used on more than one computer. In an kowski, 2006). GNOME 2.22 – the default office environment, for example, this is likely desktop environment of Ubuntu – offers support to represent a significant problem. for 46 languages. Similarly, multi-language In sum, the literature that addresses issues capabilities have been incorporated in recent related to FOSS and its adoption in develop- versions of GRASS and MapServer (Nonogaki ing countries, in terms of advantages and et al., 2004; Raghavan et al., 2006) and gvSIG disadvantages, is diverse and abundant. The is available in more than ten languages. Ad- preceding discussion is meant to provide only ditional efforts need to focus on translation of a brief overview of some well acknowledged online help files, manuals and so on. and commonly discussed advantages that FOSS Another IGIadvantage GLOBALof FOSS is that they provide PROOF and elaborate on those specific to GIT. usually do not require the newest or best com- We acknowledge that there are many more ad- puter hardware in order to function efficiently vantages to FOSS in general for the developing and correctly. In contrast, and especially in the world that are beyond the scope of this paper to GIT software domain, proprietary companies discuss and so recommend reviews by Rajani are continually releasing new versions that de- et al. (2003) and Weerawarana and Weeratunge mand the newest and most advanced hardware. (2004), or Cook and Horobin (2006) who In developing countries most users do not have examine FOSS from a public administration access to powerful, high-end computers with (government) perspective. the most current hardware available (Calde- weyher et al., 2006). Therefore, even if they could purchase sophisticated GIT software they DISCUSSION would also require computers with sufficient Considering that disasters are spatial phenom- RAM and CPU power to take advantage of the enon, it would be reasonable to assume that advanced capabilities that are offered in most GIT can be effectively employed throughout commercial GIT packages. all phases to assist the disaster management A couple of final remarks about the po- process. While this is true in the case of de- tential of FOSS for developing countries relate veloped countries, the same cannot be said for to accessibility and unlimited installations. developing countries. The use of GIT for disaster First, FOSS can be downloaded (accessed) management in developing countries is limited from anywhere with an Internet connection; by a number of well documented barriers, and where the Internet is unavailable FOSS including: a lack of financial resources (Renyi can be delivered on a CD or provided via a

& Nan, 2002), a lack of local GIT expertise/ etary GIT software. Table 4, containing search knowledge (Mohamed & Plante, 2002), insti- results from the SCOPUS database, provides tutional/political instability (Pande, 2006), and quantitative evidence of the difficulty in find- a shortfall of spatial data (ESRI, 2006; Murgia ing literature that applies FOS GIT in the field et al., 2002). The degree to which each of these of natural disaster management in developing GIT implementation barriers limits the use of countries. Although the concepts of GIT have GIT varies between developing countries, and existed for decades, and are now reflected in a within each country depending on the admin- large commercial GIT software market, those istrative level. concepts have only recently been transformed We have examined the literature and dis- into capabilities within FOS GIT, with the ex- cussed various aspects of research on natural ception being GRASS. However, GRASS was hazards and disaster management. Developing originally developed by the U.S. Army Corp of countries are highly vulnerable to natural di- Engineers in the early 1980s and is currently sasters, a vulnerability that could be addressed used in academic and commercial settings partially through GIT. However, there are nu- around the world, and by numerous government merous barriers to GIT implementation in the departments (e.g., USGS and NASA) (http:// developing world. We reviewed the application grass.itc.it/intro/general.php). As a result of of GIT in natural disaster management and its long use and development (which is now focused on describing the significant potential multi-national) history it has evolved into a so- for GIS, RS and IGIS to assist the disaster phisticated and powerful raster/vector GIS, used management phases. We discussed the concept for geospatial data management and analysis, of FOSS, and argued that they have reached image processing, graphics/map production, a stage a maturity that makes FOSS a viable spatial modeling, visualization and much more. alternative model over commercial software, Although GRASS is very powerful, its main especially in developing countries. The FOS drawback is its unfriendly user environment, GIT software domainIGI is steadily GLOBAL growing and as it functions PROOF primarily through command there are already many mature projects that cur- line operations which can be a challenge for rently offer a range of functionality (Raghavan non-expert users (Ramli et al., 2005; Steiniger et al., 2006; Steiniger & Bocher, 2009). & Bocher, 2009). Given the continued growth of FOS GIT, Furthermore, a key word search of ‘hazard’ we propose that its current availability offers and ‘disaster’, independently, of all presentation increasing potential to the local level disaster titles from the FOSS4G 2006 and 2007 http:// management practitioner community in de- www.foss4g2006(2007).org/) conference veloping countries. In this section we discuss yielded only three that contained ‘hazard’ and why and how FOS GIT can, and should, be one that contained ‘disaster.’ This is surprising implemented at the local level in developing given the increasing frequency of natural di- countries in order to improve disaster man- sasters and the critical role that GIT plays agement capacity, and propose future research within the various management phases. How- needs in this area. ever, although very few presentation titles in- The cosmopolitan existence of FOSS is clude the word ‘hazard’ and/or ‘disaster’, it is a relatively new phenomenon, as are the ap- likely that many of the topics discussed do plications that are emerging. Currently, within contain recent research or software development the literature that addresses FOS GIT, there relevant to the study and/or management of appears to be very little that examines its use in natural hazards and disasters. For example, the specific research domain of natural hazards addressing how FOS GIS software can contrib- and disaster management, and this finding may ute to SDI development is relevant to disaster be related to the fact that the entire FOS GIT management, since the required spatial data are domain is quite new compared with propri- often lacking, although an explicit connection

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 50 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

Table 4. Literature themes search results from Scopus (http://www.info.scopus.com)

LITERATURE ADVANCED SEARCH PHRASES USED # OF RE- THEMES (in: title-abstract-keyword) SULTS Natural Hazards and Disasters ((“disaster*” OR “natural disaster*” OR “natural 254,109 hazard*” OR “environment* hazard*” OR 459, 881 “flood*” OR “earthquake*” OR “storm*” OR “hurricane*” OR “tsunami*” OR “landslide*”)) + “hazard*” (which includes hazards that are not of natural origin) Natural Hazards and Disasters ((“disaster*” OR “natural disaster*” OR “natural 54,283 + hazard*” OR “environment* hazard*” OR Disaster Management “flood*” OR “earthquake*” OR “storm*” OR “hurricane*” OR “tsunami*” OR “landslide*”) AND (“disaster* manage*” OR “hazard* manage*” OR “mitigation” OR “prepare*” OR “respon*” OR “relief”)) Natural Hazards and Disasters ((“disaster*” OR “natural disaster*” OR “natural 4,994 + hazard*” OR “environment* hazard*” OR Disaster Management “flood*” OR “earthquake*” OR “storm*” OR + “hurricane*” OR “tsunami*” OR “landslide*”) Geospatial Information Technology (GIT) AND (“disaster* manage*” OR “hazard* manage*” OR “mitigation” OR “prepare*” OR “respon*” OR “relief”) AND (“geo* info* sys*” OR “GIS” OR “geoinfo*” OR “3S” OR “geospatial” OR “geo* info* tech*” OR “remote sens*” IGI GLOBALOR “RS” ORPROOF “earth observation*” OR “EO” OR “satellite” OR “spatial*”)) Natural Hazards and Disasters ((“disaster*” OR “natural disaster*” OR “natural 769 + hazard*” OR “environment* hazard*” OR Disaster Management “flood*” OR “earthquake*” OR “storm*” OR + “hurricane*” OR “tsunami*” OR “landslide*”) Geospatial Information Technology (GIT) AND (“disaster* manage*” OR “hazard* man- + age*” OR “mitigation” OR “prepare*” OR “re- Developing Countries spon*” OR “relief”) AND (“geo* info* sys*” OR “GIS” OR “geoinfo*” OR “3S” OR “geospatial” OR “geo* info* tech*” OR “remote sens*” OR “RS” OR “earth observation*” OR “EO” OR “satellite” OR “spatial*”) AND (“develop* countr*” OR “develop* region*” OR “develop* nation*” OR “transition* countr*” OR “third world” OR “less develop*” OR “East* Europe” OR “Africa” OR “Asia” OR “South America” OR “Central America” OR “China” OR “India” OR “Pakistan” OR “Indonesia” OR “Malaysia” OR “Thailand” OR “Bangladesh”))

continued on the following page

Table 4.continued

Natural Hazards and Disasters ((“disaster*” OR “natural disaster*” OR “natural 15 + hazard*” OR “environment* hazard*” OR Disaster Management “flood*” OR “earthquake*” OR “storm*” OR + “hurricane*” OR “tsunami*” OR “landslide*”) Geospatial Information Technology (GIT) AND (“disaster* manage*” OR “hazard* man- + age*” OR “mitigation” OR “prepare*” OR “re- Developing Countries spon*” OR “relief”) AND (“geo* info* sys*” OR + “GIS” OR “geoinfo*” OR “3S” OR “geospatial” Free and Open Source Software (FOSS) OR “geo* info* tech*” OR “remote sens*” OR “RS” OR “earth observation*” OR “EO” OR “satellite” OR “spatial*”) AND (“develop* countr*” OR “develop* region*” OR “develop* nation*” OR “transition* countr*” OR “third world” OR “less develop*” OR “East* Europe” OR “Africa” OR “Asia” OR “South America” OR “Central America” OR “China” OR “India” OR “Pakistan” OR “Indonesia” OR “Malaysia” OR “Thailand” OR “Bangladesh”) AND (“open source” OR “OS” OR “open source software” OR “OSS” OR “free software” OR “FS” OR “free and open source” OR “FOS” OR “free and open source software” OR “FOSS” OR “F/OSS” OR “free/libre and open source software” OR “FLOSS” OR “open source software/free software” OR “OSS/FS” OR “FOS- S4G” OR “GRASS” OR “gvSIG” OR “Quan- tum” OR “SAGA” OR “uDig” OR “ILWIS” OR IGI GLOBAL“JUMP” OR “DIVA” PROOF OR “KOSMO”)) may not necessarily be made. Coppock (1995), and more specifically, how they can be used in a study of GIS and natural hazards research, by the disaster management community in makes an analogous point, explaining that GIS developing countries. Again, we emphasize and natural hazards are multidisciplinary fields, that sufficient time has not elapsed to allow for and that much of what has been written appears the creation of a literature base in this research in the literature of those disciplines. area, yet strongly believe that more and more While a review of the application of FOS examples of successful FOSS implementation GIT in the area of natural hazards and disaster will undoubtedly emerge in the coming years. management is desirable, the lack of available This belief is based on the increasing interest literature (Table 4) makes this task nearly in FOSS and overall growth of the user and impossible. However, when you remove the development communities, through initiatives FOSS component, there is an abundance of such the OSGeo and websites like opensource- literature that utilizes proprietary GIT software, gis.org and freegis.org. as was demonstrated in the previous section on Among the broad set of Millennium De- GIT and disaster management, and in Table 4. velopment Goals that the United Nations estab- More common from the FOSS perspective is lished in 2000, one stands out: “Make available literature that emphasizes the great potential it the benefits of new technologies, especially offers in areas such as disaster management, information and communications technologies” without going into any more detail about (http://www.un.org/millenniumgoals). GIT specific applications or examples of success- are within the domain of ICT, yet peppered ful usage. Thus, further research is required throughout this current review are a number of regarding the current capabilities of FOS GIT, barriers that can limit GIT implementation in

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 52 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 developing countries. We propose that FOSS to institutional instability (Murgia et al., 2002), can reduce some of these barriers, and in doing which can result in a lack of commitment to ad- so, can also help in achieving the development dress disaster management requirements, such goals established by the UN. as providing adequate technology and training. Primarily, FOSS eliminates the need to As a result of this instability, Ramasubramanian purchase expensive software licenses and makes (1999) suggests that GIS implementation in GIT accessible to anyone, including local level developing countries is likely to be more suc- disaster managers operating on very small bud- cessful if it relies on the capacity of empowered gets. This point is of fundamental importance individuals and groups rather than solely on when considering that in many cases disaster organizational structures, such as public in- management responsibilities and duties are de- stitutions. To some extent, FOS GIT could be centralized to local governments without being harnessed to empower such groups, and allow accompanied by sufficient funding (Montoya them to benefit from technology without hav- & Masser, 2005). Second, there exists a lack of ing to rely only on institutional initiatives and local expertise and knowledge that is required financial support. to implement and maintain the use of GIT. Lastly, and perhaps most importantly, FOS GIT software could be used to learn about FOS GIT offers considerable potential to help the technology itself, and the potential ways build spatial data infrastructures required to in which it could be used for natural disaster fulfill basic disaster management mitigation management purposes. Raghavan et al. (2006) and preparation planning, and following a emphasize that previous FOS GIT technology disaster, speed response and recovery efforts. transfer was an arduous task for novice users, It has already been shown that many essential but recent developments in the availability of disaster management requirements hinge upon packaged solutions, such as FWtools (see map- the existence of required spatial datasets, such tools.org), and commercial support have helped as critical facilities, transportation infrastructure reduce implementationIGI obstacles.GLOBAL In addition, and PROOFpopulation distribution (location of towns many FOS GIT software products are available and villages), yet many of these key datasets as a single executable file, and once downloaded, are missing in developing countries. Especially are installed in the same manner as proprietary for assessing the situation in more remote re- software (e.g., point and click). Albeit, some gions, which are frequently only reached days software versions designed for Linux or other or even weeks after an event, a detailed inven- FOSS operating systems do require specific tory of settlements and infrastructure may installation instructions and a certain amount save many lives and speed up recovery (Kerle of computer knowledge. Furthermore, that & Oppenheimer, 2002). Montoya and Masser FOSS can be an effective educational/training (2005) emphasize the need to identify or de- tool is an idea that is widely acknowledged, velop cost-effective data collection methods given that they can be accessed at no cost and for producing spatially referenced information experimented with in whatever way users see in developing countries. Mansourian et al. fit. Local workforce development and capacity (2004) propose the development of a SDI as a building projects are essential for sustained GIT framework to facilitate disaster management, implementation and maintenance that does not and many currently available, and user friendly, rely on external support or funding. desktop FOS GIS (Steiniger & Bocher, 2009) In general, disaster management is depen- have the necessary functionality to assist the dent on the functional and effective operation SDI development process. of institutions, whether formal or informal, and Many of the barriers to successful GIT at the local level where it matters most (Pande, implementation in developing countries can 2006). However, government instability is potentially be addressed with FOSS. Cur- common in developing countries, which leads rently, there are many basic FOS GIT software

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 53 available and others are undergoing rapid de- this point, offer superior overall functionality velopment (Raghavan et al., 2006). However, and provide advanced analytical and visualiza- there still remains a need to identify particular tion capabilities that are beyond what current software products, establish feasible methodolo- FOS GIT software can provide, with GRASS gies and workflows, and in general document being a possible exception. However, GRASS specific examples of FOSS usage. While few is notoriously difficult to use and requires an would argue against the potential of FOSS, advanced GIT skill set, and as a result, it is not there exists a research need to clarify the role suitable for use at the local level in develop- of FOS GIT during each of the phases of the ing countries where such expertise is in short disaster management cycle. Regardless, a lack supply, and where an inability to successfully of attention to the aforementioned GIT imple- use the software may result in frustration and mentation barriers in any decision on whether discouragement. To avoid such a result, we to adopt GIT will significantly increase the risk suggest that simpler FOS GIT software could of system failures, a matter of some significance be implemented and utilized more success- in any environment characterized by a lack fully, and at the same time provide sufficient of financial, technical and human resources functionality to improve current disaster man- (Sliuzas, 1999). agement GIT requirements at the local level. This review suggests that the successful In addition, combining a number of FOS GIT application of GIT in natural disaster manage- software together can extend the functionality ment can require significant quality spatial data, in the long term. GIT usage must not be over GIT knowledge and technical skills, and/or ambitious in the early stages and should focus advanced functionality. Each of these require- on the core requirements needed to improve ments constitutes significant obstacles from a existing disaster management capabilities. developing country perspective, especially at Osti et al. (2008) make a similar point, and any administrative level other than national. suggest that low-tech systems, which are less Another noteworthyIGI fact is GLOBAL that the use of cost intensive PROOF and can be handled by the local proprietary GIT typically requires significant population, are of high interest in the context computing resources. This represents a huge of developing countries. obstacle in developing countries, which often Within the hazard and disaster management do not have access to modern (i.e., less than 2-3 literature that addresses developing countries years old) computer hardware (Caldeweyher et and GIT implementation issues there are a al., 2006). GIS users in developing countries couple additional topics that deserve mention- are likely unable to take advantage of advanced ing: the participatory approach (e.g., PGIS) spatial analysis and data visualization features and the concept of traditional/indigenous en- prevalent in current proprietary GIS software. vironmental knowledge. Both of these topics In comparison, FOS GIS software generally are also partly embedded within the technology require less processing power, and are more transfer literature, and although not within the suitable for users who do not have modern scope of this paper, we will briefly touch on computers. Thus, while advanced GIT usage these two topics since FOSS has a potentially is feasible in developed countries and within large role to play. academic/research environments (including With technology transfer there is a tendency some in developing countries), for the typical to assume that development occurs when less disaster management practitioner in developing developed countries adopt the technologies countries this type of high-end GIT software and methods of the more advanced countries usage is unachievable (Cutter, 2003). (Britton, 2000). However, this approach is Although we have argued that currently increasingly being recognized as ineffective available FOS GIT have great potential, we do (Brewer et al., 2005; Heeks, 2002). Although acknowledge that proprietary GIT software, at GIT implementation successes are often re-

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 54 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 ported in popular magazines or trade journals, suitable for the collection of spatial informa- their sustainability is seldom documented, and tion and to capture the perception of the local failures are not reported or glossed over to avoid people regarding disasters. A follow up of this criticism (Ramasubramanian, 1999). Heeks study revealed that in some communities, and (2002) explains that the introduction of GIS in without outside/donor influence, maps were developing countries has been problematic, as also used at meetings to discuss issues with a they are seen to incorporate a number of assump- spatial relevance. tions and requirements that derive from Western Generally speaking, participatory ap- rationalism. One way to possibly improve the proaches can help build local capacity to success of technology transfer initiatives is understand and implement GIT to improve through participatory approaches. Within the disaster management capabilities. Using a GIT domain, GIS has garnered considerable participatory approach local knowledge about attention from a participatory standpoint and natural hazards – their location and frequency thus will be used as an example. In terms of – can be ascertained from community members developing countries, Abbot et al. (1998) define and represented spatially within a GIS. For ex- PGIS as “an attempt to utilise GIS technology in ample, using a GPS coordinates of the location the context of the needs and capabilities of com- of previous floods can be captured, along with munities that will be involved with, and affected elevation data. Within the GIS the flood point by, development projects and programmes.” A locations can be analyzed in conjunction with central objective of the participatory approach elevation data (e.g., a DEM or contour lines), is to capture local knowledge and combine it and potential flood scenarios can be considered with more traditional spatial information, and and prepared for accordingly. Combing locally to facilitate GIS production and use which are derived hazard knowledge with other spatial community-based. Creating GIT environments data such as transportation infrastructure, criti- that pull in new ideas, and possibly new spatial cal facilities and towns/villages within a GIS methods and IGI techniques, GLOBAL rather than having environment PROOF will help the local population to them pushed in from the outside, appears to be more prepared for and help mitigate poten- be a more effective solution (Britton, 2000). tial disaster impacts. Ultimately, the goal is to Developing countries could benefit from inter- reduce disaster vulnerability, and this can be national development programs that focus on achieved with a better spatial understanding capacity building in the area of FOSS-based of the relationship between natural hazards and participatory GIT implementation. Knowledge- the elements at risk (e.g., people, infrastructure, able practitioners in developing countries could etc.) (Montoya & Masser, 2005). then further build local capacity, especially with In the past, there has been unwillingness to FOSS, which have been shown to be a good use ‘non-scientist’ or ‘indigenous knowledge’ fit for developing countries. If appropriately as data. However, there is growing recognition utilized, Rambaldi et al. (2006) emphasize that of the importance of indigenous knowledge, participatory approaches could exert profound and the need to better understand how it can be impacts on community empowerment, innova- combined with Western knowledge to reduce tion and social change. natural hazard vulnerability (Mercer et al., 2007; Kienberger and Steinbruch (2005) applied Pande, 2006). Local communities in developing a PGIS approach in Mozambique, in which the countries rely heavily on their environment; they main objectives were to acquire spatial data of depend on the subtleties of ecological cycles the communities through GPS surveys, to better and patterns, and have accumulated a body of visualize certain aspects of the community to wisdom commonly referred to as ‘traditional assist disaster risk committees, and to identify environmental knowledge’ (TEK) (Brodnig & secure locations in the case of floods. Results Mayer-Schönberger, 2000). “TEK is in essence indicate that the participatory approach was a geographical information system derived from

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 55 and embedded in the close relationship of local assessment/mapping, vulnerability assessment, people with their land and natural resources... vehicle dispatch and supply routing, damage with members of the community serving as assessment, and response resource mobiliza- repositories for different types and categories tion. Consequently, by using GIT, it is possible of data…” (Brodnig & Mayer-Schönberger, to identify and mitigate risk, be better prepared 2000). Raghavan et al. (2006) propose that FOS for and respond to disasters, and recover from GIT could help local practitioners incorporate disasters. However, we have also identified and spatial knowledge, and thus integrate TEK described a number of GIT implementation with technology use. Such an approach could barriers that are especially relevant to the ap- prove valuable for natural disaster manage- plication of GIT at local administrative levels, ment, since local knowledge of environmental where a strong disaster management initiative hazards and traditional coping strategies could is required. These barriers include, but are not be very useful. limited to: (1) a lack of financial resources, (2) a lack of spatial data, (3) political/institutional instability and (4) a lack of local GIT knowledge/ CONCLUSION expertise. Thus, until at least some of these barriers are overcome the level of ability for GIT to As the human population continues to grow, and improve overall disaster management capacity considering recent evidence of climate change at the local level will remain low. that might exacerbate meteorological related Many researchers have highlighted the natural hazards in particular, there is reason particular opportunity that free and open for concern that natural disasters may occur source software (FOSS) now provides for in the future with increasing frequency and developing countries that were previously consequence. As a result, there is an immediate unavailable. Attractive characteristics of FOSS need to utilize available technology to reduce from a developing country perspective include: natural hazard vulnerability and in general to IGI GLOBALcost, freedom, PROOF accessibility, customizability, be more prepared to effectively respond to compatibility, and software/technical capacity disasters when they occur. development opportunities. The cost aspect is This article has reviewed and examined especially significant from a GIT perspective, the use of geospatial information technology as proprietary/commercial GIT software are (GIT) in the field of natural disaster manage- expensive and a lack of financial resources is ment, with an emphasis on developing countries currently a very significant GIT implementation where natural hazard vulnerability is high and barrier. Recent growth and development in the disaster impacts can be particularly devastat- FOSS-based GIT domain has resulted in the ing. The ability of GIT to acquire, interpret, emergence of many mature, very capable and analyze, map and disseminate information, user-friendly software products that in some are essential during all phases of the natural cases offer functionality that is comparable disaster management cycle – the process of with commercial GIT software. As a result, we mitigation, preparation, response and recovery. propose that FOSS-based GIT products can be Since disasters are spatial phenomenon there is used to improve the use of GIT in developing a strong relationship between disaster manage- countries, especially at the local level. Specifi- ment requirements and the spatial information cally, FOSS-based GIT provide opportunities to and decision support capabilities offered by improve local GIT knowledge and related skill GIT. In this context, GIT includes geographic sets required for their effective application in information systems, remote sensing, global the field of natural disaster management. Addi- positioning systems and Internet GIS. Among tionally, the current capabilities of FOSS-based numerous other tasks, GIT provides a basis for GIT allow for the development of spatial data planning mitigation strategies, hazard and risk infrastructures (SDI) that are required by the

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 56 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012 disaster management practitioner community Amendola, A., Linnerooth-Bayer, J., Okada, N., to successfully implement GIT, for purposes & Shi, P. (2008). Towards integrated disaster risk such as hazard and risk assessment. Considering management: case studies and trends in Asia. Natural Hazards, 44(2), 163–168. doi:10.1007/ the lack of available spatial data in developing s11069-007-9152-z countries, which is a GIT implementation barrier, the development of SDIs will represent a Becking, I. (2004, September 6-10). How space can better support emergency management and critical major step forward. However, future research infrastructure protection. In Proceedings of the En- is required that identifies specific FOSS-based visat & ERS Symposium, Salzburg, Austria. GIT and functionality that can be effectively implemented at the local level in developing Brewer, E., Demmer, M., Du, B., Ho, M., Kam, M., & Nedevschi, S. (2005). The case for technology in countries to improve overall natural disaster developing regions. IEEE Computer Society, 38(6), management capacity. 25–38. doi:10.1109/MC.2005.204 Briceño, S. (2004). Building disaster-resilient com- ACKNOWLEDGMENTS munities: The road to the second World Conference on Disaster Reduction, Kobe, Hyogo, Japan. Natural During the course of this research, the authors Resources Forum, 28, 234–236. doi:10.1111/j.1477- 8947.2004.00093.x appreciate the discussions with Mr. Chandeep Corea, head of the GIS department of the Sri Britton, J. M. R. (2000). GIS Capacity Building in Lanka Wildlife conservation Society (SLWCS). the Pacific Island Countries: Facing the realities of technology, resources, geography and cultural dif- The critical input of Dr. Barry Wellar was es- ference. Cartographica, 37(4), 7–19. doi:10.3138/ sential in developing this paper. The authors also B267-8663-4071-2375 acknowledge financial support to Sam Herold in the form of a postgraduate scholarship from Brodnig, G., & Mayer-Schönberger, V. (2000). Bridg- ing the gap: The role of spatial information technolo- the Natural Sciences and Engineering Research gies in the integration of traditional environmental Council of Canada.IGI Further GLOBAL financial support knowledge PROOF and western science. Electronic Journal was provided by the Canadian Foundation for on Information Systems in Developing Countries, Innovation, Ontario Innovation Trust and the 1(1), 1–15. University of Ottawa. Bui, T., Cho, S., Sankaran, S., & Sovereign, M. (2000). A framework for designing a global information network for multinational humanitarian assis- REFERENCES tance/disaster relief. Information Systems Frontiers, 1(4), 427–442. doi:10.1023/A:1010074210709 Abbot, J., Chambers, R., Dunn, C., Harris, T., de Burton, I., Kates, R. W., & White, G. F. (1993). The Merode, E., & Porter, G. (1998). Participatory GIS: environment as hazard (2nd ed.). New York, NY: opportunity or oxymoron? PLA Notes International Guilford Press. Institute for Environment and Development, 33, 24–34. Caldeweyher, D., Zhang, J., & Pham, B. (2006). OpenCIS—Open Source GIS-based web commu- Alexander, D. E. (1995). A survey of the field of nity information system. International Journal of natural hazards and disaster studies . In Carrara, A., Geographical Information Systems, 20(8), 885–898. & Guzzetti, F. (Eds.), Geographical information doi:10.1080/13658810600711378 systems in assessing natural hazards (pp. 1–19). Amsterdam, Netherlands: Kluwer Academic. Câmara, G., & Onsrud, H. (2004). Open-source geographic information systems software: Myths Amdahl, G. (2001). Disaster response: GIS for public and realities. In Proceedings of the International safety. Redlands, CA: ESRI. Symposium on Open Access and the Public Domain in Digital Data and Information for Science (pp. 127-133).

Chen, P., Liew, S. C., & Kwoh, L. K. (2005, July De Paratesi, S. R. (1989). Hazards and disasters: 25-29). Tsunami damage assessment using high concepts and challenges . In Barrett, E. C., & Brown, resolution satellite imagery: A case study of Aceh, K. A. (Eds.), Remote sensing for hazard monitoring Indonesia. In Proceedings of the IEEE International and disaster assessment: marine and coastal ap- Symposium on Geoscience and Remote Sensing (Vol. plications in the Mediterranean region (pp. 1–17). 2, pp. 1405-1408). Philadelphia, PA: Gordon and Breach. Cook, I., & Horobin, G. (2006). Implementing Dewan, A. M., Islam, M. M., Kumamoto, T., & eGovernment without promoting dependence: Open Nishigaki, M. (2007). Evaluating flood hazard for source software in developing countries in Southeast land-use planning in greater Dhaka of Bangladesh Asia. Public Administration and Development, 26, using remote sensing and GIS techniques. Water Re- 279–289. doi:10.1002/pad.403 sources Management, 21, 1601–1612. doi:10.1007/ s11269-006-9116-1 Coppock, T. J. (1995). GIS and natural hazards: an overview from a GIS perspective . In Carrara, A., Dewan, A. M., Kabir, M. D. H., Islam, M. M., Ku- & Guzzetti, F. (Eds.), Geographical information mamoto, T., & Nishigaki, M. (2007). Delineating systems in assessing natural hazards (pp. 21–34). flood risk areas in Greater Dhaka of Bangladesh using Amsterdam, The Netherlands: Kluwer Academic. geoinformatics. Georisk, 1(4), 190–201. Cova, T. J. (1999). GIS in emergency management Dong, P. (2005). Development of a GIS/GPS-based . In Longley, M. A., Goodchild, M. F., Maguire, D. emergency response system. Geomatica, 59(4), J., & Rhind, D. W. (Eds.), Geographical information 427–433. systems: Principles, techniques, applications, and management (pp. 845–858). New York, NY: John El-Masri, S., & Tipple, G. (2002). Natural disaster, Wiley & Sons. mitigation and sustainability: The case of developing countries. International Planning Studies, 7(2), Currion, P., De Silva, C., & Van De Walle, B. 157–175. doi:10.1080/13563470220132236 (2007). Open source software for disaster management. Communications of the ACM, 50(3), 61–65. Engler, N. J., & Hall, G. B. (2007). The Internet, doi:10.1145/1226736.1226768 spatial data globalization and data use: The case of Tibet. The Information Society, 23, 345–359. Cutter, S. L. (1996).IGI Vulnerability GLOBAL to environmen- doi:10.1080/01972240701572822 PROOF tal hazards. Progress in Human Geography, 20, 529–539. doi:10.1177/030913259602000407 Environmental System Research Institute (ESRI). (1993). Understanding GIS: The ARC/INFO Method. Cutter, S. L. (2003). GI science, disasters, and Redlands, CA: Author. emergency management. Transactions in GIS, 7(4), 439–445. doi:10.1111/1467-9671.00157 Environmental Systems Research Institute (ESRI). (2006). GIS and emergency management in Indian Cutter, S. L., Boruff, B. J., & Shirley, W. L. Ocean earthquake/tsunami disaster. Retrieved from (2003). Social vulnerability to environmental haz- http://www.esri.com/library/whitepapers/pdfs/gis- ards. Social Science Quarterly, 84(2), 242–261. and-emergency-mgmt.pdf doi:10.1111/1540-6237.8402002 Fitzgerald, B. (2004). A critical look at open Cutter, S. L., Mitchell, J. T., & Scott, M. S. (2000). source. IEEE Computer, 37(7), 92–94. doi:10.1109/ Revealing the vulnerability of people and places: A MC.2004.38 case study of Georgetown County, South Carolina. Annals of the Association of American Geographers. Freeman, P. K., & Pflug, G. C. (2003). Infrastruc- Association of American Geographers, 90, 713–737. ture in developing and transitional countries: Risk doi:10.1111/0004-5608.00219 and protection. Risk Analysis, 23(3), 601–609. doi:10.1111/1539-6924.00340 De La Ville, N., Diaz, A. C., & Ramirez, D. (2002). Remote sensing and GIS technologies as tools to sup- Goodchild, M. F. (2006). GIS and disasters: Plan- port sustainable management of areas devastated by ning for catastrophe. Computers, Environment and landslides. Environment, Development and Sustain- Urban Systems, 3(30), 227–229. doi:10.1016/j. ability, 4, 221–229. doi:10.1023/A:1020835932757 compenvurbsys.2005.10.004

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 58 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

Guinau, M., Pallàs, R., & Vilaplana, J. M. (2005). Karimi, H. A., & Blais, J. A. R. (1996). Current and A feasible methodology for landslide susceptibility future directions in GISs. Computers, Environment assessment in developing countries: A case-study and Urban Systems, 20(2), 85–97. doi:10.1016/ of NW Nicaragua after Hurricane Mitch. Engi- S0198-9715(96)00002-6 neering Geology, 80, 316–327. doi:10.1016/j.eng- geo.2005.07.001 Kerle, N., & Oppenheimer, C. (2002). Satellite remote sensing as a tool in Lahar disaster manage- Gunes, A. E., & Koval, J. P. (2000). Using GIS in ment. Disasters, 26(2), 140–160. doi:10.1111/1467- emergency management operations. Journal of 7717.00197 Urban Planning and Development, 126(3), 136–149. doi:10.1061/(ASCE)0733-9488(2000)126:3(136) Kienberger, S., & Steinbruch, F. (2005, September 7-10). P-GIS and disaster risk management: Assess- Heeks, R. (2002). Information systems and devel- ing vulnerability with P-GIS methods – Experience oping countries: Failure, success, and local impro- from Buzi, Mozambiqu. In Proceedings of the visations. The Information Society, 18, 101–112. International Conference on Participatory Spatial doi:10.1080/01972240290075039 Information Management and Communication: Mapping for Change, Kenya, Nairobi. Henderson, L. J. (2004). Emergency and disaster: Pervasive risk and public bureaucracy in developing Kondratyev, K. Y., Grigoryev, A. A., & Varotsos, C. nations. Public Organization Review, 4(2), 103–119. A. (2002). Environmental disasters: Anthropogenic doi:10.1023/B:PORJ.0000031624.46153.b2 and natural. UK: Praxis. Hewitt, K. (1995). Excluded perspectives in the social Kraak, M.-J. (2004). The role of the map in a Web- construction of disaster. International Journal of GIS environment. Journal of Geographical Systems, Mass Emergencies and Disasters, 13(3), 317–339. 6, 83–93. doi:10.1007/s10109-004-0127-2 Hoe, N. S. (2006). Breaking barriers: The potential Kumar, V., Bugacov, A., Coutinho, M., & Neches, of free and open source software for sustainable R. (1999, November 5-6). Integrating geographic human development. A Compilation of Case Studies information systems, spatial digital libraries and from Across the World. Retrieved from http://www. information spaces for conducting humanitarian apdip.net/publications/ict4d/BreakingBarriers.pdf assistance and disaster relief operations in urban IGI GLOBALenvironments. PROOF In Proceedings of the 7th ACM Hussain, M., Arsalan, M. H., Siddiqi, K., Naseem, Symposium on Advances in Geographic Information B., & Rabab, U. (2005, June 9-11). Emerging geo- Systems, Kansas City, MO. information technologies for natural disaster management in Pakistan: An overview. In Proceedings of Laben, C. (2002). Integration of remote sensing the 2nd International Conference on Recent Advances data and geographic information system technol- in Space Technologies. Retrieved from http://ieeex- ogy for emergency managers and their applications plore.ieee.org/iel5/10107/32390/01512618.pdf?tp= at the Pacific Disaster Center. Optical Engineer- &arnumber=1512618&isnumber=32390 ing (Redondo Beach, Calif.), 41(9), 2129–2136. doi:10.1117/1.1501137 International Federation of Red Cross and Red Crescent Societies (IFRC). (2001). World disasters Lay, T., Kanamori, H., Ammon, C. J., Nettles, M., report. Geneva, Switzerland: Author. Ward, S. N., & Aster, R. C. (2005). The Great Sumatra- Andaman Earthquake of 26 December 2004. Science, International Panel on Climate Change (IPCC). 308, 1127–1133. doi:10.1126/science.1112250 (2007). Climate change 2007: The Physical Sci- ence Basis, Contribution of Working Group I to the Magsud, M., Sann Oo, K., & Rajapaksha, J. (2005, Fourth Assessment Report of the IPCC. Retrieved November 7-11). Tsunami disaster damage detec- from http://www.ipcc.ch/ipccreports/ar4-wg1.htm tion and assessment using high resolution satellite data, GIS and GPS – Case study in Sri Lanka. In Jayaraman, V., Chandrasekhar, M. G., & Rao, U. Proceedings of the Asian Conference on Remote R. (1997). Managing the natural disasters from Sensing, Hanoi, Vietnam. space technology inputs. Acta Astronautica, 40(2), 291–325. doi:10.1016/S0094-5765(97)00101-X Mansor, S., Shariah, M. A., Billa, L., Setiawan, I., & Jabar, F. (2004). Spatial technology for natural risk Johnson, R. (2000). GIS technology for disaster management. Disaster Prevention and Management, and emergency management. Redlands, CA: ESRI. 13(5), 364–373. doi:10.1108/09653560410568480

Mansourian, A., Rajabifard, A., Valadan Zoej, M. Nebert, D. (2004). Developing spatial data infra- J., & Williamson, I. (2004). Facilitating disaster structures: The SDI cookbook v.2.0. Global spatial management using SDI. Journal of Geospatial data infrastructure. Retrieved July 2005, from http:// Engineering, 6(1), 29–44. www.gsdi.org McEntire, D. A. (1999). Issues in disaster relief: Nonogaki, S., Van Anh, T., Masumoto, S., Raghavan, progress, perpetual problems and prospective solu- V., Nemoto, T., & Mori, T. (2004). Development of tions. Disaster Prevention and Management, 8(5), Vietnamese Version of GRASS GIS. Geoinformat- 351–361. doi:10.1108/09653569910298279 ics, 15, 106–107. McEntire, D. A. (2001). Triggering agents, vulner- Oloruntoba, R. (2005). A wave of destruction and abilities and disaster reduction: towards a holistic the waves of relief: issues, challenges and strate- paradigm. Disaster Prevention and Management, gies. Disaster Prevention and Management, 14(4), 10(3), 189–196. doi:10.1108/09653560110395359 506–521. doi:10.1108/09653560510618348 Mercer, J., Dominey-Howes, D., Kelman, I., & Lloyd, OSGeo. (2008). About the Open Source Geospatial K. (2007). The potential for combing indigenous and Foundation. Retrieved from http://www.osgeo.org/ western knowledge in reducing vulnerability to envi- content/foundation/about.html ronmental hazards in small island developing states. Environmental Hazards, 7, 245–256. doi:10.1016/j. Osti, R., Tanaka, S., & Tokioka, T. (2008). Flood haz- envhaz.2006.11.001 ard mapping in developing countries: problems and perspectives. Disaster Prevention and Management, Mohamed, M., & Plante, R. (2002). Remote sens- 17(1), 104–113. doi:10.1108/09653560810855919 ing and Geographic Information Systems (GIS) for developing countries. In Proceedings of the IEEE Pande, R. K. (2006). Participation in practice and International Geoscience and Remote Sensing Sym- disaster management: experience of Uttaranchal posium (Vol. 4, pp. 2285-2287). Retrieved from http:// (India). Disaster Prevention and Management, ieeexplore.ieee.org/iel5/7969/22039/01026520.pdf 15(3), 425–428. doi:10.1108/09653560610669909 Montoya, L. (2002, April 25-27). GIS and remote Peng, Z. R., & Tsou, M. H. (2003). Internet GIS. sensing in urban disaster management. In Proceed- Hoboken, NJ: John Wiley & Sons. ings of the 5th AGILEIGI Conference GLOBAL on Geographic Plewe, B. PROOF(1997). GIS Online: Information Retrieval, Information Science, Balearic Islands, Spain. Mapping, and the Internet. Albany, NY: OnWord Montoya, L., & Masser, I. (2005). Management of Press. natural hazard risk in Cartago, Costa Rica. Habitat Quarantelli, E. L. (1998). What is disaster? Perspec- International, 29, 493–509. doi:10.1016/j.habi- tives on the question. New York, NY: Routledge. tatint.2004.04.003 Rabindra, O., Tanaka, S., & Tokioka, T. (2008). Morrow, B. H. (1999). Identifying and mapping Flood hazard mapping in developing coun- community vulnerability. Disasters, 23(1), 1–18. tries: problems and perspectives. Disaster doi:10.1111/1467-7717.00102 Prevention and Management, 17(1), 104–113. Murgia, J., Amemiya, N., & Turkstra, J. (2002, May doi:10.1108/09653560810855919 15-18). Local spatial data infrastructure, Trujillo, th Radke, J., Cova, T., Sheridan, M. F., Troy, A., Lan, Peru. In Proceedings of the 6 Seminar on GIS and M., & Johnson, R. (2000). Application challenges Developing Countries. Retrieved from http://www. for geographic information science: Implications gisdevelopment.net for research, education, and policy for emergency Musinguzi, M., Bax, G., & Tickodri-Togboa, S. S. preparedness and response. URISA Journal, 12(2), (2004, June 7-9). Opportunities and challenges for 15–30. SDI development in developing countries – A case th Raghavan, V., Masumoto, S., & Hastings, D. (2006, study of Uganda. In Proceedings of the 12 Interna- November 9-11). Free and open source software for tional Conference on Geoinformatics – Geospatial geoinformatics: Present status and future prospects. Information Research: Bridging the Pacific and In Proceedings of the International Symposium on Atlantic, Gävle, Sweden. Geoinformatics for Spatial Infrastructure Develop- ment in Earth and Allied Sciences, Ho Chi Minh, Vietnam. Retrieved from http://wgrass.media.osaka- cu.ac.jp/gisideas06/viewpaper.php?id=166

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 60 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

Rajani, N., Rekola, J., & Mielonen, T. (2003). Free as Roets, R., Minnaar, M., & Wright, K. (2007). Open in education: Significance of Free/Libre Open Source Source: Towards successful systems development Software for developing countries. Retrieved from projects in developing countries. In Proceedings of http://www.itu.int/wsis/docs/background/themes/ the 9th International Conference on Social Implica- access/free_as_in_education_niranjan.pdf tions of Computers in Developing Countries, São Paulo, Brazil. Retrieved from http://www.ifipwg94. Ramasubramanian, L. (1999). GIS implementation in org.br/fullpapers/R0061-1.pdf developing countries: Learning from organisational theory and reflective practice. Transactions in GIS, Rüdenauer, H., & Schmitz, M. (2005, March 14-16). 3(4), 359–380. doi:10.1111/1467-9671.00028 Use of Landsat 7 ETM+ data as basic information for infrastructure planning. In Proceedings of the Rambaldi, G., Kwaku Kyem, P. A., McCall, M., & ISPRS Joint Conference of the 3rd International Weiner, D. (2006). Participatory spatial information Symposium on Remote Sensing and Data Fusion Over management and communication in developing Urban Areas, and the 5th International Symposium countries. The Electronic Journal on Information on Remote Sensing of Urban Areas, Tempe, AZ. Systems in Developing Countries, 25(1), 1–9. Retrieved from http://www.isprs.org/commission8/ Ramli, M. F., Sulaiman, W. N. A., Yusoff, M. K., workshop_urban/ruedenauer.pdf & Low, Y. Y. (2005). Open source geographi- San Miguel-Ayanz, J., Vogt, J., De Roo, A., & cal resources analysis support system (GRASS) Schmuck, G. (2000). Natural hazards monitoring: for landslide hazard assessment. Disaster Forest fires, droughts and floods – the examples of Prevention and Management, 11(4), 522–532. European Pilot Projects. Surveys in Geophysics, 21, doi:10.1108/09653560510618357 191–305. doi:10.1023/A:1006750412500 Ramsey, P. (2007). The state of open source GIS. In Sanyal, J., & Lu, X. X. (2003, October 13-15). Ap- Proceedings of the Global Conference on Free and plication of GIS in flood hazard mapping: a case study Open Source Software for Geospatial, Vancouver, of Gangetic West Bengal, India. In Proceedings of BC, Canada. Retrieved from http://www.refrac- the Map Asia Conference, Kuala Lumpur, Malaysia. tions.net/expertise/whitepapers/opensourcesurvey/ Retrieved from http://www.gisdevelopment.net/ap- survey-open-source-2007-12.pdf plication/natural_hazards/floods/pdf/ma03138.pdf Rautela, P., & Pande,IGI R. K. GLOBAL (2005). Implications of Schutzberg, PROOF A. (2005). MapServer Community, Au- ignoring the old disaster management plans: Lessons todesk Announce MapServer Foundation. Directions learnt from the Amparav tragety of 23 September Magazine. Retrieved from http://www.directions- 2004 in the Nainital district of Uttaranchal (India). mag.com/article.php?article_id=2037 Disaster Prevention and Management, 14(3), 388–394. doi:10.1108/09653560510605045 Showalter, P. S. (1999). Remote sensing’s use in disaster research: A review. Disaster Rego, A. J. (2001, March 21-23). National disaster Prevention and Management, 10(1), 21–29. management information systems & networks: An doi:10.1108/09653560110381796 Asian overview. In Proceedings of the Global Di- saster Information Network Conference, Canberra, Shrestha, S. (1994). Environmental issues in Asia and Australia. Retrieved from http://www.adpc.net/ the pacific - a perspective towards the information in infores/adpc-documents/paperatgdin01.pdf decision making. In Proceedings of the International Conference on Expert Systems for Development, Renyi, L., & Nan, L. (2002). Flood area and dam- Bangkok, Thailand. Retrieved from http://ieeexplore. age estimation in Zhejiang, China. Journal of ieee.org/iel2/973/7458/00302314.pdf?tp=&isnumb Environmental Management, 66, 1–8. doi:10.1006/ er=&arnumber=302314 jema.2002.0544 Simonovic, S. P. (2002). Role of remote sensing in Rivereau, J. C. (1995). Spot data applied to di- disaster management (Paper Series No. 21). London, saster prevention and damage assessment. Acta ON, Canada: University of Western Ontario, Institute Astronautica, 35(7), 467–470. doi:10.1016/0094- for Catastrophic Loss Reduction. 5765(94)00281-P Simpson, D. M., & Human, R. J. (2008). Large- scale vulnerability assessments for natural hazards. Natural Hazards, 47(2), 143–155. doi:10.1007/ s11069-007-9202-6

Sliuzas, R. (1999, March 11-12). Research issues for Van Westen, C. J., & Hofstee, P. (2001). The role of the adoption of geographic information technology remote sensing and GIS in risk mapping and damage for urban planning and management in develop- assessment for disasters in urban areas. Enschede, ing countries. In Proceedings of the International The Netherlands: International Institute for Aero- Workshop on Concepts and Paradigms of Urban space Survey and Earth Sciences (ITC). Retrieved Management in the Context of Developing Countries, from http://www.dkkv.org/forum2001/Datei61.pdf Venice, Italy. Retrieved from http://www.naerus.net/ sat/workshops/1999/paper_sliuzas.html Weber, S. (2004). The success of open source. Cam- bridge, MA: Harvard University Press. Smith, K. (2004). Environmental hazards: Assess- ing risk and reducing disaster (4th ed.). New York, Weerawarana, S., & Weeratunge, J. (2004). Open NY: Routledge. source in developing countries. Stockholm, Swedish International Development Cooperation (Sida) Agen- Space Agencies. (2005). International charter on cy. Retrieved from http://www.sida.se/publications space and major disasters. Retrieved from http:// www.disasterscharter.org/main_e.html Weichselgartner, J. (2001). Disaster mitigation: The concept of vulnerability revisited. Disas- Steiniger, S., & Bocher, E. (2009). An overview ter Prevention and Management, 10(2), 85–94. on current free and open source desktop GIS doi:10.1108/09653560110388609 developments. International Journal of Geo- graphical Information Science, 23(10), 1345–1370. Wheeler, D. A. (2007). Why Open Source Software/ doi:10.1080/13658810802634956 Free Software (OSS/FS, FLOSS, or FOSS)? Look at the Numbers! Retrieved from http://www.dwheeler. Thomson, C. N., & Hardin, P. (2000). Remote sens- com/oss_fs_why.html ing/GIS integration to identify potential low-income housing sites. Cities (London, England), 17(2), Wu, M., & Lin, Y. (2001). Open source software 97–109. doi:10.1016/S0264-2751(00)00005-6 development: An overview. IEEE Computer, 33-38. United Nations. (UN). (2006). The millennium Zerger, A., & Smith, D. I. (2003). Impediments to development goals report. Retrieved from http:// using GIS for real-time disaster decision support. mdgs.un.org/unsd/mdg/Resources/Static/Products/ Computers, Environment and Urban Systems, 27, Progress2006/MDGReport2006.pdfIGI GLOBAL123–141. PROOFdoi:10.1016/S0198-9715(01)00021-7

Sam Herold has a MSc in Geography from the University of Ottawa and is a former member of the Laboratory for Applied Geomatics and GIS Science (LAGGISS). Mr. Herold is currently employed as an analyst at the Canadian Population Health Initiative (CPHI), a branch of the Canadian Institute for Health Information (CIHI), where his work involves the use of GIS for population health research, analysis and mapping. Previously, he worked for two years as assistant cartographer for Canadian Geographic magazine.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 62 International Journal of Applied Geospatial Research, 3(2), 24-62, April-June 2012

Michael C. Sawada is member of the Hazard Mitigation and Disaster Management Research Center of the University of Ottawa and is the Director of the Laboratory for Applied Geomatics and GIS Science (LAGGISS) in the Department of Geography. Dr. Sawada's research is within the applied GIS, Remote Sensing, Global Positioning Systems (GPS), spatial analysis and 3-D visualization. His transdisciplinary work has led to the development of new spatial analytical methods that address pressing social and environmental policy issues such as soil erosion and contaminant modeling and transport, disaster mitigation technologies, air quality and population health. Dr. Sawada received the 2005 Julian M. Sciecz Award. The award is presented annually by the Canadian Association of Geographers in recognition of a significant achievement by a Canadian geographer at an early career stage. He was also named Young Researcher of the Year for the Faculty of Arts at the University of Ottawa in 2006.

IGI GLOBAL PROOF

Embracing Geographic Analysis Beyond Geography: Harvard’s Center for Geographic Analysis Enters 5th Year

Weihe (Wendy) Guan, Harvard University, USA Peter K. Bol, Harvard University, USA

ABSTRACT

Without a department of geography, Harvard University established the Center for Geographic Analysis (CGA) in 2006 to support research and teaching of all disciplines across the University with emerging geospatial technologies. In the past four and a half years, CGA built an institutional service infrastructure and unleashed an increasing demand on geographic analysis in many fields. CGA services range from helpdesk, project consultation, IGItraining, hardware/software GLOBAL administration, communityPROOF building, to system development and methodology research. Services often start as an application of existing GIS technology, eventually contributing to the study of geographic information science in many ways. As a new generation of students and researchers growing up with Google Earth and the like, their demand for geospatial services will continue to push CGA into new territories.

Keywords: Geographic Analysis, Geography, Geospatial Technology, GIS Technology, Harvard University

HARVARD RETURNED TO to geospatial analysis that combines GIS tech- GEOGRAPHY nology with quantitative analysis--could make to disciplines across the University, from public Harvard lost its Geography Department - actu- health to sociology to history. The question the ally what had become the social geography wing University faced was how most effectively, and of the Geography and Geology Department - in quickly, to make it possible for students and 1948 (Smith, 1987) and the Computer Graphics faculty to bring geospatial to bear on research and Spatial Analysis lab in 1991(Chrisman, and teaching. The Center for Geographic 2006). By the early 2000s the University rec- Analysis was established to do just that, with the ognized that it was missing out on the potential understanding that faculty appointments would contributions the new geography--the approach follow as endowment funds for new professor- ships were raised and as departments came to see the important roles geography could play DOI: 10.4018/jagr.2012040104 in their own disciplines (Reed, 2006).

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 64 International Journal of Applied Geospatial Research, 3(2), 63-71, April-June 2012

The field of geographic analysis has un- tigation and discussion led to the decision by dergone a series of revolutions over the past the University to establish the Center for Geo- half a century. The quantitative revolution graphic Analysis (CGA). The inauguration of brought quantitative analysis into geography the Center was held in May 5, 2006 (Gehrman, in the 1950s, which inspired a new generation 2006), “a new dawn for geography at Harvard” of students and scholars (Dutton, 2006). The (Waters, 2006) (Figure 2). birth of geographic information systems (GIS) in the 1960s equipped geographic analysis with computers (Niemann & Niemann, 1998). A CENTER THAT SERVES ALL Computer cartography and analysis of remotely There are two common models of GIS centers sensed images in the 1970s gave the discipline in universities. One is department based, such new dimensions. The integration of geographic as the three National Centers for Geographic analysis with numerical models gained much Analysis (NCGA), the GeoPlan Center at the ground in the 1980s, especially in the field of University of Florida, and the Environmental environmental studies. The information technol- Resources Spatial Analysis Center (ERSAC) ogy (IT) boom in the 1990s gave geographic at the University of Minnesota. Departments analysis a swift push into the IT mainstream, hosting GIS centers include geography, urban complete with relational and object-oriented planning, and forestry, among others. The other database management systems (RDBMS and common model of university GIS centers is ODBMS), programming platforms, client server library based, or library joined with university architectures, and web based implementations. IT services. Examples are abundant, such as The completion of the Global Positioning Stanford University, Yale University, MIT, or System (GPS) and the ever increasing earth Rice University. surveillance satellites resulted in an exponen- Prior to 2006, Harvard University had tial increase of spatial data in native digital both models. The Graduate School of Design format, feedingIGI content to,GLOBAL and at the same time PROOF had been offering GIS course modules as part putting pressure on, geographic information of its curriculum and spatial analysis support systems. Google released Google Maps and by its IT staff. Meanwhile, the Harvard Col- Google Earth in the mid 2000s and generated a lege Library Map Collection also had GIS heightened public interest in geographical issues staff offering spatial data and GIS support to and expanded awareness of geography in other users across the university. However, Harvard disciplines. The geotechnology revolution is faculty and administrators felt that much more still ongoing, pushing geographic analysis up to needed to be done to meet future demand for computing clouds and down to mobile devices, geographic analysis. and into all walks of life. Its continued growth To the degree that a department based center has expanded the job market for geographers, is focused on serving the hosting department increased enrollment in geography departments, or school it may have few resources left for the and brought improvement in geography educa- rest of the university. A library based center tion at all levels (Murphy, 2007). serving the entire university may be limited in In 2003, a faculty committee on spatial its service by the operational conventions of analysis was formed at Harvard University. Led libraries. CGA was established to overcome by Peter Bol (Figure 1), an historian interested these limitations. With a modest start of 3.5 in applying GIS to the study of history, the FTE, it has a broad mandate to serve the entire committee set out to find a solution to address University (Figures 3 and 4). the increasing concerns of faculty and students At its inception, the CGA was designed to with improving access to spatial data, support focuses on research and education in the field for research employing geospatial analysis, and of spatial analysis and geographic information. curriculum development. Two years of inves-

Figure 1. Professor Bol addressing the audience at the CGA inaugural workshop on May 5, 2006

The Center’s goal has been to work with enti- but by no means matching up with the rate of ties across the university to strengthen univer- growth of demand on its services. Guan and sity-wide GIS infrastructure and services; Bol (2009) provides a comprehensive summary provide a common platform for the integration of the services offered by the Center, which of spatial data from diverse sources and knowl- includes the following: edge from multipleIGI disciplines; GLOBAL enable schol- PROOF arly research that would use, improve or study • Research services geospatial analysis techniques; and improve ◦◦ Help desk the ability to teach GIS and spatial analysis at ◦◦ Consultation all levels across the university. ◦◦ Project work The early years of the Center have not been ◦◦ Contract management without challenges, and a major challenge stems ◦◦ Grant proposals from the fact that Harvard has no geography • Teaching and training or geographic information science faculty. For ◦◦ For-credit courses better or worse this has meant that the Center ◦◦ Non-credit training series does not have its own research agenda. Instead, ◦◦ Annual thematic workshop/conference its agenda is to enhance research in all disci- ◦◦ Ten-day GIS training institute plines. It does not have a particular interest to • Infrastructure support serve; its interest is to help all who can benefit ◦◦ Site licenses from geospatial technology to use it properly. ◦◦ Online data library ◦◦ Hosting of web GIS services ◦◦ GPS, plotting and scanning equipments UNSTOPPABLE GROWTH ◦◦ Computer labs on both campuses OF DEMAND • Community building Over the past 4 years, as demand kept increas- ◦◦ Newsletter and website ing, the Center has doubled its annual budget, ◦◦ Student award tripled its staff size, and quadrupled its office ◦◦ Technical seminars space, a significant growth by any standard,

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 66 International Journal of Applied Geospatial Research, 3(2), 63-71, April-June 2012

Figure 2. CGA’s inaugural theme: Harvard Returns to Geography

◦◦ Technology information exchange show the growth of services over time, in both (blogs) research and teaching. As this service infra- • Collaboration promotion structure gradually takes shape, it benefits not ◦◦ Cross disciplinary connections only Harvard scholars, but through them and ◦◦ Institutional Membership our collaborators, the world community at large. ◦◦ Visiting scholars, research fellows and One such example is the Earthquake Geospatial student interns Research Portal as described in ESRI (2008) and Gudrais (2010). Guan et al. (2010) provides a more detailed description ofIGI the CGA’s GLOBAL role in building an PROOF institutional infrastructure for geographic analysis at the University. Figures 5 through 7

Figure 3. The CGA is housed in the CGIS Knafel Building. Not to be mistaken as the “Center for GIS” building, CGIS here actually stands for the Center for Government and International Studies, and CGA occupies only a small portion of its concourse level

Figure 4. The central location of CGA on Harvard’s main campus provides convenient access for most of its users

RESEARCH WITH GIS AND WorldMap is a web based and map centric RESEARCH ON GIS data exploration system that is built on the open source geospatial technology. Its first implemen- Started without a geographicIGI GLOBAL information sci- tation, forPROOF the continent of Africa, with a beta ence faculty to drive research on GIS, the CGA release in November 2008, has proven to be an focused on using mature GIS technology to effective solution for integrating and making enhance research and teaching on all subjects accessible dispersed humanities collections, – research with GIS. However, in doing so, combined with a wealth of mapping materials, we have found that even though mature GIS to allow for cross-disciplinary inquiries from a technology is far from being utilized at its full spatial perspective. AfricaMap has since been potential, there are cases when proven technol- cloned to create portals focused on other parts ogy falls short in solving a particular problem of the world, including Boston, Vermont, Paris or providing a particular service. Such unmet and East Asia, collectively called WorldMap. needs push the envelope of geographic informa- WorldMap is designed to handle an un- tion science itself, and inevitably lead us into limited amount of data content, not only by the research on GIS. the scalability of its system servers, but more One such challenge is the need for a system importantly through its service oriented archi- that is publicly accessible, simple to use for non- tecture that allows the system to connect to technical scholars, fast in search and mapping remote systems, to search for, request for and speed, and rich in geographic content. Finding display materials hosted on them, following the no exiting tool or system that met this require- Open Geospatial Consortium (OGC) standards. ment, the CGA has made a particular effort to In return, materials hosted on the WorldMap develop a geoportal, first known as AfricaMap servers are available for other systems’ live (http://africamap.harvard.edu), which has since consumption as well. The system handles both evolved into the WorldMap system. mapping data and data in any format as long as it has a geographic location associated with

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 68 International Journal of Applied Geospatial Research, 3(2), 63-71, April-June 2012

Figure 5. Helpdesk incidences accounted for in this chart include visits to the GIS help desks on both the main campus in Cambridge and the medical campus in Boston, as well as email and phone inquiries

Figure 6. CGA staff is invited to give presentations and demos to different groups and at various events on campus, and is involved in many research projects, from a few hours of consultation (free), to a few days of data analysis or mapping (minor), to a few weeks or months of methodology research orIGI system developmentGLOBAL (major) PROOF

it. Examples include YouTube videos, Picasa rapid search of even remote locations, and photos, museum collections, library archives, quickly zoom to the selected vicinity. The wired news, selected blogs, etc. The several user may select several data layers to overlay million place names in its gazetteer allow for together, and has full control of transparency

Figure 7. CGA offers 2-hour hands-on training seminars on Friday afternoons during the school year, gradually adding more modules and increasing the number of sessions

for each layer, allowing for visual investigation of historical objects and events, uncertainty and of spatial patterns and relationships. The power nonstandard time definitions, and the verifica- of WorldMap is on its ability to integrate an tion and consolidation of multi-source scholarly ever expanding data collection stored locally inputs. Similarly, the China Historical GIS and remotely, its IGI flexibility GLOBAL on focusing on project (http://www.fas.harvard.edu/~chgis/),PROOF any particular theme (defined by the filtered another example of GIS supporting humanities data layers) on any geographic extent (from and social sciences, not only applied mature the entire world to a city neighborhood), and GIS database and mapping technologies to its interactive tools allowing users to discover, historical data, but also touched on some combine, and visualize materials from differ- thorny spatial-temporal modeling issues, such ent sources, in different formats, with different as representing hierarchical relationships of languages, but all organized through a map historical administrative units whose polygonal (Lewis & Guan, 2010). boundaries were never defined or not known, or The usefulness of the WorldMap system retrofitting historical population distribution to a has generated much enthusiasm among a broad finer scale based on modeled natural, economi- user community inside and outside of Harvard, cal and social environment over the historical as well as strong demands for its expansion landscape (Berman, 2005, 2007). and enhancement. Based on the beta release These projects often started as an applica- of its proof-of-concept implementation, CGA tion of existing GIS technology, eventually con- is actively planning for a 2nd phase of design tributing to the study of geographic information improvements to facilitate greater scholarly science in many ways. Unlike stable business collaboration. operations in the industrial or public sectors, Another project, the Digital Atlas of Ro- academic research in any discipline constantly man and Medieval Civilization (DARMC explores new approaches, new analytical means, http://darmc.harvard.edu), applied off-the-shelf and new perspectives on very diverse objects web mapping technology (ArcGIS Server), but and processes. They present a strong demand encountered a number of unconventional issues and ample testing cases for the study of new in its geodatabase design, such as the life span scientific methodology. The CGA is uniquely

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 70 International Journal of Applied Geospatial Research, 3(2), 63-71, April-June 2012 positioned at the crossing point of research system for mapping non-spatial contents to with GIS and research on GIS. It is firmly support location based inquiries and reason- based on the former, with much potential to ing, etc. Looking forward, a new generation grow in the latter. of students and researchers has grown up with Google Earth, iPhones loaded with maps and GPS, georeferenced twits, etc. Their demand WHAT COMES NEXT? for geospatial services will continue to push CGA into new territories. By tradition Harvard University is a distributed system, with each organization operating with a high degree of autonomy, described as “every tub has its own bottom.” Five years into REFERENCES its operation, the CGA has yet to reach out to Berman, M. L. (2005). Boundaries or networks in all potential users. Outreach and coordina- historical GIS: Concepts of measuring space and tion continues to be major tasks for the CGA, administrative geography in Chinese history. His- competing for limited resources with technical torical Geography, 33, 118–133. projects. Moreover, for users who are familiar Berman, M. L. (2007). Persistence or transience? and satisfied with CGA services, the CGA has Tracking the evolution of places over time with his- become their “one-stop shop” for technology torical geographic information systems. Historisches support. It is often hard for a scholar to clearly Forum 10 (Tielband II): Geschichte im Netz: Praxis, define the different technical domains that Chancen, Visionen (pp. 302-313). Berlin, Germany: Beiträge der Tagung. hist. speak to his/her project’s needs. Oftentimes one project needs data acquisition, database Chrisman, N. (2006). Charting the unknown: How design, web development, geographic analysis, computer mapping at Harvard became GIS. Red- lands, CA: ESRI. and statistical modeling. The researcher may not be familiarIGI with all GLOBALthe specialty support Dutton, PROOF G. (2006). Everything is where it is having teams scattered across the University for each moved there. A memoir on the occasion of the dedica- of these components. They may rely on CGA, tion of the Harvard Center for Geographic Analysis. Retrieved from http://isites.harvard.edu/fs/docs/icb. or whichever team that they are most familiar topic39008.files/Dutton_Memoir.pdf with, to help them define their project scope, identify technology needs, and establish initial ESRI. (2008). China earthquake geospatial research portal launched by Harvard University. ArcNews contact with the other specialty teams. Online. Retrieved from http://www.esri.com/news/ The rapidly evolving geospatial technology arcnews/fall08articles/china-earthquake.html blurs boundaries between CGA’s core technology domain - GIS, and related IT domains such Gehrman, E. (2006, May 11). Geography center launched - Center for Geographical Analysis to as database design, web design, cloud comput- explore ‘vast intellectual territory’. Harvard Ga- ing, etc. CGA is riding important current trends zette. Retrieved from http://www.news.harvard.edu/ such the open source, open standards, crowd gazette/2006/05.11/05-geography.html sourcing, and service oriented architecture, Guan, W., & Bol, P. K. (2009). Harvard revisited: pushing the technology envelopes while ap- Geography’s return as GIS. Geography Teacher, 6(2). plying the technology. Some of the projects under planning include the development of Guan, W., Burns, B., Finkelstein, J. L., & Blossom, J. (2011). Enabling geographic research across dis- a “Map OCR”, or map-based optical charac- ciplines: Building an institutional infrastructure for ter recognition system which automatically geographic analysis at Harvard University. Journal recognizes and extracts text labels and spatial of Map and Geography Libraries . Geographic Op- features from scanned maps while maintaining portunities in Medicine, 7(1), 36–60. their geographic registration and feature-label relationship; a global temporal gazetteer; a

Gudrais, E. (2010, January 19). Harvard center’s Reed, C. (2006). Hello, Geotech - “Modeling our maps help Haiti effort. Harvard Magazine. Retrieved world,” geography returns to Harvard. Harvard from http://harvardmagazine.com/breaking-news/ Magazine. Retrieved from http://harvardmagazine. harvard-center-hosts-haiti-map-portal com/2006/11/hello-geotech.html Lewis, B., & Guan, W. (in press). Jump starting the Smith, N. (1987). Academic war over the field of next level of online geospatial collaboration: Les- geography: The elimination of geography at Harvard, sons from AfricaMap . In Li, S., Dragicevic, S., & 1947-1951. Annals of the Association of American Veenendaal, B. (Eds.), Advances in Web-based GIS, Geographers. Association of American Geographers, mapping services and applications. Rotterdam, The 77(2), 155–172. doi:10.1111/j.1467-8306.1987. Netherlands: A. A. Balkema. tb00151.x Murphy, A. B. (2007). Geography’s place in Waters, N. (2006). Harvard, geography and GIS: A higher education in the United States. Journal of race to the top. Retrieved from http://www.geoplace. Geography in Higher Education, 31(1), 121–141. com/uploads/FeatureArticle/0609_EDGENODES. doi:10.1080/03098260601033068 asp Niemann, B. J., Jr., & Niemann, S. (1998). GIS innovator - Allan H. Schmidt: GIS journeyman. GeoInfoSystems.

Weihe (Wendy) GuanIGI came toGLOBAL Harvard in 2006 as the DirectorPROOF of GIS Research Services for the newly established Center for Geographic Analysis. Prior to that, she managed professional services at a GIS consulting firm in Washington, headed the geospatial information technology department for a multinational forestry corporation, and supervised GIS teams in a Florida government agency. Wendy has a PhD in ecology and GIS; a MA and MS in geography and resource management, and a BS in biology. She taught GIS in various universities, including the Harvard Extension School. Peter K. Bol is a Harvard College Professor and the Charles H. Carswell Professor of East Asian Languages and Civilizations. He led Harvard’s university-wide effort to establish support for geospatial analysis in teaching and research; in 2005 he was named the first director of the Center for Geographic Analysis. He also directs the China Historical Geographic Information Systems project, collaboration between Harvard and Fudan University in Shanghai to create a GIS for 2000 years of Chinese history, and is involved in other projects aimed at enhancing digital information linkages between East Asian and Western scholars.

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 72 International Journal of Applied Geospatial Research, 3(2), 72-77, April-June 2012

A Reflection on the PhD Program in Spatially Integrated Social Science at the University of Toledo

Bhuiyan Monwar Alam, The University of Toledo, USA Jeanette Eckert, The University of Toledo, USA Peter S. Lindquist, The University of Toledo, USA

ABSTRACT

The use of spatial analysis tools is on the rise in many academic fields and practical applications. These tools enhance the abilityIGI to examine GLOBAL data from spatial perspectives. PROOF Though the study of place and space has traditionally been the domain of the field of geography, growing numbers of researchers are turning to these tools in the social sciences and beyond. The University of Toledo has established a unique Ph.D. granting program to encompass the theories, tools, and applications of spatially integrated social science. In the first couple of years of its inception the program has attracted students from different places and diverse backgrounds. It is expected that the program will continue to thrive in attracting diverse students, securing external grants, and positively impacting on the economy of Northwest Ohio. This paper is a personal reflection of the views of the authors on the Ph.D. program in Spatially Integrated Social Science at the University of Toledo two years after its inception in fall 2009. The views, by no means, are of the University of Toledo, its SISS program, or any of the participating departments and faculty members.

Keywords: Geographic Information Systems, Northwest Ohio, Ph.D. Program, Spatially Integrated Social Science, The University of Toledo

INTRODUCTION prevalence of cutting-edge technologies like Geographic Information Systems (GIS) and The use of a variety of spatial analysis tools Global Positioning Systems (GPS) in the social is increasing in social science research. While sciences is facilitating the study of the role of social problems and study areas span multiple place in the society. Thus, geography can be a disciplines, the question of location has his- unifying field for the social sciences, wherein torically been the domain of geographers. The traditional data analysis can be conducted within the context of place.

DOI: 10.4018/jagr.2012040105

While each discipline has historically pos- sciences, GIS itself does not fully account for sessed its own set of analytic tools, the increasing or measure the complexities and relationships ability to measure processes in a spatiotemporal inherent in spatial data (Páez & Scott, 2004). context facilitates sharing of these new tools, a Páez and Scott (2004) argue that current GIS general development that can then be applied to software, even with the recent inclusion of local and unique conditions (Goodchild et al., spatial modeling, do not adequately account 2000). Human demographic data, for example, for heterogeneity, interdependence or spatial has historically been presented in a series of association, and thus are not as accurate or useful tables. However, Weeks (2004) argues that as it could be if more advanced spatial statistics such data is spatial in nature because it deals were used. Logan et al. (2010) argue that GIS with characteristics of human populations in gives way to more complex spatial analysis tools specific regions. It varies based on time, space, when patterns on a map lead to additional ques- and scale. The recent shift towards locational tions that simple visualization cannot answer. information in data sets and displaying data in For example, a traditional chloropleth map made interactive maps in addition to traditional tables in GIS does not take into account the outliers or and charts has brought to light the importance an uneven population. The longstanding method and usefulness of spatial ways of looking at of collecting data in tables and translating it to data in the social sciences. maps only scratches the surface of displaying As GIS tools have become more power- and analyzing spatial data. ful and more widely available, they have been The last decade has seen the development of increasingly used in the application side of a cohesive shift towards spatial ways of thinking the social sciences. More and more it is being about social data and a demand for tools that realized that these tools are equally powerful facilitate this approach (Voss, 2007). This area for theoretical queries and social research. As of study is increasingly referred to as Spatially Sui (2004) asks, GIS is the answer, but what is Integrated Social Science (SISS). A few key the question? Sui (2004)IGI goes GLOBAL on to argue that programs PROOF have progressed SISS techniques, advances in geocomputation will continue to including the University of California at Santa bring social science disciplines into the trend of Barbara (UCSB), which is home to many cen- spatializing previously non-spatial data. ters and programs related to the field, perhaps Spatial analysis is both an inductive and most notably the Center for Spatially Integrated deductive tool, as it can reveal unforeseen pat- Social Sciences (CSISS, 2011). CSISS focuses terns and test existing theories about expected on integrating the social sciences via the context patterns (Goodchild & Janelle, 2004). Anselin of place, and also holds workshops for students, (2006) categorizes spatial analysis techniques researchers, and instructors (CSISS, 2011). Also into three general groups: exploratory spatial at UCSB is the National Center for Geographic data analysis (searching for patterns), visual- Information and Analysis and the UCSB Spatial ization (methods of showing the patterns), and Studies Center, making the University a vital hub spatial modeling (methods of explaining and of spatial information research. CSISS reports predicting patterns). Recently, the development that the majority of participants in their offered of the GeoDa software has created a program workshops, for which they receive far more ap- capable of both advanced spatial statistics and plicants than they can accept, are from the field mapping (Anselin et al., 2006). of geography, though the ratio is decreasing, indicating greater interest from other disciplines (CSISS, 2003). CSISS also produces sample WHY SPATIALLY INTEGRATED syllabi to be used to teach spatial analysis in a SOCIAL SCIENCE? variety of disciplines. The Pennsylvania State University (PSU) As much as GIS has revolutionized many has a strong GIS component as well, hosting disciplines and practices, including the social the multidisciplinary Geospatial Information

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 74 International Journal of Applied Geospatial Research, 3(2), 72-77, April-June 2012

Systems Council. PSU’s Population Research public health to civil engineering, continuously Institute (PRI) focuses on demography in the enroll in GIS classes, demonstrating the wide context of a variety of disciplines, and conducts application of this technology. Contrarily, the statistical and GIS analyses of population data. SISS program is an interdisciplinary program The PRI also offers an interdisciplinary dual incorporating the departments of Geography Ph.D. degree in demography (http://www. and Planning, Economics, Political Science pop.psu.edu/). PSU’s partnership with CSISS and Public Administration, and Sociology and has encouraged participation in workshops on Anthropology (Lindquist, 2009). The SISS pro- spatial analysis by demography students and gram at the University of Toledo incorporates faculty, helping to further the integration of many of the approaches outlined by CSISS. It the social sciences, GIS, and spatial statistics. is centered on spatial analysis tools, theories, The University of Illinois at Urbana- and problems as well as spatial information Champaign has established the University processing technologies, including GIS, remote Consortium of Geographic Information Science, sensing, and digital image analysis (SISS Ph.D. another prominent GIS program to promote the Graduate Program Handbook, 2009). interdisciplinary use of geospatial techniques. The Department of Geography and Plan- The GeoDa Center for Spatial Analysis and ning at the University of Toledo was already Computation at Arizona State University is the home to the Center for Geographic Information current home of the GeoDa software (http:// Sciences and Applied Geographics (GISAG), a geodacenter.asu.edu/). Another spatial analysis well-equipped research center which conducts software tool is CrimeStat, developed at the much of the University of Toledo’s GIS work Inter-University Consortium for Political and (Czajkowski et al., 2003). Through the GISAG, Social Research at the University of Michigan. there have been many joint projects between CrimeStat is in use for practical application by the departments of Geography and Planning, police departments as well as by criminal justice Economics, Political Science, Environmental and social scienceIGI researchers GLOBAL (Inter-University Sciences, PROOF Business Administration, and Civil Consortium for Political and Social Research Engineering. Thus, the Ph.D. program in SISS at the University of Michigan, 2011). This is a is a natural extension of this collaboration and short sampling of some of the centers that have the increasing applicability of spatial analysis created spatial analysis tools and are focusing (SISS, 2011). It is expected that this degree on an interdisciplinary approach to furthering program will solidify such research focus, spatial research. Many universities across the attracting faculty and students of diverse back- United States offer classes, certificates, and ground, and strengthen the university’s role as degrees in GIS. The SISS program at the Uni- a research institution. versity of Toledo establishes a Ph.D. degree that The original proposal for the SISS program focuses on the theories, tools, and approaches was submitted to the Ohio Board of Regents of advanced spatial analysis. in 2007, and the program was approved by the Board of Regents and the University of Toledo’s Board of Trustees in September 2008 (Lindquist, THE SISS PROGRAM AT THE 2009). The first and second classes of four and UNIVERSITY OF TOLEDO five students entered into the program in fall 2009 and fall 2010. Although the application The SISS program is the only Ph.D. granting process for admission is on the move, it is program of this nature in the Midwest. The anticipated that the number of students who Master’s Certificate Program in GIS at the would enroll into the program in fall 2011 will University of Toledo is housed within the De- be stable compared to last two years. As such partment of Geography and Planning, however one can be optimistic about the sustainability students from several disciplines, ranging from of the program. The admitted students come

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. International Journal of Applied Geospatial Research, 3(2), 72-77, April-June 2012 75 from a diverse background: geography, plan- EXPECTED CONTRIBUTION ning, health education, engineering, sociology, TO NORTHWEST economics, and such. The applicants have to OHIO’S ECONOMY meet three pre-requisites for entering into the program. The first of those is that they have Northwest Ohio is, like many older industrial to have a master’s or equivalent in a Social areas, in a period of self-reinvention. The City Science discipline. The second and third of Toledo has a strong legacy in manufacturing pre-requites include prior successful comple- and supplying glass and parts to Detroit auto tion of two courses in GIS and one course in industry. The region continues to shift away graduate-level multivariate statistics. A typical from traditional manufacturing due to decline student would successfully complete six core in that industry, but is hopeful to make use of courses (18 credit hours), three advanced the skilled labor pool to focus on high tech seminar courses (nine credit hours) and three manufacturing and alternative energies. The elective courses (nine credit hours) to fulfill the University of Toledo is very involved in this requirement of 36 credit hours of coursework. transition, and plays a vital role in the regional The core courses consist of Spatial Statistics, economy. As such, the creation of the Ph.D. Geographical Information Science in SISS, program in SISS at the University of Toledo Foundations of SISS, SISS Theory, Advanced is expected to contribute to this by attracting a Spatial Data Analysis, and Research Design. unique group of students, faculty, and collabo- The advanced seminar courses, on the other rators. The SISS program is also expected to hand, include Geo-Computation, Advanced contribute to the local economy by working as Qualitative Analysis in SISS, Policy Evalu- a research center assisting local organizations, ation and SISS, Space and Society: Critical governments, and businesses. Theory in SISS, Discrete Choice Spatial Process One specific area where the SISS program Modeling, Advanced Modeling Methods and expects to have a significant impact is in the Techniques in SISS,IGI Spatial PerspectivesGLOBAL on the area of transportationPROOF research. The Greater Environment, Spatial Transport Modeling and Toledo region is located on the Lake Erie and Planning, Directed Readings in SISS, Seminar in contains the intersection of the Ohio Turnpike Special Topics, and Doctoral Dissertation. The and I-75, making it an ideal transportation hub. elective courses can be taken from any of the The region has embraced this possibility, and the participating departments. In addition, she/he SISS program has enough room to collaborate has to complete 24 credit hours of dissertation with local government agencies, businesses, research and successfully defend the research and the Intermodal Transportation Institute at as the requirements to receive the Ph.D. degree the University of Toledo. (Lindquist, 2009; SISS, 2011). Currently the core instructors are com- prised of nine faculty members representing CONCLUSION the participating departments. Because the core As spatial analysis approaches gain popularity faculty members have diverse background and and the processing abilities of spatial software expertise, and that the admitted students come grow, we can expect to see continued increases with similar diverse backgrounds, the students in interest in the philosophy and methodology have wide open options to expertise in differ- of spatial analysis theory and applications. ent subfields of SISS: transportation, hous- This unique program at the University of To- ing, environment, economics, public health, ledo should prove to contribute not only to the demography, and such. The list of core faculty economy of Northwest Ohio, but to the literature members and their expertise are available from and discipline of advanced spatial analysis. the SISS program website (SISS, 2011). The program is designed to train and educate

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. 76 International Journal of Applied Geospatial Research, 3(2), 72-77, April-June 2012 students on the cutting edge technology and Goodchild, M., Anselin, L., Appelbaum, R., & new ways of thinking about data, giving them Harthorn, B. H. (2000). Towards spatially integrated the skills to contribute in both theoretical and social science. International Regional Science Re- view, 23(2), 139–159. practical applications of spatial analysis in the social sciences. It envisions attracting students Goodchild, M., & Janelle, D. (2004). Thinking spa- from outside the Northwest Ohio region due to tially in the social sciences . In Goodchild, M., & Janelle, D. (Eds.), Spatially integrated social science its distinctive curriculum. It is expected that (pp. 3–17). New York, NY: Oxford University Press. the students will both help the region grow and participate in research that will further promote Inter-University Consortium for Political and Social economic development and policymaking. Research at the University of Michigan. (2011). CrimeStat III: A spatial statistics program for the One may expect, then, that the program analysis of crime incident locations. Retrieved will help establish the University of Toledo as April 27, 2011, from http://www.icpsr.umich.edu/ a significant center of SISS applications and CrimeStat/ research. The use of spatial analysis tools is on Lindquist, P. (2009). Incorporating intermodal the rise, and as we study more facets of spatial transportation into the spatially integrated social data, from demography to public health to land sciences. Toledo, OH: University of Toledo – Uni- use to transportation, it will be important to be versity Transportation Center. Retrieved April 27, able to identify patterns when taking several 2011, from http://www.utoledo.edu/research/ututc/ unique variables into account, as well as to be docs/UTUTC_SISS__FinalReport.pdf able to measure and understand the relation- Logan, J., Zhang, W., & Xu, H. (2010). Applying ships and connections of different places and spatial thinking in social science research. GeoJour- scales. There are several reasons, therefore, to nal, 75(1), 15–27. doi:10.1007/s10708-010-9343-0 be optimistic that the Ph.D. program in SISS at Páez, A., & Scott, D. (2004). Spatial statistics for the University of Toledo will help further the urban analysis: A review of techniques with ex- abilities of several disciplines to do just that. amples. GeoJournal, 61(1), 53–67. doi:10.1007/ IGI GLOBALs10708-004-0877-x PROOF Sui, D. (2004). GIS, cartography, and the ‘‘Third REFERENCES Culture’’: Geographic imaginations in the computer age. The Professional Geographer, 56(1), 62–72. Anselin, L. (2006). How (not) to lie with spatial The University of Toledo. (2009). Spatially Inte- statistics. American Journal of Preventive Medicine, grated Social Science Ph.D. Graduate Program 30(2), 3–6. doi:10.1016/j.amepre.2005.09.015 Handbook 2009-2010. Toledo, OH: Author. Anselin, L., Syabri, I., & Kho, Y. (2006). GeoDa: The University of Toledo. (2011). Ph.D. Program An introduction to spatial data analysis. Geo- in Spatially Integrated Social Science. Toledo, OH: graphical Analysis, 38(1), 5–22. doi:10.1111/j.0016- Author. Retrieved April 27, 2011, from http://www. 7363.2005.00671.x utoledo.edu/as/siss/ Center for Spatially Integrated Social Science. Voss, P. (2007). Demography as a spatial social (2003). Survey of CSISS program applicants and science. Population Research and Policy Review, participants 2000-2002. Santa Barbara, CA: UCSB 26(5), 457–476. doi:10.1007/s11113-007-9047-4 Social Science Research Center. Weeks, J. (2004). The role of spatial analysis in Center for Spatially Integrated Social Science. (2011). demographic research . In Goodchild, M., & Janelle, Spatial resources for the social sciences. Retrieved D. (Eds.), Spatially integrated social science (pp. April 27, 2011, from http://www.csiss.org 381–399). New York, NY: Oxford University Press. Czajkowski, K., Lawrence, P., Attoh, S., Reid, N., & Lindquist, P. (2003). Acquisition of instrumentation in support of the Center for Geographic Information Science and Applied Geographics (GISAG). Arling- ton, VA: National Science Foundation.

Bhuiyan Monwar Alam is an assistant professor of urban planning in the Department of Geog- raphy & Planning, and a core faculty in the PhD program in SISS at the University of Toledo, Ohio. He teaches “spatial statistics” course in the SISS PhD program. He also teaches courses on transportation planning, quantitative techniques, land development & planning, community planning workshop, seminar in spatial analysis, urban & regional planning techniques, world regional geography, and world cities in the graduate and undergraduate programs in geography & planning at the same university. His research interests focus on transportation planning, GIS and spatial analysis, environmental planning, brownfield redevelopment, and planning practices in South Asia. He holds a PhD degree in transportation planning and a Master’s in transportation engineering from Florida State University, another Master’s in regional & rural development planning from the Asian Institute of Technology, and a Bachelor’s in civil engineering from Bangladesh University of Engineering & Technology.

Jeanette Eckert is a survey research center manager and research assistant in the Urban Affairs Center at the University of Toledo, Ohio. She holds a Master’s in geography & planning from the same university. Her research interests focus on community gardening, spatial analysis, and brownfield redevelopment.

Peter S. Lindquist is an Associate Professor of Geography and is the director of the Spatially Integrated Social Science PhD Program at The University of Toledo. He is also director of the university’s Center for Geographic Information Science and Applied Geographics. He teaches courses in Geographic Information Systems/Science, Location Analysis, Geocomputing and Quan- titative Methods. His current research is devoted to the acquisition and analysis of data within a comprehensive repository dealing with freight movements in the Upper Midwest. In conjunction with these efforts heIGI has developed GLOBAL a set of web-based GIS systemsPROOF for viewing freight movements. He has also participated in studies linking economic development to the freight transportation infrastructure. He has developed a variety of operations research software applications geared toward location analysis, vehicle routing, and performance evaluation of the transportation system. He holds a BS degree from the University of Wisconsin-Eau Claire, an MS degree from the University of Wisconsin-Madison and a PhD from The University of Wisconsin-Milwaukee.

An official publication of the Information Resources Management Association

Mission The International Journal of Applied Geospatial Research (IJAGR) publishes research that exemplifies the usage of geographic information science and technology (GIS&T) to explore and resolve geographical issues from various application domains within the social and/or physical sciences. IJAGR is designed to provide planners and policy analysts, practitioners, academicians, and others using GIS&T useful studies that might support decision-making activities.

Subscription Information IJAGR is published quarterly: January-March; April-June; July-September; October-December by IGI Global. Full subscription information may be found at www.igi-global.com/ijagr. The journal is available in print and electronic formats.

Institutions may also purchase a site license providing access to the full IGI Global journal collection fea- turing more than 100 topical journals in information/computer science and technology applied to business & public administration, engineering, education, medical & healthcare, and social science. For information visit www.igi-global.com/isj or contact IGI at [email protected].

Copyright The International Journal of Applied Geospatial Research (ISSN 1947-9654; eISSN 1947-9662). Copy- right © 2012 IGI Global. All rights, including translation into other languages reserved by the publisher. No part of this journal may be reproduced or used in any form or by any means without written permission from the publisher,IGI except GLOBALfor noncommercial, educational PROOF use including classroom teaching purposes. Product or company names used in this journal are for identification purposes only. Inclusion of the names of the products or companies does not indicate a claim of ownership by IGI Global of the trademark or registered trademark. The views expressed in this journal are those of the authors but not necessarily of IGI Global.

Correspondence and questions: Editorial: Donald Patrick Albert Subscriber Info: IGI Global Editor-in-Chief, IJAGR Customer Service Department of Geography and Geology, LDB #332 701 E Chocolate Avenue Sam Houston State University, 1900 Ave I Hershey PA 17033-1240, USA Huntsville, Texas 77341-2148, USA Tel: 717/533-8845 x100 Tel: 1-936-294-1453 E-mail: [email protected] Fax: 1-936-294-4203 E-mail: [email protected]

The International Journal of Applied Geospatial Research is currently listed or indexed in: Bacon’s Media Directory; DBLP; Google Scholar; INSPEC; Library and Information Science Abstracts (LISA); MediaFinder; The Standard Periodical Directory; Ulrich’s Periodicals Directory