CONTENTS. Volume 13 Number 2 Spring Journal of the Urban and Regional Information Systems Association

Transcription

1 Volume 13 Number 2 Spring 2001 Journal of the Urban and Regional Information Systems Association CONTENTS REFEREED 5 GIS Education Today: From GI Science to GI Engineering Andrew U. Frank and Martin Raubal 11 Exploring County-level Production of Framework Data: Analysis of the National Framework Data Survey David Tulloch and Jennifer Fuld 23 Linking Transit Operational Data to Road Network with a Transportation Object- Oriented GIS Martin Trépanier and Robert Chapleau PLUS 31 Comparative Book Review 33 Book Review: Remote Sensing and Urban Analysis 35 Monitoring Land Supply with Geographic Information Systems: Theory, Practice, and Parcel-Based Approaches 37 Book Review: Valuing the Built Environment: GIS and House Price Analysis 39 Book Review: Information, Place, and Cyberspace 41 Book Review: Geographical Information Systems, Second Edition. Volume 1. Principles and Technical Issues and Volume 2. Management Issues and Applications 45 GIS in Public Policy: Using Geographic Information for More Effective Government 47 Database Review: American FactFinder by the U.S. Census Bureau 51 Software Review: ArcPad 5.0.1: A Heavy-Hitting Lightweight On the Cover Choice. It is the spark that flames free will. The more choices we have in life the more freedom and opportunity we have stretched out in front of us. A college or university graduate in the field of geoscience now has a number of options presented to them upon commencement. The old standard of graduates cutting their teeth in large organizations, with all-inclusive GIS systems is being supplanted with the idea that the GIS post-graduate now may enter the workforce through a small, contracted, private sector firm. This Big GIS vs. Small GI decision and its impact on the global GIS community is the subject of an article by Andrew U. Frank, a Professor of Geoinformation, and Martin Raubal, a Ph.D. candidate, both at the Technical University of Vienna in Austria. This burgeoning situation along with a GIS education model to help explain it, are presented in their paper which highlights the Spring Issue of the URISA Journal.

2 Journal Publisher: Editor-in-Chief: Journal Coordinator: Executive Director: Electronic Journal: Urban and Regional Information Systems Association Harlan Onsrud Scott A. Grams William A. Gentes username: URISAMEMBER password: LONGBEACH EDITORIAL OFFICE: Urban and Regional Information Systems Association, 1460 Renaissance Drive, Suite 305, Park Ridge, Illinois ; Voice (847) ; Fax (847) ; SUBMISSIONS: This publication accepts from authors an exclusive right of first publication to their article plus an accompanying grant of nonexclusive full rights. The publisher requires that full credit for first publication in the URISA Journal is provided in any subsequent electronic or print publications. For more information, the Manuscript Submission Guidelines for Refereed Articles is available on our web site, or by calling (847) SUBSCRIPTION AND ADVERTISING: All correspondence about advertising, subscriptions, and URISA memberships should be directed to: Urban and Regional Information Systems Association, 1460 Renaissance Dr., Suite 305, Park Ridge, Illinois, ; Voice (847) ; Fax (847) ; URISA Journal is published four times a year by the Urban and Regional Information Systems Association by the Urban and Regional Information Systems Association. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by permission of the Urban and Regional Information Systems Association. Educational programs planned and presented by URISA provide attendees with relevant and rewarding continuing education experience. However, neither the content (whether written or oral) of any course, seminar, or other presentation, nor the use of a specific product in conjunction therewith, nor the exhibition of any materials by any party coincident with the educational event, should be construed as indicating endorsement or approval of the views presented, the products used, or the materials exhibited by URISA, or by its committees, Special Interest Groups, Chapters, or other commissions. SUBSCRIPTION RATE: One year: $295 business, libraries, government agencies, and public institutions. Individuals interested in subscriptions should contact URISA for membership information. US ISSN URISA Journal Vol. 13, No. 2 Spring 2001

3 EDITORS AND REVIEW BOARD URISA Journal Editor Editor-in-Chief Harlan Onsrud, Spatial Information Science and Engineering, University of Maine Thematic Editors Editor-Urban and Regional Information Science Lewis Hopkins, Department of Planning, University of Illinois-Champaign/Urbana Editor-Applications Research Lyna Wiggins, Department of Planning, Rutgers University Editor-Social, Organizational, Legal, and Economic Sciences Ian Masser, Department of Urban Planning and Management, ITC (Netherlands) Editor-Geographic Information Science Michael Goodchild, Department of Geography, University of California-Santa Barbara Editor-Information and Media Sciences Michael Shiffer, Department of Planning, Massachusetts Institute of Technology Editor-Spatial Data Acquisition and Integration Gary Hunter, Department of Geomatics, University of Melbourne (Australia) Editor-Geography, Cartography, and Cognitive Science David Mark, Department of Geography, SUNY-Buffalo Editor-Education Karen Kemp, Department of Geography, University of California-Berkeley Section Editors Software Review Editor Jay Lee, Geography, Kent State University Book Review Editor Rebecca Somers, Somers-St. Clair Literature Review Editor Zorica Nedovic, University of Illinois- Champaign/Urbana Article Review Board Peggy Agouris, Department of Spatial Information Science and Engineering, University of Maine Michael Batty, Centre for Advanced Spatial Analysis, University College London (United Kingdom) Kate Beard, Department of Spatial Information Science and Engineering, University of Maine Yvan Bédard, Centre for Research in Geomatics, Laval University (Canada) Barbara P. Buttenfield, Department of Geography, University of Colorado Keith C. Clarke, Department of Geography, University of California-Santa Barbara David Coleman, Department of Geodesy and Geomatics Engineering, University of New Brunswick (Canada) David J. Cowen, Department of Geography, University of South Carolina Massimo Craglia, Department of Town & Regional Planning, University of Sheffield (United Kingdom) William J. Craig, Center for Urban and Regional Affairs, University of Minnesota Robert G. Cromley, Department of Geography, University of Connecticut Kenneth J. Dueker, Urban Studies and Planning, Portland State University Geoffrey Dutton, Spatial Effects Max J. Egenhofer, Department of Spatial Information Science and Engineering, University of Maine Manfred Ehlers, Geoinformatics and Institute for Environmental Sciences, University of Vechta (Germany) Manfred M. Fischer, Economics, Geography & Geoinformatics, Vienna University of Economics and Business Administration (Austria) Myke Gluck, School of Information Studies and Geography, Florida State University Michael Gould, Department of Science, Experimentales Universitat (Spain) Daniel A. Griffith, Department of Geography, Syracuse University Francis J. Harvey, Department of Geography, University of Kentucky Kingsley E. Haynes, Public Policy and Geography, George Mason University Eric J. Heikkila, School of Policy, Planning, and Development, University of Southern California Stephen C. Hirtle, Department of Information Science and Telecommunications, University of Pittsburgh Dr. Gary Jeffress, Department of Geographic Information Science, Texas A&M University- Corpus Christi Richard E. Klosterman, Department of Geography and Planning, University of Akron Robert Laurini, Claude Bernard University of Lyon (France) Thomas M. Lillesand, Environmental Remote Sensing Center, University of Wisconsin Xavier R. Lopez, Oracle Corporation David Maguire, Environmental Systems Research Institute John McLaughlin, Research and International Cooperation, University of New Brunswick (Canada) Harvey J. Miller, Department of Geography, University of Utah Joel L. Morrison, Center for Mapping, Ohio State University Atsuyuki Okabe, Department of Urban Engineering, University of Tokyo (Japan) Jeffrey K. Pinto, School of Business, Penn State Erie Gerard Rushton, Department of Geography, University of Iowa Bruce D. Spear, Geographic Information Services Bureau of Transportation Statistics, Washington, D.C. Jonathan Sperling, Geography Division, U.S. Census Bureau David J. Unwin, School of Geography, Birkbeck College, London (United Kingdom) Stephen J. Ventura, Environmental Studies and Soil Science, University of Wisconsin-Madison Nancy von Meyer, Fairview Industries Barry Wellar, Department of Geography, University of Ottawa (Canada) Michael F. Worboys, Department of Computer Science, Keele University (United Kingdom) URISA Journal Vol. 13, No. 2 Spring

4

5 GIS Education Today: From GI Science to GI Engineering Andrew U. Frank and Martin Raubal Abstract: Discussions about geographic information system (GIS) education should include content as well as methodology and technology. Currently, the trend has moved from Big GIS, as built and used by large organizations, to Small GI (geographic information) enterprises that sell small pieces of information to individuals. GIS education needs to adapt to this trend by extending its content to include legal and the business aspects of geographic data. A GIS education model based on the three pillars of Geo, Info, and Business is proposed. It is further argued that education curricula move from a GI science perspective to a GI engineering approach to satisfy industrial demands for the required skills of graduates. Introduction It is critical that educational programs and teaching methods be adapted to the current climate: The methods must use the best technology to help students learn, and the content must cover what graduates need to know. There is extensive debate on the best use of modern technology for education at all levels - for example, various methods from Virtual Field Trips (EUGISES 2000) to distance learning using the Web (Johnson 2000) were discussed during the recent 2nd European GIS Education Seminar in Budapest. Discussion of technology must not distract from the primary concern in education: What should students learn? Discussion about the content of geographic information system (GIS) education is not new. Efforts include the development of the National Center for Geographic Information and Analysis core curriculum (Kemp 1990), a Delphi study to identify the expected knowledge and skills for GIS managers (Kemp et al. 1993), and a dialogue between potential employers from industry and major user organizations (Timpf 1998). Similar debates have been held in other countries. Based on these discussions, a standard curriculum content has evolved that is covered in most recognized textbooks on GIS such as DeMers (1999), Bernhardsen (1999), Burrough and McDonnell (1998), and Worboys (1995). The growth of standard GIS applications is slowly approaching saturation; convincing state agencies of the need for new GIS installations is coming up against empty state funds and a general trend toward lean governments. However, interest in using GIS commercial applications is rapidly gaining steam. This new use of commercial GI (geographic information) is - as most efforts in the new technology sector - limited by the number of well-educated and competent professionals available. The main argument here is that the commercial use of GIS has a different quality than the traditional uses that we have seen up to this time. We will contrast the traditional Big GIS with the new commercial Small GI and discuss the knowledge and skills of the professionals who will make this new trend a reality. The Change in the GIS Field: From Big GIS to Small GI In the past, GIS was mostly built for and used by large organizations that needed spatial information on a regular basis to make decisions. These organizations collected data, managed them in their own databases, and produced reports and maps for various internal uses. The cost-benefit analysis is notoriously difficult within a single organization; it is difficult to assess the extent that a service contributes. The planning of systems was usually influenced more by internal politics than by deep analyses of requirements or business re-engineering (Hammer 1990). Numerous examples demonstrate that the systems are beneficial to the organization, but often not for the reasons for which they were originally designed. A cost-benefit analysis may typically show benefits in the form of a reduced workforce, but actual implementation later demonstrates that the major benefits are, for example, in better and faster services to the customers. We want to contrast such systems built for public utilities, towns, and regions for planning purposes, highway departments, etc., with GIS established by service providers who build a GIS and collect the necessary data in order to sell the information produced with the system to many users in small quantities. We call this Small GI, because the amount of information sold in each case is very small. It is clear that the systems are not necessarily smaller or require less data collection and management than Big GISs. The difference is in the separation between the organization that operates the GIS and the user of the information produced. Small GI is a commercial enterprise: the service provider is paid to operate the GIS. The income comes from either the users of the GI to which it is sold or by some other organization that indirectly benefits from the information. For example, location- URISA Journal Frank, Raubal 5

6 based services, where vendors such as hotels and restaurants are willing to pay a fee to have potential users efficiently informed about their location and offerings; this is advertising in a new, more direct form. The trend to Small GI service providers is advanced by an increased need to acquire spatial information. In the past, spatial information has been found in the environment when needed, learned over the course of a lifetime, and never paid for. In today s world of high mobility, we often find ourselves in unfamiliar environments. To plan for future activities, we need to acquire the information before actual travel. Consider, for example, a business trip using different modes of transportation, where the traveler needs to know in advance the connection times needed so as to meet for an appointment. At the same time, today s environment is transformed by new construction and transportation technology, which makes navigation more difficult. Commonsense is no longer sufficient, and counter-intuitive moves are often necessary to reach a goal. Providing spatial information is becoming a business. To estimate the size of the market for GI, one can start with the well-known statement that 80% of all decisions are in some way spatial (Albaredes 1992); this can be combined with the observation that, in all cases, where the benefits of GI can be assessed economically (e.g., in the use of GI for routing), increases in efficiency of about 20% are observed. This suggests that up to 16% of the Gross National Product is created from GI; this is an enormous market potential that can only be realized over an extended period of time. Small GI produces information for sale to others. The value of the information to the potential clients must be larger than the price charged (Krek and Frank 2000). The use of GI in business has been very difficult in the past. If the only distribution channel for geographic data is in the form of paper maps, a just in time delivery of information is not possible. Mobile phone technology creates a channel through which the required information can be transported to the user on demand and just in time. Mobile phones and Personal Digital Assistants are the first steps to computer systems that can be carried around all the time; see also an article on wearable computers produced by the Massachusetts Institute of Technology (2000). A great number of possible application areas exist: services to inform potential clients of the locations of hotels, restaurants, cultural sites, etc., but also such specialty areas as automatic teller machines (Kottman 1998). The combination of Yellow Page information with location-based technology creates opportunities for businesses. However, services are also available for assisting persons in the process of buying or renting a new home or apartment and for users of public transportation systems. This latter case is used in the next section to give a concrete example for such an application that is technically feasible today. Information for Users of Public Transportation Systems The user of public transportation requires surprising amounts of information. Most users are habitual and once they have acquired this information, they use it on a daily basis and are not aware of the amount of detailed knowledge they use. However, when a person arrives in a foreign country, they are painfully aware of the information needs and the effort needed to use the public transportation system - constantly asking for help, often to no avail, because of a lack of knowledge of the local language. Similarly, handicapped persons, particularly those who are site-challenged, have a difficult time using the public transportation system because they cannot acquire the necessary information about the spatial environment through visual perception (Golledge et al. 1998). The following are needs for users of public transportation systems: Information must be collected to make a routing decision and to decide which transportation service is used for which part of the journey. Knowledge about tickets and reservations is necessary. Knowledge of the location of departure points and where to buy tickets and make reservations is needed. In the world that we live in today, the friendly and usually helpful person in the ticket office is being increasingly replaced by ticket vending machines. This poses such interesting challenges to the user as instructions are often only in the local language; payment is accepted only with a restricted set of local currency tokens. More user-friendly and therefore better systems are feasible today: Let us assume that a traveler has booked an airplane ticket and a hotel through a travel agent. After arrival at the airport and claiming his or her baggage, the traveler is informed through the Global System for Mobile (GSM) communication phone about the best route to the hotel, and all of the information necessary to carry out this trip is provided step by step. It is not effective to deliver all information ahead of time and in printed form, because this is not flexible enough. New instructions due to wrong actions of the user or changes in the environment (e.g., air planes and trains being late) are not possible; also, the memory of the traveler is not sufficient to take in all of this information, which is only useful for a single decision at a particular moment in time. Only for those trips that we undertake frequently are we willing to invest time and energy in the learning of the route; for all other trips, we navigate as well as we can without learning. Using the GSM communication phone, the traveler can be assisted with necessary tickets - ticketing using mobile communication is working today with the Austrian Railway System (for further details in German see 14.html), and it is expected that ticketing in all forms (e.g., for rail, public transportation in cities, and movie theaters) will be an interesting business aspect of future mobile communication systems. 6 URISA Journal Vol. 13, No. 2 Spring 2001

7 from another GIS operator) and provides a pay-per-use service (Wenzl 2000), providing small pieces of information to users precisely when and where the information is needed. The fee must be less than the value of the information for the user. To advance such business, competent GI specialists must establish the GIS, design the user interfaces, organize the business, and be aware of legal issues such as privacy and copyright of data. In the next section, we look at the question of what knowledge and skills are needed by the professionals. Figure 1: Information on modes of transport through wireless application protocol (WAP) service. Figure 2: Information on how to go from the City-Air-Terminal to the subway station. With information technology, it is possible to guide a user continuously on a trip and provide all information just in time and accurately for the situation of the traveler (Figures 1 and 2), even if mistakes are made and parts of the transportation systems malfunction. The traveler can be saved all efforts to collect the necessary information. For a 70-minute trip from Vienna International Airport to a hotel in the city, we observed that a total of 30 minutes was used in information collection (Pontikakis et al. 2000). The value of the information is certainly a reasonable percentage of the cost of the ticket (in this case $6.50) - say $1.00 to $2.50. Every day, approximately 16,400 visitors arrive at the airport; if only 10% use the information service, the service provider s income will be between $600,000 and $1,500,000. Obviously, many other uses of GI are possible, all leading to the same business structure. A service provider collects data (often, most of the data are already collected or can be acquired Education Needs Graduates must be equally prepared for Big GIS and Small GI. The content taught in the past enabled graduates to establish GIS and to collect and manage data for organizations that operate their own GIS. These skills continue to be important for the service providers in the Small GI business. In addition to technical questions, the integration of data from different sources becomes more important, and limits to meaningful combinations require attention. Data quality and the manner in which it translates to quality in the GI produced becomes a central issue in a business, where clients buy information for a fee and expect quality (Goodchild and Jeansoulin 1998). Furthermore, customers may file liability claims if the information is grossly and negligently wrong (Perritt, Jr. 1996). For such demands, a GI science curriculum is not sufficient. The curriculum content is described in this section, and the difference between a science curriculum and an engineering approach is discussed in the next section. Discussion with industry has led to a curriculum design, which is based on three pillars (Figure 3). Geo The Geo pillar should provide students with a thorough feeling of how spatial situations can be observed, measured, analyzed, and represented. In addition to the fundamental concepts of physical and human geography (primarily the concept of processes in space), it includes the understanding of spatial data collection (e.g., surveying, photogrammetry, and remote sensing). Not only is it necessary to model and analyze spatial processes, we must also communicate the results. Digital cartography, another component of the Geo pillar, deals with principles of spatial visualization. Info Graduates must have knowledge of modern information technology and the necessary skills to use it ( Info ). This includes knowledge of programming languages and database technology, principles of user interface design, and the organization of information systems. Furthermore, students need to learn different aspects of networking and mobile technologies. The special problems of spatial information, such as spatial indexing and computational geometry, must be known. Graduates also need to understand information science - the logical structuring of information, the assessment of data quality, and the problems of inte- URISA Journal Frank, Raubal 7

8 Figure 3: Education of a GI professional is based on three pillars: Geo, Info, and Business. grating data from different sources and potentially different semantics. Business It is not sufficient that today s graduates know how to technically build GISs. The GI professional must understand the principles of marketing, understand the user needs, and see how they can be satisfied. Knowledge about the connection between technology and marketing is vital for the success of a Small GI enterprise. Furthermore, it is necessary to teach students the principles of e-commerce so that they know about such things as the mechanisms for collecting user fees, the organization of help desks and their costs, and marketing. Legal aspects concerning information systems and data are another major component of the Business pillar. Nowadays, the majority of data are transmitted electronically, collected by one person, distributed by someone else, and used by yet another person. This procedure is very sensitive to legal aspects, such as privacy rights, copyright of data, or legal liability (Perritt, Jr. 1996). Graduates must not only know about the legal impacts on the use of databases and spatial datasets, but also about legal options to deal with conflicts (Onsrud and Rushton 1995). Figure 4 shows our proposed curriculum in a generic form (details are available at It is a German-style 4-year program (after high school) and does not contain any general knowledge areas such as are in U.S. undergraduate programs. Students will acquire a practical degree that is comparable to a non-thesis M.S. degree and that will lead them to professional practice. We propose adding a sizable chunk of business-related teaching to the curricula (e.g., business law and marketing) because industry leaders have indicated that successful engineers are increasingly required to understand the business implications of technical solutions or to find technical solutions for business opportunities. Integration is only achievable if one person can oversee the project Figure 4: Generic model curriculum. - possibly asking specialists to contribute their individual pieces. Education is a zero sum game. Adding content is only possible if content is reduced at a different front. Industry is clearly telling us that they want young graduates, and that adding more years to the curriculum is not an option that should be considered! In the next section, we argue for an engineering approach and, thus, reduce the length of the study program compared to a scientifically oriented program. Form: Engineering Instead of Science We begin by describing our concept of engineering and science. Science is the search for knowledge (new knowledge, to be pre- 8 URISA Journal Vol. 13, No. 2 Spring 2001

9 cise), and engineering is the systematic application of the results of scientific research to solve real-world problems in a predictably successful way. The outcome of a scientific experiment cannot be guaranteed. If we know for sure what the result will be, then it is not a meaningful scientific experiment. An experiment that does not confirm the hypothesis is as valid as one that does. The outcome of an engineering project must predictably work; engineers build bridges and only very seldom does a bridge fail. Failures are not acceptable and are definitely detrimental to an engineer s career! Engineering reduces to practice the results of scientific research. Scientific laws are combined with the results of long-term experience with useful guidelines that codify the state of the art as standards, which engineers follow. Many GIS courses are situated in the departments of scientific geography or computer science. The primary goal of teaching in these departments, at least according to the theory, is oriented toward the student becoming a scientist. This is obviously not the goal of the majority of students, and curricula have been adapted to provide better service to the students leaving the university after having obtained a B.S. or an M.S. degree. It is suggested that this movement be pushed to a conclusion; that is, separating GI science (i.e., the efforts to advance our understanding of geoinformation in all forms) from GI engineering, which consists of science-based predictive rules on how to build working systems. In addition to introducing a change in the style of teaching GI, great effort is required to establish a GI engineering science (i.e., the scientific efforts to establish the rules and heuristics which engineers can use to build GI systems that predictably work). The scientific results that we have produced over the past decades are substantial and cover most aspects of GI. However, they are not yet reduced to practice to be usable to design systems that predictably work. Conclusions A discussion of education must not only focus on the use of new technology in the teaching environment, but must always and primarily consider the substantive issue of what the appropriate content is. Our teaching methods must be technically adequate and follow the development of technology, but content must also track changes in the real world. As the GIS environment changes with respect to technology, there is also change in the business organization. We have argued that, in addition to the well-known Big GIS installations, which are operated by large organizations and provide various bits of geoinformation to different parts of this organization, we are seeing a Small GI business emerge, where a service provider operates the GIS and sells small amounts of information to various users. These users buy the information as needed and with delivery at the instant they are making a spatial decision. In this article, we have demonstrated using Small GI for step-by-step guidance of users of a public transportation system on their way to their destination ( door to door ), combining spatial information regarding the location of service points, operational information such as train schedules, and the organization of business (such as ticketing). A curriculum organized to equip students with the necessary knowledge and skills to design, develop, and manage Small GI businesses adds Business to the two classical pillars of GI science - namely Geo (e.g., geography, surveying, and cartography) and Info (e.g., computer science and information technology) - as an equally important part. Understanding the concepts of business from marketing (from product design and user studies) to legal issues is indispensable in this realm. Welldesigned systems, where technology and business logic fit together, are only possible if a team leader can see the complete picture. As we propose to add new content to a GI curriculum, we must also indicate where reductions are possible. The professionals working with GIS processes are closer to engineers than to scientists. They must reliably solve real-world problems and design systems that work. They cannot take a scientific approach, where only novel problems are of interest and hypotheses that cannot be confirmed are as interesting as hypotheses that experiments confirm. This means that the results of the past decades of GI science research must be reduced to practice and simplified into useful engineering rules. Graduates must understand the rules and the scientific background that has led to them, and they must understand the limitations and the built-in assumptions. They must be GI engineers but they do not need to be GI scientists! GI engineering is necessary in order for GI to achieve an important place in our world. A large part of the Gross National Product is produced with the help of geoinformation, and future economic growth is best achieved by making business processes more efficient. Geographic knowledge is often the key to this. About The Authors Andrew U. Frank has been Professor of Geoinformation at the Technical University Vienna since Prior to this, Frank was a Professor of Land Information Studies at the Department of Surveying Engineering at the University of Maine. He was also head of the Maine branch of the National Center for Geographic Information and Analysis. He received a Ph.D. from the Swiss Federal Institute of Technology, Zurich, Switzerland in His research interests are in formal models of spatial problems, spatial cognition, user interfaces for geographic information systems, and the economical and organizational aspects of collection, management, and use of geographic information. Martin Raubal is a Graduate Research Assistant and Ph.D. candidate at the Institute for Geoinformation, Technical University Vienna. He earned an M.S. in Spatial Information Science and Engineering from the University of Maine (1997) and a Dipl.-Ing. in Surveying Engineering from the Technical University Vienna (1998). His research interests URISA Journal Frank, Raubal 9

10 are in formal models and simulation of human wayfinding, spatial cognition, and artificial intelligence. References Albaredes, G., 1992, A New Approach: User Oriented GIS. In Proceedings of EGIS 92, Munich, Bernhardsen, T., 1999, Geographic Information Systems: An Introduction (John Wiley & Sons). Burrough, P. and R. McDonnell, 1998, Principles of Geographical Information Systems. (Oxford: Oxford University Press). DeMers, M., 1999, Fundamentals of Geographic Information Systems (John Wiley & Sons). EUGISES, 2000, Second European GIS Education Seminar. In Markus, B. (Ed.), EUGISES 2000, Budapest, HU, CD- Rom. Golledge, R., R. Klatzky, J. Loomis, J. Speigle, and J. Tietz, 1998, A geographical information system for a GPS based personal guidance system. IJGIS, 12(7), Goodchild, M. and R. Jeansoulin (Eds.), 1998, Data Quality in Geographic Information - From Error to Uncertainty (Paris: Hermes). Hammer, M., 1990, Re-engineering Work: Don t Automate, Obliterate. Harvard Business Review, July - August 1990, Johnson, A., 2000, ESRI Virtual Campus Developments. In Markus, B. (Ed.), EUGISES 2000, Budapest, HU, CD- Rom. Kemp, K., 1990, A Review and Assessment of the NCGIA Core Curriculum Evaluation Program. In Brassel, K. (Ed.), 4th International Symposium on Spatial Data Handling, Zurich, Switzerland, Kemp, K., W. Kuhn, and A. Frank, 1993, Making High-Quality GIS Education Accessible: A European Initiative. Geo Info Systems, 3, Kottman, C., 1998, Progress Toward Interoperability and a Geospatial Infrastructure at the Open GIS Consortium. In Peckham, R. (Ed.), 4th EC-GIS Workshop, June 1998, Budapest. Krek, A. and A. Frank, 2000, The Economic Value of Geo Information. Geo-Informations-Systeme - Journal for Spatial Information and Decision Making, 13(3), The MIT Wearable Computing Web Page (May 2000), Onsrud, H. and G. Rushton, 1995, Sharing Geographic Information (Rutgers: CUPR Press). Perritt, Jr., H., 1996, Law and the Information Superhighway (New York: John Wiley & Sons). Pontikakis, E., A. Frank, and M. van der Vlugt, 2000, Information support for public transport, Technical Report for Ministry of Transport and Science (Vienna: Institute for Geoinformation, TU Vienna). Timpf, S., 1998, Workshop on the development of a curriculum for the Fachhochschule in Villach, Technical Report (Vienna: Department of Geoinformation, TU Vienna). Wenzl, P., 2000, A Technical Concept for Pay-per-Use in Geomarketing Services (Vienna: Dept. of Geoinformation, TU Vienna). Worboys, M., 1995, GIS: A Computing Perspective (London: Taylor & Francis). 10 URISA Journal Vol. 13, No. 2 Spring 2001

11 Exploring County-level Production of Framework Data: Analysis of the National Framework Data Survey David Tulloch and Jennifer Fuld Abstract: Throughout the United States, an enormous amount of geospatial data is produced at local levels. While these data serve the local efforts for which they are created, questions remain about whether the data can serve a larger audience as well. This article reports the results of a national survey on the state of local Framework data in the U.S., demonstrating that these data are a rich and diverse resource. The survey was conducted by the Federal Geographic Data Committee and the National States Geographic Information Council. Data analysis is based on a sample of 836 county-level data producers. We find that the quality, quantity, and very nature of Framework data are diverse, suggesting that regional or national Framework efforts are quite complicated. The analysis also finds that many data producers may not be fully prepared to share data in a manner consistent with regional or national spatial data infrastructure. Background A variety of research has been conducted into the nature and extent of various institutional practices in local government. Traditionally, much of this research has been performed under the broad umbrella of adoption and diffusion of geospatial technologies (e.g., Masser and Onsrud 1993). These studies examined the presence of technology and the patterns of adoption of the technologies (Campbell and Masser 1992, Budic 1993, Masser and Campbell 1994, Sprecher 1994, Cullis 1995, Chan and Williamson 1999), using surveys to assess the barriers to development (Onsrud and Pinto 1993, Tulloch 1997) and consider the ultimate goals and outcomes of the implementation process (Masser and Campbell 1993, Budic 1994, Campbell and Masser 1995). Overall, research has focused on explorations of issues impacting inter- and intra-organizational cooperation (Nedovic- Budic and Pinto 1999). The development of the Framework concept has been a major emphasis by the Federal Geographic Data Committee (FGDC) to assemble a National Spatial Data Infrastructure (NSDI). Shortly after the FGDC was formally created as an office intended to foster a nationwide network of publicly available spatial data, the agency identified several major concepts as central to supporting an NSDI (Clinton 1994). These specific areas of emphasis include metadata, clearinghouses, standards, stakeholders, and framework. While each of these areas is deemed important to facilitate the large-scale sharing of spatial data that would comprise an NSDI, it has been the Framework concept that has emerged as particularly important for investigation. Framework is a set of common themes of spatial data that could serve as a crucial foundation upon which other data types could be built. The NSDI framework is an initiative to develop a readily available set of basic geographic data. It includes the information, operational environment, and technology to provide access to these data, and the institutional setting to sustain its development. (FGDC 1997:70) The FGDC encourages a broader acceptance of the Framework concept among the GIS community, and one of its priorities has been to explore organizations and agencies, both governmental and private, that generate Framework data. As the Framework data concept was developing, the FGDC sponsored a survey of likely Framework participants to explore user needs (Frank et al 1996). In the late 1990s, the National States Geographic Information Council (NSGIC) and the FGDC developed and distributed an extensive national survey to explore and assess the status of publicly available geospatial data from a wide variety of data producers in the U.S. This significant process involved the design of the survey instrument, the distribution of thousands of surveys, and the subsequent construction of a directory of Framework data users. The survey methods and design are quite complex and have been detailed in both a Web-accessible outlet ( and in the recently published article, A progress report on a US National Survey of Geospatial Framework Data (Tulloch and Robinson 2000). The foremost goal of the survey was to measure the current status of Framework data throughout the U.S., including quantity, quality, and utilization. The Framework Survey was distributed throughout the entire country to approximately 15,000 potential respondents (including federal, state, local, private, and other GIS data producers), and more than 5200 surveys were completed and returned. In order to distribute the survey within all 50 states, the FGDC and NSGIC worked cooperatively to identify a survey coordinator in each state who would determine the actual list of respondents for their state. Since the sampling frame for each state was not determined before instructing the state coordinators on survey distribution, the state coordinators created a list of URISA Journal Tulloch, Fuld 11

12 potential geographic information system (GIS) users, including state, local private, tribal, and other GIS data producers and sent surveys to those respondents. While these lists became the default sampling frames within each state, the initial task of the survey coordinator was to conduct a census of all possible Framework data users. In other words, rather than choosing a specific sample of respondents, the survey coordinator was to send a survey to every person on the list that he/she had generated. However, while each state coordinator was instructed on a uniform survey distribution process and collection, in reality there was no uniform sampling design across states. Not all of these potential Framework data users returned the survey, and there was no follow-up on nonrespondents. The various processes by which surveys were distributed within each state as well as the nonresponse rate for each state impacted the analysis of the collected data. Specifically, most of the data collected provides exploratory information concerning specific Framework data users; however, state comparisons of GIS Framework data as well as empirical generalizations to the country-wide level of Framework data are severely limited. At the present time, there is no report or publication on the research design limitations and subsequent limitations for analysis, but one important issue is that isolated respondents may not make their work well known. A lesser-known county office is difficult to target if it does not communicate well with other agencies within its jurisdiction or participate in state or regional groups. It is the organizational and communication patterns of officials that makes any regional or national GIS census so difficult. While the county-level responses reported here parallel some municipal, state, and federal responses, it is important to note that there are also significant variations and, therefore, the findings reported here cannot be generalized to other governmental levels. Overall Response Patterns From the data collected, we can begin to see specific patterns among Framework users; while these data provide a portrait of a subset of organizations utilizing Framework data and presents some patterns of Framework data users, they do not fully describe the larger population of GIS data producers in the U.S. Of 15,000 distributed surveys, 5348 were completed; 69% of the respondents indicated that their organization was involved in creating, integrating, updating, or disseminating digital geographic data. Of 3688 respondents producing digital geographic data, 73% indicated that their organization was specifically involved in creating, integrating, updating, or disseminating Framework data. Of the overall 3688 data producers (including both Framework and non-framework data producers), 33% reported that their jurisdiction was best described as a county, 21% responded that they worked with a state-level jurisdiction, and 19% with a municipal jurisdiction (Table 1). A total of 14% reported that their jurisdiction was multi-county, including organizations such as regional planning agencies and districts of state and federal agencies, 5% national jurisdiction, 3% tribal jurisdiction, and 5% other. From these data, it is clear that county jurisdiction data producers are a noteworthy portion of all of the data producers who responded to the survey. This reflects an emphasis in the data collection techniques that focused on county governments. In addition, nongovernmental organizations and privatesector organizations that focus largely on a single county area are also included within the category of county-level data producers. Selected Survey Responses The survey results in this analysis are drawn from a publicly accessible dataset. These data are available for download at the FGDC website: download.html. This site includes access to a complete copy of the survey instrument, data in multiple formats, and a description of the survey process. Additional resources continue to be added to the site to assist users in analyzing these data. Due to the varying sampling frames by state and nonresponse rates within a state, it is crucial that any analysis take into consideration the limitations of the data. Because state coordinators were instructed to send a survey to every county in their state, the overall county response rate is somewhat higher than the other agencies and organizations surveyed. Therefore, we analyze county-level data for this article and refer readers to the site above for further noncounty jurisdiction level analysis. For this reason and to facilitate meaningful exploration of the data, 26 states were identified for analysis in this article. After examining the county-level response rate for each state, it was Table 1 Percent of respondents producing data by jurisdiction type* Jurisdiction Type Frequency Percent of Data Producing Respondents County % State % Municipality % Multi-County % National 183 5% Tribal 98 3% Other 105 3% Sub-County 51 1% Multi-State 41 1% * Counties comprised one-third of the total response. A total of 3688 responded. 12 URISA Journal Vol. 13, No. 2 Spring 2001

13 Figure 1: A Map of County-Level Responses from Each State Analyzed. States Shown in White were Discarded for Analytical Purposes regarded as necessary to eliminate those states with low response rates before presenting an analysis of results. Therefore, states were eliminated on the basis of state response rates of less than 50% of all organizations surveyed and/or states with noticeably biased response rates; for example, response rates correlated with population county size. As a result, the 26 states in this analysis include those states with response rates above 50% and without biased nonresponse rates. Furthermore, based on this sample of 26 states, respondents who indicated that they were not in any way involved with Framework data were eliminated from this analysis. As a result, 836 county-level responses (out of a nationwide group of 1600 county responses) are analyzed in this article (Figure 1). It is important to note that the jurisdiction type not be confused with a description of whether the respondent organization is a public, private, or nongovernmental entity, but is simply a reflection of the extent of their work area (e.g., countylevel responses might include nonprofit organizations interested in a county s transportation system, or a private organization that sells orthophotography products for a single county). It is also important to recognize that the 836 responses analyzed within the 26 high-response states are not generalizable to national-level comparisons of Framework activity. These data represent only the 26 states included in this sample; due to the numerous limitations in the overall nationwide sampling design, it is not valid to draw broader conclusions. In other words, the responses discussed here, while representative of the 26 states in the sample, are not necessarily representative of all 50 states; however, they do present a relatively strong cross-section geographically and politically. Notable omissions include the two most populous states (California and Texas) and the states of New England. Readers must be careful not to over-generalize the results discussed here and should recognize the methodological limitations of the study overall. Nevertheless, the findings from the states included are substantive and begin to provide a portrait of Framework activity throughout the different regions of the country. Except where noted, the following analysis represents 836 county-level responses within 26 states (Table 1). The department(s) at the respondents agency currently engaged in GIS work from a list of 20 possible departments was measured: Within your organization, please indicate which departments (or groups) are creating, maintaining, inventoring, distributing, or using digital geographic data. The possible departments include the following: GIS/mapping Land records (assessor, recorder, clerk, etc.) Community Development (planning, zoning, economic development, etc.) Public safety (fire, police, etc.) Transportation Water Wastewater Other public works (excluding water, wastewater, transportation) Engineering (surveying, street lights, etc.) Health and human services Elections Education (school districts) Libraries Administrative and finance Information services (data processing) Environmental Historical preservation and archeology Natural resources Agriculture A total of 78% of respondents stated that they employ a GIS or mapping office to conduct their GIS work (Figure 2), 69% reported utilization of land records departments, 60% community development and planning, 50% engineering, 38% transportation, and 36% public safety. The large number of respondents reporting a GIS department or office is an unexpected finding and suggests that GIS work has become crucial for many county organizations. It is not as clear whether these GIS departments are producing data or using data from other departments to produce map products. In addition, the vast majority of respondents are using multiple agencies for their counties GIS work (Figure 3). It was noted that 61% of respondents reported that four or more departments within the county agency are engaged in Framework GIS work. This reflects an important shift in GIS organizations from single agencies to multiple agencies, supporting the traditional multipurpose cadastre and multipurpose land information system (MPLIS) concepts (McLaughlin 1975, National Research Council 1980, 1983, Brown and Moyer 1989). The concept of the MPLIS focuses on systems consisting of multiple departments or groups constructed with shared (often transactional) data used for a variety of purposes. In the future, therefore, it is likely that GIS work will continue to be shared among multiple departments. This does not tell us, however, the mechanisms by which data are utilized and shared among departments within a single organization. An additional measure of resources comes from the following question regarding staff size: Please estimate the range that best reflects the number of full time employees (equivalents) in URISA Journal Tulloch, Fuld 13

14 GIS/Mapping Land Records Community Development Engineering (Incl. Surveying) Transportation Public Safety Environmental Elections Information Services Water Natural Resources Agriculture Other Public Works Education (School Districts) Wastewater Health and Human Services Figure 2: Percent of Respondents whose Organizations Include Various Departments or Groups that are Creating, Maintaining, Inventoring, Distributing, or Using Digital Geographic Data. Nearly 80% of Respondents (n = 836) Indicated that Their Organization had an Active GIS or Mapping Department Figure 3: The Number of Departments or Groups Actively Using GIS Within Individual Organizations (n = 833). The Mean is 5.7 Departments or Groups per Organization 14 URISA Journal Vol. 13, No. 2 Spring 2001

15 Table 2 Percent of data producing organizations by staff size.* The most common response was that of organizations with staff sizes between one and five No. Full Time Staff Frequency Percentage < % % % % > % * A response of less than one suggests staff members contributing on a part-time basis. A total of 674 responded. your organization currently involved in development, inventory, coordination, integration, or distribution of Framework data. Of the respondents, 62% indicated having between one and five full-time employees (Table 2). Almost 20% of respondents indicated that they had the equivalent of less than one full-time employee. This could mean that the only staff member working on Framework has time split between these GIS-related activities and some other non-framework tasks (which might include GIS work on non-framework-themed data). Overall, these data demonstrate that digital data users may be present in multiple departments within the county organization as well as indicate that more than one employee may be engaged in Framework data work. Framework Data In addition to questions about the types of organization and its practices, the focus of the survey was on specific data types produced by the organizations. The seven Framework themes of geographic data are those produced and used by most organizations. Many organizations use one or more of the seven themes as either a base layer or theme fundamental to the work of their agency. The themes form a critical foundation for the NSDI and have widespread usefulness. In addition, a cooperative approach to producing and sharing these common data themes benefits most organizations that use geographic data (FGDC 1995). Specifically, these themes have been identified as: geodetic control orthoimagery elevation transportation hydrography governmental units cadastral information For the purpose of analyses, the cadastral information theme is being limited to cadastral reference data even though similar questions were asked about cadastral data for public lands and one question was asked about private land ownership. The information about private land ownership is excluded because it is not within the definition of Framework data established by the GFDC (FGDC 1995, 1997). The public lands questions are not included because of the limited relevance in many parts of the country. Respondents were asked whether they produced each of the seven separate Framework themes. For each theme, an additional question was then asked: Do you create, update, integrate or distribute Framework data specific to each of the individual themes. The theme of governmental units was the most highly reported theme activity, with 58% of respondents stating that they collect geospatial data on government boundaries (Figure 4). As expected, a similar finding was reported for transportation data, with 57% of respondents responding that they collect transportation data. Only 28% of respondents reported collecting elevation data, which may reflect a heavy reliance on federally produced elevation data. In addition, 34% reported using digital orthoimagery, perhaps because it can be very costly and is only rarely done in small sets. A separate set of questions inquired about other activities revolving around the individual Framework data themes. For the more common data themes (particularly government boundaries, transportation, cadastral reference), a significant portion of the respondents actively producing thematic data were also revising and updating those data themes. This offers promise for systems in which maintenance has become a planned part of the data production process. Other data themes (particularly elevation and digital orthoimagery), which are seldom produced on an in-house basis, had much lower levels of being revised or updated, but these themes were distributed to others (Figure 4). Figure 5 illustrates the percent of respondent organizations creating zero to seven themes. While a considerable portion of respondents (24%) indicated that they were not producing any Framework data themes, more than 30% indicated that their organization was producing five or more different Framework themes. There is considerable variability in the production of specific themes in county organizations, ranging from zero themes to 8% of organizations creating all seven themes. Respondents who indicated activity in any of the seven Framework themes were also asked whether they maintained metadata for that theme. Metadata is specific information explaining how data have been produced and maintained. It also includes contact information for the person(s) responsible for the data. Overall, a large number of respondents did not maintain metadata for any of the Framework themes in which they produced data (Figure 6). Specific themes were found to be more likely than others to have metadata maintained. For example, URISA Journal Tulloch, Fuld 15

16 Boundaries Transportation Cadastral Reference Hydrography Geodetic Control Digital Orthos Elevation Figure 4: Percent of Respondents Engaged in Three Different Activities (Data Production, Revision, and Distribution) for each Framework Theme. Some Data Themes, Particularly Elevation and Digital Orthoimagery, are Distributed More Commonly Than They Are Updated Figure 5: The Number of Framework Themes Created, Updated, Integrated, or Distributed by Respondent Organizations (n = 836). Fewer than 10% of the Responding Organizations Indicated Activity with All Seven Themes. 16 URISA Journal Vol. 13, No. 2 Spring 2001

17 Boundaries Transportation Cadastral Reference Hydrography Geodetic Control Digital Orthos Elevation Figure 6: The Percent of Organizations Producing Framework Data and Metadata whose Organizations are Active in the Production of Seven Different Themes of Framework Data (dark gray) and Those Who were Actively Maintaining Metadata for Each of the Themes (light gray) Table 3 The percent of respondent organizations sharing data. The vast majority of respondents share data Data Sharing Frequency Percent Yes % No 99 12% N = 836, 30 missing responses only 15% of the respondents who were producing digital orthoimagery were maintaining metadata; in contrast, 30% of the respondents producing geodetic control were maintaining metadata. These findings raise questions about whether certain themes have low metadata development rates because of the difficulty of developing the metadata or because of a lack of need in using the metadata. Data Dissemination Much of the NSDI s work is based on the exchange or sharing of data. To explore the readiness of county-level jurisdictions for sharing data, the survey measured several issues regarding different aspects of exchanging or sharing data. In response to the following question, Does your organization share your data with other organizations? 88% of county-level organizations stated that they shared data (Table 3). This finding serves as a relatively basic measure of an organization s ability and willingness to allow other data users to access their data. However, the wording of this question does not measure the mechanisms behind sharing practices and whether these sharing activities occur with other caveats or conditions attached. Thus, among the population of organizations that are sharing data, there might be restrictions including charging for data access or allowing some agencies access to data while denying access to others. A more specific measure of sharing data needs to be investigated in order to fully undertsand agenicies sharing policies and practices and the impact on creating the NSDI. A specific example of the complexities of data sharing is reflected in the various ways in which organizations institutionalize data-sharing policies. The survey included a measure of redistribution of data, Does your organization permit others to redistribute its data? Only 30% of organizations sharing data allowed the unrestricted redistribution of their data (Table 4), while the majority of organizations (70%) implemented some form of restricted redistribution; 25% indicated that they did URISA Journal Tulloch, Fuld 17

18 Table 4 The percent of respondent organizations redistributing data Redistribution Practices Frequency Percent Yes, the data can be redistributed by others with no restrictions by us % Yes, the data can be redistributed by others but only under conditions set by us % No, the data may not be distributed under any condition % n=836, 58 missing responses Table 5 The percent of respondent organizations participating in coordinating councils Coordinating Council Frequency Percent Yes % No % N=836, 27 missing responses Table 6 The percent of respondent organizations with policies on data dissemination Dissemination Policy Frequency Percent Yes % No % N=836, 28 missing responses not allow any redistribution and another 40% indicated that they allowed redistribution under certain restrictions. While some of these restrictions might simply include a requirement about crediting the original source of the data, others could be strict to the point of preventing a free flow of data. Since this number includes some nongovernmental and private organizations, this is not an unreasonable expectation. Redistribution is going to continue to be a significant issue because data and information are commodities that can be easily reproduced. As participation in NSDI increases, sharing and redistribution issues will become more important. A limitation of the survey data, however, is that there are no details about sharing requirements. A common mechanism for encouraging sharing and more general data development, is the coordinating council. To the question Does your organization participate with a geographic data coordinating council or group? 42% of the respondents indicated participation in such a coordinating group (Table 5). A total of 58% of the respondents indicated that their organization did not participate with a geographic data coordinating council or group. Considering the relatively little cost or effort required to participate in one of these groups, the low response rate raises serious concerns about the underlying causes for the low participation rate. When asked to list the two geographic data coordinating councils or groups with which respondents were most active, a fairly robust list was produced. The open-ended question made it difficult to analyze in detail, but a descriptive analysis indicated that 50% of respondents listed a state-level organization. Examples include the Wisconsin Land Information Board, the New Jersey State Mapping Advisory Committee, and the Colordao Counties GIS/LIS Committee. While a few respondents mentioned national orgnizations (including the U.S. Census Bureau and the Urban and Regional Information Systems Association), the remainder were mostly regional and local organizations. Some of these coordinating councils included formal institutions with some authority over participants, such as MetroGIS or the Cincinnati Area GIS (CAGIS). Others were voluntary organizations (e.g., the Eastern Oregon GIS User Group), which were more loosely bound together professionals working in a variety of settings, including the public and private sector. The predominance of state organizations indicates both the success of such groups and the regional perspective shared by many local data producers. A more systematic method of data sharing can be insured with a specific policy; however, this was found not to be as common as might be expected. While 40% of the survey respondents indicated that they had a policy describing data distribution, while 60% said that they had no policy (Table 6). This suggests that many of the organizations that actively disseminate data are doing so without any formal policy describing with whom to share data, the conditions of the data sharing, and the liability for any problems with the data. Analysis shows that the potential role of data-sharing policy is related to whether or not the organizations engage in sharing data (Table 7). Responses to the survey indicate that having a data-sharing policy (even though some policies prohibit sharing) increases the likelihood of an organization sharing data with other organizations. Specifically, 95% of respondents whose organizations had data-sharing policies indicated that their organization was sharing data. In contrast to those without a policy, only 78% were actively sharing data with other organizations. While this may not seem surprising, it shows how significant the formalized nature of policies can be to ensure specific organizational activities. A total of 88% of organizations shared data, 75% allowed some form of redistribution, 40% had policies regarding data 18 URISA Journal Vol. 13, No. 2 Spring 2001

19 Table 7 A cross-tabulation of respondent organizations sharing data by data-sharing policy Data-sharing Policy Sharing Data with Other Organizations Yes No (or not Applicable) Yes 304 (95%) 403 (78%) No (or not applicable) 16 (5%) 113 (22%) Total 320 (100%) 516 (100%) Chi square = , df = 1, P < Table 8 The percent of respondent organizations that advertise their data in catalogs or clearinghouses Advertise Data Frequency Percent Yes 69 9% No % N=386, 39 missing responses distribution, and only 9% indicated that they advertised their data in catalogs or clearinghouses (Table 8). Advertising data is an active means of sharing data, often requiring significant changes in practices and policy. Such an active approach may not be possible for organizations with limited resources. With the recent growth of both private and public clearinghouse nodes and catalogs, this response may be significantly different than any other if the survey were conducted again today. Another interesting finding concerns the sharing of public data. In order for the FGDC s Framework concept to be realized, many agencies will need to share their geospatial data with other organizations. Fortunately, 88% of the county-level respondents indicated that they do share data (Table 3). However, the ways in which organizations share may indicate that some issues remain to be addressed; 40% of organizations have a policy describing how they should share data and only 9% of county organizations actively advertise their data on catalogues or through clearinghouses. This finding should be considered important to the FGDC with regard to their clearinghouse effort because of the central role that geospatial data clearinghouses are expected to play in the NSDI. Certainly, the low level of activity through clearinghouses might reflect the timeliness of the survey response data (collected largely in the winter of ). With the recent growth of clearinghouses and online data catalogs, it is possible that the percent advertising in clearinghouses would be much higher if the survey was replicated today. Analysis Of particular interest to the FGDC should be the relatively low level of more active and sophisticated forms of data sharing and metadata production reflected in these survey results. It is important to note, however, the timeliness of the survey results, which largely reflect the state of technology in It seems likely that these conditions have changed since the survey s application, particularly in the area of advertising data in clearinghouses. An example of the growth is that there were only 40 registered clearinghouse nodes online in October of 1997 (Nebert 1997) and 88 in September of 1998 (Nebert 1998). A recent check of the FGDC clearinghouse server (at /serverstatus.html) showed 205 registered nodes (FGDC 2000). Still, the survey raises questions about whether respondents can be drawn together as a network of data providers in their current form. The questions about data-sharing practices also serve as reminders of the somewhat unique nature of the approach of the U.S. to building an NSDI. Both intra- and interorganizational data sharing (as broadly discussed throughout Onsrud and Rushton 1995) raise new concerns, but offer opportunities for increased benefits of system development (Azad and Wiggins 1995, Lopez 1996). While some emphasis falls on the benefits described as efficiency and effectiveness (Budic 1994), data sharing also makes possible the benefit labeled equity (Tulloch 1999). Since the FGDC s current model for the NSDI relies heavily on voluntary sharing as the mechanism for populating the Framework with data, it is also reliant on public agencies with open attitudes toward records. Despite open records laws requiring agencies in many parts of the U.S. to make data available at the cost of reproduction, a number of public-sector organizations continue to pursue cost-recovery polices that make assembling larger data infrastructures difficult, if not impossible. Continued emphasis of the aforementioned agency-wide and communitywide benefits of sharing is important if data dissemination practices are going to continue in the manner necessary to support an NSDI. One solution to the problems around data sharing being pursued by the FGDC is the Geo Data Alliance ( The Geo Data Alliance is a self-organizing approach to facilitating a flow of geographic information among individuals, communities and organizations built on the chaordic principles of Dee Hock (Hock 1999). This group is intended to offer a more fluid and less traditional structure for data sharing across many organizations. Still in very preliminary stages, this organization is better equipped than most traditional institutions to deal with the enormous variety that is represented in these survey results. However, this flexibility, which might loosely be compared to some Internet file sharing approaches, might also bring strong resistance from more traditional data production organizations. URISA Journal Tulloch, Fuld 19

20 Another crucial issue that emerged in the survey process is that of the state-level coordination of Framework data organizations. The simple process of conducting the survey presented problems that are emblematic of the difficulties that face NSDI efforts. States have varying political or institutional structures, making a nationwide survey process difficult. Some states had very active counties, while some New England states have effectively eliminated counties as a governing body. Not all states had clear state-level leaders to help distribute the survey instrument, and some states offered a limited level of participation that resulted in very low survey response rates. It is hard to determine whether these patterns are indicative of similar difficulties facing a NSDI approach, but they do serve as potential hazards of which the FGDC should be aware. Certainly, the challenges facing the FGDC are daunting. The complex patterns of data availability and data quality combined with attitudes, policies, and abilities regarding data sharing are going to make difficult any attempt to weave together a basic, coherent network of data producers, much less a comprehensive NSDI. But, these data suggest that data producers are falling into patterns of activities that are conducive to an NSDI. Conclusions As the condition of these organizations continues to change rapidly, the FGDC should take this enormous survey effort and use it as a learning experience. Interest in continuing the survey has emerged in ways that suggest that it would be warmly received (Tulloch 2000); however, this requires some significant improvements over the previous survey development and distribution process. These changes in the process should include a much more rigorous research design to ensure the validity and reliability of collected data. In addition, the sampling design must be improved to capture a wider sample of Framework data users. Finally, the survey should be offered as a digital, on-line survey designed to facilitate a systematic data-collection process that collects longitudinal data and measures changes in the population of data producers. This article presents an overview of the results of the Framework Data Survey. These data demonstrate that a considerable amount of spatial data is being produced, but under conditions that seem less than ideal. Rather than simply presenting a clear picture of how the FGDC should proceed with its Framework initiatives and with the larger NSDI effort, these results raise some important questions. Do the existing pieces of Framework data represent enough of a structure upon which to build an NSDI-like system? Are organizations willing to engage in the sorts of actions - such as data sharing and metadata maintenance - that are necessary for the development of a bottom-up NSDI? Do the less commonly produced Framework data themes represent a barrier to the production of a sufficient framework to support regional or national data infrastructures? Participants in the recent Framework Survey should consider the implications of these findings presented in this article. As this analysis illustrates, the data landscape is a complex mosaic across which the coordination of data creation, maintenance, and sharing is a great challenge. Framework data exist in considerable quantity, but it also exists in many forms, varying accuracy, and with a diverse degree of accessibility. A daunting task faces the FGDC: the challenge of finding ways to bring together these disparate data sources into a coherent single framework. About The Authors David Tulloch is an Assistant Professor in Landscape Architecture at Cook College, Rutgers University. He also serves as Associate Director for Program Development at the Grant F. Walton Center for Remote Sensing and Spatial Analysis. David Tulloch Department of Landscape Architecture 93 Lipman Drive Cook College Rutgers, The State University of New Jersey New Brunswick, NJ dtulloch@crssa.rutgers.edu p f Jennifer Fuld is a doctoral candidate in Sociology at Rutgers, The State University of New Jersey. She also serves as a senior research associate at the Center for Drug Use and HIV Research at the National Development & Research Institutes in New York City. Acknowledgments This work has been made possible primarily through a grant by the Universities Consortium for Geographic Information Science on behalf of the Federal Geographic Data Committee. Additional support has been provided by the New Jersey Agricultural Experiment Station (Hatch No ). While countless individuals were involved in the questionnaire development, survey distribution and compilation of results, the authors would like to thank Steve Gillespie and John Calkins for their compilation and quality control of the survey responses. 20 URISA Journal Vol. 13, No. 2 Spring 2001

21 References Azad, B., and L. L. Wiggins, 1995, Dynamics of Inter-Organizational Geographic Data Sharing: A Conceptual Framework for Research. In H. J. Onsrud, and G. Rushton (Eds.), Sharing Geographic Information (New Brunswick, NJ: Center for Urban Policy Research). Brown, P. M., and D. D. Moyer (Eds.), 1989, Multipurpose Land Information Systems: The Guidebook (Washington, DC: Federal Geodetic Control Committee). Budic, Z., 1993, GIS Use Among Southeastern Local Governments. Journal of the Urban and Regional Information Systems Association, 5 (1), Budic, Z., 1994, Effectiveness of Geographic Information Systems in Local Government Planning. Journal of the American Planning Association, 60 (2), Campbell, H., and I. Masser, 1995, GIS and Organizations: How Effective Are GIS in Practice? (London: Taylor and Francis). Chan, T. O., and I. P. Williamson, 1999, The Different Identities of GIS and GIS Diffusion, International Journal of Geographical Information Science, 13 (3), Clinton, W. J., 1994, Coordinating Geographic Data Acquisition and Access the National Spatial Data Infrastructure. Executive Order Cullis, B. J., 1995, An Exploratory Analysis of Responses to Geographic Information System Adoption on Tri-Service Military Installations (Vicksburg, MS: U.S. Army Engineer Waterways Experiment Station). Federal Geographic Data Committee, 1995, Development of a National Digital Geospatial Data Framework (Washington, DC: Federal Geographic Data Committee). Federal Geographic Data Committee, 1997, Framework Introduction and Guide (Washington, DC: Federal Geographic Data Committee). Federal Geographic Data Committee, 2000, Server status page, ( Frank, S., M. Goodchild, H. J. Onsrud, and J. K. Pinto, 1996, User Requirements for Framework Geospatial Data. Journal of the Urban and Regional Information Systems Association, 7 (2), Hock, D., 1999, Birth of the Chaordic Age (San Francisco, CA: Berrett-Koehler Publishers). Lopez, X. R., 1996, Stimulating GIS Innovation Through the Dissemination of Geographic Information. Journal of the Urban and Regional Information Systems Association, 8 (2), Masser, I., and H. Campbell, 1994, Monitoring the Take-Up of GIS in British Local Government. In M. Salling (Ed.), 1994 URISA Annual Conference Proceedings, Milwaukee, WI, August 1994, 1, pp Masser, I., and H. J. Onsrud (Eds.), 1993, An Introduction. Diffusion and Use of Geographic Information Technologies (Dordrecht: Kluwer Academic Press). Masser, I., and H. Campbell, 1993, The Impact of GIS on Local Government in Great Britain. In P. M. Mather (Ed.), Geographical Information Handling - Research and Applications (New York: John Wiley & Sons). McLaughlin, J., 1975, The Nature, Design, and Development of Multi-Purpose Cadastres, Ph.D. Dissertation, (Madison, WI: University of Wisconsin). National Research Council, 1980, Need for a Multipurpose Cadastre (Washington, DC: National Academy Press). National Research Council, 1983, Procedures and Standards for a Multipurpose Cadastre (Washington, DC: National Academy Press). Nebert, D., 1997, The U.S. National Spatial Data Infrastructure: An Overview. PowerPoint Presentation File ( /powerpoint/NEWNSDI.PPT). Nebert, D., 1998, Clearinghouse Status and Architecture. PowerPoint Presentation File ( Nedovic-Budic, Z., and J. K. Pinto, 1999, Understanding Interorganizational GIS Activities: A Conceptual Framework. Journal of the Urban and Regional Information Systems Association, 11 (1), Onsrud, H. J., and J. K. Pinto, 1993, Evaluating Correlates of GIS Adoption Success and the Decision Process of GIS Acquisition. Journal of the Urban and Regional Information Systems Association, 5 (1), Sprecher, M. H., 1994, ICMA Survey Dissects Local Government IT Use. Journal of the Urban and Regional Information Systems Association, 6 (2), Tulloch, D. L., 1999, Theoretical Model of Multipurpose Land Information Systems Development. Transactions in Geographic Information Systems, 3 (3), Tulloch, D. L., 2000, Looking Back and Looking Ahead: The State of Framework Data. In M. J. Salling (Ed.), URISA 2000 Conference Proceedings (CD-ROM), pp Tulloch, D. L., and M. Robinson, 2000, A Progress Report on a US National Survey of Geospatial Framework Data. Journal of Government Information Systems, 27, URISA Journal Tulloch, Fuld 21

22

23 Linking Transit Operational Data to Road Network with a Transportation Object-Oriented GIS Martin Trépanier and Robert Chapleau Abstract: Geographic information system (GIS) software is widely available to urban planners; however, it seems that traditional transportation models do not adapt well to these tools. The purpose of this article is to demonstrate the benefits of a comprehensive linking of transit operational data to a road network within transportation object-oriented GIS. Based on the totally disaggregate approach and object-oriented modeling, the proposed method has been tested through many applications in the Montreal area: user information, bus interlining, origin-destination surveys, and infrastructure management. With interactive planning, the system gathers data from each application and can transfer the knowledge from one to another. The results of the challenges of bus interlining are described. They imply an appropriate merge of road and transit networks within the GIS. Primary conclusions show that the system can also be developed to integrate real-time data gathered by global positioning system technologies. Introduction Geographic information systems (GISs) are currently used by many urban public transportation authorities (Zhong-Ren et al. 1998). The systems offer quick storing, displaying, manipulating, and analyzing of spatial data. Planners and operators often build their system from sources of commercial and government data (Nedovic-Budic 2000). Unfortunately, these data blocks are not always suitable to specific transportation applications such as transit trip calculation or on-street vehicle-path modeling. Therefore, data must be edited and adapted. Moreover, other sources of data can be easily gathered. Operational data on bus stops, bus routes, and schedules can contribute to the development of GIS and vice versa. In 1996, Sutton noted that transportation planning based on modeling is hardly adaptable to GISs, which are datadriven. Through the years, many efforts have been made to adapt GIS to transportation modeling, beginning with the first efforts by Vonderohe et al. (1993) to the latest developments by Dueker and Butler (2000), who proposed an enterprise GIS-T (GIS-transportation) data model and addressed transportation data sharing. The purpose of this article is to demonstrate the benefits of a comprehensive linking of transit operational data to a road network within transportation object-oriented GIS (OO- GIS). Transit authorities need such function, especially for global positioning system (GPS) data integration. The GIS developed in this research relies on several years of experience involving an origin-destination survey in the Montreal area using a totally disaggregate approach (Chapleau 1986). The approach has been used since 1995 for transit path calculation software at the Montreal Urban Community Transit Corporation (MUCTC) Information Centre (Chapleau et al. 1996). In addition, telephone interviewers of a 1998 origindestination survey in the Montreal area used specially packaged software supported by the same OO-GIS. Object-oriented modeling is a natural way to address common problems. A vast majority of computer programmers use object-oriented languages to construct programs. The method is also used for noncomputed modeling (Rumbaugh et al. 1991). At this time, the use of an object-oriented design in transportation is rare and seems to be limited to software design and to simulation models (Tattegrain-Veste and Bellet 1999, Rafanelli 1998). The difficult data-model integration in GIS and other programming contexts are possible causes of this situation. The article presents some essential concepts of object-oriented integration in GIS systems. It describes road network and transit network objects as part of the urban object-oriented model. It then elaborates on some experiments conducted in the Montreal area to completely merge the transit operational data with existing road network information. This is done to improve the use of GIS in many applications, such as mapping, bus interlining, user information, origin-destination survey, and infrastructure management. Research Objectives This research project is part of a continuous effort in the Montreal area to improve transportation demand modeling and data integration by using a totally disaggregate approach to replace other traditional methods. It originated from the use of a family of software modules called MADITUC (Modèle d Analyse Désagrégée des Itinéraires de Transport Urbain Collectif or Model for the Disaggregate Analysis of Urban Transit Itineraries). The use of the totally disaggregate approach has resulted in many changes in transit planners habits in the region (Chapleau 1991, Chapleau et al. 1987), including: URISA Journal Tulloch, Fuld 23

24 A no analysis zone has been used for transit modeling at MUCTC since Every piece of information is coded at an X-Y coordinate level. The large origin-destination surveys in the Montreal area (more than 50,000 households in 1978, 1982, 1987, 1993, and 1998) generated trip files in which every trip is maintained on an individual level, with information on household, person, mode, purpose, transit routes, and other derived variables. The level of resolution of transportation data has been considerably improved over the years. Smaller scales are used in space, time, and sociodemographic attributes such as age, gender, and revenue. Data on territory, demand, network, and operation are adapted to different roles: planning, user information, the Internet, cartography, market analysis, network evaluation, and other transportation tasks. With this research project, the GIS uses the totally disaggregate approach to migrate toward object-oriented modeling, first by integrating actual data and second by linking pieces of data to generate new information to add to the existing system. Informodels are created from the merge of fresh data and established models. Object-oriented GIS Although currently available commercial GIS software offers several possibilities and functions, it is our view that they are not as adaptable to operational transportation problems as object-oriented GIS (Chapleau et al. 1998). The Dueker and Butler (2000) enterprise GIS-T data model is based on an object-oriented approach and addresses several issues related to transportation data sharing such as identification, spatial accuracy, database completeness, network definition, and interoperability. When the GIS is based on transportation data coming from georeferenced origindestination surveys, it must also integrate daily operational data and consider dynamic objects, which are the transportation actors, in the heart of the data model (Trépanier and Chapleau 2001). This permits a closer look at transportation infrastructure to obtain information such as the categories of people who use a subway line between two given stations or the origins of people crossing a bridge. An OO-GIS relies on two major bases: the object definitions of each element of the system, and comprehensive data integration reflecting the object model. Transportation Object-oriented Definitions Transportation object-oriented modeling (TOOM) is based on existing object-oriented methods described by Rumbaugh, Booch, Coad, Colbert, and Yourdon. These methods define the object as a unique entity based on a given template, the class. Classes give properties (or attributes) to objects to represent the object s state and methods (or services) and to represent its behavior (interactions with other objects). Therefore, TOOM identifies transportation objects in addition to their properties and methods. Figure 1: Transportation Object Metaclasses Transportation objects are special components intended to facilitate the modeling, the observation, the planning, and the analysis of the transportation system. These objects are from four metaclasses (Trépanier and Chapleau 2001) (Figure 1): Immobile (static) objects have a fixed location in time and space. Their role is to describe the territory and to serve as transportation movement beacons. Some examples include the TripGenerator, PostalCode, CensusTract, and Zone objects. Dynamic objects are the transportation actors. These objects decide and contribute to their movements. They represent a group of persons (household, person objects), a moving object (bus, car), or even moving objects (goods). Kinetic objects are the movement describers. Some examples include trip, transitlink (simple kinetics), path, and transitroute (composed kinetics) objects. System objects are a group of embedded objects with their set of relations. They can be operational (transitnetwork, roadnetwork), informational (survey, census), or globally comprehensive (urban system object). The main use of TOOM is not a software design or database structure; the modeling is not a database issue. At first, it is a way of thinking about the role and the specific use of every piece of information in the system. With adequate object diagrams, objects can be rapidly identified, along with their properties and methods engaged in the analysis. Comprehensive Data Integration An OO-GIS requires having the elements correctly defined in relational context (what, where, when, and why). In addition, data must be georeferenced, classified, and related. For example, a transit route is bound in space (geometric sequence of stops) and time (service date interval and schedules). In an object-oriented approach, the way the information is displayed in GIS soft- 24 URISA Journal Vol. 13, No. 2 Spring 2001

25 Figure 2: Overview of the Montreal OO-GIS Figure 3: Road Network Object Model ware is independent from the data: it could be on a single layer or multiple layers. The GIS display feature is a method like any other like path calculation or network resources analysis. Over several years, objects have been added to the GIS in the Montreal area, including: street objects (to go beyond the link/node/arc definition), civic addresses and intersection objects (for user information and geocoding), postal code objects (representing each blockface in Canada), trip-generator objects (as major places for origin-destination survey processing and other analysis), and territorial definitions (districts, municipalities, regions, census divisions, and postal divisions). The system is now mature enough to accept more complex information such as transit operational data, automobile data (individual car ownership), scholar registration data (to feed origindestination surveys), and other low-level characterized information. Figure 2 is an example of actual data integration for the Montreal OO-GIS used at the MUCTC. Road Network Object System Current research uses two major components of the OO-GIS: the road network object and the transit network object. Definition In TOOM, the road network is the set of kinetic objects on which dynamic objects such as cars, trucks, buses, taxis, and pedestrians can circulate. In the urban example, freeways, streets, and some pedestrian alleys define the road network. It has been completely generated from OO-GIS data. Object Model The object model is a diagram that presents the numerous objects of a given system and shows the relation between these objects. In this article, the object models are shown in a special TOOM notation based on common OMT and Unified notations (Rumbaugh et al. 1995). The road network object is a collection of super-street objects (see dark shapes in Figure 3). A super-street is a discontinuous sequence of likely named street objects crossing several municipalities. Each street object is an oriented collection of street arcs, composed of link and node objects. The other objects (white shapes) accompany the road network. The car, transit operational, and walk network system objects are collections of street objects, depending on their hierarchy. Path and tree objects (in path calculation) are collections of street sections. Municipality, block, and postal codes static objects are all related to the road network. Road Network Properties The geometric representation of the road network relies on its modeling usage. For transportation macroscopic modeling purposes, street centerlines are sufficient at some conditions. The connectivity of the network must be assured at each intersection, and no movements are permitted at grade crossings. This overrides some problems related to traditional geographic data files, in which every arc intersection is represented by a node (Sutton 1996). The streets are defined as whole objects in the GIS system, not just a sequence of arcs (like in traditional GIS). The system maintains the information on each section, but considers the entire street as a single object. Finally, the road network responds to a service hierarchy, defined for the multiple purposes. Table 1 enumerates the street subclasses along with their use in different system objects. The hierarchy permits separate operational, road, and pedestrian networks to impose preferences to some street classes and to generate a multilayered GIS representation. URISA Journal Trépanier, Chapleau 25

26 Table 1: Road network hierarchy (ONO = operational network object, RNO = road network object, PNO = pedestrian network object) Transit Network Object System Definition The transit network is a spatialized and temporized set of kinetic objects that describes the trip circumstances of transit users. One or more operators provide the transit service. A transit network can regroup several transportation modes (e.g., bus, tramway, subway, trolleys, train, and taxis). Object Model The transit network object model representation depends on the view from which it is observed (Figure 4). The transit user (and transit planners) would see the route, trace (direction), and route stops. On the other side, the operator has to offer the service with a set of vehicles and drivers assigned to its garages. Journey objects are defined to package a driver s day and are comprised of runs (productive route sections) and deadhead trips (with no passengers). The system object ( real-world representation) is presented in Figure 5. The transit network and the road network objects are linked at bus stops (along the street objects). The bus-stop object is also the interface between the transit network and the user, which circulates on the pedestrian network before boarding. Level of Resolution In transit network modeling, the level of resolution can vary depending on the desired model or application (Figure 6). At the timepoint level (A), buses are required to arrive at a specific location at a given time. The transit nodes (B) are used for network modeling because they consider all transfer possibilities between routes. The bus-stop level (C) permits user information, bus shelter maintenance, and bus-stop sign management. Finally, the complete onroad bus path (D) ( gauche-droite or right-left ) is defined in some specific applications, like cartography and interlining. Figure 4: Transit Network Object Model Figure 5: Transit Operation System Object Figure 6: The Level of Resolution in Transit Route Definition 26 URISA Journal Vol. 13, No. 2 Spring 2001

27 Pick-Out Sheet (Operational Time Table) The pick-out sheet ( table des temps de marche ) contains the vehicle (and drivers) assignment to the transit network. For a given route, it contains the timepoint definition and series of hours that indicate the scheduled bus arrivals at these points. First, the table is generated for driver assignment (for a given level of service on routes) and then the route schedules are derived from the timepoint attributes. Methodology and Experiments Some experiments have been conducted to verify the interest of a linking an operational timetable to road network and vice versa. Road and Transit Network Updates The update of the road networks at the MUCTC relies on five major activities. The transit network management is also based on these applications. Transit network planning from origin-destination surveys with the totally disaggregate approach, which necessitate a sufficient cartographic base for transit nodes and routes coding (Chapleau 1992); Bus-stop management is based on the same GIS, including an intensive inventory of every bus-stop site (16,000 entries); Origin-destination surveys data geocoded at a disaggregate level need an up-to-date street network with identification of civic addresses and intersections (42,000 intersection possibilities for Montreal island only); User information provided at the MUCTC Information Centre uses a GIS defined at a pedestrian network level, and updates are provided daily by users who inform the agent about new streets and addresses (3500 daily calls) (Chapleau et al. 1996); Interlining between bus routes is defined on a special road network, where MUCTC-owned roads are added to correctly model bus link-to-link shortest paths between two route timepoints (36,000 interlining paths) (Chapleau et al. 1998). With these applications feeding in, the OO-GIS is continuously updated. The system is currently maintained on personal computer technology in a blend of AutoCAD, FoxPro, Visual Basic and commercial GIS software. Linking Road and Transit Network The linkage of the road and network objects uses a combination of cold, warm, and hot links as defined by Sutton (1996): Cold linkage requires the manual hard coding of spatial elements to be used in the bus-stop inventory system to locate them in the street system. This has been done over many years at the MUCTC, so it does not require extensive updating. Figure 7: Transit network object over road network object Figure 8: Transit network object melted into road network object Warm linkage implies that the model network definition is entirely within the GIS. In this project, it has been used to create operational road network objects from GIS topology starting with a set of existing streets. Hot linkage is done by automating the exchange of data between a network model and the GIS. It has been used to determine the part of this operational road network used by transit vehicle. Figure 7 shows the part of a transit network covering a municipality in the Montreal area. Because the routes are not yet linked to the road network, the path between timepoints appears as straight lines. Timepoints are roughly spatialized for representation purpose. This critical step is used to determine the correct path of the buses in operation. The bus paths could be coded manually, but there are more than 600 timepoints and 250 routes for the URISA Journal Trépanier, Chapleau 27

28 MUCTC, for a total of about 30,000 paths. Instead, bus routes and paths are calculated between each scheduled bus-stop pair and then compared to actual bus routes. The path calculation model, a link-to-link shortest path with street-transfer penalties, is calibrated to achieve the best reproduction of the real bus path without necessitating too many adjustments. A link-to-link calculation is necessary because there can be numerous bus positions at start and arrival points. The operational road network is used for the calculation. It includes public and private streets but excludes pedestrian overpasses. Freeways and ramps are not usually used for the calculation of local route paths, but are kept for express routes. Figure 8 shows the result of the melting of the bus network, immediately showing which streets are used by buses in operation. The complete method can also be used for deadhead bus paths. In a recent interlining project at the MUCTC, bus paths were calculated to estimate the driving times between all network timepoints including depots (more than 25,000 paths). GIS-Aided Interactive Planning A completely automated calculation of a deadhead bus path on a street network is hardly achievable because several aspects cannot be synthesized in mathematical or algorithm operations without a database support in the background. For example, some street segments can only be used by some types of vehicles or only at specific times. In addition, a bus operator may prefer to use a different path between two points (to avoid noise or other disturbances). To ensure the best reproduction of the deadhead bus paths, the operator is asked to provide some feedback on precalculated paths; this is called interactive planning. Each correction made to a specific path allows the OO-GIS to learn the new parameters and intervene on future and previous paths in the system. Figure 9 shows software used at the MUCTC Operation Centre in Montreal. The MADINTER software (Module D acquisition de Données pour les Interlignes) architecture is based on the OO-GIS approach. While origin-destination surveys in a totally disaggregate approach can benefit from respondent declaration (especially on trip-generator data), interactive planning instructs the operator Figure 9: An OO-GIS Tool for Bus Path Interactive Planning Table 2 Exchangeable properties between road and transit objects Figure 10: Transit route superposition on a street link to tell the system what choices must be made for path calculation. As mentioned earlier, the OO-GIS generated a map and a left-right sequence of streets for each interlining path. In addition, it is not necessary that the system be complete at the start, reducing the data-development costs brought by the traditional GIS. Data can be coded only as needed. Transit Link Properties Transfer The association of transit and road network objects (for busroute operation or interlining purposes) permits some exchange of properties between the two systems. Table 2 notes some possibilities. Property transfer is not a simple task, and some of the challenges can be identified. The superposition of two or more transit links on a single road link (Figure 10) must be examined with care because of the multiple sources of information that are involved. For some reason, certain links have different travel times depending on the garage operator and must be reconciled. The operational speed of a bus on a route varies with time, passenger load, and other external factors. Moreover, speed does not usually remain constant between timepoints. The planning of a fixed schedule transit network (at the one used for this experimentation) relies on long experience. Some routes have existed for 20 years or more, and the timepoints have been updated and refined several times. With adequate assumptions, the bus-speed variation during a day for a given street link (with or without passengers) could depend on the bus congestion on that link. In addition, several statistics can be derived from the exchange of property. 28 URISA Journal Vol. 13, No. 2 Spring 2001

29 Figure 11: Load profile of the usage of the street network by deadhead bus paths Multiplicity of OO-GIS Usage The linkage of transit and road networks leads to the development of several applications: Transit usage impacts on road infrastructures. The knowledge of road usage by buses is useful for geometric design, pavement refection, lighting, synchronization of traffic lights, etc. Figure 11 is an example of a street-load profile generated by bus deadhead paths. The operator can visualize the impact of bus passages and utilize the database to know which buses run on a given street. Interlining and service resources studies. A good knowledge of travel times on street network can lead to a better resource usage. The OO-GIS is used to calculate new paths before the operation of a route begins. Impact of a road closure. The calculator, which is part of the OO-GIS, is used to determine side effects of a specific street closure over the network. Market analysis. The traditional features of the GIS (band analysis, layers overlay) are added to road and network objects to more precisely analyze census and origin-destination survey data related to a specific route. Feeding other systems. One major task of the OO-GIS is to feed other corporate systems with data. For example, the planner uses the GIS to generate the data that he or she needs for demand forecasting. Conclusion The urban planning context brings complexity, especially for transit planners who must deal with territorial and operational data. Merging a bus network definition to a road network object brings a comprehensive representation of a transit operational network and, with the help of an object-oriented GIS, leads to new ways to explore the close interrelations between bus operation and traffic and road systems in an urban context. The totally disaggregate approach has helped defined the context and provided basic data and methods for the system. Object-oriented modeling helped in software programming and data structure and is still reflected in methods usage. Every user can benefit from such a system. With an OO- GIS, the operator obtains more precise information regarding street usage of its vehicle and can intervene more rapidly if a problem occurs on the road. The planner examines the present and future impact of bus service on neighborhoods and can directly feed the models with GIS data. Finally, the transit user obtains detailed information at bus-stop levels. With OO-GIS interactive planning, all of these results can be achieved at a fraction of the efforts required by traditional GIS. Some issues remain to be examined about transportation data integration in GIS. The integration of data gathered using new technologies such as an on-board GPS or a Smart Card for users is now a key issue, but it seems that the tools developed in this research project could assist. In addition to the spatial components of transportation problems, the management of temporal data (e.g., schedules and network morphology over time) within GIS is also examined. About the Authors Martin Trépanier, P. is assistant professor of Industrial Engineering at Ecole Polytechnique de Montreal (engineering school affiliated to University of Montreal). He collaborates with the MADITUC research group. His main works are related to object-oriented modeling in civil and industrial applications (enterprise information systems, transportation and hydraulics), GIS development, database systems development, and Internet applications. mtrepanier@polymtl.ca Corresponding Address: Martin Trépanier École Polytechnique, département MAGI P.O. Box 6079, station Centre-Ville Montréal, Qc, Canada H3C 3A7 phone: (514) #4911 fax: (514) Robert Chapleau, P. is professor of Civil Engineering and the founder-director of MADITUC group, Ecole Polytechnique de Montreal (engineering school affiliated to University of Montreal). He developed the totally dissagregated approach in transportation, and participates in several research projects in the Greater Montreal area. His works include travel behavior research, urban mobility, transit planning, and GIS development. rchapleau@polymtl.ca URISA Journal Trépanier, Chapleau 29

30 References Affum, J. K., and M. A. P. Taylor, 1998, Integration of Geographic Information Systems and Models for Transport Planning, Land Use and Environmental Analysis. Proceedings of the 1st Asia Pacific Conference on Transportation and the Environment, 2, Chapleau, R., 1992, La Modélisation de la Demande de Transport Urbain avec une Approche Totalement Désagrégée. Selected Proceddings of The World Conference on Transportation Research, WCTR Society, Lyon, II, Chapleau, R., 1991, La Planification et l Analyse des Systèmes de Transport Urbain: un Bilan des Méthodes et Modèles Disponibles avec L approche Désagrégée. Les Cahiers Scientifiques du Transport, Éditions Paradigme, No. 24, Chapleau, R., 1986, Transit Network Analysis and Evaluation with a Totally Disaggregate Approach. Centre de Recherches sur les Transports, Montréal, No Chapleau, R., M. Trépanier, and B. Allard, 1998, Practical Implementations of Object-oriented GIS-T. World Conference on Transportation Research Proceedings, Antwerpen, Belgium (unpublished). Chapleau, R., B. Allard, and M. Trépanier, 1996, Transit Path Calculation Supported by a Special GIS-Transit Information System. Transportation Research Record No, 1521, Washington, DC, p. 98. Chapleau, R., P. Lavigueur, and K. G. Baass, 1987, A Posteriori Impact Analysis of a Subway Extension in Montréal. Transportation Research Record, Washington, DC, No. 1152, Crowson, J. L., D. E. Leasure, R. W. Smith, and F. P. Worthen, 1997, A GIS for Public Transit. ESRI International User Conference Proceedings, San Diego, Dueker, K. J., and J.A. Butler, 2000, A Geographic Information System Framework for Transportation Data Sharing. Transportation Research Part C, 8, Miles, S. B., and C. L. Ho, 1999, Applications and Issues of GIS as Tools for Civil Engineering Modeling. Journal of Computing in Civil Engineering, July, Nedovic-Budic, Z., 2000, Geographic Information Science Implications for Urban and Regional Planning. Journal of the Urban and Regional Information Systems Association, 12 (2), Nielsen, O. A., T. Israelsen, and E. R. Nielsen, 1998, Handling Traffic Modelling Network in GIS - Conflicts, Solutions and Applications. 8th World Conference on Transportation Research Proceedings, Antwerpen, Belgium (unpublished). Rafanelli, M., 1998, A Graphical Interface to Define and Store Data on Transportation using an Object-Oriented Geographic Information System. 8th World Conference on Transportation Research Proceedings, Antwerpen, Belgium (unpublished). Rumbaugh, J., M. Blaha, W. Premerlani, F. Eddy, and W. Lorensen, 1991, Object-Oriented Modeling and Design (Prentice-Hall), 500 pp. Sargeant, P., 1999, OO-GIS: The Future of GIS Database Management. GeoWorld, February, Slavin, H., 1996, An Integrated, Dynamic Approach to Travel Demand Forecasting. Transportation, 23, Smith, B. L., 2000, Using Geographic Information Systems and the World Wide Web for Interactive Transit-Trip Itinerary Planning. Journal of Public Transportation, Center for Urban Transportation Research, 3 (2), Sutton, J. C., 1996, Role of Geographic Information System in Regional Transportation Planning. Transportation Research Record No. 1518, Transportation Research Board, Washington, DC, pp Tattegrain-Veste, H., and T. Bellet, 1999, A Framework for Representing Driving Knowledge. International Journal of Cognitive Ergonomics, 3 (1), Trépanier, M., and R. Chapleau, 2001, Analyse Orientée-Objet des Données D enquêtes Origine-Destination. Canadian Journal of Civil Engineering, 28 (1), Trépanier, M., and R. Chapleau, 1997, SIG-TOO: Applications (Pragmatiques) sur des Technologies Légères. 32e Congrès de l Association Québécoise du Transport et des Routes Proceedings, Trois-Rivières, 32(1), U.S. Department of Transportation, 1994, Glossary of Transportation Terms. Washington, DC, 333 pp. Vonderohe, A. P., L. Travis, R. L. Smith, and V. Tsai, 1993, Adaptation of Geographic information Systems for Transportation, National Cooperative Highway Research Programs (NCHRP) Report 359, Washington, DC, 69 pp. Zhong-Ren, P., J. N. Groff, and K. J. Dueker, 1998, An Enterprise GIS Database Design for Agency-Wide Transit Applications. Journal of the Urban and Regional Information Systems Association, 10 (2). 30 URISA Journal Vol. 13, No. 2 Spring 2001

31 COMPARATIVE BOOK REVIEW Statistical Analysis with ArcView GIS J. Lee, and D. W. S. Wong. (New York, NY: John Wiley & Sons Inc.), pages. A Casebook of Spatial Statistical Data Analysis: a Compilation of Analyses of Different Thematic Data Sets. D. A. Giffith, and L. J. Layne. (New York, NY: Oxford University Press Inc.), pages. Spatial Analysis, GIS and Remote Sensing Applications in the Health Sciences Edited by: D. P. Albert, W. M. Gesler, and B. Levergood. (Chelsea, MI: Ann Arbor Press), pages. These three books have been chosen for a comparative review because they treat spatial analysis at different depths and they range from a generic approach to specific applications. Although each book is aimed at a specific market segment, it is likely that readers may wish to range across the texts, either to seek further depth from an introductory stage or to apply spatial data analysis in their professional area. They are, however, quite different books. Statistical Analysis with ArcView GIS by Lee and Wong is a very practical introductory book. It aims to bring spatial statistical analysis into GIS both from a theoretical standpoint and through practical exercises the authors have implemented on the popular ArcView software from ESRI. This is a major strength of this book. The implementation is through Avenue scripts, although, as this reviewer found, they do require version 3.2 to run. Rather than include a CD-ROM with the book, the scripts and sample data must be downloaded from the publisher s Web site. While no assumptions are made regarding the statistical background of the reader, a familiarity with ArcView is necessary for the practical exercises. The book is organized into separate chapters dealing with relevant techniques for attribute, point, line, and area data. The chapter on attribute data takes the traditional nonspatial approach in univariate and bivariate analysis of data fields within a table. After introducing scales of measurement, the authors describe measures of a central tendency, dispersion, relationship (correlation), and trend (regression). Worked examples are provided throughout, with panels giving instructions on how to achieve them in ArcView. Point features are covered in two chapters, the first looking at descriptors (e.g., spatial mean and standard distance) and the second dealing with pattern detectors. The latter includes quadrant analysis, nearest neighbor analysis, and spatial autocorrelation (Geary s G and Moran s I); a comparative study using all of these for environmental data is provided. An odd omission in the section on quadrat analysis is the Index of Cluster Size which, although easy to calculate, is not considered. The chapter on line descriptors include topics on connectivity, directional statistics, and network analysis. The final chapter on area features focuses exclusively on spatial autocorrelation as descriptors of pattern: joint count statistics, Moran and Geary indices, general G-statistic, local indicators of spatial association (LISA), and the Moran scatterplot. Overall, the book is very a clear and readable introduction, although it is constrained by what is possible using a shapefile-based GIS. The overall quality of the production is high, although some figures (e.g., 3.10 and 5.13) have degraded. A Casebook of Spatial Statistical Data Analysis: A Compilation of Analyses of Different Thematic Data Sets by Griffith and URISA Journal Book Review 31

32 Layne is a very technical book which, compared with Lee and Wong, assumes that the reader has a substantial knowledge of the techniques and literature on spatial autocorrelation, geostatistics, and conditional autoregressive (CAR) models. This is not intended to be an introductory text. The primary objective is to motivate and illustrate sound analyses of georeferenced data sets exhibiting spatial dependence. The book is laid out in three parts: Theoretical Background, Georeferenced Data Set Case Studies, and Visualizing What Is Not Observed. The Theoretical Background covers the main modeling assumptions in conventional statistics and their relative influences. These are in order of importance: constant variance, absence of spatial autocorrelation, and normality. Violation of one or all of these assumptions is inevitable with spatial data and can result in a 10 to 20% overestimate of the R 2 in a simple regression model. The workings of popular spatial autoregressive and geostatistical models are given in detail with SAS and SPSS codes for their implementation. The second part applies these techniques to a range of data types including socioeconomic variables, natural resources data, agricultural yields, pollution, and epidemiological variables. These worked examples draw on published data sets that, for the most part, are not reproduced in the book. The reader is therefore not in a position to follow these through using the SAS and SPSS codes without considerable research. I view this as a weakness detracting from what is otherwise a very welcome book to the literature. Part Three is by far the shortest and concerns itself with spatial interpolation and the estimation of missing data values. In the Conclusion, the authors set out their views on proper spatial statistical analysis as a series of steps that all spatial scientists should be taking with their data. Overall, this book is a challenging read of a subject that is important to those wishing to analyze spatial data. It is well illustrated throughout, although many of the plots suffered in legibility due to overreduction in size. Spatial Analysis, GIS and Remote Sensing Applications in the Health Sciences by Albert et al. have brought together a structured set of papers around the spatial sciences applied to medical geography and health sciences. Included are nine chapters, a master bibliography resource guide, a glossary, and both subject and geographical indexes. The Introduction sets out the structure of the text and provides a cautionary note, worth re-iterating here as it applies widely to GIS: know your subject first, because without such in-depth knowledge, it is difficult to develop GIS applications of any sophistication. Chapters 2 to 6 focus on the use of GIS and spatial analysis. In Chapter 2, Gesler and Albert provide a tour of how spatial analysis has been used in medical geography structured around point, line, area, and surface patterns with discussion on theory building and the appropriate role of technique. Albert, Gesler, and Wittie in Chapter 3 and Albert, Gesler, and Horner in Chapter 4 consider the application of GIS to medical geography and health services research, respectively. Both chapters show the enormous potential of GIS for research in these areas but that it is still at an early stage of diffusion. In Chapter 5, Fellers moves to the problems and pitfalls of GIS-aided investigations of environment-health relationships. This chapter provides plenty discussion of such issues as data quality, confidentiality, aggregation bias, principles of valid ecological study design, risk evaluation, and mapping contamination zones. It provides a summary of known spatial relationships between exposure and health outcomes for pesticides, lead, toxic wastes, and radiation and finishes with a critical look at modeling issues. In Chapter 6, Albert examines infectious diseases and GIS with a review of the literature on dracunculiasis, Lyme disease, human babesiosis, LaCrosse encephalitis, and malaria. Chapters 7 and 8, both by Messina and Crews-Meyer, turn their attention to the use of remote sensing: the historical development of systems, resolution and trade-offs, and the application of imagery analysis in environmental studies of disease vectors. These two chapters have a series of plates, many in color. The Master GIS/RS Bibliography Resource Guide is organized predominantly by disease type and environmental influences. It has a cut-off date of 1 st January It does, however, usefully complement the Bibliography compiled by Rushton et al. (2000) which is organized by GIS function. Overall, the book is readable and well printed. One criticism is that most of the chapters appear dated, with references only up to Although the time from manuscript to print is sometimes long, the chapters could have benefited from updating before finally going to print. Also, I am personally not in favor of the term geotechniques used by the editors, as I feel that it is too close to the well-established engineering subject of geotechnics. Reference Rushton, G., G. Elmes, and R. McMaster, 2000, Considerations for Improving Geographical Information Systems Research in Public Health. URISA Journal 12 (2), Reviewed by: Allan Brimicombe, BA (Hons), MPhil, Ph.D. Professor and Head, Centre for Geo-Information Studies University of East London London, United Kingdom. a.j.brimicombe@uel.ac.uk 32 URISA Journal Vol. 13, No. 2 Spring 2001

33 BOOK REVIEW Remote Sensing and Urban Analysis Edited by: Jean-Paul Donnay, Mike J. Barnsley, and Paul A. Longley (GISDATA 9, Series Editors: Ian Masser and François Salgé) (London, UK: Taylor and Francis), pages. The use of remote sensing in urban analysis presents multiple problems when compared with image classification in natural areas. The size of the pixel, the mixture of different land covers, and the irregularity of shapes ensure the difficulty of per-pixel classification methods. This explains the difficulty of defining land cover and use inside the urban boundary and justifies some of the lack of interaction between professionals in urban planning and analysis and those in remote sensing. In light of this gap, Remote Sensing and Urban Analysis is a welcome addition to the scarce literature on urban analysis and remote sensing. The publication of this book was motivated by the recent wave of high-resolution satellites (<5 meters) and the assumption that there is a need for more interaction between professionals involved in the urban realm. The book has its roots in a specialist meeting on remote sensing and urban analysis held at the University of Strasbourg, France and sponsored by the European Science Foundation s (ESF) GISDATA Program. The intent of the book is to unveil what modern digital remote sensing can offer both urban planners and society in general. The method is to explore the potential of the technology in general, and remote sensing in particular. Remote Sensing and Urban Analysis is an assemblage of superbly written chapters on theory and practice. It is a resource for remote sensing, urbanism, geography, and planning, and includes contributions from many prominent experts on remote sensing, urban analysis, and GIS. It features themes that range from the continuous need to advance image processing, to go beyond the simple mapping of the physical form of urban areas, to conclude the importance of defining new models and spatial analysis tools, and to the improvement of the communication between the different specialists. The authors are effective at catching our attention with the broad idea, exploring the multiple implications, and then detailing their contributions in a coherent manner. The book is a reflective and well-written piece of work that is needed and that highlights the new opportunities available for students and professionals, as a doubt taker for those working on a daily basis and an important landmark in the advent of this third generation of very high spatial resolution of satellite sensors. The book is divided into four parts, and the topics intertwine without losing the flow and the attention of the reader. The authors seem to speak with each other throughout the chapters, sharing problems, acknowledgments, and doubts, as well as making this a comprehensive source. Part 1 begins with an Introduction by Donnay, Barnsley, and Longley that details the major issues to be discussed. The state of the art in urban remote sensing, the demand for higher spatial resolution data, the enhancement of the classification process, and the importance of going from urban morphology/physical dimension to urban/human dimension are some of the subjects presented. Ranchin et al. (Chapter 2) explored the fact that even though a new generation of sensors is available, other methods are available for merging old and new datasets. He explores a method based on the wavelet transform. Brvio et al. (Chapter 3) uses geostatistical techniques of fitting parameters resulting from semivarigrams to the data sets to distinguish different urban en- URISA Journal Book Review 33

34 vironments. Bianchin (Chapter 4) explores the use of mathematical morphology operators to distinguish urban areas and their form. Part 2 devotes considerable attention to the use of traditional per-pixel classification based in spectral signature and important new techniques such as image segmentation (broadly accepted by the authors as one of best techniques, but still not widely used). It begins with a chapter by Mesev (Chapter 5) that reviews the problems associated with the more traditional standard techniques of statistical classification and a number of practical ways to overcome those problems. Bahr (Chapter 6) explores the importance of segmentation techniques. Barnsley (Chapter 7) takes it further, detailing the techniques of structural pattern recognition. It flows from the readings that the identification of the morphological properties of urban areas is hardly identifiable through spectral signature and per-pixel analysis alone. Urban complexity implies important relations of neighborhood that cannot be ignored. As soon as we understand the importance of urban morphology and its different facets, the authors open another chapter that goes beyond. The importance of the urban/human dimension, the possibility of using remote sensing to infer the character of the place, and the possibility of challenging theories and methods are raised. Part IV ( Defining Urban Populations Over Space and Time ) begins with Longly and Messev (Chapter 9) addressing the importance of urban morphologies described in the previous chapter to derive indicators of urban shape and form. With that comes the questioning of data-led theories leading to theory-led data analysis. Issues such as fractal dimension/ allometry should be part of our urban agendas. Batty and Howes (Chapter 10) explain how time is included in remote sensing and urban analysis; they seek to understand and predict temporal patterns, using a method to classify development by age with remote sensing and ancillary data. Using time discontinuities regarding differences in densities, building materials, and transport systems, three important time frames are proposed: the 1930s, 1960s, and 1990s (p. 187). Donnay and Unwin (Chapter 11) direct our attention to the possibility of disclosing different population densities, also with the help of ancillary data. Finally, Baudout (Chapter 12) puts geography into perspective: do not forget that different continents have different realities, different objectives, and certainly different methods. This sincere and clear disclosure of information throughout the book telling us of what we need to be aware or the unfolding opportunities that we will find along the way makes this book interesting. Chapter 13, A Research Agenda in Remote Sensing and Urban Analysis, is a highlight of the book. As the editors point out, it is important to challenge remote sensing analysis, by going beyond per-pixel analysis into questions of configuration, syntax, structure, and function. Contrary to what is stated in the Epilogue, the book did not come out too early (p. 245), but just in time! It prepares the readers for what we should expect (potentialities, problems, advantages, questions still unsolved) in different opinions, methods, and how the different fields need to be related. As the authors conclude, there is still much to do on some real problems. The possibilities of this new wave of satellites also presages new problems, and there is a danger that remote sensing specialists will expend more effort addressing the opportunities and problems posed by each new technological advance, rather than seeking operational solutions to the use of existing systems, mainly if we want to go further than the simple urban form and try to extract other functional properties that are to the interest of urban planners and cannot be observed directly (p. xxiii). Much needs to be done in terms of the outdateness of some ancillary data to monitor fast changing urban development trends, or the balance between coarse resolution and confidentiality, and the incompleteness of some of the data sets (p. 252). These are presented and explored as a tool to help us realize what needs to be done. The authors see solutions through collaboration, with this book as proof. As stated in the Conclusion the outcome of research will determine whether urban remote sensing is to fulfill only a technical role in corroborating and updating other data sources, or whether it might fulfill a more central role in terms of quality of life studies, and data-rich modeling of form and function (p. 255). Reading this book reveals the encouraging optimism of the authors. Overall, the aim of this book is well accomplished. It is indeed a balanced exposition of detailed and broad views, and the bibliography helps the reader who wants to go further. Reviewed by: Elisabete A. Silva Department of Landscape Architecture and Regional Planning University of Massachusetts Amherst, Massachusetts bete@larp.umass.edu 34 URISA Journal Vol. 13, No. 2 Spring 2001

35 BOOK REVIEW Monitoring Land Supply with Geographic Information Systems: Theory, Practice, and Parcel-Based Approaches Edited by: Anne Vernez Moudon, and Michael Hubner. (New York: John Wiley & Sons, Inc.), pages. Urban sprawl is a significant problem. While national efforts are underway to deal with urban sprawl, some states have enacted legislation requiring local jurisdictions to contain sprawl within specified areas. This has increased the need for better information and methodologies regarding land supply and capacity. Monitoring Land Supply with Geographic Information Systems was written to offer a user s critical perspective on how geographic information systems (GIS) are used for land supply and capacity monitoring (LSCM). It does this by providing the reader with a general framework for LSCM and then building on that framework using case studies and thematic issues. The material for the book was developed with assistance from the Lincoln Institute on Land Policy and includes some chapters that originated as papers presented at a May 1998 seminar at the University of Washington. The book has three parts: Overview, Case Studies, and Thematic Issues. Each chapter includes one or more commentaries from experts in the fields of land-use planning and GIS, which help to broaden the perspective on each set of ideas presented. The book also includes extensive appendices, which provide detailed information including definitions and methods of analysis. The book is a hybrid between the fields of land-use planning and GIS. Instead of focusing on how GIS can be applied to land-use planning, it is more about the growing reliance of landuse planning on GIS to answer complex questions, especially those related to the spatiotemporal nature of urban growth management. While there are works on automated land supply information systems, this is the first comprehensive book on the use of GIS for monitoring land supply and capacity. It is suited for an upper-division or graduate-level course in urban and regional information science or land-use planning. It is also appropriate reading for GIS practitioners who work with land-use planners and for academics doing research in the field of urban growth management. In the Introduction, the editors establish the focus for this book and the use of parcel-based GIS as an increasingly critical component of LSCM. Along with a brief history of automated land supply information systems and a review of the book s structure, the editors identify five issues that recur throughout the book. They also provide an appreciation for the complexity and scope of this subject as well as its importance now and in the future. Both chapters in Part 1 ( Overview ) were written by the editors. Chapter 1 provides an extensive discussion on LSCM, describing it more as a set of practices than as a cohesive theory. They go into detail about the characteristics of these practices and private sector concerns before discussing the growing role of GIS in government and parcel-based GIS in particular. They identify six challenges: definitions, methods, issues regarding data, market conditions affecting land activity, environmental constraints, and political and financial support. Lastly, the authors identify opportunities through advancements in GIS and a national movement that should eventually lead to standards in LSCM. Chapter 2 establishes a general framework for LSCM, providing extensive material on definitions and framework elements as well as numerous drawings and models. This chapter is URISA Journal Book Review 35

36 especially helpful to GIS practitioners who have a limited background in land-use planning. Part 2 ( Case Studies ) includes three chapters and five commentaries written by various authors. Chapter 3 provides a case study on Portland, Oregon and lays out an inventory approach to LSCM along with the implications for database design. The concepts of homogenous geographic areas and multiple spatial units are also introduced. While the beginning of the chapter included numerous drawings and tables, only one table is provided as well as few examples to help in grasping these relatively abstract ideas. Fortunately, the first commentary provides another perspective on them, allowing the reader to have a greater appreciation for how they might be used. The second commentary provides a much-needed private sector perspective to the entire process of LSCM. Chapter 4 provides a case study of Montgomery County, Maryland. This chapter provides a contrast to the Portland, Oregon case and demonstrates how a completely different approach to urban growth management can create a very different landscape. The commentary on this chapter provides a criticism of the Portland, Oregon growth management model, identifying that it does not specify a logical urban form (other than a growing blob). Chapter 5 provides a case study of the supply and demand of industrial land in the Puget Sound region of Washington. It also reveals much about the unique nature of industrial land as compared to residential and commercial land. The two commentaries provide further elaboration on the unique nature of industrial land and implications for designing a methodology for monitoring its supply and capacity. Part 3 ( Thematic Issues ) includes three chapters and five commentaries. Chapter 6 is a concise and clear history of LSCM as well as technological (GIS) advances that have resulted in an expanded user base. The author identifies that increased exposure to land-use monitoring brings pressure for improved monitoring methods and more refined data models. This becomes a never-ending cycle of improved technology resulting in changes in public mandates that pose even greater demands on technology, including the ability to monitor transitional and mixed land uses. This chapter introduces a spatiotemporal model that is used to identify land-use transitions with what the author calls a kind of temporal topology. The commentaries delve deeper into the issues surrounding the limitations of how GIS technology is applied to LSCM and the evolving nature of land use, the environmental issues, and other factors affecting LSCM. Chapter 7 provides a broad discussion of data sharing and organizational issues affecting LSCM. The authors point out the need for crossjurisdictional data sharing and the complex organizational structures needed for coordination. The commentary following provides an in-depth look at organizational issues affecting successful data sharing, particularly the need for institutional capacity to change. Two examples of advanced GIS-based integrated systems are also provided as evidence of what can be done in LSCM. Chapter 8 provides a simulation model (UrbanSim) that uses variables for both land supply and demand. The model was implemented in Lane County, Oregon, and the author provides a detailed discussion of how urban modeling can be relevant to planning. The first commentary examines some of the weaknesses of UrbanSim (ignored variables), while the second commentary examines how models like UrbanSim can be used to test the practicality of urban planning concepts. This book is quite comprehensive. It is both broad and deep, requiring the reader to grasp a variety of concepts ranging from accounting, economics, land-use planning, and GIS. The reader is thereby able to understand the inherent complexity of LSCM and how GIS plays an increasingly critical role in modeling it. Practitioners will find this book immensely valuable in adapting GIS to long-range planning. Academicians will find it an excellent text for classroom discussion and essential reading in urban and regional information science, given the challenges of urban growth over the next century. Reviewed by: Gregory A. Newkirk, AICP GIS Coordinator for the City of Vancouver Vancouver, Washington greg.newkirk@ci.vancouver.wa.us 36 URISA Journal Vol. 13, No. 2 Spring 2001

37 BOOK REVIEW Valuing the Built Environment: GIS and House Price Analysis Scott Orford. (Hants, United Kingdom: Ashgate Publishing Ltd.), pages. In Valuing the Built Environment, GIS and House Price Analysis, Orford highlights the importance of including the impact of externalities on house prices. This is done by exploring the techniques that can be used to measure the impact of externalities within the context of the hedonic house price index literature. This book is a valuable introduction to major issues confronting the creation of hedonic house price models using the spatial expansion method for modeling externalities, particularly with regard to developments through the early 1990s. As discussed below, however, there are notable omissions in the author s treatment of the literature and of ongoing debates regarding the construction of house price models. Thus, this book should not be relied upon as a single source of information for this field. Rather, it will be useful in combination with other works that specifically address other types of house price models and other techniques for measuring spatial effects. This book will be of primary interest to novices in this field perhaps to current users of the geographic information system (GIS) interested in exploring hedonic house price models. Potential readers should be aware that the author uses a case study of Cardiff, United Kingdom to develop the GIS and house price models. While some of the context from the UK may be universal to the construction of GIS and house price models, others issues, such as data sources and geographic units, are more specific to that country. As a result, translation issues may confront readers hoping to apply lessons from this book to other housing markets. Unfortunately, the author does not facilitate this translation by identifying Cardiff-specific or UK-specific issues or discussions. While sections of the book are quite well written and are valuable additions to the house price literature, the same cannot be said of other sections. Instances of poor editing occur throughout the book tend to detract from Orford s message. The overall impression that the reader gets from the book is that it is quite dated: for instance, most of the references to existing literature are from the 1970s and 1980s, although many extremely relevant developments in the literature emerged in the 1990s. At a more superficial level, the maps and other visual supporting materials are quite rudimentary. Given the strides made in presentation and graphics technology in the past few years, this seems to be a curious anomaly, particularly for a book with a primary focus on GIS. Chapter 1 provides an introduction to the housing market literature and the microeconomic theory therein. The discussion is broad in scope, from the perfectly competitive housing market to the trade-off model to the bid-rent function and the notion of housing submarkets. Chapter 2 ( Hedonic House Price Function ) provides a very good historical discussion; the theoretical underpinnings are discussed as well as the major themes in the literature until the mid 1990s. Chapter 2, as well as Chapter 3 ( Housing Attributes and Spatial Data ), would serve as an extremely useful introduction and historical survey of the hedonic house price literature up to the early 1990s. There are two serious drawbacks to this prescription, however. The discussion of the literature on house prices and on externalities is dated. For instance, the latest reference in Chapter 2 is to a 1994 article. Moreover, coverage to 1994 is selective; while URISA Journal Book Review 37

38 the 1992 Can article is included in the discussion of spatial expansion model, the 1992 Dubin article that provides an alternate way (via kriging) to model spatial effects within the hedonic specification is left out. Moreover, there is no reference made to the heteroskedasticity problem that might be encountered in the spatial expansion method (Anselin (1990)) or the solutions thereof. Notable contributions in the development of hedonic models such as an article by Can and Megbolugbe (1997) that provides an explicit modeling of spatial effects using GIS, an article by Dubin (1998) that further develops kriging, or the numerous attempts to resolve and/or avoid the functional form issue (e.g., the article by Meese and Wallace (1991) on nonparametric estimation of hedonic models) are not cited. The second drawback with the discussion in Chapters 2 and 3, and indeed elsewhere in the book, is that no mention is made of alternatives to the hedonic house price specification such as repeat sales estimation. Since debate over the two methods of estimation is lively, this is a serious omission. Moreover, most of the focus in the literature since the mid-1990s (especially in the United States) has been on hybrid techniques that combine elements of both hedonic and repeat sales estimation (Meese and Wallace (1997) is a good example of this strand of literature). Developments in modeling of spatial autocorrelation and spatial dependence have also been in the context of all three types of specifications (Pace et al. (1998) provide a survey of spatial statistical applications in the real estate literature). Chapter 4 uses a case study of the Cardiff, UK housing market to walk the reader through the construction of a context-sensitive GIS. The discussion of the GIS itself is very rudimentary. Moreover, the discussion is somewhat disorganized; the discussion of the construction of a context-sensitive GIS should be streamlined as it is hard to stay interested in the details. For instance the discussion on linking address-based datasets on p. 88 is quite confusing. Instead of providing a general discussion of the kind of issues that are likely to arise in the linking of databases using the Cardiff housing case study as an illustration, the discussion instead provides details on issues in the Cardiff housing case study without indicating how universal these are. Maps relating to the Cardiff case study are presented in Chapter 4 and appear crude in this day of high-resolution maps. All the maps in the book share this drawback. This is really unfortunate because the ease of visual representation is an advantage of GIS techniques. Chapter 5 ( The Spatial Dynamics of an Urban Housing Market ) provides a competent and interesting discussion of the application of the various hedonic specifications outlined in Chapter 3. The model results in this chapter highlight the importance of locational context and spatial effects on the implicit price of structural attributes. The resultant heteroskedasticity and the spatial autocorrelation among residuals that arise if these effects are ignored in the construction of the implicit price of structural attributes are outlined well. Chapter 6 explores the influence of specific locational externalities using the multilevel hedonic specification that was shown to be superior in Chapter 5. This is a very interesting chapter and would be valuable to house price practitioners in helping them conceptualize the interaction between different externalities and in interpreting the measured impacts. As mentioned earlier, the graphs and maps in this chapter could be improved very easily. Chapter 7, the concluding chapter, provides a summary of the main results. Overall, the book is a good overview of the historical development of the modeling of externalities until the early 1990s within hedonic house price models. Since it is somewhat dated in its literature survey, and therefore its scope, it is not very relevant to house price practitioners or to GIS practitioners. Rather it is more suited to novice students of housing markets. The major contribution of this book is that it provides a good theoretical discussion of externalities and the spatial effects in the house price literature. The discussion of the results of the Cardiff case study is also of some value to the newcomer to the hedonic house price literature. References Anselin, L., 1990, Some Robust Approaches to Testing and Estimation in Spatial Econometrics. Regional Science and Urban Economics, 20, Can, A., 1992, Specification and Estimation of Hedonic Housing Price Models. Regional Science and Urban Economics, 22, Can, A., and I. Megbolugbe, 1997, Spatial Dependence and House Price Index Construction. Journal of Real Estate Finance and Economics, 14, Dubin, R., 1992, Spatial Autocorrelation and Neighborhood Quality. Regional Science and Urban Economics, 22, Dubin, R., 1998, Predicting House Prices Using Multiple Listings Data. Journal of Real Estate Finance and Economics, 17 (1), Meese, R., and N. Wallace, 1991, Nonparametric Estimation of Dynamic Hedonic Price Models and the Construction of Residential Housing Price Indices. Journal of the American Real Estate and Urban Economics Association, 19, Meese, R., and N. Wallace, 1997, The Construction of Residential Housing Price Indices: A comparison of Repeat-Sales, Hedonic-Regression and Hybrid Approaches. Journal of Real Estate Finance and Economics, 14, Pace, R. K., R. Barry, and C. F. Sirmans, 1998, Spatial Statistics and Real Estate. Journal of Real Estate Finance and Economics, 17 (1), Reviewed by: Padmasini S. Raman, Ph.D. Economics, Rutgers University Fannie Mae, Washington, DC padmasini_raman@fanniemae.com 38 URISA Journal Vol. 13, No. 2 Spring 2001

39 BOOK REVIEW Information, Place, and Cyberspace Edited by: D.G. Janelle, and D. C. Hodge. (Springer Verlag) pages. In 1998, the National Center for Geographic Information and Analysis (NCGIA) convened a group of academics to present their research on the interrelationship of information, place, and the still-emerging notion of cyberspace. The resulting book is a collection of conference presentations that address aspects of accessibility. Such aspects include how conceptualizations of accessibility are currently influenced by the advent and development of modern information technology. Nearly all of the books contributors have backgrounds in geography, planning, or spatial analysis. This sets the tone and parameters for the discussion. In my estimation, there is far less debate across disciplines as to what constitutes information and cyberspace, but is there more debate related to estimating and ascertaining place. Individual chapters raise interesting questions and research possibilities, especially for readers who are geographers or planners in academic settings. For others, there may be less appeal because several popular perspectives regarding the manner in which information technologies are currently reshaping spatial interaction and accessibility are not addressed. This should not be seen as a shortcoming of the volume. One challenge to reviewing an edited volume such as Information, Place, and Cyberspace is the diversity of topics and perspectives addressed by the 28 contributors. The volume is divided into three parts: Conceptualization and Measurement, Visualization and Representation, and Societal Issues. In terms of conceptualizing and measuring accessibility, Couclelis and Getis (pp ) point out that information technologies (i.e., cyberspace) are changing the way we think about accessibility. Traditionally, acces- sibility and mobility were generally limited to the likelihood of physical interaction and the capabilities of individuals or groups to interact. New communication and information technologies are allowing certain types of interaction to be unlimited in terms of distance, location, time, and hence, cost. This represents a particular analytical challenge: how do we measure or quantify levels of access in virtual space where it can occur any where, any time, and at little or no marginal cost? Scott (Chapter 3), Shen (Chapter 4), and Forer and Huisman (Chapter 5) consider this in terms of urban accessibility and employment, where integrating communications and information technologies in the form of telecommuting may come to represent a widespread mode of work-related travel. Telecommuting represents the ability to commute any where, any time, and at little cost. However, as some authors acknowledge, there is little evidence that telecommuting has significantly impacted the nature of work travel. In contrast to the tangible aspects of traditional accessibility, Heikkila (Chapter 6) discusses the fuzziness associated with virtual access. He defines virtual access as combining both the geographic and nongeographic realms and presents fuzzy set theory and fuzzy logic as a potential way to understand these associations. In Chapter 7, Sui takes a more traditional view of the emerging virtual world by mapping user and host locations. It is unclear whether identifying these locations will continue to be of interest in the future, as the technology is more pervasive. The next section focuses on representing and visualizing geographic and nongeographic space the former being more easily accomplished than the latter. Batty and Miller (p. 133) URISA Journal Book Review 39

40 imply that the means to representing and visualizing virtual space stems from traditional approaches. For example, while metaphors exist for distance, space, time, and accessibility in the virtual world, they are actually not relevant to virtual space nor are they easily measured in virtual space. Shortest paths, least cost travel, nearest neighbors, etc., have little or no meaning to perceived travel in virtual space. For instance, few care if an message bounces around the globe before it is sent to a colleague across campus as long as it arrives within minutes. Likewise, few would care if an on-line book order is filled in Greenland or the United States, as long as it arrives within 3 to 7 business days. For these reasons, it is unclear whether the authors references to geographic and temporal accessibility based on time zones (Harvey and Macnab, Chapter 9) or cities and regions (Moss and Townsend, Chapter 10) will endure. Will these geographic aggregations continue to be meaningful in the future given the changing meaning of realtime? This point is partially illustrated by the map titled Information spaces on the Internet in the chapter by Dodge (p. 189), where location and morphology are used to depict the relationships between services such as Telnet, file transfer protocol (FTP), and . On the other hand, this may not be the case for virtual social spaces such as those discussed by Dodge, where people can assume identifities (avatars) and interact with others virtually. These on-line parallel worlds may only be electronic forms of entertainment and recreation. However, the author argues that location decisions and place manifest themselves despite being contained within virtual space. Some of the reasons mentioned above illustrate the difficulties of representing information spaces graphically as discussed by Murnion (Chapter 12), Adams (Chapter 13), and Kwan (Chapter 14). The authors contend with representing how someone can be constrained by physical space while simultaneously having the capability of nearly unlimited virtual mobility. Their discussions highlight the hyper-dimensionality of combined geographic and nongeographic space. The final five chapters of the volume deal with the social implications of rapidly developing information technologies. Questions of social equity arise in terms of access to technology as well as access through technology. Access can be limited by social, economic, and political forces. Hansen (p. 240) points out that it is somewhat naive to think these new technologies will act as equalizing forces capable of providing opportunity to social and economic classes that have been previously denied. Technology, by itself, may not go very far toward equalizing what is an imbalanced yet established power structure. Other authors such as Ocelli (Chapter 17), Onsrud (Chapter 18), and Mugerauer (Chapter 19) identify important research issues relevant to evaluating whether new information technologies will serve to increase the existing digital divide between technological haves and have-nots. All of the social impacts related to the rapid development of information technology are not yet evident. What is evident, however, is that the impact perceived thus far has the potential to divide and segregate individuals similar to the segregating effects of a group s physical proximity and geographic accessibility to social and economic opportunity. Overall, this volume hints at the possibilities for an understanding of evolving communications and information technology. At first, the topics seem somewhat disjointed, but upon further consideration they are intriguing and thought provoking. Because information technologies (e.g., the World Wide Web) are relatively new, there is still a high level of uncertainty about how they will influence our lives and decision-making. The one element not explicitly addressed by the authors is how communication and information technology will continue to merge and emerge as an indistinguishable mode of facilitating communication. The combined technologies of wireless communications, global positioning systems, and hand-held personal computing are upon us and will achieve the ubiquity of television and telephone. Some of these applications (e.g., location-based services) have potentially vast implications for data infrastructure and industrial changes that will likely continue to be driven by commercial endeavors. After reading Information, Place, and Cyberspace, I became more aware of the complexities of the new accessibility. In the future, perhaps it will be the computer scientists that will define place and not the geographers or planners. Reviewed by: Thomas W. Sanchez School of Urban Studies and Planning Portland State University Portland, Oregon 40 URISA Journal Vol. 13, No. 2 Spring 2001

41 BOOK REVIEW Geographical Information Systems, Second Edition. Volume 1. Principles and Technical Issues and Volume 2. Management Issues and Applications Edited by: Paul A. Longley, Michael F. Goodchild, David J. Maguire, and David W. Rhind. (New York: John Wiley & Sons, Inc.), Volume 1 (pp 1 580), Volume 2 (pp ) This two-volume set of Geographical Information Systems by Longley, Goodchild, Maguire, and Rhind is a must have in the library for both practitioners and academics. Its comprehensive contents include more than 1000 pages and 72 chapters, with close to 100 eminent contributors in the geographic information system (GIS) and related fields. The high quality of both the individual chapters and the overall text is then hardly surprising. The editors and the majority of authors are affiliated with North American and European academic, government, and private sector institutions, with a few authors from other continents five from Australia and New Zealand, four from South Africa, and one from Asia. While this geographic distribution does not necessarily translate directly into the significance and nature of GIS activities, it suggests the regions dominating in GIS thought, invention, research, and practice. With unlikely competition in the genre, Geographical Information Systems is justifiably called the Big Book of GIS. The two volumes offer a state-of-the-art overview of the field and, similar to other disciplinary big books, * the new edition reflects the evolution of the field. Even those who have read the first edition will find a look at the second edition to be a worthwhile experience. In fact, in the making the second edition, the editorial team joined by Paul Longley started from scratch and the result is new organization and a new cast of authors. In the Introduction, the editors state that their intention, for example in presenting GIS history, is on updating the story with a brief account of major events and trends since 1991 [when the first edition of this book appeared] (p. 2) rather than on summarizing an already extensive publishing output. Other topics are also treated by building on existing knowledge and characterizing the recent developments and trends. The contents of Volumes 1 and 2 in the two editions are displayed in the table below. Volume 1, Principles and Technical Issues, dwells primarily on spatial data collection, development, and analysis. Following a theoretical introduction, a selection of topics addressed includes error and uncertainty in spatial data; spatial statistics and modeling; GIS interoperability; methods of data access; principles database design; spatial referencing; remote sensing and global of positioning systems as means of data capture; transformation of spatial data (e.g., interpolation); and virtual GIS. Volume 2, Management Issues and Applications, discusses various aspects of GIS introduction into organizational settings and requirements for its effective use. Among other topics, the authors present institutional issues, legal implications, standardization, privacy concerns, and geospatial data policies. Examples of GIS applications are presented under the two categories of operational and social and environmental, the latter being a rather coarse grouping of quite diverse ways of employing GIS. The main text of the book is complemented by 66 colored plates; the author and subject indexes are thorough and useful. The text also contains a 79 page consolidated bibliography. The contents, coherency, and overall organization of the two volumes are, to an extent, a reflection of the state of the discipli ne at the time the manuscripts were prepared. However, credit for improvements goes to the chapter contributors who, with many others, URISA Journal Book Review 41

42 First Edition Volume 1 Section I: Overview Section II: Principles a. Nature of spatial data b. Digital representation c. Functional issues d. Display issues e. Operational issues Volume 2 Section III: Applications a. National and international GIS programs b. Socioeconomic applications c. Environmental applications d. Management applications Section IV: Epilogue Second Edition Volume 1 Part 1: Principles a. Space and time in GIS b. Data quality c. Spatial analysis Part 2: Technical Issues a. GIS architecture issues b. Spatial databases c. Technical aspects of GIS data collection d. Data transformation and linkage Volume 2 Part 3: Management Issues a. Making the GIS efficient, effective, and safe to use b. Data as a management issue c. GIS as a management tool d. The impact of broad societal issues on GIS Part 4: Applications a. Operational applications b. Social and environmental applications participated in advancing the GIS field and to the editors who managed to capture those extensive developments and to provide a new framework for the field that flourished and began to mature during the 1990s. The opening sections of each volume, one on GIS principles and the other on management issues, illustrate this evolution of the field. The GIS principles are introduced with a general discussion of space and time in GIS, which draws on philosophy and history of science (geography, mathematics, and physics in particular); cognitive and social theory; and theoretical and methodological bases for representation, visualization, and substantial generalization of spatial data. This is a departure from a more pragmatic overview of GIS definitions and implementation settings (technological, commercial, government, and academic) that invited the readers of the first edition. The management theme is an enhanced version of the Operational Issues section in the first edition. A somewhat awkward introduction to the theme reveals that management is still aside from the mainstream. The quality of individual contributions notwithstanding, this section could have more explicitly conveyed that GIS management is only one aspect of a more complex process of GIS technology transfer, organizational change, and general information resource management. In addition, the three sample applications of GIS as a Management Tool cover an interesting scope, ranging from business and service planning to public discourse and to state economy (Chapters 51 53), however, their grouping within Part 3 seems a bit contrived and somewhat arbitrary. With appreciation for different ways in which not only a section but a whole book could be organized, and despite the minor criticism, this section offers a wealth of information and insight.** Common difficulties associated with compilations (such as ensuring systematic and complete coverage of relevant topics, repetition in contents, and flow and coherency of various book sections) are successfully mastered by the editors and only rarely experienced throughout the two volumes. As for the substantive critique, the editors themselves alert to the omissions from the first edition and the challenges of the second edition of the book. In some respects, the self-critique seems too harsh; for example, in not foreseeing the Internet-related developments and their implications on GIS software and future directions. The pace of technological change against the publication lag time makes it almost impossible to present it in a timely manner and quite difficult to envision the future. Even from 1999 when this book appeared (or , when it was most likely worked on), new technological solutions would make some of the text in the new edition out of date. Otherwise, the editors self-critique and commentary are insightful and thought provoking. For example, they accurately observe that, despite recognition in the early 1990s of the primary importance of GIS implementation and institutionalization challenges, there is a continued need for innovative technical solutions. Perhaps extending a bit on that comment or looking from a somewhat different perspective, this assertion of technical developments may contain an element of self-fulfilling prophecy 42 URISA Journal Vol. 13, No. 2 Spring 2001

43 of the predominantly technical and system-oriented academic and professional community. Another perceptive remark is about invigorating information management and data-related aspects of GIS research and development as the key for wider adoption of GIS, opposed to an earlier call for enhanced analytical functionality. The return to data issues has indeed been strong. Perhaps less strong but equally important have been the efforts toward the other end of the data-system-function continuum. The efforts aim to extend the utility of GIS from data storage, manipulation, and presentation to decision-making tools. The only chapter that relates GIS to decision-support systems is Yeh s Urban Planning and GIS (Chapter 62). This topic, however, warrants more extensive treatment. Finally, related to the introductory note about the regional affiliation of the contributors to the book, it should be noted that there is a lack of explicit presentation of international trends, issues, and policies. While many chapters, particularly the nontechnical ones, inevitably carry a specific regional perspective and are socioculturally embedded only one chapter in the book discusses national and international geospatial data policies. In the first edition a section National and International GIS Programmes contains six chapters with case examples from the United States (United States Geological Survey-USGS), United Kingdom (Ordinance Survey), Sweden, Japan, Australia, and developing nations. I applaud the editors for finding the wealth of theoretical and conceptual material accumulated in the 1990s more of a publishing priority and exciting content over descriptive case studies of GIS as used in particular countries or regions. However, an overview and equally conceptually rich presentation of the international arena and trends that go beyond European/Anglo-Saxon environment would be a valuable contribution to a future edition. The chapters on National and International Data Standards (Chapter 50), National and International Geospatial Data Policies (Chapter 56), GIS in Land Administration (Chapter 61), and a case study of South Africa (Chapter 65) partially fulfill this purpose. The editors have upheld their promise to give the current awareness of the field and to look at the future. A figure The changing domain of GIS (p. 749) and a table Factors supporting the GIS paradigm shift (p. 1020) illustrate and summarize the editors view of the evolving field of GIS. In the epilogue, they state the interaction between policies, institutional factors, technologies, applications, and social context as the guide for understanding and forging future endeavors. They review the areas and goals of research agendas in the US (the National Center for Geographic Information & Analysis-NCGIA and the University Consortium of Geographic Information Science-UCGIS) and Europe (GISDATA) to find considerable progress made toward more useful, integrated and networked but yet transparent GIS environments. From the public and private user sectors, they support the expectation for semantic and geometric integration; data encapsulation; intuitive, human cognition-based systems; standardization (de facto or other); developed and accessible spatial data infrastructures; geocomputation; GIS-based elementary education; and community empowerment. While GIS has moved away from map metaphor toward multimedia, virtual reality, and temporal GIS, and software and hardware inventions bring wonders of wireless and wearable computing, the editors expect visible GIS is here to stay. I would highly recommend this unique book of encyclopedic proportion and value to our GIS community. The organization of the topics by Longley et al. makes the book s framework transparent and easy to grasp. Because the two volumes are not written in textbook style and because of an advanced level of treatment of topics (with some variation), I expect the contributions in the book be used as supplementary teaching materials to other books covering the GIS fundamentals. While those who closely follow the many streams of GIS literature have probably encountered some of the contributions in various other books or journals, it would be hard to find a comparable compilation of all the current significant material in one place. Many readers will encounter new and stimulating concepts, such as goecomputation, data mining, virtual GIS, semantics, interoperability, and ideas such as encapsulation of data with methods (driven by object-oriented environments). This is a shelf reference for every GIS developer, user, and researcher to be joined by third edition in due time. Reviewed by: Zorica Nedovic-Budic, Associate Professor University of Illinois at Urbana-Champaign Department of Urban and Regional Planning *Urban planning, for instance, had the third edition of its Green Book The Practice of Local Government Planning published in year **The management theme can be used as an example of a tension characteristic for most of the GIS field. This tension arises from a somewhat awkward connection to other disciplines, to the broader system of knowledge, and to application contexts. While the GIS field draws on many traditional sciences (as explained by Couclelis in Chapter 2 Space, Time, Geography ), its further developments are sometimes disjointed or even negligent of the source disciplines. Perhaps that is inevitable for an integrative field in the process of establishing its location and identity within the system of science and societal practice. The tension may be the manifestation of that struggle. URISA Journal Book Review 43

44

45 BOOK REVIEW GIS in Public Policy: Using Geographic Information for More Effective Government R.W. Greene 2000 ESRI (380 New York St, Redlands, CA, ) 100pp, ISBN , paper cover GIS in Public Policy is part of the ESRI Press Case Studies Series. In this series, the applications of ESRI geographic information system (GIS) software in different fields, such as natural resources, energy and landscape architecture are explored. As in other books in the series, the graphic design of GIS in Public Policy is beautifully rendered to form a visually stunning book. It is full of colorful photos and maps perfect for those short on time, attention, or both. Also like the other books in the series, GIS in Public Policy can be seen a marketing tool to showcase the many uses of ESRI software. The intended audience for the book seems to be practitioners in government and their consultants who are not currently using GIS and who are interested in how to use GIS to assist with policy problems. This book would not directly help GIS professionals in government because that group does not need convincing about the usefulness of GIS, and the book does not go into technical details about the analyses used in the cases. GIS professionals could benefit from the book if decision-makers in government were to read it, get excited about the possibilities of GIS, and increase funding for it. This may be the most effective use of the book, which, in the introduction, touches on an important lesson learned by case study participants for GIS to make a difference for an organization, its power must be recognized by the leadership. When I first picked up the book, I was hoping for a broader look at how GIS has affected public policy studies. In my studies of public policy at the University of Chicago, the quantitative methods of choice were statistics and economics; spatial analysis was not taught. Some questions I was hoping the book would address are: Has GIS made a difference in the practice of public policy? What kind of problems does it solve that were difficult or impossible to solve before? What are the limitations? How widely is it being used? What level of investment and staff time is needed, and is it worth the effort? Unfortunately, the book did not address any of these questions, except to give specific examples of some of the problems in which GIS was useful. Another minor critique I had about the book was its tone, which is a bit self-congratulatory. Each section begins with a stubborn public policy problem that miraculously gets solved with ESRI products. Despite its tone, the book contains some interesting case studies that should inspire even the most technophobic reader. The book is split into six chapters, with two or three case studies per chapter. The chapters cover education, health and safety, public services, environment, social services and international. Within the education section is a remarkable case involving Fresno County s Office of Education in California. They had been using outdated nine-track tapes and the maps-on-the-floor method of calculating the number of low-income students each school had in order to estimate how to distribute federal funds. After converting to ArcView GIS shapefiles and joining the databases from several sources, 550 hours of staff time and $22,000 were saved. What is remarkable is not only how much they saved, but how many other school districts are likely doing it the oldfashioned way. URISA Journal Book Review 45

46 Another good case study in education is The Weed Project in a rural Idaho school. The students are using global positioning systems to map a noxious weed that is damaging watersheds, wetlands, and range land. The students are compiling their information into a weed basemap and giving it to the County Weed Superintendent, who is in charge of eradicating the weeds. Prior to this project, the Superintendent did not know where the weeds were located and lacked the resources to gather this information. An interesting note about this project is how fast the students picked up ArcView GIS and other computer skills; the students not only taught ArcView to one another but also to the teachers. These case studies demonstrate the manner in which GIS can be useful for school planning as well as in the classroom. In the health and safety chapter, the author documents case studies in crime mapping and public health. GIS-based crime mapping can be seen as taking pin mapping (paper map on the wall with stickpins for each crime) to the next level. In addition, police are using GIS to solve multiple crimes at once by confronting a suspect with a map of crimes. They have found that this induces the suspect to confess to additional unsolved crimes. GIS also enabled a county in Washington State identify the extent to which a Medicaid program called First Steps was being used. The program s goal is to prevent low birth rates by providing extra prenatal services to pregnant women. Yakima County staff geocoded its Medicaid births and all First Steps births to see the percentage of mothers the program was reaching in different areas of the county. They identified three cities with lower than average First Steps usage and made plans to target these cities with more education about the free program. Each example shows how GIS was used in daily public safety and also to assess and redirect the implementation of a pubic health program. In terms of public service, the city/county of Indianapolis GIS program is a good example of the way in which a metro government can utilize GIS to increase the effectiveness of its services and communication. Indianapolis was having difficulty in managing snow removal after big storms. This resulted in wasting money because of such problems as some streets being plowed twice and some none at all, which was a perennial source of embarrassment for the city. The Mayor pushed to find a solution, which resulted in a GIS application called Snow Fighter that keeps track of snow removal progress and cost. Another GIS solution is the Data Viewer, which allows all staff to view and use GIS data. Additionally, the GIS is available to the public through the IndyGov Web site and citizens are allowed access to services such as custom bus routes, permits, and reservations to parks. In the environment chapter, the author describes how the drinking water of Boston, Massachusetts has improved. For years, state legislators tried to pass the Watershed Protection Act, a bill that would preserve land near the rivers and streams leading into Boston s water reservoirs. However, citizens in western Massachusetts who lived near the reservoirs blocked the legislation out of fear that it would lead to widespread land-use restrictions. Once the 200- and 400-foot buffers were displayed graphically using GIS, the opposition evaporated and the measure passed. The map communicated better than words that the buffers affected a minuscule amount of land. An interesting case study is presented in the social services chapter of a ride-share program set up to provide nondriving senior citizens the mobility they need. Elderly people call in and specify where they are going, what time, and how many passengers they are willing to ride with. The operators use GIS to analyze all the inputs and optimize the route and the cost. The program is self-supporting and very popular. This is an example of how GIS can resolve complicated logistics problems. The international chapter contains an example of the manner in which the USGS responded to the Hurricane Mitch emergency by creating the Digital Atlas of Central America and distributing it for free to relief agencies. In addition, the USGS used their successful use of GIS to lobby for $1 billion in disaster prevention funds, including the mapping of floodplains. This case illustrates how geographic information is vital to efficient response to disasters. It also showcases one of the most important applications of GIS it can be used to minimize loss of life and property from natural disasters through more informed land-use planning. Reviewed by: Laura Pinnas received a Master s of Environmental Management from Duke University and a Bachelor of Arts in Public Policy from the University of Chicago. She works as the GIS Analyst for the Town of Oro Valley, Arizona. She serves on the Technical Advisory Committee for the Pima Association of Government and the Regional Electronic Plan Submittal Committee, an effort to set standards and protocols for electronic plan submission. 46 URISA Journal Vol. 13, No. 2 Spring 2001

47 DATABASE REVIEW American FactFinder by the U.S. Census Bureau By Rosmarie Kelly and Jay Lee As anyone who uses data from the United States Decennial Census knows, the Census Bureau is releasing much of the 2000 Census data by state on a rolling basis. American FactFinder ( /factfinder.census.gov), the Census Bureau s online data delivery system that became fully operational in March 2001, will serve as the primary means of distribution for the latest Census data. The goal of this system is for the demographic and economic data collected by the Census Bureau to be easy to find, easy to get, and easy to use. The outcome will be that the Census Bureau will be able to disseminate more information more quickly than at any other time in its history. The FactFinder site provides access to all available data from the 1990 and 2000 censuses, the American Community Survey for each year it has been conducted (1996 to 1999), and the Economic Census for Population datasets are (or will be) available for each of the 50 states by county, census tract, and census block. American FactFinder makes it easy to look up summary statistics by address or by place name. In addition, data can be summarized to a table that can be printed or downloaded in a spreadsheet format, or a downloadable map can be created from the data. The home page for American FactFinder is set out in an easy-to-use manner, especially considering the amount of data that are available. For ease of navigation, the FactFinder site contains direct links to various options for extracting data. Users can extract the desired data in tabular form or have the data displayed in maps. Searching for data is straightforward, as FactFinder allows users to extract data by using topic links, by searching with a keyword or a place name, or by using the Basic Fact user interface. These options are: Search functions to help quickly locate any type of information that is available in FactFinder; Basic Facts, for Quick Tables and Geographic Comparison Tables for population and housing data, Quick Reports for economic data, and predefined thematic maps; Geographic Comparison Tables for comparing data for different geographic areas; Detailed Tables that provide easy access to all tables and maps for each summary file; Data Sets to access all available tables for the Decennial Censuses for 1990 and 2000, the American Community Survey for each year it has been conducted, and the Economic Census for 1997; and Reference Maps that display the boundaries of Census geographic areas, and thematic maps that display data items graphically. URISA Journal Database Review 47

48 For the Spanish-speaking user, data for Puerto Rico are accessible on a Spanish language page (Puerto Rico data en español). For children ages 7 through 11, there is a fun and educational site with information about the Census Bureau and facts about all 50 states (Kid s Corner). For all users, items of interest include summary files with tables on age, sex, households, families, and housing for smaller areas available on a state-by-state basis; housing unit counts for states, counties, and places; demographic profiles with age, sex, race, and Hispanic/Latino counts, and information on housing, households and families; and ranking, comparisons, and summaries of the population change for the United States as a whole, states, counties, metropolitan areas, and large places. All of these data can easily be extracted for online display or downloaded. reports for economic data. Once the user has found the desired report, table, map, or product, the FactFinder menu bar can be used to make additional selections. The menus vary depending upon the location within FactFinder, but may include Change selections, Print/Download; and/or related items. Additional information about the data and the method of collection is also available, and where appropriate, there are disclaimers about the data, including information on confidentiality protection, nonsampling error, and definitions. While navigating through FactFinder, the user can click Help on the banner and a menu will open to help with completing the current task. All of the topics can be accessed through their hyperlinks in the Table of Contents. However, since each topic is contextsensitive, some may be fully understood only when on the appropriate page in FactFinder. In addition to tables, FactFinder can be used to create, print, and download two types of maps reference maps and thematic maps. Reference maps display the location and geographic boundaries for Census tabulation areas (e.g., the boundaries of a Census tract). Thematic maps display interesting facts about places by using colors or patterns to shade areas on the map. They are used to display statistical data, such as population or median income, on a map. Some of the data easily accessed using FactFinder are facts about people, business, and geography at the state or county level; geographic comparison tables, detailed tables, summary and percent change tables for the American Community Survey; and data sets, industry quick reports, and geography quick In addition to actual data, a great deal of information is available at the FactFinder site. Included are reports on how data are collected, comparisons of questions asked in the 1990 and 2000 Censuses, information about the Economic Census, general information about the 2000 Census, information for elected officials, lesson plans and information for teachers, and special reports about specific data. There are also disclaimers about the validity of the data collected by the Census Bureau. In addition, several articles are available containing information about the incompatibility of data on race in Census 2000 compared to those collected in previous Censuses, including: 48 URISA Journal Vol. 13, No. 2 Spring 2001

49 Census 2000 Brief, Overview of Race and Hispanic Origin Press Release, Census 2000 Shows America s Diversity Questions and Answers for Census 2000 Data on Race From the FactFinder home page, users can click the link of Rankings, Comparisons, and Summaries to start a process of extracting tabulated data. The resulting data can be displayed on screen or downloaded in Adobe PDF format, in MicroSoft s Excel file format, or in comma-delimited ASCII text format. Users can select to rank or compare data on many population, demographic and housing topics by state, county, metropolitan area, or incorporated place. The process of extracting and downloading data was smooth and easy as tested by the reviewers. Before importing these downloaded files into a geographic information system (GIS) or a mapping program, it would be necessary to edit out the header information that includes title, data source, and other contact information. Many spreadsheet programs will be able to read the comma-delimited text file and convert it to the dbase file format required by GIS/mapping programs. In doing so, users should keep the attribute that contains Federal Information Processing System (FPIS) codes for the geographic units associated with the data. Perhaps the best feature of FactFinder for GIS professionals is that the data are free and downloadable in a format compatible with most mapping packages. While volumes of Census data will still be printed and distributed to repository libraries, the Census Bureau s American FactFinder makes it possible for the public to access these data in a timely fashion and in a useful format. Compared to the tedious process needed before when ordering Census data, the direct access to Census data via the American FactFinder Website provides a much more efficient and economic way for school teachers, researchers, planners, local government, and the general public to learn of the population and housing distributions in our nation. There are also links to other information and products available from the US Census Bureau, including press releases, a subject contact list, a schedule of data releases, and selected historical data. Using American FactFinder, it will be possible for the Census Bureau to disseminate more information to more users than ever before in the history of the Census. About the Authors Rosmarie Kelly is with the Department of Health for the City of Akron, Akron, Ohio. Jay Lee is with the Department of Geography, Kent State University, Kent, Ohio. URISA Journal Database Review 49

50 M A R K Y O U R C A L E N D A R Help URISA Celebrate its 40th Anniversary! URISA s 40th Annual Conference & Exposition October 26-30, 2002 Hyatt Regency Chicago Photo courtesy of the Chicago Convention and Tourism Bureau Chicago, Illinois Watch for the Call for Presentations to be distributed in December!

51 SOFTWARE REVIEW ArcPad 5.0.1: A Heavy-Hitting Lightweight By Gregory Zimmerman Environmental Systems Research Institute, Inc. (ESRI) introduced ArcPad 5.0 in March and in June 2001 as an answer to the growing need for handheld software for the mobile geographic information system (GIS). The lightweight (± 1.0 MB for Windows CE) program can be used either as a stand-alone program or as a field data collection device for a personal computer (PC) or network-based system. ArcPad allows users to create, edit, and view GIS files in real time while in the field. ArcPad 5.0 is primarily aimed at current GIS users looking for a low-cost ad-on for field GIS data capture and editing. then uploaded back into the master database Although developed for Windows CE, ArcPad ships with versions for Windows 95, 98, NT, and CE. This becomes an important aspect of the software as it allows for full portability across different platforms. ArcPad software is installed easily onto a Compaq Aero 2100 with 24 MB of RAM via MicroSoft ActiveSync 3.0. When memory is limited, users should disable unnecessary ActiveSync service, but keep the File Synchronization for file transfers. A key feature of ArcPad is that it uses the same vector-based shapefile format as ESRI s popular ArcView GIS, ArcInfo, and other programs, but in the Windows CE environment. Shapefiles and many formats of graphic files from an ArcView project can be brought into ArcPad directly without conversion, edited, and URISA Journal Software Review 51

52 The ArcPad environment, although a little disconcerting for ArcView or ArcInfo users because of a lack of sophisticated functions, is well designed. The controls are easy to see and intuitive to use. It possesses excellent qualities in a PC with a screen as small as 2.5 x 3.5 inches. The key function of ArcPad is its ability to work with a Global Positioning System (GPS) receiver. A user can see his or her exact position in the context of a map in real time via a red crosshair on the screen. Coordinates are displayed in Universal Transverse Mercator (UTM) and latitude and longitude. Dropdown navigational tools include speed, heading, compass, and distance/bearing to a waypoint. ArcPad 5.0 will only accept Trimble Standard Interface Protocol (TSIP), a proprietary Trimble GPS output format, from a GPS within the United States with a sublicense from Trimble. Subsequent to the release of 5.0, ESRI obtained a sublicense from Magellan that enabled later versions of ArcPad to work with GPS receivers from Ashtech and Magellan. Outside the U.S., ArcPad is compatible with most GPS receiver brands with National Marine Electronics Association (NMEA) output. The newest version of ArcPad, ArcPad 5.0.1, adds support for a number of GPS receivers (Trimble, Magellan, and Ashtech) and is available as a free download for those who own the 5.0 Version of ArcPad. The Version has the ability to receive data from an ArcIMS server using wireless technology. One potential problem with the GPS capability of ArcPad is the manner in which ArcPad records point data. Currently, ArcPad records one pair of coordinates for point capture. In most highend GPS units, multiple coordinates are recorded, and calculation of all the readings is used to capture one point to increase accuracy and reduce error. For applications that require a high degree of GPS precision, such as wetland delineations, this may be unacceptable. My initial attempt to connect ArcPad with a GPS was an educational experience. Merely adding shapefiles to ArcPad does not permit activation of its connection to a GPS. ArcPad does not enable the GPS buttons until the projection of your project is defined. The layers used within ArcPad must be in a supported projection and must have an associated projection (or.prj) file. For these reasons, it is extremely important to make sure that the GPS unit you are planning to buy is compatible with ArcPad. One of the best sources of technical information on ArcPad is the ESRI Web site ( The ESRI homepage contains a link to an ArcPad Community, where questions and problems with ArcPad are posted and other ArcPad users or, in many cases, employees of ESRI answer questions. 52 URISA Journal Vol. 13, No. 2 Spring 2001

53 Another cornerstone feature of ArcPad is the ability of the user to create custom data collection forms. Custom forms can be created either directly in ArcPad, or by using the Dialog Designer extension in ArcView on a desktop computer. ArcPad 5.0 supports date/time fields, combo boxes, check boxes, text boxes, and edit boxes. Unfortunately, the program does not support slider bars, one of my personal favorites for pen-based systems. The ArcPad Tools extension for ArcView GIS is extremely useful. The most useful feature may be the exporting georeferenced JPEG button, which outputs the screen view into a lightweight JPEG format and writes a world file for the image. The image and world file can then be copied to the handheld device and it is ready for use in fieldwork. The tool gives you the option of choosing an output quality (1 to 100) so that users can conserve valuable system memory. This tool is especially useful in converting high-resolution images, such as DOQ. tif files, into a portable format. The ArcPad Tools extension also allows a user to clip shapefiles from an active theme in an ArcView project and save the results in a new folder. The clipped files can then be edited in the field and returned to the project. ArcPad Tools can utilize legend symbology from ArcView GIS by creating an ArcPad Symbology File (*.aps), which can utilize Simple and Unique value legends. ArcPad 5.0 only supports labeling of point features. It does not support labeling of lines or polygons. However, this problem is expected to be resolved in the 6.0 upgrade. Hardware can be a limiting factor with ArcPad. Often, in direct sunlight, the screen of the handheld PC, such as the Compaq Aero 2100 that I used, tends to become unreadable. Rugged, tablet-type computers are much better suited for fieldwork than many of the Palm PCs on the market. Some of the new-generation palm-size computers, such as the Cassiopeia, have developed larger, brighter displays and boast more RAM and faster processors. Users with needs to do a large amount of fieldwork may want to explore newly released color pocket PCs that have excellent active display under sunlight. Overall, our experience with ArcPad 5.0 has been very positive. ArcPad is a good choice for light- to medium-volume field GIS data collection applications. One limiting factor is that ArcPad does not support multiple data collectors on a common database without some creative scriptwriting. The program is relatively simple to use the user is familiar with the concepts of ESRI s other programs. Once the customized data forms are created and the layers and projection set up, anyone can use the system in the field. Most of the problems we experienced were related to the initial GPS setup, a lack of program memory (before expanding from 8 MB to 24 MB RAM), and a few minor glitches in the program. Many of these problems could have been avoided by browsing through the ArcPad discussions on ESRI s Website or by downloading the free version of ArcPad The 6.0 Version of ArcPad, available since the Second Quarter of 2001, promises to provide greater GPS point accuracy along with a large number of other substantial improvements. URISA Journal Software Review 53

54 Overview of ArcPad 5.0 Performance Very good Performance can vary widely based on hardware. Medium to large files require lots of RAM and speed. Tools are easy to use. GPS tracking in real time on a GIS project is excellent. Manuals Fair Tutorial was very helpful but manual was lacking in a few important areas. The ESRI Website was the best source of information. GPS translation Fair ArcPad 5.0 will not work with NMEA GPS receivers in U.S. ArcPad and later versions will accept NMEA GPS receivers. Most functions work well but need to average point data for greater accuracy. Ease of learning/use Fair to good Once set up, program is simple to use. Set up and data form creation can be challenging. Data forms need to support slider bars for faster data collection. Support Excellent On-line community support is a great resource. Value Excellent Excellent value for the cost. Summary Very good Affordable, lightweight, and relatively easy to use with many field applications. Most problems are solved in the free download. The 6.0 Version, which was released in the Second Quarter, 2001, has many more improvements. System Requirements ArcPad 5.0 supports the following four chips for Windows CE devices: Hitachi SH3 and SH4, Intel StrongARM, and MIPS CPU. These chips make up most of the Windows CE devices currently available. ESRI recommends a minimum of 8 MB RAM (ArcPad will run on 4 MB RAM, but that does not leave room for much else), but suggests a computer with Windows CE 2.11 or higher with 24 to 32 MB RAM and ActiveSync 3.0. ArcPad 5.0 does not support Palm OS computers. Rugged, tablet-type Windows CE computers will likely be the most effective for heavy field use or large data files. System Dealers Environmental Systems Research Institute, Inc. 380 New York Street Redland, CA (714) About the Author: Greg Zimmerman is an Aquatic Biologist/Project Manager at EnviroScience, Inc., Stow, Ohio. 54 URISA Journal Vol. 13, No. 2 Spring 2001

55 URISA Quick Study Series 10 Titles Now Available! URISA QUICK STUDY SERIES Since their inception, the extremely popular Quick Study Series has assisted numerous professionals who are in need of important information, but without the time to read extensive literature or attend expensive seminars. These tutorial-type books provide the reader with... subject matter fundamentals term definitions negotiating and purchasing tips recommendations flowcharts, diagrams, photos and other examples and much more to bring you up to speed on these important and timely topics. All publications are from pages in length and are prepared by recognized experts in their fields. With three newlypublished titles, there are now ten publications in the Quick Study Series. NEW! JUMPSTARTING GIS: HOW LOCAL GOVERNMENTS CAN GET STARTED IN GIS WITH LIMITED RESOURCES This publication is a must for small local governments who are just starting their GISs with limited funds and time. It is a resource guide for GIS system and data components that can be acquired and put into use quickly and inexpensively. The document describes alternative funding opportunities and starter kits available from government agencies and vendors. Discussion also covers building local and regional partnerships and alliances and how they can be helpful to jumpstarting a GIS. Edited by Allen Ibaugh N E W! GIS PROCUREMENT AND RFP DEVELOPMENT This publication provides detailed steps to help local governments and other organizations procure GISrelated products and services. It discusses the development of effective RFPs that facilitate the development of a solid platform for GIS technology purchase and implementation. Edited by Allen Ibaugh N E W L Y U P D A T E D AERIAL IMAGERY GUIDELINES The intent of this publication is to define the basic terms and to discuss and explain the common issues surrounding aerial imagery acquisition. In addition, negotiating tips and minimum standards for imagery are suggested, along with a listing of resources for learning more about this subject. Edited by Mary Tsui DEVELOPMENT MANAGEMENT SYSTEMS Provides an overview of system selection to meet needs in planning and regulation. This URISA publication includes sections on benefits, types of systems, system features, advance system functions, implementation and a listing of software products. Edited by Dennis Sandquist, AICP DATA MODELING This book details the role of data modeling in overall system planning and the relationship to database design. This 40-page publication includes definitions, reasons for data modeling, how a relational database works, and the steps involved in actually doing data modeling. Edited by Mike Walls INTERNET, WEB AND E-COMMERCE This Quick Study will provide you with the information necessary to 1) get your organization on-line, 2) develop a useful World Wide Web site and 3) begin conducting commerce via electronic transactions (e-commerce). Edited by David J. Martin & David Reid INTERNET BASED GEOGRAPHIC INFORMATION SYSTEMS AND DECISION SUPPORT TOOLS This Quick Study provides readers with an understanding of the availability and potential of existing Internet resources, and prepares them to use the next generation of Internet based decision support tools. Edited by Shilpam Pandey, Jon Harbor and Bernard Engel N E W! A PRIMER ON REDISTRICTING This Primer includes everything you and your organization need to know to create effective redistricting plans. To set the stage, the primer begins with a discussion of the data you will be using -Census TIGER 2000, P.L population and race data, political data. Edited by Howard Simkowitz GIS DATABASE CONCEPTS GIS DATABASE CONCEPTS reviews all types of data repositories, data collection, development and conversion approaches, discussions on scale, accuracy, data quality and metadata, and the management of a relational spatial database. Edited by Peter Croswell GIS GLOSSARY OF TERMS This handy reference guide was written in order to equip the IT/GIS professional with the definitions, terms and concepts that are central to the GIS community. Edited by Daniel Parr ALL URISA PUBLICATIONS ARE AVAILABLE FOR PURCHASE AT THE CONFERENCE. STOP BY THE URISA MEMBER BOOTH IN THE EXHIBIT HALL OR ACROSS FROM REGISTRATION TO CHECK OUT THE EXTENSIVE LIBRARY OF RESOURCES.

56