ontents - urisa · 45 geographic information science: critical issues in an emerging...

Volume 12 • Number 1 • Winter 2000 R

Journal of the Urban and Regional Information Systems Association

CONTENTS

REFEREED

7 An Empirical Approach to Estimating GIS BenefitsStephen R. Gillespie

15 GIS Implementation in the GrassrootsR. E. Sieber

31 Beyond City Limits: The Multi-Jurisdictional Applications of GISMichael J. Greenwald

SPECIAL REPORT

45 Geographic Information Science: Critical Issues in an Emerging Cross-DisciplinaryResearch Domain

Edited by David M. Mark

55 Review of Current Journal LiteratureCompiled by Zorica Nedovic

On the Cover

Saving time and money are among the benefits extolled for imple-menting geographic information systems. A model for predictingthe economic value of geographically referenced data is discussedin this issue in an article by Stephen R. Gillespie of the U.S. Geo-logical Survey. URISA is pleased to introduce a new and updatedcover design for the Journal to begin the 21st Century. We hopeyou enjoy the new “look.”

2 URISA Journal • Vol. 12, No. 1 • Winter 2000

R

Journal

Publisher: Urban and Regional Information Systems Association

Editor-in-Chief: Harlan Onsrud

Editor Emeritus: Kenneth J. Dueker

Managing Editor: Robert Berry

Executive Director: David J. Martin, CAE

Electronic Journal: http://www.urisa.org/journal.htmusername = URISA2000password = ORLANDO

EDITORIAL OFFICE: Urban and Regional Information Systems Association, 1460 Renaissance Drive, Suite 305, Park Ridge, Illinois 60068-1348; Voice (847) 824-6300; Fax (847) 824-6363; E-mail [email protected].

SUBMISSIONS: This publication accepts from authors an exclusive right of first publication to their article plus an accompanying grant of non-exclusive full rights. The publisher requires that full credit for first publication in the URISA Journal is provided in any subsequent electronic or printpublications. For more information, the “Manuscript Submission Guidelines for Refereed Articles” is available on our web site, www.urisa.org, or bycalling (847) 824-6300.

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, 60068-1348; Voice (847) 824-6300;Fax (847) 824-6363; E-mail [email protected].

URISA Journal is published four times a year by the Urban and Regional Information Systems Association.

© 2000 by the Urban and Regional Information Systems Association. Authorization to photocopy items for internal or personal use, or the internalor 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 conjunctiontherewith, nor the exhibition of any materials by any party coincident with the educational event, should be construed as indicating endorsement orapproval of the views presented, the products used, or the materials exhibited by URISA, or by its committees, Special Interest Groups, Chapters, orother commissions.

SUBSCRIPTION RATE: One year: $495 business, libraries, government agencies, and public institutions. Individuals interested in subscriptionsshould contact URISA for membership information.

US ISSN 1045-8077

URISA Journal • Vol. 12, No. 1 • Winter 2000 3

EDITORIAL BOARD

Robert Aangeenbrug Robert LaMacchiaUniversity of South Florida U.S. Bureau of the Census

David Arbeit John McLaughlinOffice of Minnesota Planning University of New Brunswick

Marc Armstrong Gilbert MitchellUniversity of Iowa National Geodetic Survey

Kate Beard Joel MorrisonUniversity of Maine U.S. Bureau of the Census

Richard Brail Harlan OnsrudRutgers University University of Maine

William Craig David PhillipsUniversity of Minnesota University of Virginia

Peter Croswell Carl ReedPlanGraphics, Inc. Carl Reed & Associates

Earl Epstein Mark SallingOhio State University Cleveland State University

Joseph Ferreira Allan SchmidtMassachusetts Institute of Technology Schmidt Associates

Steve French K. Stuart SheaGeorgia Institute of Technology TASC

Lewis Hopkins Larry SugarbakerUniversity of Illinois Washington Department of Natural Resources

William Huxhold Nancy TostaUniversity of Wisconsin-Milwaukee Tosta Enterprises

Charles Kindleberger Peter Van DemarkCity of St. Louis Caliper Corporation

Richard Klosterman Barry WellarUniversity of Akron University of Ottawa

Kenneth Kraemer Peter ZwartUniversity of California-Irvine University of Tasmania


Editorial: Revised Mission andNew Appointments

With our first issue of the year 2000, we are pleased to introduce an expanded mission forthe URISA Journal as well as newly appointed Editors and Article Review Board.

Mission StatementThe URISA Journal is a peer-reviewed journal published both in print and online by

the Urban and Regional Information Systems Association, a non-profit internationalassociation for information technology professionals. The Journal is published quarterlyand adheres to access and copyright policies designed to further the advancement ofknowledge and science. Articles are accepted for publication in the following areas:

Urban and Regional Information Science - sciences which advance research in thespatial and temporal relationships of phenomena in the natural and human-modifiedenvironment, including advances in planning, urban modeling and in environmental,transportation and engineering information systems.

Applications - applied research advances relating to information system develop-ments in areas such as public health, emergency response, crime analysis, marketing,cadastral mapping, vehicle routing, infrastructure development, environmental assess-ment and similar applications.

Social, Organizational, Legal and Economic Sciences - advances in understandingthe social, organizational, institutional, legal, ethical and economic environments affect-ing the design and use of information technologies in urban and regional settings.

Geographic Information Science - tools, techniques and methods for analyzing,displaying, visualizing, and communicating spatial data, including the tools, techniquesand methods of geographic information systems (GIS), spatial statistics, spatial analysisand computer science.

Information and Media Sciences - emerging and related areas pertaining to multi-media, virtual environments, spatial simulation, digital libraries, human computer inter-action, and web-based GIS.

Spatial Data Acquisition and Integration - tools, techniques and methods of geodesy, surveying, photogrammetry, global posi-tioning systems, remote sensing, engineering and computer science to acquire, manage and integrate spatial data.

Geography, Cartography and Cognitive Science - advances in understanding the manner in which people think about and representtheir geographic surroundings, including advances in geography, cartography, cognitive science, computer science and related sciences.

Education - advances relating to the teaching and learning of material in any of the above described areas.As shown on the following page, the Journal has put in place a solid team of academic editors and a newly appointed Article

Review Board (ARB) to appropriately manage the expanded coverage of the Journal. Serving on the ARB requires an applicationprocess and submittal of a vita for evaluation by the editors (see http://www.urisa.org/journal.htm). While all individuals appointedmust have a significant peer-reviewed publication record within the preceding five years, the application process to serve on the ARBis relatively open and democratic. The editors will be adding new members to the ARB as demand for additional reviewers in someareas becomes evident and as appointments expire. We invite you to review at the web site the vitas of the individuals who have beenappointed to date and we think you will agree that their credentials are very impressive.

Other recent changes to the Journal have included institution of a copyright policy that allows authors to retain full but non-exclusive copyright in their works, provision of both paper and online versions of the Journal and implementation of a peer reviewprocess that is now entirely electronic. Combined with newly appointed editorial and review boards that represent the very best of thefield, we believe the URISA Journal is positioned to make lasting and expanding contributions to the advancement of knowledge andscience as we enter the new millennium.

Harlan Onsrud

In this Issue…

Three articles, a special report, anda review of literature are presented in theWinter 2000 issue.

A model for predicting the benefitsof using geographic information systems(GIS) technology is described by StephenR. Gillespie. The multiple regressionequations in the model were developedby the U.S. Geological Survey.

The results of studying GIS imple-mentation by four grassroots organiza-tions are explored by R. E. Sieber toinvestigate alternatives to traditional mod-els of GIS implementation.

Michael J. Greenwald examines or-ganizational and technological support is-sues involved in creatingmulti-jurisdictional GIS by contrasting anearlier federal project with a current re-gional support project.

A special report on critical issues ingeographic information science resultedfrom a National Science Foundation(NSF) workshop on the needs for basicresearch in this area. David M. Mark ed-ited the report that details the recommen-dations participants made to the NSF.

Zorica Nedovic contributes a reviewof current journal literature. The selectedarticles are assigned to one of nine cat-egories. Also contributing to this sectionare Ted Koch and Ken Dueker.

URISA Journal • Vol. 12, No. 1 • Winter 2000 5

URISA Journal Editors

Editor-in-ChiefHarlan Onsrud, Spatial Information Science andEngineering, University of Maine

Managing EditorRobert Berry, URISA Staff

Thematic Editors

Editor-Urban and Regional InformationScience

Lewis Hopkins, Department of Planning,University of Illinois-Champaign/Urbana

Editor-Applications ResearchLyna Wiggins, Department of Planning,Rutgers University

Editor-Social, Organizational, Legal, andEconomic Sciences

Ian Masser, Department of Urban Planningand Management, ITC (Europe)

Editor-Geographic Information ScienceMichael Goodchild, Department of Geography,University of California-Santa Barbara

Editor-Information and Media SciencesMichael Shiffer, Department of Planning,Massachusetts Institute of Technology

Editor-Spatial Data Acquisition and IntegrationGary Hunter, Department of Geomatics,University of Melbourne (Australia)

Editor-Geography, Cartography, andCognitive Science

David Mark, Department of Geography,SUNY-Buffalo

Editor-EducationKaren Kemp, Department of Geography,University of California-Berkeley

Section Editors

Software Review Editor Jay Lee, Geography, Kent State University

Book Review EditorRebecca Sommers, Somers-St. Clair

Literature Review Editor Zorica Nedovic, University of Illinois-Champaign/Urbana

Article Review BoardPeggy Agouris, Department of Spatial Informa-tion Science and Engineering, University of Maine

Michael Batty, Centre for Advanced SpatialAnalysis, University College London

Kate Beard, Department of SpatialInformation Science and Engineering,University of Maine

Yvan Bédard, Centre for Research inGeomatics, Laval University

Barbara P. Buttenfield, Department of Geog-raphy, University of Colorado

Keith C. Clarke, Department of Geography,University of California-Santa Barbara

David Coleman, Department of Geodesy andGeomatics Engineering, University of NewBrunswick

David J. Cowen, Department of Geography,University of South Carolina

Massimo Craglia, Department of Town &Regional Planning, University of Sheffield

William J. Craig, Center for Urban and Re-gional Affairs, University of Minnesota

Robert G. Cromley, Department of Geogra-phy, University of Connecticut

Kenneth J. Dueker, Urban Studies and Plan-ning, Oregon State University

Geoffrey Dutton, Spatial Effects

Max J. Egenhofer, Department of Spatial Informa-tion Science and Engineering, University of Maine

Manfred Ehlers, Geoinformatics and Institutefor Environmental Sciences, University of Vechta

Manfred M. Fischer, Economics, Geography& Geoinformatics, Vienna University of Eco-nomics and Business Administration

Myke Gluck, School of Information Studiesand Geography, Florida State University

Michael Gould, Department of Science,Experimentales Universitat

Daniel A. Griffith, Department of Geography,Syracuse University

Francis J. Harvey, Department of Geography,University of Kentucky

Kingsley E. Haynes, Public Policy and Geog-raphy, George Mason University

Eric J. Heikkila, School of Policy, Planning,and Development, University of SouthernCalifornia

Stephen C. Hirtle, Department of Informa-tion Science and Telecommunications, Univer-sity of Pittsburgh

Richard E. Klosterman, Department of Geog-raphy and Planning, University of Akron

Robert Laurini, Claude Bernard Universityof Lyon

Thomas M. Lillesand, EnvironmentalRemote Sensing Center, University ofWisconsin

Xavier R. Lopez, Oracle Corporation

David Maguire, Environmental Systems Re-search Institute

John McLaughlin, Research and InternationalCooperation, University of New Brunswick

Harvey J. Miller, Department of Geography,University of Utah

Joel L. Morrison, Center for Mapping, OhioState University

Atsuyuki Okabe, Department of Urban En-gineering, University of Tokyo

Jeffrey K. Pinto, School of Business, PennState Erie

Gerard Rushton, Department of Geography,University of Iowa

Bruce D. Spear, Geographic InformationServices Bureau of Transportation Statistics,Washington, D.C.

Jonathan Sperling, Geography Division, U.S.Census Bureau

David J. Unwin, School of Geography,Birkbeck College, London

Stephen J. Ventura, Environmental Studies andSoil Science, University of Wisconsin-Madison

Nancy von Meyer, Fairview Industries,Wisconsin

Barry Wellar, Department of Geography,University of Ottawa

Michael F. Worboys, Department ofComputer Science, Keele University

Mark Your Calendar!○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

4th Annual

IntegratingGIS & CAMAConferenceApril 16-19, 2000Fontainebleau Hilton ResortMiami Beach, FL

37th Annual

URISA Conference& Exhibition

August 19-23, 2000Omni Rosen HotelOrlando, FL

2nd Annual

Street Smart &Address Savvy ConferenceOctober 25-27, 2000Omni Inner Harbor HotelBaltimore, MD w w w . u r i s a . o r g

URISA Journal ■ Gillespie 7

IntroductionAll current and potential users of geographic information system(GIS) technology must deal with the issue of the costs and ben-efits of their activities. It has long been recognized that the onlyjustification for any organization’s expenditures on digital data isthat the data’s benefits exceed their cost (Dickinson & Calkins1988). Nonetheless, accurate data on benefits generated by GIStechnology are rare.

The 1994 Urban and Regional Information Systems Asso-ciation (URISA) Conference dramatically illustrated that the GISuser community recognizes both the importance and the currentpaucity of benefits information. The theme of the conferencewas “Integrating Information and Technology: IT Makes $ense”(Tsui 1994). By this, the conference coordinators meant that thetechnology must be cost effective. Despite this stated objective,GIS management consultant Rebecca Somers, reviewing the con-ference, wrote, “A notable absence was that of any real discussionabout the actual costs and benefits of GIS ... conference attend-ees would expect a range of presentations presenting real figuresand results, and perhaps even some guidelines-something thatwe desperately need, but the dearth of information in this areapersists” (Somers 1994).

The lack of reliable benefits estimates can have a real cost.Failure to adequately quantify potential benefits can lead to un-dervaluing GIS technology in costs/benefits studies designed to

An Empirical Approach To Estimating GIS Benefits

By Stephen R. Gillespie

Abstract: Data on the benefits of using geographic information system (GIS) technology are rare. The U.S. Geological Survey(USGS) has developed a model to predict the benefits of using GIS technology. The USGS model focuses on the complexity of aGIS application as the key factor influencing the level of benefits. Three different aspects of complexity are input to a modelconsisting of a pair of multiple regression equations. The equations explain from one-half to three-fourths of the measuredvariation in GIS benefits and present a powerful tool for improving the quality of GIS costs/benefits studies.

Stephen R. Gillespie is an economist in the Office of Strategic Plan-ning & Analysis of the U.S. Geological Survey. He received his Ph.D.in economics from George Mason University. He came to the Geo-logical Survey in 1989 following 17 years of work at the Bureau ofLabor Statistics. Dr. Gillespie’s primary research interest is the eco-nomic value of geographically referenced data.

justify its implementation or expansion. Too conservative an es-timate of net benefits can cause the delay or cancellation of in-vestment in a technology that might be seen as highly cost effectiveif benefits were measured more thoroughly.

The problem is not fundamentally a theoretical one; theissues involved in accurate benefits measurement are wellknown (Obermeyer 1999). Theoretically, benefits estimatesshould be based on the societal marginal willingness to payfor GIS-provided improvements over the present system(Peterson & Sorg 1987). In the absence of externalities andmonopoly, societal willingness to pay is measured by marketprices. However, it is in the nature of government involve-ment in GIS operation that markets cannot set meaningfulprices for many of the changes. There is extensive literatureon the theoretical valuation of nonpriced and nonpriceablegoods. Contingent valuation studies rely on surveys to deter-mine how much respondents would be willing to pay(Cummings et al. 1986). Travel costs have been used to esti-mate the value of recreational resources (Clawson & Knetsch1966). Hedonic models infer values not directly observablefrom related markets where values are directly observable(Brookshire et al. 1982). Methods include the use of propertyvalues and wage differentials in labor markets (Viscusi 1993).

USGS research published in the Fall 1994 URISA Jour-nal (Gillespie 1994) demonstrates that there are practicaltechniques for measuring benefits that might initially ap-pear to be nonquantifiable. The real difficulty in applyingsuch techniques is that they can be time consuming and ex-pensive. Converting from qualitative to quantitative benefitsmeasurement can easily double or triple the cost of a costs/benefits study. It would be very useful to have a relativelyquick and inexpensive method for making ballpark estimates


of the likely benefits an organization would gain from theuse of GIS technology.

One way to avoid the expensive process of directly measur-ing the benefits of using a GIS is to identify and stress factorsthat contribute to a successful GIS application. Numerous pub-lished studies address the question of how to successfully imple-ment GIS technology in an organization. They concentrate onorganizational factors, such as “selling” the technology to high-level management, involving users, designing effective pilotprojects, and consensually creating a vision for the organization’sGIS (Anderson 1992). The social interactionist approach(Campbell 1999) stresses a focus on the traditions, values, andskill bases of individual organizations to ensure successful exploi-tation of a GIS.

Less common is the identification of factors that influencethe success of particular GIS applications. Aronoff (1989) dis-cusses how the usefulness of existing spatial data (factors suchas correctness, comparability, and consistency) affects the suc-cess of a GIS. The more useful the existing data, the greater thelikelihood that the GIS can be successfully used. The causativelink is cost avoidance; that is, when the existing data are good,the user does not have to spend as much to provide good datainput for the GIS. Ripple (1987) identifies the rate of changefor existing data and the likelihood of legal challenges to deci-sions in which GIS applications were important factors. Thefaster the rate of change, the greater the value of the GIS, thevalue stemming from the relative ease of updating computerfiles. The greater the likelihood of legal challenge, the greaterthe value of the GIS, the value coming from the appearance ofprofessionalism and rigor of GIS outputs. The Bureau of In-dian Affairs (1988) identifies the existence of repetitive work asa key to a successful GIS application. USGS research has ex-tended this early work by creating and applying a comprehen-sive framework for analyzing the factors that influence the valueof GIS technology for particular applications. The resultingmodel greatly simplifies the task of quantifying benefits for abroad range of Federal GIS applications.

General Framework for GIS BenefitsMany different taxonomies of benefits are available for use incosts-benefits studies (Smith & Tomlinson 1992). A particularlyuseful distinction for measuring benefits from the use of GIStechnology is between efficiency benefits and effectiveness ben-efits. Efficiency benefits result when a GIS is used to do a taskpreviously done without a GIS; the same quality of output isproduced, but at lower cost. For example, cut and fill calcula-tions can be made by applying planimetric techniques to con-tour lines on a graphic map or by manipulating digital elevationdata in a GIS. Both methods yield the same results, but a GIS ismuch faster and easier.

Effectiveness benefits result when a GIS is used to improvethe quality of a current output or to produce an output not pre-viously available; that is, the GIS is used to do something that

could not or would not be done without it. For example, a GIScan quickly and easily produce maps showing how the proposedroute for a new road would affect a series of environmentallysensitive resources. Such maps could be manually drafted, butthe process would be so expensive that they probably would notbe prepared. A GIS also can overlay a large number of separateenvironmental themes and calculate an overall impact. When thereare more than just a few overlays, this task simply is not feasibleusing non-GIS techniques.

The level of benefits realized when using a GIS to run anapplication is determined by comparing the cost of using the GISmethod with the cost of using the non-GIS method, and by com-paring the value of the outputs produced by the two methods.

Benefits of GIS = (Value of outputGIS

- Value of outputNON-GIS

) +(Cost

NON-GIS - Cost

GIS)

Pure efficiency benefits and pure effectiveness benefits canbe seen as special cases of this general formula. When the GISoutputs are equivalent to the non-GIS outputs, the first termvanishes, leaving Benefits = (Cost

NON-GIS - Cost

GIS ), or pure effi-

ciency benefits. When the costs of the two methods are the same,the second term vanishes, leaving Benefits = (Value of output

GIS -

Value of outputNON-GIS

), or pure effectiveness benefits.The general formula shows why benefit measurement of a

proposed GIS application is expensive. Of the four terms in theformula, Cost

NON-GIS is the only one for which a government

agency is likely to have reasonably accurate information. Estima-tion of Cost

GIS could require an extensive pilot test. Estimation

of the value of outputs requires identification of users and uses ofthe outputs, impacts of changes in outputs on the users and uses,and dollar valuations of the impacts, none of which is likely to beeasy. Because this is such a daunting task, it is not surprising thatquantitative measurement of GIS benefits is so rare.

Factors Influencing Level of GIS BenefitsUSGS research focuses on the complexity of a GIS application asthe key factor influencing the level of benefits realized from theapplication. Complexity was chosen as the key factor because (1)the theoretical direction of its influence on efficiency and effec-tiveness benefits is clear, (2) it is identified as being an importantGIS success factor in the previously cited studies, and (3) it provedto be a useful and measurable concept in a series of USGS casestudies of specific Federal GIS applications.

There are three aspects to the complexity of an application:1. Input complexity concerns the data themes needed to per-

form the application. It involves such things as the numberand diversity of data themes, the total volume of input data,and the areal extent of the application.

2. Analysis complexity concerns how the data themes are ma-nipulated inside the application. It involves such things asthe maximum number of concurrent overlays, the numberof steps in the analysis, the number of intermediate data


themes created, and the number of potential interactionsbetween data themes.

3. Output complexity concerns the products of the applica-tion. It involves such things as the number of distinct usesfor the outputs and the likelihood that the outputs will beused in adversarial hearings.

Each complexity factor can be expected to influence the levelof efficiency and effectiveness benefits in a predictable way.

Among the measures for input complexity, both the arealextent of the application and the volume of input data are ex-pected to be positively related to the level of efficiency benefits.All other things being equal, the larger the study area or the greaterthe amount of physical data, the greater the manual inputs re-quired. Having greater manual inputs implies a larger potentialfor efficiency benefits by using GIS technology. Both also areexpected to be positively related to the level of effectiveness ben-efits. All other things being equal, the larger the study area, thegreater the value of outputs. A greater volume of input data im-plies a larger information content in the outputs; a greater valueof outputs implies a larger potential for effectiveness benefits.

In fact, it is expected that both of these input complexitymeasures would have a log linear relationship to the level ofGIS benefits. This is because there are economies of scale indealing with inputs, so that a doubling of the volume of in-puts does not double the complexity of the application. Thereare two general types of economies of scale that operate withinput complexity. Both types are illustrated by an applicationto find an optimal route.

For example, assume that the best route must be found forshipping something from point A to point B. The road networkwould be an input to this application. The complexity of theinput would be affected by the level of detail sought about theroad network. If only interstate highways are relevant to the analy-sis, then the input is not very complex. As more levels of detailare needed (for example, primary roads, secondary roads, andunpaved roads) the input becomes more complex. However, thecomplexity does not increase as quickly as does the volume ofinput data. There may be five times as many miles of secondaryroads as primary roads, but their inclusion only raises the inputcomplexity by one level.

Another economy of scale comes about because much of theinput data is not relevant to the problem. For example, mostsecondary roads are clearly not on the optimal route and quicklycan be eliminated from further consideration. Adding the entiresecondary road network could double or triple the volume ofinput data but probably would add only slightly to the volumethat must be seriously considered.

Among the measures for analysis complexity, both the num-ber of concurrent overlays and the number of potential interac-tions between data themes are expected to be positively related tothe level of benefits. All other things being equal, the greater thenumber of themes overlaid, the greater the manual inputs re-quired, and the larger the potential for efficiency benefits. Simi-

larly, the greater the information content in the outputs, the largerthe potential for effectiveness benefits.

The number of data themes overlaid is expected to have alinear relationship with the level of efficiency benefits. There areno economies of scale with analysis complexity, however, it isexpected to have a curvilinear relationship to the level of effec-tiveness benefits. It is true that diminishing returns apply to thesimple addition of data themes. For example, assume that it isnecessary to predict what effect increased logging in a nationalforest would have on an endangered species. Expanded loggingwould create various environmental stresses that could affect theendangered species. To find the single most dangerous stress, onewould examine each stress independently. As more and more sepa-rate stresses were examined, diminishing returns would quicklyset in.

However, the concurrent examination of multiple datathemes also involves the ever-increasing complexity of interac-tion effects. Interaction effects can be very important. For ex-ample, perhaps no one environmental stress would have a seriouseffect on the endangered species, but the cumulative effect ofmany stresses would be fatal. The interaction effects created asmore themes are overlaid could make a major contribution tothe value of the output. The number of interactions betweendata themes increases geometrically as the number of data themesincreases arithmetically.

The measures for output complexity are expected to have alinear relationship to the level of both efficiency benefits and ef-fectiveness benefits. There are no economies of scale with thesemeasures. Increasing any of them is likely to result in a propor-tionate increase in the complexity of the application. Likewise,there are no significant diminishing returns to the number ofdifferent uses for the output or to the probability of the outputbeing used in adversarial hearings. Increasing either of these mea-sures is likely to result in a proportionate increase in the value ofthe output.

It also is likely that there are interaction effects between thedifferent aspects of complexity. An application’s overall complex-ity is more than just the sum of its input, analysis, and outputcomplexity; these aspects are more likely to be multiplicative thanadditive. The impact of overall complexity on the level of GISbenefits is expected to vary depending on the relative strengthsof the three different aspects of complexity.

A Model to Estimate BenefitsThe USGS has linked the complexity factors discussed in theprevious section from a theoretical perspective to the generalframework for GIS benefits to produce a quantitative model forestimating these benefits. The model is specified as a pair of ordi-nary least squares multiple regression equations. Input to themodel is provided by a series of 62 case studies of Federal GISapplications conducted by the USGS in 1990 and 1991 (Gillespie1991). The model estimates efficiency and effectiveness benefitsindependently.


Pure effectiveness benefits (that is, where CostNON-GIS

= CostGIS

) are estimated by the equation

LT = 3.752 + 0.673 INPLEX1 + 0.045 INTERACT + 0.429 OUTPLEX + 3.147 SMALL + residual(3.5) (5.7) (1.6) (2.3) (2.8)

where LT = Natural log of the dollar value of the pure effectiveness benefitsINPLEX1 = Measure of input complexityINTERACT = Measure of analysis complexityOUTPLEX = Measure of output complexitySMALL = Dummy variable reflecting overall complexity of application

The equation has an R2 of 0.592, an F value of 11.250, andis based on 36 observations. The t statistics are in parenthesesbelow each coefficient.

The R2 value means that the equation explains about three-fifths of the measured variation in the level of effectiveness ben-efits across the 36 applications studied. The F statistic tests thehypothesis that all of the coefficients except the intercept are 0.There is less than 1 chance in 10,000 of obtaining an F value thishigh if all of the coefficients are 0. The t statistics test if eachcoefficient individually is equal to 0. All of the variables exceptINTERACT are significant at the 99 percent confidence level.This means that there is less than 1 chance in 100 that the coef-ficient is 0. INTERACT is significant at the 80 percent level.

The dollar estimate of pure effectiveness benefits is found by taking the antilog of the estimated natural log. For example,

when LT = 5, the dollar value = $148;when LT = 7, the dollar value = $1,097;when LT = 9, the dollar value = $8,103.

If the effectiveness benefits are not pure (that is, if CostGIS

≠CostNON-GIS

), then the difference between CostGIS

and CostNON-GIS

must besubtracted from the estimated total. For example, if estimated pure effectiveness benefits = $5,000, Cost

GIS = $2,000, and Cost

NON-GIS

= $500, then estimated net effectiveness benefits are $5,000 - ($2,000 - $500) = $3,500.

Pure efficiency benefits are estimated by the equation

RATIO = 0.477 + 0.100 INPLEX2 - 0.001 INTERACT + 0.051 OUTPLEX + 0.377 SMALL (7.9) (6.5) (-0.4) (4.3) (6.2)

+ 0.232 COST - 0.186 LAND + residual (4.4) (-4.1)

where RATIO = Ratio of efficiency benefits from GIS to manual cost of running the applicationINPLEX2 = Measure of input complexityINTERACT = Measure of analysis complexityOUTPLEX = Measure of output complexitySMALL = Dummy variable reflecting overall complexity of applicationCOST = Dummy variable reflecting cost of performing application with manual methodsLAND = Dummy variable reflecting subject area of application

The equation predicts the natural log of pure effectivenessbenefits, implying that unit changes in the independent variablescause percentage changes in the level of effectiveness benefits.For example, consider the effect of a one-unit increase in theoutput complexity factor of the number of distinct uses. Thelevel of effectiveness benefits increases by the value of the out-puts to the new class of users. Lacking other information, thebest estimate of the value to the new class of users is the meanvalue to the previous classes of users. The increase in the level ofeffectiveness benefits depends on the previous level; the increaseis a constant percentage, not a constant dollar amount. That is,the marginal effect of each of the independent variables on thedollar amount of effectiveness benefits increases with the level ofeffectiveness benefits.


The equation has an R2 of 0.742, an F value of 11.531, andis based on 31 observations. The t statistics are in parenthesesbelow each coefficient. The equation explains about three-quar-ters of the measured variation in the ratio of efficiency benefits tomanual cost across the 31 applications studied.

The efficiency equation has some structural differences fromthe effectiveness equation. Rather than estimating the absolutelevel of efficiency benefits, the equation estimates the fraction ofthe manual cost of running the application that is saved by theuse of GIS technology. Because the manual cost restricts the effi-ciency benefits to a maximum value, manual cost is an impor-tant factor to include in any model. Incorporating the manualcost into the dependent variable eliminates the need to include itas an independent variable. This brings the influences of the othervariables into clearer view.

The two additional dummy variables are included becauseof the above change in the dependent variable. COST flags ap-plications that are neither very expensive nor very inexpensive torun manually. It is expected that applications in the midrange ofmanual cost will tend to save a larger percentage of their manualcost than would be estimated solely on the basis of the values ofthe other variables. This is due to the frequency with which thistype of application is run; more expensive applications tend tobe run less frequently. An agency wouldn’t have to save a verylarge percentage of the manual cost of a less expensive applica-tion to make it valuable to use a GIS. Less expensive applicationsare run very frequently, and the sheer volume makes the totalefficiency benefits large. An agency wouldn’t have to save a verylarge percentage of the manual cost of an expensive applicationto make the use of a GIS valuable. Such applications are so ex-pensive that the efficiency benefits are large in absolute termsanyway. However, an agency does have to save a large percentageof the manual cost of a moderately expensive application to makethe use of a GIS valuable. Such applications cannot be justifiedon the basis of volume (because they are not run very frequently)or on the basis of large absolute savings (because the manual costsare not extremely large), and so they require a larger percentageof savings.

LAND flags applications that are primarily concerned withthe economic value of the land (for example, forestry, soils, waterresources) rather than with the land as the location of other hu-man activity (for example, transportation, emergency prepared-ness, urban planning). It is expected that such applications willtend to save a smaller percentage of their manual cost than wouldbe estimated solely on the basis of the values of the other vari-ables. This is because LAND applications are more expensive torun (both manually and with a GIS) than are non-LAND appli-cations because they are more likely to involve continuous vari-ables (for example, soil conditions change incrementally over ageographic area), and non-LAND applications are more likely toinvolve discrete variables (for example, political units changeabruptly at defined boundaries). The fuzziness of continuousvariables can increase the difficulty of both processing and analy-sis and thus raise the cost of running an application. The higher

level of both types of costs reduces the ratio of efficiency benefitsto manual cost.

All of the variables are significant at the 99 percent levelexcept for INTERACT. The low significance (and negative coef-ficient) for INTERACT also is due to estimating the fraction ofsavings from using a GIS. Because a GIS can handle additionalconcurrent overlays very easily, it was expected that Cost

GIS would

increase very little when analysis complexity increased. This inturn would lead to an increase in the fraction of the manual costsaved by using a GIS. The equation contradicts this expectation.It appears that there is a significant increase in Cost

GIS associated

with an increase in analysis complexity. The explanation for thisis probably that, even though the marginal cost of physically over-laying another data theme is trivial with a GIS, the marginal costof interpreting the results is not trivial. Whether the overlays aredone manually or with a GIS, it is considerably more difficult tointerpret the results of overlaying a larger number of themes.

This does not mean that a GIS is not valuable for handlingincreased analysis complexity; all other things being equal, thelevel of efficiency benefits will increase when the analysis com-plexity of the application increases. However, the effect of in-creased analysis complexity on the ratio of efficiency benefits tomanual cost is indeterminate. That is, there is no firm theoreticalexpectation as to the direction of the effect; the direction be-comes an empirical question.

The dollar estimate of pure efficiency benefits is found bymultiplying the estimated ratio times the manual cost of run-ning the application. For example,

when RATIO = 75.0 and CostNON-GIS

= $1,000, the dollar value = $750;

when RATIO = 80.0 and CostNON-GIS

= $200, the dollar value = $160.

When an application generates both effectiveness benefitsand efficiency benefits, then the estimate of GIS benefits is thesum of the estimates from the two equations.

How to Use the ModelThe GIS benefits estimation model can be a powerful tool for im-proving GIS costs/benefits studies. The model can produce rea-sonable estimates of the likely level of benefits for a fraction of thecost of direct benefits measurement. There are 10 steps to follow:

1. Identify the different types of applications that will be runusing a GIS.

For each type of GIS application:

2. Identify the major source of benefits:a. Efficiency benefits: that is, lower cost to run the appli-

cationb. Effectiveness benefits: that is, higher value output from

the application


c. Both types of benefits are important.3. Estimate how frequently the application will be run.

For each application where efficiency benefits are expected to beimportant:

4. Estimate the information needed to run the equation:a. The values of the complexity variables used in the effi-

ciency equation. (Details on the construction of thesevariables are provided in the appendix.)

b. The cost of running the application using the existing(non-GIS) method.

5. Enter the estimated values for the variables into the equa-tion.■ The result is an estimate of the fraction of Cost

NON-GIS

that will be saved by the use of GIS technology.6. Convert the fraction to dollars, and aggregate across appli-

cations.a. Multiply the fraction by the estimated manual cost.

■ The result is an estimate of the dollar value of GISefficiency benefits for running the application.

b. Multiply the estimated efficiency benefits times the fre-quency with which the application will be run.■ The result is the total dollar value of efficiency sav-

ings for the application.c. Sum these totals across all efficiency applications.

■ The result is the total dollar value of efficiency ben-efits for the use of GIS technology.

For each application where effectiveness benefits are expected tobe important:

7. Estimate the information needed to run the equation.a. The values of the complexity variables used in the ef-

fectiveness equation.b. If it is more expensive to run the application using a

GIS, the amount by which CostGIS

is greater thanCost

NON-GIS.

8. Enter the estimated values for the variables into the equa-tion.■ The result is an estimate of the natural log of the dollar

value of the new or improved outputs the GIS will pro-duce.

9. Convert the estimate to dollars and aggregate across appli-cations.a. Take the antilog of the estimated natural log.

■ The result is the dollar value of pure effectivenessbenefits for running the application.

b. Subtract the excess of CostGIS

over CostNON-GIS

.■ The result is the dollar value of the net effectiveness

benefits for running the application.c. Multiply the net effectiveness benefits times the fre-

quency with which the application will be run.

■ The result is the total dollar value of the net effec-tiveness benefits for the application.

d. Sum these totals across all effectiveness applications.■ The result is the total dollar value of the effective-

ness benefits for the use of GIS technology.10. Verify the reasonableness of the benefit estimates by select-

ing a small number of applications and performing a tradi-tional benefit measurement on them.

The ten-step process produces a suite of outputs that togethertell a compelling story about the potential value of GIS technol-ogy.1. Quantitative estimates of GIS benefits. Impressive on their

own, they can be combined with cost data to produce costs/benefits ratios, net present values, internal rates of return,and project breakeven dates.

2. Case studies of selected applications. These demonstrate inconcrete terms that the estimated benefits are real.

3. Ratio of the dollar values of effectiveness to efficiency ben-efits. This dramatically demonstrates where the value of aGIS truly lies. Typically the ratio will be large, making itclear that GIS is an enabling technology, primarily impor-tant because it helps agencies work better, not because ithelps them work cheaper.

The Montana Geographic Information Council (MGIC)followed these ten steps to analyze GIS implementations in stateand county governments of Montana (McInnis & Blundell 1998).They report that “the model is an excellent tool for assessing thebenefits of GIS installations.” The full report of the MGIC isavailable online at: http://www.mt.gov/isd/groups/eacba/eacba_cba.htm.

ConclusionThe general framework for GIS benefits is broad enough to sup-port many different models for estimating those benefits. TheUSGS tactic of concentrating on complexity factors is not theonly possible approach, but it has proven to be a fruitful one.Within the broad categories of input, analysis, and output com-plexity, there is room for much experimentation concerning whichvariables to include and how to combine them. The specific formsof the complexity variable used in the USGS model work wellfor the particular set of highly diverse Federal GIS applicationsstudied. Alternative formulations of the variables might be moreappropriate for specific types of applications or for applicationsrun by non-Federal agencies. There is much useful work still tobe done. The USGS research provides a firm foundation uponwhich to build a better knowledge of where and why GIS tech-nology is valuable.


Appendix: Construction of IndependentVariables in USGS Model

Input ComplexityInput complexity concerns the data themes needed to performthe application. It is modeled as a combination of the total vol-ume of input data and the areal extent of the application.

The total volume of input data (MB) is measured as thenumber of megabytes of computer memory required to hold thedata used during a single occurrence of the GIS application. Theareal extent of the application (MU) is measured in map units.One map unit is the area that can be represented on a map sheetwith the physical dimensions of a USGS 1:24,000-scale quad-rangle (18 inches by 23 inches). The actual square miles of areaincluded in a map unit varies according to the map scale used.

Both MB and MU have a log-linear relationship to GIS ben-efits. The log forms of these variables are highly correlated and sosubject to multicollinearity. The presence of multicollinearityreduces the precision of the estimates, making it difficult to dis-entangle the relative influences of the affected variables. This prob-lem can be eliminated either by combining the two variables intoa single measure of input complexity, or by dropping one of thevariables from the model.

INPLEX1 (used in the effectiveness equation) is definedas the natural log of MB plus the natural log of MU.INPLEX2 (used in the efficiency equation) is defined as thenatural log of MU.

The alternative treatment of input complexity in the effi-ciency equation is required by the choice of the dependent vari-able. Although MB is positively related to the absolute level ofefficiency benefits, the effect of MB on the ratio of efficiencybenefits to manual cost is theoretically indeterminant. Empiricalresults show a slight negative relationship. The best solution tothe multicollinearity problem, therefore, simply is to drop MBfrom the efficiency equation.

Analysis ComplexityAnalysis complexity concerns how the data themes are manipu-lated inside the application. It is modeled by the number of po-tential interactions between data themes. If MAX is the maximumnumber of distinct data themes overlaid at any one step in theanalysis, then the number of potential interactions is given bythe formula (MAX2 - MAX) / 2.

Output ComplexityOutput complexity concerns the products of the application. Itis modeled as a combination of the number of distinct uses forthe outputs and the likelihood that the outputs will be used inadversarial hearings.

Variety of uses (VU) is measured as the total number of sepa-rate concerns that must be kept in mind by the GIS staff whenrunning the application, and that must be addressed by the out-

put of the application. It is closely related to the number of spe-cial interest groups that will be watching the agency’s actions.

The likelihood of use in adversarial hearings (AH) is mea-sured by the probability that the output from a typical singleoccurrence of the GIS application will be used to support theposition of one (or both) sides in a formal adversarial setting.This could be in a legal setting (before a jury, judge, or regula-tory commission) where the decision reached is legally bind-ing. It could be in an administrative setting (at a public hearingof some sort) where the expressed opinions are advisory only.“Adversarial hearings” definitely refers to something more struc-tured than internal agency disputes over the most appropriatemanagement policy. It refers to situations where, after the agencyhas reached a decision, some group external to the agency chal-lenges that decision in some structured setting. “Adversarial”implies controversy; there are winners and losers. Tempers arelikely to rise. This does not refer to mere informational presen-tations, but to meetings where at least some of the participantshave staked out opposing positions.

Both VU and AH have a linear relationship to GIS benefits.The two variables are highly correlated and so subject tomulticollinearity. The problem is eliminated by combining thetwo variables into a single measure of output complexity.

OUTPLEX = (VU / 3) + (AH / 25). The values for thedenominators weight the two variables approximately equally.

Overall ComplexityThe overall complexity of the application (that is, interactioneffects between the three aspects of complexity) is measured bythe dummy variable SMALL. SMALL has a value of 1 when allaspects of complexity are low, and a value of 0 otherwise.

Operationally, SMALL is determined by appropriatelyweighting and then summing the five factors used to model thecomplexity factors.■ MU and MB are weighted by taking the natural log of the

raw value, and rounding up to 0 when the natural log is lessthan 0.

■ MAX and VU are weighted by dividing the raw value by 3and rounding up to the next highest integer, with a maxi-mum value of 4.

■ AH is weighted as follows:

Raw percentage = 0 (never used): weighted value = 0Raw percentage = 1 to 49 (sometimes used): weighted value = 1Raw percentage = 50 to 99 (often used): weighted value = 2Raw percentage = 100 (always used): weighted value = 3

In the effectiveness equation, SMALL=1 when the sum ofthe weighted factors is less than 6. In the efficiency equation,SMALL=1 when the sum of the weighted factors is less than 5.There are natural breaks in the data from the USGS case studiesat these points. In addition, these values make sense intuitively asreasonable definitions of a “simple” GIS application generatingeach type of benefit.


Level of Manual CostThe dummy variable COST has a value of 1 when the cost ofperforming a single occurrence of a GIS application using non-GIS methods is between $20,000 and $50,000, and a value of 0otherwise.

Subject Area of ApplicationThe dummy variable LAND has a value of 1 when the primaryfunction of the GIS application is in the areas of agriculture, fishand wildlife management, forestry, land management, soils, orwater resources. LAND has a value of 0 when the primary func-tion of the GIS application is in the areas of commerce and eco-nomic development, defense law enforcement and emergencypreparedness, energy and mineral management, environmentalprotection, geological surveys, library and academic research, parksand recreation, taxation and revenue, transportation, and urbanand regional planning.

References

Anderson, Carrie S. 1992. “GIS Development Process: A Proac-tive Approach to the Introduction of GIS Technology.” GIS/LIS Proceedings, pp. 1-10.

Aronoff, Stan. 1989. Geographic Information Systems: A Manage-ment Perspective. WDL Publications, Ottawa, Canada.

Brookshire, D.S., M.S. Thayer, W.D. Schulze, and R.C. d’Arge.1982. “Valuing Public Goods: A Comparison of Survey andHedonic Approaches.” The American Economic Review, 72,pp. 165-177.

Bureau of Indian Affairs. 1988. Final Report on Cost and BenefitAnalysis of Geographic Information System Implementation toBureau of Indian Affairs.

Campbell, H.J. 1999. “Institutional Consequences of the Use ofGIS.” Longley, Paul A., Michael F. Goodchild, David J.Maguire, and David W. Rhind, eds. Geographical Informa-tion Systems, pp. 621-631.

Clawson, Marion and Jack Knetsch. 1966. Economics of OutdoorRecreation. Johns Hopkins University Press for Resources forthe Future, Baltimore, MD.

Cummings, Ronald G., David S. Brookshire, and William D.Schulze. 1986. Valuing Environmental Goods: An Assessmentof the Contingent Valuation Method. Rowman and Littlefield,Totowa, NJ.

Dickinson, Holly J. and Hugh W. Calkins. 1988. “The EconomicEvaluation of Implementing a GIS.” International Journal ofGeographical Information Systems, 2:4, pp. 307-327.

Gillespie, S. 1991. “Measuring the Benefits of GIS Use.” Techni-cal Papers, 1991 ACSM-ASPRS Fall Convention, pp. 84-94.

Gillespie, S. 1994. “Measuring the Benefits of GIS Use: Two Trans-portation Case Studies.” URISA Journal, 6:2, pp. 62-67.

McInnis, Logan and Stuart Blundell.1998. Analysis of GeographicInformation Systems (GIS) Implementations in State andCounty Governments of Montana, http://www.mt.gov/isd/groups/eacba/eacba_cba.htm.

Obermeyer, N.J. 1999. “Measuring the Benefits and Costs ofGIS.” Longley, Paul A., Michael F. Goodchild, David J.Maguire, and David W. Rhind, eds. Geographical Informa-tion Systems, pp. 601-610.

Peterson, George L. and Cindy F. Sorg. 1987. Toward the Mea-surement of Total Economic Value. USDA Forest Service Gen-eral Technical Report RM-148.

Ripple, William J., editor. 1987. Geographic Information Systemsfor Resource Management. American Society for Photogram-metry and Remote Sensing and American Congress on Sur-veying and Mapping.

Smith, Douglas A. and Roger F. Tomlinson. 1992. “AssessingCosts and Benefits of Geographical Information Systems:Methodological and Implementation Issues.” InternationalJournal of Geographical Information Systems, 6:3, pp. 247-256.

Somers, Rebecca. 1994. “URISA ‘94: Trying To Keep Up.”GeoInfoSystems, October 1994, pp. 20-22.

Tsui, Mary. 1994. “URISA ‘94: A Conference Overview.”GeoInfoSystems, July 1994, pp. 42-45.

Viscusi, W. Kip. 1993. “The Value of Risks to Life and Health.”Journal of Economic Literature, 31:4, pp. 1912-1946.

www.urisa.orgC H E C K I T O U T !

URISA Journal ■ Sieber 15

IntroductionIncreasingly nonprofits are following the lead of public agenciesand private industry by implementing a GIS. They are drawn toa GIS because it can combine large amounts of data from dispar-ate sources and on different media, order them into layers orthemes, and analyze or display various relationships. The greaterpromise of the technology, to nonprofits such as grassroots orga-nizations (GROs) and nongovernmental organizations (NGOs)1,is that it may assist in influencing public policy, through the so-phistication of analysis and presentation of powerful images. Withincreasing visibility that successful GIS adoption is possible (e.g.,see the growing presence of GROs/NGOs in the annual mapbooks produced by Environmental Sciences Resource Institute[ESRI]), GIS skills in the grassroots are being viewed as usefuland even indispensable (Aberley 1993).

The potential of geographic information systems to empowerthe GROs has led to the emergence of supportive institutionalstructures, such as vendor foundations and technological assis-tance programs (Sawicki and Craig 1996, Barndt 1998, Leitneret al. 1998). At the same time that studies explore the process ofGIS diffusion to GROs by external organizations (Leitner et al.1998, Sawicki and Peterman 1998), a growing debate questionsthe appropriateness of geographic information systems in thegrassroots: whether it empowers or marginalizes GRO impact(Pickles 1995, see also http://www.nciga.ucsb.edu/varenius/ppgis/papers/index.html). However, these studies and arguments ig-nore the process of GIS implementation within the GRO itself.

GIS Implementation in the Grassroots

R. E. Sieber

As increasing numbers of grassroots organizations use geographic information system (GIS) programs, the debate grows onwhether or how they should adopt the technology. However, we know little about the GIS implementation processes within theseorganizations. This paper presents the results of five-year case study research of GIS implementation patterns by grassrootsconservation organizations in northern California. This paper investigates alternatives that four cases have found to traditionalmodels of GIS implementation and factors of successful usage. It also explores how these strategies might extend the traditionalviews of implementation and inform other institutional users, such as local governments, with innovations in GIS implementa-tion and use.

This paper describes the processes, requisite resources, andorganizational characteristics of grassroots GIS implementation.It considers how GRO implementation mirrors or contrasts withthe existing implementation literature, which to date has beenconducted solely in the public sector. The nonprofit and grassrootsliteratures suggest that experiences in public sector agencies, suchas municipalities, provide a poor barometer for GRO activities;their literatures are scanned for likely approaches to GIS imple-mentation. A case study methodology allows for the framing ofdistinct implementation models and a list of implementation fac-tors specific to GROs. Four conservation GROs from northernCalifornia are presented in the context of these models and fac-tors. Their experiences allow us to reassess existing implementa-tion approaches and provide valuable lessons to others in theirimplementation of a GIS.

GIS Implementation in GovernmentGovernmental agencies have discovered that how a GIS is imple-mented influences its successful usage (Onsrud and Pinto 1993).Implementation, defined as the “activities necessary to put theinnovation into practice and incorporate it into existing and de-veloping operations” (Onsrud and Pinto 1993: 21), encompassesa complex array of organizational, contextual, and technical is-sues that spans the initial hardware installation to the desired endof routinization of tasks involving the GIS (Azad 1993, Rogers1993). Although implementation involves a considerable degreeof technical issues, such as system design and installation, thetechnical issues are equaled or surpassed by organizational issues(Croswell 1991). GIS implementation tends to alter the organi-zation substantively because it is expensive and complex. De-pending on the scope of the system and the promisedorganization-wide utility, implementation frequently crosses de-

R. E. Sieber is an assistant professor at McGill University, jointlyappointed to the department of Geography and the School of theEnvironment. There she teaches GIS and researches the use of GISand other computer technologies in social movement organizations.


partmental/subunit lines and alters power relations as the con-trol of information changes (described in Pinto and Onsrud1995). In the past, many systems have failed because manage-ment focuses on technical installation and does not realize thescope of change and additional expenses for implementation.

Effects of Organizational Structure and CultureDuring the dynamic process of any innovation, an organizationadapts the technology to suit its needs; however, the organiza-tion also is modified by the technology. Certainly, innovationmust be sufficiently flexible to fit the organization. Within theorganization, GIS implementation suggests that, for the effectiveadoption of an innovation—that is, a nonrountinized technol-ogy—an organization also must be sufficiently flexible to re-in-vent itself. Simply laying an innovation on top of old processeswill not induce the implementation to succeed (Markus 1983).Organizational structures must be able to withstand a state ofinstability—to “unfreeze” while integrating the technology andthen “freeze” into an adapted structure (Kwon and Zmud 1987).This limits the implementation capacity of organizations withstructures that are rigid, extremely hierarchical, highly central-ized or formalized, multi-layered (with management), and com-plex (Rogers 1995). Conversely, sufficient structure is needed forcommunications channels as well as for resolution of power andcontrol conflicts (Campbell 1991).

Because an innovation is inherently destabilizing (Rogers1995), organizations subject to high levels of internal and exter-nal uncertainty (a city government with a contentious electedcouncil, for example) will suffer under GIS introduction(Campbell 1991). Recognizing the shockwaves induced by in-novation, Budic (1997) emphasized internal stability as the nec-essary structural element for effective GIS implementation in: 1)resources (in terms of staff size, tenure, and organizational bud-gets); 2) degree of centralization, complexity, and formalization(in varying degrees dependent on the stage of adoption); and 3)internal politics. Budic (1997) added that these elements vary inimportance—some may even become detrimental—dependingupon the stage of the implementation process. This rather con-tradictory message—an organization need for flexibility/“instability” and a simultaneous need for stability—plus the time-dependent application of structural elements, alerts us to the needto more fully understand the role of organization structure in theprocess of implementation. Implementation researchers seek buthave yet to find the equilibrium point, primarily because the lim-ited amount of time available for organizational research mayallow the capture of only one implementation stage.

Structure presents one dimension for analyzing organizationalimpacts and impediments; the organizational culture reveals yetanother clue to the dynamics of GIS implementation. Whereasorganizational structure comprises the formal arrangement of theorganization, such as mission statements and organizational charts,organizational culture comprises the informal beliefs and valuesinherent in organizational units and how they shape attitudesand practices (Goldhaber 1990, p.71). Regardless of the

organization’s intent, an individual department’s culture may notcomplement the diffusion of GIS innovation. For instance, de-partmental values and goals, cognitive styles, levels of commit-ment, previous computing/innovation experiences, and the styleof leadership may inhibit successful incorporation into depart-mental practices (Croswell 1991, Onsrud and Pinto 1993). Su-perimposed on these are what Feldman (1993) considereddistinctly American cultural icons that can enhance or impedeorganizational change: idealism, conformity, and selfishness. Ide-alism and selfishness, for instance, may nurture innovative be-havior yet breed mistrust and territoriality.

Effects of Employee PerceptionsLike many innovations in information technology, a GIS canresult in substantial change to job descriptions and organizationalunits. People naturally are apprehensive of organizational change(Kanter 1983). GIS implementation further exacerbates this anxi-ety in government because so many departments collect geo-graphic data and are thus impacted by an organization-wideimplementation (Perkins 1990; Antenucci et al. 1991). It shouldbe noted that people often have reason to fear change, becauseauthority/status, control over work, career opportunities, and jobsatisfaction can be diminished. These dynamics may or may notimprove with the addition of GIS responsibilities (Crain 1990).Employee attitudes toward GIS implementation run the gamutfrom the desire to sabotage adoption, through resistance, igno-rance, apathy, passive acceptance, active participation, and facili-tation, to leadership and even evangelism (Brown and Friedley1988, Croswell 1991). Often, success can hinge on the perceivedadvantage a GIS may bestow upon the individual as well as thealignment of personal values and experiences with the technol-ogy (Budic and Godschalk 1996). Therefore, employee attitudesare more complex than just ensuring “happy” staff. Employeeacceptance of the technology has been correlated to successfulusage (Igbaria and Nachman 1990); conversely, employee re-sistance has suspended the most technically advanced system(Er 1989).

Factors of Successful ImplementationGiven the impacts of structure and perception, just how doesone implement a GIS? Overall, researchers have concluded thatsuccessful implementation of a GIS in local government is de-pendent on a number of well-documented factors. They include:1) evaluation of user needs; 2) long-term upper management com-mitment to the project; 3) sufficient allocation of resources; 4)adequate staffing; 5) timely and sufficient training; 6) someone,called a “GIS champion,” who will shepherd the project from ac-quisition to use; and 7) organizational communication or diffu-sion to smooth the transition to full utilization (Croswell 1991,Huxhold 1991, Azad 1993, Onsrud and Pinto 1993, Budic 1994).

Clearly, management plays a key role in achieving GIS adop-tion. In addition to commitment to the system and the securingof resources, implementation depends on the upper echelon’sability to articulate organizational goals for the GIS, including


management of possibly conflicting departmental goals (Budicand Godschalk 1994, Huxhold and Levinsohn 1995). Enthusi-asm for a GIS must be balanced with realistic expectations(Antenucci et al. 1991), particularly when vendors are willing tomanipulate the image of the product to fit any desire. Moreover,management must involve users of innovation from the begin-ning (Cheney and Dickson 1982). Management may nurturestaff to engage in innovation and risk-taking, through rewardsand recognition; alternately, the structure may espouse change,but in reality organizations may not recognize new responsibili-ties and reward changes in employees’ job descriptions.

GIS champions function as the innovators, the first wave ofdiffusion in an organization. They introduce the technology, pro-mote its use throughout the organization, gather informationabout it, and join technology user groups. They likely will createthe first project with the software. Paradoxically, these same cham-pions, who provide a service to the organization (because theiractivities typically exceed the bounds of their jobs), are generallyin conflict with the norms of the organization and may symbol-ize to other organization members “social deviants” (Rogers 1995).For example, by using technical (i.e., ill-understood) jargon cham-pions may alienate themselves from their own organizations.Huxhold and Levinsohn (1995) contended that the role of cham-pion is unnecessary with increased GIS diffusion, yet the diffu-sion literature suggests that for sectors of organizations with littleexposure to the technology, the need remains.

GIS Implementation within GrassrootsOrganizationsThe literature on GIS implementation suggests that these ge-neric factors, employee perceptions, and structural elementsshould extend to all organizations, regardless of function or size.The nonprofit literature shows that similarities abound; resourcesand politics will affect both government and GROs. Much of theliterature on the smaller GROs challenges traditional implemen-tation approaches, which are based on larger and more stableorganizations.

Similarities to Governmental OrganizationsThe literature on the general use of computers by nonprofit groupssuggests that GIS implementation will mirror those observed inmunicipalities. Rubinyi (1985), in his study of computer use incommunity-based organizations, identified a mixture of techni-cal and organizational barriers to extensive use, including unco-ordinated efforts, lack of time, technical problems, and budgetconstraints. Schoech (1982) found that, historically, the cost ofequipment has been a major limitation to the full implementa-tion of information technologies (ITs).

Schoech (1982) identified numerous technical barriers toIT adoption by nonprofits but still traced successful develop-ment back to the existence of organizational elements: a key leader,executive involvement and commitment of resources, policies andprocedures geared toward easy installation, and user participa-

tion throughout the process. Prospective adopters of computingtechnology are advised by nonprofit consultants and other tech-nical experts to put someone in charge (ideally someone whoexpresses a “passion for computers”) before system acquisition,remember the organizational mission (and that you work forpeople/clients), seek expert advice (including resources offeredby universities, such as student projects, interns, or faculty assis-tance), do not minimize maintenance and support, be cautiousof donations that can be costly to maintain, build a system fromthe software and the data, plan for the costs, conduct a user needsassessment, and understand the extra demands on staff (Cohenand Perrault 1991, Smith 1991, Nonprofit World 1994).

Differences to Governmental OrganizationsNotwithstanding similarities to governmental experience withgeographic information systems, GROs likely will face their ownset of challenges in implementation. The primary difference liesin the much more fragile nature of the grassroots group, both inits ability to attract and retain resources (Gittell 1980, Wolch1990, McCullough 1991) and its capacity to hold together a manytimes loosely knit assemblage of individuals with diverse goalsand varying strategies to accomplish them (Crowfoot andWondolleck 1990). This instability may inhibit the adoption ofa new technology. As one nonprofit consultant reported inRubinyi (1985, p. 92): “When you’re living at the edge of exist-ence, innovation is not a top priority.”

Also their role in society is different. GROs are unlikely tooperate with mandates or require accountability. They generallydo not deliver public services. This looser arrangement, in termsof structures (decentralized, informal, participatory) and culture(receptive to innovation), actually might promote innovationwithin GROs. Indeed, Wolch (1990) found flexibility to be theprimary reason why governments contract out to nonprofits,because the latter are viewed as efficient alternatives to rigid bu-reaucracies. However, excessive organizational flexibility also mayconstrain adoption. Organizations that are very democratic indecision making may lack sufficient structure upon which to es-tablish procedures/rules and communications channels for inte-grating an organization-wide sophisticated information system.

Organizational culture, as witnessed in governmental de-partments, can be antithetical to adoption. GROs may ex-hibit a culture dramatically more hostile to innovation.Among social movements, the conservation movement has along history of anti-progress, anti-capitalism, and anti-tech-nology sentiment. This attitude is expressed by some mod-ern-day environmentalists, such as Earth First!, the deepecology movement, and even among members of the GIS-using Wildlands Project (Manes 1990, Coveny 1992, Snow1992).2 Many members of these organizations reject tech-nology because it defines and then consigns the environmentto an exploitable resource; accordingly technology must beopposed in order to regain the balance with nature. Frans(1993) observed “ideological antipathies” to computers inhuman services nonprofits; that is, some staff found the use


of computers incompatible with the humane provision ofservices to disadvantaged individuals.

Staff provides another flashpoint. Part of GROs’ fragilityresults from their heavy reliance on the support and skills pro-vided by volunteers and by underpaid staff. It has long been knownthat volunteers bring with them varying levels of commitmentand quality of work (Schindler-Rainman and Lippit 1975, Cliftonand Dahms 1980). For GIS work, even high levels of commit-ment are insufficient; quality and continuity of work are essen-tial. Because a GIS is only as accurate as the data entered—onthe scale of most natural resource maps, misplacing a line byone-eighth inch might misclassify data by one-half mile—gooddata-entry personnel are crucial. Even if digital spatial data isprocured from other sources, technical knowledge is required touse the data properly in a spatial analysis.

Staff-led GROs, irrespective of size, tend to attain greaterinstitutional stability than their all-volunteer counterparts (Snow1992). Attracting expert staff still may be difficult. A prime rea-son lies in the fact that the nonprofit sector as a whole has laggedfar behind other sectors in its ability to compensate staff (Barbeitoet al. 1998). Expertise in computers enhances employability, whichmay mean their esteem inside the nonprofit may soar (NonprofitWorld 1994), but so will their attractiveness to other higher-pay-ing organizations. Conservation organizations already experiencea high turnover rate (Snow 1992); GIS may exacerbate that rate.Reported the vice president of the NGO, Conservation Interna-tional: “It’s been my experience that as soon as we trained some-one in the GIS and they became fairly good at it, that personwould be offered a salary three times higher by someone in theprivate sector.” (Specht 1996, p. 43.)

MethodologyTo compare and contrast the experiences of GROs relative totraditional implementation strategies, this research employs a casestudy methodology. Researchers consider a case study methodol-ogy to be the richest way to examine the nature and use of asophisticated technology that is implemented within a complexmedium (Yin, Bateman, and Moore 1985, Lee 1989) and ideallysuited for GIS implementation (Onsrud and Pinto 1991, Onsrud,Pinto, and Azad 1992). The intricate combination of environ-mental activism and innovative GIS applications produces a de-tailed model that can best be analyzed with this type ofmethodology. A scan of the public participation GIS literature(e.g., see the Cartography and GIS special issue on Public Partici-pation GIS 1998, 25 (2)) shows that case studies are the pre-ferred approach to analyze GIS in GROs.

To identify a sufficiently large pool of cases and to providecontrols on political issues and group character, the research fo-cused on a specific geographic area: northern California (splithorizontally approximately at San Luis Obispo). Northern Cali-fornia was chosen for four reasons: 1) California, especially thenorthern portion, is a state at the forefront of environmental policyand activism (Benenson 1990, Fulton 1991, Hall and Kerr 1991,Walton 1992); 2) California is an innovative state with regard to

implementing geographic information systems and other infor-mation technologies (Sprecher 1994); 3) because it has beenshown that environmentalists, on the whole, quickly adopt—ifnot surpass—the technology utilized by influential governmentor target agencies (Hays 1987), there likely will be higher GISadoption rates among environmentalists in this region; and 4)preliminary research (Sieber 1997b) with environmental organi-zations indicates that a sufficient number of GIS adopters existsin this geographic area to conduct case study work.

This last point suggests extensive GIS use in the grassroots.Survey research (Sieber 1997b) revealed that 20 percent of envi-ronmental/conservation nonprofits surveyed in a random samplemail survey used or accessed GIS. GIS was adopted despite sig-nificant resource scarcity; 60 percent had no paid staff. Estimatesof GROs range from 12,000 to 15,000 environmental and con-servation groups in the United States (Snow 1992, Wikle 1995),which implies large numbers of GROs using GIS software. Thisadds a fifth reason. Comparison of adoption rates with othergrassroots GIS users, such as community development corpora-tions (Obermeyer 1998), suggests that environmentalists and con-servationists play an early adopter role in GROs’ GIS diffusion.The study of early adopters is important because they represent abellwether of the use of innovation in similar groups (Rogers1995).

To further refine the selection of cases, a set of expert inter-views was conducted at the ESRI Users Conference in 1994, where25 of the approximately 30 conservation GROs present at themeeting were interviewed. These individuals were asked to iden-tify organizational characteristics and requisite resources of suc-cessful grassroots GIS adopters. This was followed by consultationin 1994 and 1995 with representatives from GIS vendors, pri-vate firms, academics, conservation scientists, and GROs (somenew and some from the original set). These interviews identifiedfour best-practice cases of successful GIS implementation.

Successful implementation has been defined by Clapp et al.(1989) and subsequently used in other work (Budic 1994, Sieber1997a) as a hierarchy of benefits that build upon each other.Benefits range from operational efficiencies, such as increasedcartographic capacity; operational effectiveness, such as improvedinformation access, program effectiveness such as augmenteddecision making; and contribution to well-being, such as thedelivery of social justice. Successful implementation was gaugedby interviewing case study participants on their satisfaction ofthe system; user satisfaction is correlated with successful usage(Igbaria and Nachman 1990). Overall, respondents were verysatisfied with the benefits of the system (reported in Sieber 1997a),and reported the greatest benefit to be the highest, the contribu-tion of GIS to social change.

Onsrud, Pinto, and Azad (1992) reported that case studymethodology often is criticized for its apparent lack of rigor, hy-pothesis generation capacity, cross comparability, validity, andreplication. However, they and others (Lee 1989, Yin 1989) dem-onstrated that rigor can be achieved through the use of formal-ized measurement protocols. Therefore, the case study


methodology combined observations structured with a case studyprotocol, interview instruments for open-ended interviews withkey personnel, and pattern matching (discussed below) as com-ponents to fortify and triangulate each case study. Hypotheseswere generated and tested through pattern matching. At the sametime, flexibility was retained to balance rigor with the explor-atory nature of the research.

Approximately one week was spent on-site in each of thefour case studies at the end of 1995. This was followed by “progressreport” phone or in-person interviews at the end of 1996 and1997. Instead of a producing a snapshot of these organizations,this approach provided a critical longitudinal view of the fragileand constantly evolving GROs.

Models of ImplementationAccording to Lee (1989), as reported in Onsrud, Pinto, andAzad (1992), generalization in case study methodology isachieved by repeatedly testing theory across a range of circum-stances. One investigates groups that have similar goals but dis-similar attributes, groups that express “natural variations in thereal world” (Onsrud, Pinto and Azad 1992, p. 35). The selectedcases hold successful GIS implementation as a common goalbut employ different tactics to achieve that implementation. Awide spread of strategies allowed for a stringent test of tradi-tional successful implementation factors and a way to extendcase study experiences to a larger community.

The expert interviews revealed that alternate models of imple-mentation exist in the GROs compared to organizations docu-mented in the literature. These counter the traditional model ofimplementation, which implies that hardware, software, trainedtechnical staff, and data are internal to the organization. Such anin-house model requires considerable financial investment andmay be cost-prohibitive to small organizations. The models also

challenge the notion that implementation assumes some thresh-old of system usage by the organization. A GRO may incorpo-rate a GIS into routine activities without owning the technology.

The key points and differences among four models that wereidentified by the expert interviews are highlighted (Table 1).

Wants GIS. This model reflects an organization’s desire tohave the representational and analytic capacity of the GIS in-house. In that sense, the model represents the traditional modelof GIS implementation: the user agency acquires and maintainsmost of the hardware, software, and data. Paid staff generallymanages the system. An in-house model offers greater controlthan outsourcing (see below) over the design and timely deliveryof analysis and output, and is very important when the technol-ogy is considered a strategic function (Nam et al. 1996). Theindividual(s) who develops the system is a paid or volunteer staffmember. The predominant “end-user” of GIS services is withinthe organization.

Wants Map. The second model reflects a desire to possessGIS output (and possibly limited analytic or thematic mappingcapacity) but neither the hardware/system nor the technical ex-pertise to maintain a system. The predominant “end-user” of GISservices is within the organization.

The prime difference between this and the previous modelis that this type of organization outsources to one or more exter-nal contractors for its GIS needs, including data, system, trainedpersonnel, and expertise. Outsourcing represents an attractivealternative to organizations for several reasons: When employeeskills are insufficient or scarce, a technology is rapidly changingto make it cost-prohibitive to keep up, or the external environ-ment is uncertain and forces a firm to concentrate on its corefunctions (Slaughter and Ang 1996). Historically, environmen-talists have relied on outside experts to conduct scientific analy-ses and create environmental models (Hays 1987, Gottlieb 1993).

Table 1 GIS implementation model types

Model Name Description Technical Details GIS Developer Location End-user

Wants GIS Wants presentation/ Owns/accesses hardware, In organization or Members/staffanalytic capacity software, data, organizational subset of organizationin-house trained staff

Wants Map Wants output, Outsources for all Outside consultant Members/stafflimited analysis system components of organization

Wants Consortium Wants to share Owns/accesses hardware, In member Members/staffsystem costs OR software, data, trained staff organization of organization,Wants to operate outside clientsas service bureau

Wants Independence Individual wants Member owns/accesses Individual member Members/staff ofcomplete capacity hardware, software, data, of organization organization,

is trained Individual


Snow (1992, p. 25) noted: “The smaller [conservation] groupsoften increase their technical firepower by reaching outside: Theyrecruit volunteers or paid consultants who are the same kinds ofspecialists employed by the larger groups.” By outsourcing forGIS services, both large and small groups likely will benefit fromthe additional and timely expertise, while remaining focused ontheir core functions of activism.

Wants Consortium. A GRO in this model operates predomi-nantly as a service center to support the GIS needs of other orga-nizations. The model may take several forms and serve differentend-users: existing solely to support other GROs within a for-malized structure, acting as a third-party nonprofit consultingfirm seeking clients for its GIS services, or functioning as a tech-nology assistance center or a data center. The organization mayevolve into this role, spin off from another organization, or becreated precisely for this GIS service. GIS staffing, digital data,and system equipment reside within this organization. Servicecenters allow user GROs to pool resources (including expertiseand data), feel safer about their data (because it is off-site), main-tain computing support, and build capacity to keep up with thelatest versions of software (Nonprofit World 1990).

Conservation-specific service centers are emerging as a domi-nant model for conservationists (Ferber 1992, Specht 1996, Con-servation GIS Consortium 1998, CTSP 1999). GROs shouldbenefit as GIS resources and skills would be tailored to theirneeds—according to Rogers (1995), a reciprocal fitness of tech-nology to the organization—in an affordable and more distrib-uted manner. The greatest beneficiaries would be the smallestGROs that might otherwise lack the skills and financial resourcesto adopt GIS. Conservationists also may view service centers asan alternative to individual organization use—viewed as inher-ently “undemocratic”—because only a few conservation organi-zations might afford a full GIS (Ferber 1992).

Wants Independence. The last model refers to a sole indi-vidual who is an expert in GIS technical issues. The individualowns most, if not all of the hardware and software and can ob-tain/enter data (or already may possess much of the data). Farfrom an isolated instance, expert interviews revealed five of theinterviewees as this type of “environmental entrepreneur,” whowish to make a vocation out of an avocation. The individual mayjoin an existing organization (or leave and join another) or createa new organization to ensure personal compatibility with a mis-sion. The individual can best be described as an in-house techni-cal consultant, with ties to a mission if not an organization, wholikely requires reimbursement for GIS skills. The predominant“end-user” of GIS services is the current organization and theindividual who builds a portfolio of equipment and products.

Pattern MatchingTo compare and contrast GRO implementation with government,each of these models was tested against the factors of implemen-tation. These diverse models test the assumption that factors rep-resent universals across organizational type (e.g., nonprofit andgovernment) and implementation strategy. This was effected

through a technique called pattern matching. Pattern matchingalso allowed me to establish operational implementation andensure consistency among cases.

GIS pattern matching compares observed factors of imple-mentation to the theory of GIS implementation (Lee 1989,Onsrud, Pinto, and Azad 1992). As an example, if upper man-agement commitment is important to effective implementation,one can assign a threshold amount of necessary commitment (e.g.,50 percent or more of the board members express support) andthen observe or interview for that behavior in the case. If thesame degree of commitment is observed, the pattern is effectivelymatched. Pattern matching in this research serves two purposes.First, it determines whether these effective GIS-using GROsmatch the implementation patterns of documented success sto-ries (in government). Second, the method offers a way to testrival explanations for successful implementation in these cases.

The base pattern set was drawn from the review of the GISimplementation, management information system (MIS), andnonprofit literature. Possible values for the matches (+ for match, -for no match, or m for mixed) were assigned to items in the set.This follows other case study research of GIS implementation(Budic 1994, Azad 1998). Direct observation, group member in-terviews, and document examination for comparisons of cases tothe set of theoretical propositions was then utilized. Results werecompared to determine whether they matched existing conditionsor established rival patterns. Factors are shown in Table 2.

CasesThe following briefly describes each case and furnishes some ex-amples of GROs’ response to specific factors of GIS implemen-tation.

Case 1: Wants GISThe Greenbelt Alliance, in San Francisco, is a moderate-sizedmetropolitan nonprofit dedicated to preserving open space inthe nine-county Bay area. Formed more than 20 years ago, itemploys 13 staff, spread among 3 field offices. This GRO buildson its already substantial resource base of computing, funding(including support from major corporations), and passionate staffand volunteers to operate GIS programs.

The predominant application has been the biennial nine-county Greenbelt Mapping Assessment Program (GMAP), iden-tifying “Open Space At Risk.” Begun in 1988, two employees(the current and previous director) would draw land use patternsonto U.S. Geological Survey topographic maps and thenoutsource the GIS portion to the local universities. Volunteerswould assist staff in collecting data from each municipality andgenerating maps from the GIS output.

In the decade since GMAP’s inception, Greenbelt hasamassed considerable knowledge about GIS operations; none-theless, understanding of the mechanics remains with just a fewindividuals. This caused one staff member to comment, “TheGIS is pretty much an isolated braintrust. They keep it tight in-side the ‘beltway.’” In its desire to integrate the GIS into other


activities, upper management spends substantial time demon-strating the GIS to staffers and members. Staffers have respondedwith interest and are eager to incorporate a GIS into their activi-ties. Simultaneously they have felt short of time and other re-sources: “It’s not resistance to the GIS [that prevents me fromlearning] but inertia . . . . We are thin already on our policy andadvocacy work.” Questions have arisen regarding the future ofGIS decentralization. The past director would like rudimentarydiffusion, not of feature creation per se, but simple spatial analy-sis and map production. Some staffers have appreciated the util-ity of the GIS maps but not the need for software training; theywant to be consumers of output but not analyzers of data.

During the course of this research, Greenbelt went frombriefly possessing in-house capacity to spinning off its capacityinto a GIS technical assistance center. This center functions asthe technology transfer hub of a regional GIS nonprofit consor-tium. In this way, the new organization could concentrate onhelping not only Greenbelt, but assisting other nonprofits as well.

Case 2: Wants MapThe Nature Conservancy of Lanphere-Christensen Dunes

Preserve (TNC-Dunes), in Arcata, is an autonomous chapter ofa national land trust organization dedicated to preservation ofrare and endangered habitats on a 450 acre parcel of coastal dunesthat they own. The chapter’s two employees conduct ecologicalresearch and monitoring. TNC-Dunes is closely allied with an-other organization, Friends of the Dunes, that conducts publicoutreach and restoration of native vegetation. TNC-Dunes hasapplied the GIS to inventory vegetation on the dunes and trackmitigation of non-native species.

TNC-Dunes outsources all GIS programs to the local uni-versity. This follows the cultural ethic of national organization.According to interviews, TNC discourages the development ofin-house GIS capacity at the regional and chapter levels, prefer-ring instead to build capacity in like-minded institutions, such aspublic agencies. When a chapter wants the analytic capacity andoutput, it hires out for the service.3

Table 2 Implementation factors used in pattern matching method

Implementation Factor Description

Upper Management Commitment The support and commitment offered by board members, directors, and otherdecision makers

Allocation of Resources Adequate allocation by organizational decision makers of time, money, equip-ment, and personnel to GIS operation

Sufficient Training, Understanding Timely and sufficient training about GIS to increase user understanding andcarry out GIS tasks or adequate understanding extant in organization

GIS Champion A person who takes over direction of GIS development in organization

System Use Ease of data entry and output production; quality of user interface

Organizational Communication/ Coordination Communication/diffusion of GIS knowledge between organizational deci-sion makers and GIS users, GIS developers; coordination among participantsof GIS-related activities

Lack of Resistance Participants’ lack of resistance to, or apathy about, GIS implementation

Voluntarism Participation of volunteers in the implementation and utilization of GIS

Scarcity of Resources Lack of, fragility, and/or unevenness of resources in the grassroots groups thatimpact GIS implementation

External Sources of Funding Influence of external funding sources on the implementation and use of a GIS

Tension between Passion and Progress Members who resist, or missions that run contrary to, GIS implementationbecause the GIS technology represents “progress”


By outsourcing to the local university, TNC-Dunes’s direc-tor relies on its resources: equipment, data, students, and profes-sors. TNC-Dunes benefits from the university’s multimilliondollar GIS lab and extensive geo-registered spatial database. Thedirector has been satisfied with the quality and the accuracy ofmost student work: “All the Master’s projects I’ve been thrilledwith. The class projects, like trail maps, we didn’t even keep them,because the quality was low.” It is inexpensive, too: “We’re fortu-nate that we don’t have to pay professional consultants at $30 anhour.” One distinct disadvantage about the arrangement is thatnon-funded students must use the lab during off-hours and yieldto the funded/formally-arranged projects. Further, work fluctu-ates with the semester schedule, “so that slows us down and putsus at the mercy of the traffic in the lab.” Work also depends onthe interests of the professors. As one reminded me, professorsmust remain mindful of the researchable quality of the contractand its contribution to “RTP-retention, tenure, and promotion.”Therefore, contract projects should transcend the prosaic—be-come an article in a peer-reviewed journal—even if all TNC-Dunes needs is another thematic map.

TNC-Dunes has been successful in using its GIS to directits mitigation efforts and obtain additional funding from theparent organization. Unfortunately, in 1999 TNC transferred thepreserve to the U.S. Fish and Wildlife Service, which has ex-pressed little interest in continuing GIS analysis on the preserve.The Friends of the Dunes board have attempted to interest mem-bers in data collection and entry, but to no avail.

Case 3: Wants ConsortiumTrinity Community GIS was established by an economic devel-opment NGO in Hayfork as a separate GIS service center tosupport its programs and programs of affiliated groups. Operat-ing with three employees, Trinity’s applications have been in iden-tification of non-timber forest products, contract work for areapublic agencies, and training of unemployed loggers in GIS op-erations, global positioning system (GPS) operations, and datacollection techniques.

Affording and using the technology remain a vexing prob-lem for Trinity. Allocation of resources to GIS has been promis-ing, although the group has survived from one grant or contractpayment to the next. Trinity owns three personal computers withGIS software; all were obtained from grants. For peripherals anddata conversion, the organization has relied on a university thatis a five-hour drive from Hayfork. The direction toward moresophisticated contract work clearly demands a workstation-basedsystem. According to the director, “If we’re really going to be ableto fulfill our potential as a repository for GIS data, we’re going toneed to be able to make that jump.” The transition from per-sonal computers to a workstation means a more complex operat-ing system and greater resource outlays. The director continued,“What happens if it crashes? We certainly can’t afford the kind ofsystem manager who knows UNIX.” Staff offered these frustrat-ing details about current conditions: “The conversion fromAUTOCAD to GIS left me in tears one midnight.” “The manual

said ‘transform’ but it doesn’t tell you what steps you have to do[prior to] that and [in] what order you have to do [them]. They[data suppliers] had a projection on [only] one part of it; I’vebanged my head against the wall for three days”.

The internal diffusion of GIS expertise has formed an inte-gral component of Trinity’s vision. Staff receives continual expo-sure to the mechanics and application of the technology. Thedirector epitomizes the GIS champion but the staff, trained bythe director, “are more techies than I am.” Despite the extensivetraining they have received, respondents in 1995 were rightfullyconcerned about the director’s gradual departure (by 1998 only10 percent of her time was spent at Trinity). One employee won-dered “whether we can strategically replace that [expertise] withbits and pieces. That’s one of the reasons that we’re trying so hardto learn.”

The community of Hayfork is slowly healing from the battlebetween loggers and environmentalists. Trinity originated as amiddle ground, serving both populations. Unfortunately, someaffiliate GROs have viewed Trinity (including its non-native di-rector) as allied with the opposition because it accepts contractsfrom public agencies. This view offers one explanation for whysome of the affiliates have shown up “late at night,” after thedirector is gone, to use the GIS.

Case 4: Wants IndependenceIn 1995, Samuel Jones4 was leaving one organization (Friends ofthe Tecate Cypress) in Orange County to form the San AndreasLand Conservancy (SALC), a land trust headquartered in Jones’shome just north of Santa Cruz. His applications have includedthematic maps of the proposed land trust for SALC, mountainlion movements, and the impacts of a toll road for Tecate.

Jones embodies GIS implementation. He personally ownsthe workstation and software, has received formal GIS training,and has entered most of the data. In his organizations, GIS ex-pertise has remained concentrated with little internal diffusion.Tecate board members know little about system functionality andcapability. Jones has not attempted to train SALC board mem-bers in GIS. Involving others in the GIS would just slow himdown and cloud his agenda for GIS use in the organization.

Jones has adopted an uncompromising stance to conserva-tion that has limited his GIS implementation. Orange County,in cooperation with a developer, had launched a conservationplan to avoid the listing of a bird species as endangered by thefederal government. Because of Jones’s refusal to negotiate withthe developer, Tecate was shut out of those meetings—otherGROs were not—and was unable to obtain the county-collectedspecies data. Tecate was not entirely pleased with his actions. Joneshas remedied this conflict in his new organization through thecareful selection of board members who agree with his goals.

Jones has survived on the odd GIS job and temporary workin computer consulting (experience he gained largely from hisGIS work). Full-time employment would detract from his GISactivism. He has overextended his credit cards and has enduredoccasional eviction. Jones’s uncompromising stance may place a


further barrier in the path of financial security. Jones has desireda funded position within his organization, yet his unwaveringvision may have precluded him from attracting donations, be-cause donors may desire participation in the organization’s deci-sion making. Alternately, a charismatic vision may draw donorsto him.

DiscussionThe pattern matching of the cases to factors observed in

documented organizations (“Upper Management Commitment”to “Lack of Resistance”) and factors likely to be found in GROs(“Voluntarism” to “Tension between Passion and Progress”) isshown below (Table 3). Case results are reported in columns.

Description of Factors Mirroring DocumentedOrganizationsUpper Management Commitment. Upper management com-mitment emerged as the highest rated factor in a prior mail sur-vey (Sieber 1997b) among GROs when asked about their generaluse of IT. Nonetheless, this factor received mixed reviews in casestudies. Greenbelt exemplified upper-level commitment, withdirect involvement from both its current and past directors. Con-versely, TNC-Dunes has received no encouragement from itsparent organization, despite internal supports for GIS develop-ment. The Jones’s case exhibited both sides: the current boardhas been chosen for its acquiescence; conversely, one of Tecate’sboard members showed open hostility to the GIS. Clearly, thisfactor proved important but not essential to successful environ-mentalist implementation.

Allocation of Resources. Despite importance in the litera-ture, this factor was mixed in the cases. For instance, Trinity’sdirector and staff not only have appreciated the need for suffi-cient resources but also have been aware of the perils of expan-sion with an insufficient computing infrastructure. Jones has

allocated what meager resources are available. TNC-Dunes hasallocated sufficient resources, although it has encountered resis-tance in acquiring grants for GIS-specific projects.5 Adequate al-location sped GIS implementation; however, as will be discussed,a case’s “resourcefulness” could substitute for missing essentials.

Sufficient Training, Understanding. Consistent with thattraditional factor, Trinity’s director has instructed her employeesin the technical details of GIS and GPS operations; these em-ployees, in turn, have been training others. Otherwise, findingsrevealed case training experiences that were patchy, informal,concentrated, on the job, or outsourced. At one extreme, the ex-tent of TNC-Dunes’s knowledge lay in a single GIS overviewshort course. At the other end of the spectrum, the least stableorganization (Jones) received formal GIS training from ESRI.These two cases called into question the definition of “sufficient”extant in the literature and gave mixed results on training as afactor of implementation success.

A GIS Champion. GIS champions were present in all four;excepting TNC-Dunes, the champions also were technically in-volved in system development. This factor emerged as essentialto furthering GIS adoption, insofar as Jones’s departure from oneGRO signified the end of its GIS.

System Use. Cases rated system use almost uniformly poor.In Greenbelt’s case, poor cartographic-handling system capacityhas caused the organization to import GIS output into AdobePhotoshopTM. Difficult system use may complicate GIS imple-mentation but does not prevent GIS adoption.

Organizational Communication/Coordination. Casesranged from strong diffusion programs (Trinity), attempts at dif-fusion (Greenbelt), and minimal diffusion of GIS skills or GISinformation (TNC-Dunes and Jones). With the exception ofTrinity, research revealed the isolated nature of GIS knowledgewithin cases.

Lack of Resistance. I had anticipated greater resistance thanwas found in the mixed pattern matching. Attitudes range from

Table 3 Results of pattern matching for cases

Factor Greenbelt TNC-Dunes Trinity JonesFactors Mirroring Documented OrganizationsUpper Management Commitment + — + mAllocation of Resources + — + —Sufficient Training, Understanding m — + +GIS Champion + — + +System Use — N/A — —Organizational Communication/Coordination m — + —Lack of Resistance — m — m

Factors Unique to GroupsVoluntarism + + m +Scarcity of Resources — — + +External Sources of Funding + + + —Tension between Passion and Progress — — m +

(+ = positive match, — = negative match, m = mixed)


aversion (a Tecate board member) and distaste (a TNC-Dunesstaff member), through ignorance (Tecate and some Trinity boardmembers) and passive acceptance (a few Greenbelt board mem-bers and staff ), to enthusiasm (all the champions) and devotion(Jones). Consternation may result from staff resentment over theconcentration of resources and fear of changing job descriptions(Greenbelt) or a lingering suspicion of academics and remnantsof old animosities (Trinity). Notwithstanding some mixed feel-ings, individual resistance was limited and only marginally im-pacted GIS development.

Description of Factors Unique to GroupsHaving compared case study experiences to factors well docu-mented in the implementation literature; I turn now to moreexploratory factors suggested in the nonprofit literature.

Voluntarism. All cases relied on volunteers, presumably tocreate a more fluid, dedicated, and cost-effective workforce. EvenTrinity, which operates as a staffed GIS enterprise, has been as-sisted by students from its training courses in data collection, useof GPS, and digitizing. Volunteers have been utilized to reduceimplementation costs and allow Trinity and Greenbelt’s spin-offto offer low-cost consulting rates to other organizations. Thedownside of voluntarism—turnover and burnout—has been lim-ited but not avoided as students leave with expanded résumés.

Scarcity of Resources. Experiences of these cases falsifiedthis factor as groups found effective alternatives to resources;nonetheless, scarcity slowed implementation as the case partici-pants scrambled to maintain their systems. Two cases (Trinityand Jones) have experienced difficulty in, for example, obtaininggrants and data, which adversely affect their ability to completecontracts. Conversely, Greenbelt’s implementation demonstratedthe advantages of a more secure base of institutional and fiscalsupport.

External Sources of Funding. External resources may greatlyenhance GRO capacity and a GRO’s own influence; alternately,such support may control or divert its activities. Most cases wereaffected to some degree by the conditions and constraints of theirexternal funding, however benign. Respondents stated that thestature of Greenbelt and Trinity was augmented by their suc-cesses in grant acquisition and institutional cooperation. In turn,Greenbelt grant writers believe that they must avoid appearingtoo technical in its approach so as not to alienate funders. Toensure continued access to data, equipment, and contracts, Trin-ity must continue to conform to the overall goals of its partneragencies. TNC-Dunes’s implementation model and project fund-ing proceeded directly from its relations with the parent organi-zation. Only Jones has seemed immune (except to the overalllack of funding).

Tension between Passion and Progress. Observed to vary-ing degrees in three of the cases, Jones exemplified the factor inexperiencing personal struggles between sitting in front of a com-puter and communing with nature. To a much lesser extent, thefactor manifested itself among Hayfork environmentalists whodesired GIS functionality but were suspicious of Trinity’s associa-

tions. Alternately, some Greenbelt and TNC-Dunes membershad expressed personal resistance to innovation and its resultantchanges, but they did not oppose the technology spiritually. Iexpected to find abundant evidence of conflicts, but discoveredthat the passion and progress tension was largely absent. Respon-dents viewed computing technology as a useful tool for complet-ing work tasks and aggressively pursued implementation.

Cases Do Not Mirror Traditional Factors ofImplementationCase study groups neither uniformly followed nor opposed thefactors found in the literature for governmental agencies (Table3). Therefore, these best practice cases used GIS implementa-tions to their satisfaction but did not comply with all of the rec-ommendations cited in the implementation literature. One couldpostulate that groups would have preferred to conform to all thesenorms (i.e., groups might desire greater organizational diffusionof GIS skills) but were prevented by various limitations. How-ever, little evidence supported this assertion; overall, traditionalfactors were irrelevant as case respondents improvised when re-sources or support were absent. In the case of TNC-Dunes, tra-ditional factors offered an inadequate prediction of successfulusage. More important, these factors did not necessarily describecase study problems because many problems were still techni-cal—for example, handling unusual data formats—and not or-ganizational.

Issues in Conceptualizing Factors. One problem in usingtraditional factors is that GROs do not resemble larger institu-tions. Government agencies are sufficiently large so that uppermanagement must coordinate multi-departmental system devel-opment/funding, and policymakers (administrators or electedofficials) generally are separated from GIS champions or the de-tails of GIS implementation. Therefore, factors such as organiza-tional communication/coordination and upper managementcommitment do not hold the same meaning. In three of fourcase studies (Greenbelt, Trinity, Jones), GIS champions form anintegral component of upper management, leading the organiza-tions and the technical innovation.

Further, factors actually can oppose each other. For instance,a GIS champion concentrates technical expertise, GIS develop-ment decisions, and catalytic action; success provides the cham-pion with a sense of worth and power. This factor appeared toconflict with organizational communication/coordination andtraining, which can produce more experts and thus dilute indi-vidual power. The presence of strong GIS champions likely ex-plains the limited diffusion found in several cases. The GISimplementation literature presents factors as a normative list ofbest practices distilled from government cases but neglects tocompare and critique the factors as a set.

Another problem is that the implementation model largelyoffers an apolitical and mechanistic view of implementation andworkers. Much of that literature has neglected the social and po-litical attributes of the implementing environment. Instead, ithas been based on an idealized view of how organizations and


procedures should operate and assumes that logical managementstrategies will enable effective utilization (Campbell 1996). Asnoted in Eason (1993, p. 29), the use of rational and technicallanguage “makes it difficult to express all the needs and issuesthat may exist in the organisation[s] which do not lend them-selves to this kind of representation.” Many GROs—many ofwhich are rich in member passion, frequently exist in a dynamic(even hostile) political arena, and try to survive with limited re-sources—do not readily submit to this type of analysis.

This dissonance between government and GRO implemen-tation emphasizes the problems in generating theory based on alimited sample of organizations. Initial criticisms of the explora-tions into GIS implementation showed that, although rich inanecdotal evidence, they lacked “grounded” theory for a frame-work of implementation (Onsrud and Pinto 1993, Campbell1996); this gap was addressed by Onsrud, Pinto, and Azad (1992).Theory construction is a necessary step in the maturation of theliterature. Nonetheless, this grounded theory has implied a nor-mative universality in GIS implementation, even though theuniverse of study has been quite circumscribed. While the studyof governmental agencies and their departments might indeedform the backbone of a model of implementation, it is by nomeans all-inclusive. Identifying factors of success implies that onlyone process of implementation is the right one; that is, an orga-nization must match all (or a large percentage) of the factors orelse it will fail. Likewise, the inference is that only one outcome(accept/not accept) is possible (Azad 1998). Indeed, this researchdemonstrated that concentrating on a subset of institutions ob-scured assumptions or overlooked findings in the research, suchas the importance of volunteers, that might prove fundamentalto other types of organizations.

The Dynamic and Fragile Nature of GIS Implementation.Unexpectedly, three of the cases were undergoing substantive shiftsin their GIS capacity during this research. Greenbelt spun off itsGIS capacity; ownership of TNC-Dunes was transferred; the di-rector of SALC left Tecate and removed its GIS capacity. Changemay represent growth: Greenbelt and its spin-off may actuallyextend functionality to other nonprofits. In other cases, it mayresult in a loss of GIS-related advocacy. Even in Trinity, concernshave arisen over the long-term sustainability of the organizationwhen its director leaves. Indeed, these changes offer new mean-ing to Eason’s (1993) phrase “partial implementation”—if theorganization ceases to exist during a successful implementation—and lends support to his assertion that implementation of com-plex IT innovations should not be measured along one diffusioncurve.

Additionally these cases illustrate the problematic nature ofassigning what are dynamic processes to factors of successfulimplementation. Factors imply non-longitudinal toggle switches:once “ON” (or “+”) then normatively solved. Therefore, a re-searcher may assume groups’ implementation needs are solvedonly to miss what will remain ongoing challenges to implemen-tation for any organization. This finding also reinforces the propo-sition, advanced by Azad (1993) and others, that implementation

itself is a process and not simply a task to be completed. Thepicture only emerges over time.

Neither Are Differences Strictly a Function ofFactors Unique to Grassroots GroupsIf implementation in GROs does not explicitly match experi-ences in documented organizations, it may differ by issues ofvoluntarism, scarcity of resources, external sources of funding,and tensions between advocates of traditional activities and sup-porters of this more “progressive” technology. As a set, this didnot happen. Voluntarism cut across all models and drove systemoperation and implementation. External funding exerted a defi-nite impact (this influence was not necessarily negative) as casesminded the watchful eyes of their funders. Scarcity and tensionwere otherwise mixed. Scarcity may have slowed implementa-tion, but these GROs still utilized GIS implementations effec-tively. They accomplished this through what this author termsresource substitution and passionate commitment.

Resource Substitution. Cases circumvented scarcity and tra-ditional factors largely because they learned to substitute an abun-dance of one resource for an absent or deficient resource. Mailsurvey research on the general use of computers (Sieber 1997b)supported case study findings that GROs were improvising withlimited resources and that even the “poorer” groups—operatingwith no paid staff—could “make do.” They did it in three ways.First, donations of hardware/software and access to borrowedsystems (e.g., from members, universities, public agencies, andvendors) allowed groups to substitute for equipment they other-wise could not afford. All four cases, for instance, received freeGIS software from ESRI’s Conservation Program.

Second, volunteers and pre-trained staff offered these groupsexperienced workers for minimal operating outlay compared topublic agencies or private firms. Volunteers provided groups, suchas Tecate and Trinity, with teams of primary data collectors whogenerated detailed and ground-true data, tailored to group needs.Several groups have utilized student interns, supported by theexpertise of a supervising faculty member; these volunteers haveprovided on-the-cutting-edge technical and scientific skills morecompatible with rapidly changing GIS technology.

Third, cases implemented geographic information systemswithin a rich network of supportive institutions and groups. Thisnetwork compensated for the lack of resources, such as incomeand access to spatial data. Thus, the cases benefited from a univer-sity offering the use of its expensive peripherals and a sympatheticengineering firm disseminating its data (for an extensive discus-sion on universities and GROs see Sieber, 1997c); in turn, GROsextended their GIS experience to others. The network served asthe conduit through which the substitution of equipment, person-nel, data, experience, and encouragement was effected.

Substitution or not, these GROs could not escape the re-source demands of GIS. GIS use commanded significant effortsof at least one individual, consumed time for grant writing andon-the-job learning, and required extensive data collection/en-try/correction. Even outsourcing required administration. More-


over, as a short-term or long-term strategy, resource substitutionexhibited a limited elasticity. For instance, dependence on thelargesse of foundations or the expertise of grant writers mightrender free yet unwieldy software (Greenbelt with GRASS andgroups with ArcInfo). Reliance on university staff and students(Greenbelt, TNC-Dunes) and low-paid staff/volunteers (Trinity,Jones) could provide expertise and workers, yet group projectsmust compete with regular university courses, well-funded con-tracts, differing priorities, and loss of skilled volunteers. Hard-ware might be available, yet staff from two organizations musttravel eight hours to gain access to peripheral equipment (Trin-ity, Jones). This last example suggests that abundant resourcessuch as time or volunteers can be stretched only so far before thesubstitution becomes ineffective.6 Passion may exhibit the leastelasticity: The Friends of the Dunes volunteers mitigated non-native dunes vegetation by manually extracting plants, yet theirdedication was not transferable to computing activities, such asdata entry.

Resource substitution may exact a greater price. As GROsincrease their dependency either on the technology or the exter-nal support network, they may effect a diversion or co-optationof goals. Greenbelt spun off its GIS capacity precisely because itrepresented an unacceptable diversion—the means replacing theends—from the core mission. Co-optation may occur if GROsmust conform inordinately to the values of institutions or indi-viduals. Trinity has employed a compromising and negotiationstance when dealing with public agencies and therefore has en-sured continued access to data and contracts. Conversely, Joneshas found access blocked because he prefers to confront his op-ponents. Overall, diversion or co-optation appeared to exert alimited effect because most cases already had shifted their strate-gies as a precondition for accommodating GIS. Nonetheless, thepotential shift underscores the need to understand the price thatthis substitution holds for “nonconformers.”

Passionate Commitment. Grassroots organizations’ experi-ences suggest that some factors might be distilled to their essen-tial elements: commitment; in terms of upper managementcommitment, the presence of a GIS champion; and passion fromthe groups’ members and associates toward GIS. If strong com-mitment arises from any or all of these sources, then the GISimplementation will likely succeed. Hence, Greenbelt and Trin-ity were supported and guided in their system development withboth upper management commitment and a strong GIS cham-pion. In TNC-Dunes, faculty and students supplied the com-mitment, as well as all the labor and equipment. Alternately, Jones’spassion enabled implementation, despite scarce resources andsome resistant members. Indeed, the literature on GIS imple-mentation emphasizes the need for upper management commit-ment and the sufficient allocation of resources but could not havepredicted this interpretation: that a GIS user would sacrifice hispersonal comfort to maximize GIS functionality. To these cases,upper-level commitment may be low and resources may be scarce;however, if the will exists, then GIS can be implemented.

ConclusionThis paper has presented GIS implementation in the grassrootsand placed GROs in the context of strategies found in local gov-ernment. It has shown that organizationally, GROs did not fol-low traditional factors because they could substitute resourcesand employ different implementation models that suited eachorganizational culture while minimizing or outsourcing the im-pacts. At the same time GROs were not fully represented by thenonprofit literature, which implied that GIS might be out ofreach of these fragile entities.

These lessons hold significant validity for implementation inlarger organizations and public sector agencies. GROs engaged incomplex arrangements that account for an interdependency ofequipment, staff, and data. They utilized several alternatives to thetraditional in-house model, which could be tailored to fit organi-zational criteria and could evolve with changing needs. These re-sults match newer arrangements that suggest (Budic and Pinto1999) the need for greater interorganization cooperation and en-terprise solutions. The Wants Consortium or Wants Map modelsmay benefit initial GIS development in smaller towns, coping withtheir own set of fragile resources. Both large and small agenciesmay benefit from factors such as the use of volunteers to conductGIS-related activities, like spatial data collection.

The paper challenged the construction of factors of GIS imple-mentation. Instead of a mechanistic and normative list, GROsshowed implementation to be a colorful, contradictory, evolving,and highly political process. Factors provide the framework butshould be tested and retested in light of new types of organizationsadapting a GIS to their needs and adapting to a GIS.

Clearly, this study would benefit from further research. Howwell do GRO-specific factors survive against additional testing?What combination of factors does each model support? Canunsuccessful implementation be attributed to a failure to heedcertain factors of implementation? How does the importance offactors vary by time? This research highlighted numerous diffi-culties in the fluid environment of changing individuals and strat-egies. TNC-Dunes demonstrated that even best-practice casesneed not survive. It is quite possible that the dynamic nature ofGRO implementation will prohibit most GROs from ever“rountinizing” a GIS; instead organizations may exist in a per-petual state of implementation. Presumably, the most stable GROswill mature in their GIS usage as they build data sets and interactwith the larger conservation and GIS community. But they alsoincreasingly may resemble the institutions from which they ob-tain resources and decreasingly like the activists from which theydrew their strengths. Therefore, how do GROs and other organi-zations change in response to GIS implementation? Exploringthe smallest implementers can produce large results.


Notes

1. I utilize the Edwards and Hulme’s (1995) definition ofgrassroots organizations as membership organizations inter-ested in social transformation. This definition distinguishesGROs from NGOs, which are intermediary organizationsoffering funding and other forms of support to communi-ties and other organizations

2. Certainly this sentiment varies within the conservationistand environmentalist movement. The dissonance betweenenthusiastic adoption of and resistance to a GIS by conser-vationists was explored in Sieber (1997a).

3. Recently (August 1999) the Californian office has contractedout some work to Greenbelt’s spin-off. Such is the small worldof conservation NGO/GRO GIS users.

4. Not his real name.5. Ironically, including GIS output in applications improves

the likelihood of obtaining more general grants.6. It should be emphsized that North American GROs exist in

a wealth of resources and networks relative to GROs in thedeveloping and Third World. No amount of passion willsubstitute for a lack of software (although the GRO mightconsider a paper GIS).

Acknowledgements

The author would like to thank the members of her dissertationcommittee and the participating conservation nonprofits for theirsupport and participation. She also would like to thank BarbaraPoore for her helpful comments.

References

Aberely, D. (Ed.), 1993, Boundaries of Home: Mapping for LocalEmpowerment (Philadelphia, PA: New Society Publishers).

Antenucci, J.C., K. Brown, P.L. Croswell, M. J. Kevany, with H.Archer, 1991, Geographic Information Systems: A Guide tothe Technology (New York: Van Nostrand Reinhold).

Azad, B., 1998, Management of Enterprise-wide GIS Implementa-tion: Lessons from Exploration of Five Case Studies, DoctoralDissertation, MIT.

Azad, B., 1993, Organizational Aspects of GIS Implementation:Preliminary Results from a Dozen Cases, Paper presented atthe Urban and Regional Information Systems Association Con-ference, Atlanta, Georgia.

Barbeito, C.L., Bowman, J.P. and Applied Research and Devel-opment Institute International, Inc., 1998, Nonprofit Com-pensation and Benefits Practices (New York: Wiley).

Barndt, M., 1998, A Model for Evaluating Public ParticipationGIS Programs, NCGIA Special Meeting: Empowerment,Marginalization and Public Participation GIS, Santa Barbara,California, http://www.nciga.ucsb.edu/varenius/ppgis/pa-pers/index.html.

Benenson, B., 1990, In the Cradle of Environmentalism, thePolitical Fight Is Hot, Congressional Weekly Quarterly Report,48(3) (January 20), 189-191.

Brown, P. and D. Friedley, 1988, Assessing Organizational Pre-paredness for a Comprehensive, Distributed LIS/GIS, Pro-ceedings of the Urban and Regional Information SystemsAssociation, 3: 22-36.

Budic, Z.D., 1997, GIS Technology and Organizational Con-text: Interaction and Adaptation, In Massimo Craglia andHelen Couclelis (Eds.), Geographic Information Research—Bridging the Atlantic (London: Taylor and Francis), 165-184.

Budic, Z.D., 1994, Effectiveness of Geographic Information Sys-tems in Local Planning, Journal of the American PlanningAssociation, 60(2), 244-263.

Budic, Z.D. and David R. Godschalk, 1996, Human Factors inAdoption of Geographic Information Systems (GIS): A Lo-cal Government Case Study. Public Administration Review,56(6), 554-567.

Nevodic-Budic, Z. and J. Pinto, 1999, UnderstandingInterorganizational GIS Activities: A Conceptual Framework.Journal of the Urban and Regional Information Systems Asso-ciation, 11(1), 53-64.

Campbell, H.J., 1996, A Social Interactionist Perspective onComputer Implementation. Journal of the American Plan-ning Association, 62(1), 99-107.

Campbell, H.J., 1991, Impact of Geographic Information Systemson Local Government, TRP 101 (Sheffield, UK: The Universityof Sheffield, Department of Town and Regional Planning).

Cheney, P.H. and G.W. Dickson, 1982, Organizational Charac-teristics and Information Systems: An Exploratory Investi-gation. Academy of Management Journal, 35(1), 170-184.

Clifton, R.L. and A.M. Dahms, 1980, Grassroots Administration:A Handbook for Staff and Directors of Small Community-basedSocial-service Agencies (Monterey, CA: Brooks/Cole Publish-ing Co).

Clapp, J.L., J.D. McLaughlin, J.G. Sullivan, and A.P. Vonderohe,1989, Toward a Method for the Evaluation of MultipurposeLand Information Systems. Journal of the Urban and RegionalInformation Systems Association, 1(1), 39-45.

Cohen, J.A. and R.A. Perrault, 1991, Nonprofits’ Most Com-mon Computer Pitfalls—and How to Avoid Them. Non-profit World, 9(6), 8-10.

Conservation GIS Consortium, 1999, Conservation GIS StarterKit (Portland, OR: Conservation GIS Consortium).

Conservation Technology Support Program, 1999, Program De-scription (August 1999), http://www.ctsp.org.

Coveny, S., 1992, Technology Isn’t Entirely Evil: GeographicInformation Systems in the Services of Wildlands, Wild Earth(Special Issue), 81-82.


Crain, I.K., 1990, Analysis of Some Management Issues in GISImplementation. Proceedings of the Urban and Regional In-formation Systems Association, 2, 14-22.

Croswell, P., 1991, Obstacles to GIS Implementation and Guide-lines to Increase the Opportunities for Success. Journal ofthe Urban and Regional Information Systems Association, 3(1),43-56.

Crowfoot, J.E., and J.M. Wondolleck, 1990, Environmental Dis-putes: Community Involvement in Conflict Resolution (Wash-ington, D.C.: Island Press).

Eason, K.D., 1993, Gaining User and Organisational Acceptancefor Advanced Information Systems, In Ian Masser and HarlanOnsrud (Eds.), Diffusion and Use of Geographic InformationSystems (London: Kluwer Academic Publishers), 27-44.

Edwards, M. and D. Hulme (Eds.), 1995, Non-governmentalOrganizations: Performance and Accountability (London:EarthScan Publications).

Er, M.C., 1989, A Critical Review of the Literature on the Orga-nizational Impact of Information Technology, IEEE Tech-nology and Society (June), 17-23.

Feldman, S.P., 1993, How Organizational Culture Affect Inno-vation. In Larry Hirchhorn and Carole K. Barnett (Eds.),The Psychodynamics of Organizations (Philadelphia, PA:Temple University Press), 85-98.

Ferber, D.L., 1992, Geographic Information Systems Support forEnvironmental Groups: An Investigative Study (Bellingham,WA: Greater Ecosystem Alliance).

Frans, D.J., 1993, Computer Diffusion and Worker Empower-ment. Computers in Human Services ,10(1), 15-34.

Fulton, W., 1991, Guide to California Planning (Point Arena,CA: Solano Press Books).

Gittell, M., 1980, Limits to Citizen Participation: The Decline ofCommunity Organizations (Beverly Hills, CA: Sage Publica-tion).

Goldhaber, G.M., 1990, Organizational Communication, 5thEdition (Dubuque, IA: William C. Brown).

Gottlieb, R., 1993, Forcing the Spring: The Transformation of theAmerican Environmental Movement (Washington, D.C.: Is-land Press).

Hall, B. and M.L. Kerr, 1991, The 1991/1992 Green Index (Wash-ington, D.C.: Island Press).

Hays, S.P., 1987, Beauty, Health, and Permanence: EnvironmentalPolitics in the United States, 1955-1985 (New York: Cam-bridge University Press).

Huxhold, W.E., 1991, An Introduction to Urban Geographic In-formation Systems (New York: Oxford University Press).

Huxhold, W.E. and A.G. Levinsohn, 1995, Managing GeographicInformation System Projects (New York: Oxford UniversityPress).

Igbaria, M. and S.A. Nachman, 1990, Correlates of User Satis-faction with End User Computing: An Exploratory Study.Information and Management, 19, 73-82.

Kanter, R.M., 1983, The Change Masters (New York: Simon andSchuster).

Kwon, T.H. and R. W. Zmud, 1987, Unifying the FragmentedModels of Information System Implementation. In J. Boland,Jr. and R.A. Hirschheim (Eds.), Critical Issues in Informa-tion System Research (New York: John Wiley and Sons), 227-251.

Lee, A.S., 1989, A Scientific Methodology for MIS Case Studies.MIS Quarterly (March), 33-50.

Leitner, H., R. McMaster, S. Elwood, S. McMaster, and E.Sheppard, 1998, Models for Making GIS Available to Com-munity Organization: Dimensions of Difference and Ap-propriateness, NCGIA Special Meeting: Empowerment,Marginalization and Public Participation GIS, Santa Barbara,California, http://www.nciga.ucsb.edu/varenius/ppgis/pa-pers/index.html.

Manes, C., 1990, Green Rage: Radical Environmentalism and theUnmaking of Civilization (Boston, MA: Little, Brown andCompany).

Markus M.L., 1983, Power, Politics, and MIS Implementation.Communications of the ACM, 26(6), 430-444.

McCullough, M.F., 1991, Democratic Questions for the Com-puter Age. Computers in Human Services, 8(1), 9-18.

Nam, K., S. Rajagopalan, H. Raghav Rao, and A. Chaudhury,1996, A Two-level Investigation of Information SystemsOutsourcing. Communications of the ACM, 39(1), 36-44.

Nonprofit World, 1994, Ten Secrets of Computer Wizardry.Nonprofit World, 12(3), 24-32.

Nonprofit World, 1990, Streamlining Nonprofits through Af-fordable Computer Resources, Nonprofit World, 8(2), 19-22.

Obermeyer, N.J., 1998, HUD’s Community Connection forLocal Empowerment, NCGIA Special Meeting: Empower-ment, Marginalization and Public Participation GIS, SantaBarbara, California, http://www.nciga.ucsb.edu/varenius/ppgis/papers/index.html.

Onsrud, H.J. and J.K. Pinto, 1993, Evaluating Correlates of GISAdoption Success and the Decision Process of GIS Acquisi-tion. Journal of the Urban and Regional Information SystemsAssociation, 5(1), 18-39.

Onsrud, H.J., 1991, Diffusion of Geographic Information In-novations. International Journal of Geographic InformationSystems, 5(4), 447-467.

Onsrud, H.J. and B. Azad, 1992, Case Study Research Methodsfor Geographic Information Systems. Journal of the Urbanand Regional Information Systems Association, 4(1), 32-44.

Perkins, S., 1990, Down with Deja-vu: Making Changes Stickin your Organization, Proceedings of Urban and Regional In-formation Systems Association, 2, 1-13.

Pickles, J., (Ed.), 1995, Ground Truth: The Social Implications ofGeographic Information Systems (New York: Guilford Press).

Pinto, J.K. and H.J. Onsrud, 1995, Sharing Geographic Infor-mation across Organization Boundaries: A Research Frame-work. In Harlan J. Onsrud. and Gerard Rushton (Eds.),Sharing Geographic Information, (New Brunswick, NJ: Cen-ter for Urban Policy Research (CUPR) Press), 44-64.


Rogers, E.M., 1995, Diffusion of Innovations, 4th Edition (NewYork: The Free Press).

Rogers, E., 1993, The Diffusion of Innovations Model. In IanMasser and Harlan Onsrud (Eds.), Diffusion and Use of Geo-graphic Information Systems (London: Kluwer Academic Pub-lishers), 9-23.

Rubinyi, R.M., 1985, The Effects of the Introduction of Microcom-puters on the Communication and Information Behavior ofCommunity-Based Nonprofit Organizations (Ann Arbor, MI:University Microfilms).

Sawicki, D.S. and W.J. Craig, 1996, The Democratization ofData: Bridging the Gap for Community Groups. Journal ofthe American Planning Association, 62(4), 512-523.

Sawicki, D.S. and R. Peterman, 1998, Understanding the Breadthand Depth of PPGIS Supply. NCGIA Special Meeting: Em-powerment, Marginalization and Public Participation GIS,Santa Barbara, California, http://www.nciga.ucsb.edu/varenius/ppgis/papers/index.html.

Schindler-Rainman, E. and R. Lippitt, 1975, The Volunteer Com-munity: Creative Use of Human Resources, 2nd Edition (SanDiego, CA: University Associates, Inc.).

Schoech, D., 1982, Computer Use in Human Services: A Guide toInformation Management (New York: Human Sciences Press).

Sieber, R.E., 1997a, Computers in the Grassroots: Environmental-ists, GIS, and Public Policy. Ph.D. Dissertation, RutgersUniversity.

Sieber, R.E., 1997b, Computers in the Grassroots: Environmen-talists’ Use of GIS and Information Technology, CUPR Work-ing Paper #121, New Brunswick, NJ: Center for Urban PolicyResearch, 26 pp.

Sieber, R.E. 1997c, Geographic Information Systems in theGrassroots: The Role for Universities, Paper presented at theUniversity Consortium for Geographic Information Science, BarHarbor, Maine.

Slaughter, S. and S. Ang, 1996, Employment Outsourcing inInformation Systems. Communications of the ACM, 39(7),47-54.

Smith, M., 1991, Adventures in Dysfunctional Computerland.Nonprofit World, 9(4), 6-7, 30.

Snow, D., 1992, Inside the Environmental Movement (Covelo, CA:Island Press).

Specht, J., 1996, Mapping Earth’s Endangered Biodiversity. GISWorld, 9(3), 42-46.

Sprecher, M.H., 1994, ICMA Survey Dissects Local GovernmentIT Use. Journal of the Urban and Regional Information Sys-tems Association, 6(2), 92-95.

Walton, J., 1992, Western Times and Water Wars: State, Culture,and Rebellion in California (Berkeley: University of Califor-nia Press).

Wikle, T.A., 1995, Geographical Patterns of Membership in U.S.Environmental Organizations. Professional Geographer, 47(1),41-48.

Wolch, J.R., 1990, The Shadow State: Government and VoluntarySector in Transition (New York: The Foundation Center).

Yin, R.K., 1989, Case Study Research—Design and Methods, Re-vised Edition (Beverly Hills, CA: Sage Publications).

Yin, R.K., P.G. Bateman, and G.B. Moore, 1985, Case Studiesand Organizational Innovation, Knowledge: Creation, Diffu-sion, Utilization, 6(3), 249-260.

See you inOrlando!

Has your organization improved the delivery and quality of

government services through the application of information

technology? If so, that achievement should be recognized and

shared with your peers. Nominate your organization for a presti-

gious URISA Exemplary Systems in GovernmentTM (ESIGTM) Award.

All submissions will be reviewed by the ESIGTM Committee and

winners will be notified in May, 2000. Winners will be recog-

nized during the Awards Ceremony at URISA 2000 in Orlando,

and one person from each winning system will receive a com-

plimentary full registration for the conference. Following the

conference, winners will receive additional recognition in URISA

publications and an announcement of their accomplishment

will be made to media representatives around the world.

The application deadline is: March 31, 2000

Join the exclusive list of ESIGTM Award winners. If you’ve success-

fully improved the way in which government operates, through

the use of information technology, you should proudly complete

the application process for a 2000 URISA ESIGTM Award. The

application is available on the URISA website, http://

www.urisa.org/2000conference/urisa_2000_esig_award.htm or

by contacting URISA Headquarters at (847) 824-6300 or

[email protected].

Apply today for a URISAExemplary Systems inGovernmentTM (ESIGTM) Award!

URISA Journal ■ Greenwald 31

IntroductionDiscussion of geographic information systems (GIS) has takenprofessional and academic planning forums by storm within thelast decade. Examples of fundamental organizational topics forunderstanding a GIS include system definition (Leno 1989), im-proved efficiency (Innes and Simpson 1993; Huxhold 1995), andimplementation issues (Budic and Godschalk 1996). These dis-cussions have been supplemented by discussions of GIS in termsof individual systems applied to specific tasks (e.g., land supplymonitoring and management [Bollens and Godschalk 1987],facilities location for undesirable land uses [Lober 1995], urbanlandscaping [Miller 1995], storm water management [Shamsi1996], and office tenancy [Howland & Lindsay 1997]. Whathas been lacking is a discussion of GIS in a multi-jurisdictionalcontext, a situation where several vertically organized governmentinstitutions (e.g., cities, counties, state, and federal agencies) usethe same GIS software and databases for completion of their ownorganizational missions and simultaneously for the purposes ofregional analysis and related decision making between variousmembers of the same level of government (e.g., two cities usingthe same database for the purposes of reconciling incompatibleland uses on the border of their respective jurisdictions).

Examination of previous attempts at multi-jurisdictional GISimplementation seeks to answer some of the more daunting ques-tions regarding GIS implementation beyond the realm of a single

Beyond City Limits:The Multi-Jurisdictional Applications of GIS

Michael J. Greenwald

Abstract: Much has been written about specific geographic information systems (GIS) applications, but as cooperative relation-ships between various levels of government evolve, GIS analysis and systems development will need to adapt. The author suggestsmulti-jurisdictional GIS development, a system where different levels of government use the same GIS software and databasesfor completion of their own organizational missions and simultaneously for the purpose of larger regional analysis. This articleexamines organizational and technological support issues involved in creating multi-jurisdictional GIS through a contrast of aprevious attempt by the U.S. Department of Housing and Urban Development and a current project supported by the SouthernCalifornia Association of Governments.

agency or level of government. Most cooperative GIS implemen-tations have been confined to relatively small geographic areaswith participants having the same level of jurisdiction. This pa-per discusses some of the technological and organizational inter-actions and assumptions that must be addressed before acooperative GIS can be executed in the context of multiple juris-dictions with varying levels of political power and responsibility:What is the underlying goal of the system? Who should partici-pate in the development of the system? Why develop such a sys-tem now? What type of technical and organizational issues mightarise when creating a multiple-user system? After addressing thesequestions in the context of existing local projects, the focus isshifted to contrast two case studies in cooperative computingbeyond the local level.

The first is the attempt of the Urban Information SystemsInter-Agency Committee (USAC), an effort during the 1970s bythe federal government to develop large-scale computing capac-ity at the municipal level for the dual purposes of national policyresearch and municipal records organization. Although USACwas not a GIS by today’s standards (lacking the ability to con-duct distance analysis or graphic manipulation/selection of recordsby selection on a base map), it was one of the first attempts in-volving multi-jurisdictional computing at the local level, makingit historically important. The discussion of USAC demonstrateshow the interaction between technical limits and organizationalgoals must be considered in the development of any multi-juris-dictional computing system.

The second example is an examination of the Southern Cali-fornia Association of Governments (SCAG) Access Project (AC-CESS). The discussion of ACCESS serves three purposes. First,it contrasts with USAC the advances in technology and integra-tion of organizational missions under an organizationally consis-

Michael J. Greenwald is a Ph.D. student in the Department ofUrban and Regional Planning at the University of California-Irvine.He received his BA in Political Science from UCLA in 1995 and hisMaster of Urban and Regional Planning from UC Irvine in 1997.His research interests include economic development, transportationand inter-jurisdictional cooperation.


tent computing system, demonstrating how cooperative comput-ing arrangements can be employed for environmental manage-ment purposes (e.g., urban planning, travel forecasting, resourceinventories, pollution control, etc.) at several levels simultaneously.Second, because ACCESS is a GIS-based example, it can answersome of the questions based specifically in multi-jurisdictionalGIS implementation that USAC begs: How could a GIS be de-veloped, from the beginning, with more than one end user in mind?How will the costs be distributed equitably across participants?What level of detail in the underlying data is needed to make thesystem work? Third, an analysis of end user sophistication withcomputers and associated ACCESS usage suggest that althoughACCESS does represent significant gains over USAC, it still fallsshort of achieving its own goals, due to its own technological andorganizational impediments. The paper concludes with proposalsof what research questions multi-jurisdictional systems raise, andhow these questions might be pursued in the future.

Underlying Assumptions and Issues

What Would a Multi-Jurisdictional GIS Do?A multi-jurisdictional application of GIS attempts to expand thelevel of detail and range of tasks for which GIS analysis can beuseful to individual organizations by facilitating the transmis-sion, disaggregation, and analysis of spatially based informationbetween various levels of government. Howland and Lindsay’sresearch on office tenancy and commuting behavior demonstratesthe need for such communication and data disaggregation. (1997)The authors noted that no previous research sufficiently similarto their own work was available for literature review, because datatracking movement of office users below the county level of analy-sis had not previously been available. Their use of GIS to collect,organize, and analyze their investigation indicates a need for agreater level of detail in data collected, and spanning a widergeographic region; it is a pioneering step in urban planning re-search.

An important first step in developing a multi-jurisdictionalGIS is a justification of why such a system design is necessary.While the case-specific analysis of GIS extols the virtues and ex-poses the weaknesses of using GIS for their specific purposes,they do not explore how a GIS with different attributes mightyield different, or better, results. Such tasks are left to those whowould review GIS software (Levine and Landis 1989) or associ-ated software extensions (Levine 1996), and as a result gives thegeneral topic of GIS in the planning profession a short shrift.This lack of discussion comparing GIS capabilities and the tasksto which it will be applied inevitably leads to fragmentation be-tween systems, and the levels of government that use them. Suchfragmentation is best avoided by bringing in all parties who havea stake at the inception of the system.

How Would Participants in a Multi-JurisdictionalSystem Relate?Of primary concern is how the GIS is viewed and used by thevarious project participants. Heikkila (1998) draws a distinctionbetween how GIS is used in the planning profession and in theacademic environment. The discussion is important because inthe pursuit of professional degrees there is an endogenous rela-tionship between the state of the practice and instruction given;what is practiced serves as the basis for academic analysis andinstruction, which graduates then take into the field and use topush their profession in new directions. The professional pointof view, according to Heikkila, is to use GIS as a reference andstorage device for spatially coded information such as propertyrecords, building permits, and zoning maps. The academic per-spective attempts to broaden the GIS pallet by not only teachingthe software mechanics of GIS research for specific bureaucraticgoals, but also by taking the information sources listed previ-ously (and others) to use the system as a modeling tool based onthe needs of various academic disciplines. Regardless of the out-look adopted, executing large-scale spatial research and analysison public issues requires massive amounts of staff time and money,and detailed technical knowledge about GIS data collection andforecasting procedures. For these reasons, most municipal plan-ning departments receive regional forecasts from higher levels ofgovernment, such as the county, state, or metropolitan planningorganization (MPO), with the primary role of lower level agen-cies being enforcement and compliance. It is only necessary, then,for all but the largest jurisdictions to be able to make projectionsfrom data that is given to them. This leaves municipalities andcounties in the situation where they simply use the data-analysistools provided by GIS and do not develop full modeling capa-bilities, even when they are readily available.

This relationship results in an organizational problem: Thevariety of GIS used by different governments (or organizationswithin the same government unit) creates a technological impedi-ment to data exchange and coordination of institutions. In re-sponse to these problems, a multi-jurisdictional GIS would use astandardized system of software and map formats to pass informa-tion in both directions between levels of government. A multi-jurisdictional GIS could serve as a commonly accepted basis fordecision making, because all the information contained within andanalysis based on it could be quickly and identically replicated.

The initial development of a multi-jurisdictional GIS projectwould be in the domain of the branch of government that hasprimary jurisdiction over the geographic area in question. Theseinitial projects would be delivered to local jurisdictions for re-view and comparison to their own records. This is done to allowa contrast of a single local jurisdiction to an integrated wholelarger than itself, and to correct any errors that may exist in thedata sets serving as the project reference. These analyses and cor-rections will then be submitted to the higher levels of govern-ment for inclusion and correction in subsequent GIS projects,and the cycle begins again. Alternatively, local jurisdictions candevelop their own project interests, through cooperative arrange-


ments such as councils of governments or public/private part-nerships, and submit the project to higher jurisdictions for infor-mation purposes or necessary consent, as the case may warrant.

From this arrangement, local jurisdictions can see moreclearly how higher levels of government regard their needs indeveloping regional forecasts, in addition to having the possibil-ity of doing their own forecasting. State and federal agencies, aswell as MPOs, would be able to collect more accurate and de-tailed data. Constant dialogue between system participants wouldensure everyone was using the same data and forecast strategies,minimizing inter-jurisdictional conflicts. Ideally, there would beno impediments, technical or otherwise, to having local jurisdic-tions customize their individual systems with specialized localdata so long as they also contained common GIS data betweenall system participants.

Why Develop a Multi-jurisdictional System?Just because a framework can be designed doesn’t mean that itwill, or even should, be implemented. Why attempt developinga multi-jurisdictional GIS now? The answer lies in the technicaland organizational advances relating to computing and GIS inthe past few years, and the forthcoming opportunities in localplanning, which may very well require such cooperation in orderto fully take advantage of them. Examples include the U.S. Cen-sus in the year 2000 specifically and the proliferation of coopera-tive government relationships in general.

TechnicalPart of past difficulties in developing a multi-jurisdictional GISwas that the costs to distribute the necessary technology widelyenough were prohibitive. Most GIS work best on a 32-bit oper-ating system (Huxhold 1991). Prior to the development of 32-bit operating systems, this restricted GIS to the realm of expensive,UNIX-based workstations, requiring large investment in capitaland training. In addition, GIS software has become more versa-tile for the individual user. Spatial analysis capabilities have beensteadily incorporated into the mainstream of GIS tasks since theearly 1990s. In his review of supplemental spatial statistical mod-ules for GIS, Levine (1996) concluded that there would be anincrease in the number of statistical programs interfacing withGIS, and a growth in the ability of GIS to connect with standardstatistical packages. Levine was partly correct about the increasedconnection between statistical analysis and GIS. Rather than es-tablish links to outside application software, developers have in-cluded spatial statistical capabilities in software extension packagesand later releases of their own systems (Intergraph Corp. 1997).Such developments enhance modeling capabilities for local juris-dictions with each successive upgrade, allowing them to rebutforecasts they consider inaccurate or incomplete. These develop-ments enhancing local GIS capabilities make a multi-jurisdic-tional context all the more important to ensure that the interestsof both large and small organizations are balanced in terms ofefficiency and sovereignty.

OrganizationalCalkins and Weatherbe (1995) go so far as to say that techno-logical advances such as those described are “removing the tech-nical barriers to spatial data sharing” (Calkins and Weatherbe, inOnsrud and Rushton 1995). They state that the remaining con-straints to spatial data sharing (a necessary first step in develop-ing a multi-jurisdictional GIS) are mainly organizational in nature:Does the underlying political and bureaucratic support exist fordeveloping cooperative a GIS? The answer appears to be yes.Current organizational research connected to GIS developmentsuggests that concurrent with this leap in technology is the in-creased acceptance of GIS in the realm of traditional planningand engineering forums, and the examination of social factorsthat accelerate or hinder the development of GIS in local govern-ments. Evans and Ferreira (1995) pointed out that the overlapbetween technological innovation and organizational behaviorpatterns is a potentially rich area of research, as both impinge onthe ability of multiple users to cooperate and maintain coopera-tive GIS (Evans and Ferreira, in Onsrud and Rushton 1995). Onthe technical side, Evans and Ferreira suggested future researchfocus on the “messy transition period when not all the relevantorganizations are fully equipped or conforming to new standardsand theories for spatial data sharing.” As mitigation to the difficul-ties encountered during this transition, the authors propose that“loosely coupled, cooperating modules of software and hardware”be employed instead of large scale, specially designed systems.

Evans and Ferreira’s call was not ignored. Budic andGodschalk (1996) tested the acceptance of GIS among intendedusers in several departments in Cumberland County, North Caro-lina (a school district transportation department, the mappingsection of a county tax assessor’s office, the community assistanceand comprehensive planning sections of the county planningdepartment). The results of their investigation are enlightening.The authors found that an “absence of perceived concrete per-sonal benefits” for those who would use GIS was a hindrance inadoption of the system, while the existence of that benefit didnot necessarily ensure that the system would be adopted. Threefactors greatly enhanced the likelihood that GIS would be ac-cepted in normal operation: a higher degree of computer experi-ence overall (and a greater familiarity with GIS technologyspecifically), the ability to communicate with one’s peers on theuse of GIS (closely related to familiarity with the technology),and personal acceptance of work-related change. In addition tothe personal factors regarding GIS acceptance, Budic (1994) iden-tified six elements that are critical to developing a local GIS: po-litical support for incorporation of the technology, staff supportfor its implementation, length of time GIS has been used/experi-ence with the technology, system sharing capabilities, compre-hensiveness of the GIS database, and the number and types ofapplications for which the GIS can be used. Adopting a multi-jurisdictional approach to GIS can generate political support bydemonstrating how a GIS can enhance a local jurisdiction’s pre-dictive and enforcement capabilities. Database comprehensive-ness and systems sharing capabilities are a requirement for a


successful multi-jurisdictional GIS, and may be developed at re-duced cost through economies of scale, jurisdictional coopera-tion, and specification of data transfer protocols. Indeed, it isbecoming less necessary to devote large amounts of staff time todeveloping these databases, as external sources for them are be-ginning to proliferate (e.g., U.S. Bureau of the Census, commer-cial databases, and public utility CAD network drawings). Thus,the number of tasks for which a multi-jurisdictional GIS can beused is limited only by the amount of information in the data-base and the technical knowledge and creativity of the GIS user.

Multi-Jurisdictional OpportunitiesIn order for a multi-jurisdictional effort to succeed, it must becentered on a set of commonly accepted goals. The developmentof these goals is usually imposed by the existence of requirementsor opportunities beyond the local context. This definition is allthe more important in terms of a cooperative GIS, because thepossibility exists that group resources may end up being co-optedby individual members for their own purposes without any con-nection to the group effort. This result is particularly likely whendifferent multi-jurisdictional participants are at various stages ofdeveloping GIS capabilities in house. These factors will be ana-lyzed in greater detail in the next section, but for the momentthey give rise to the question: What set of current circumstancescould overcome organizational impediments to a GIS-based co-operative relationship beyond local levels of government?

The first factor that provides the impetus for multi-jurisdic-tional GIS is the proliferation in recent years of other coopera-tive efforts between various levels of government. Nunn andRosentraub (1997) provide examples of inter-jurisdictional effortsto provide common services, administer projects, and identifyproblems beyond immediately local control. Inter-jurisdictionalcooperation, as Nunn and Rosentraub have defined it, is restrictedto cooperation between similar political entities (i.e., cities withcities, counties with counties). They identify four basic areas ofimprovement that inter-jurisdictional cooperation can address:economic development, delivery of municipal services, improve-ment and preservation of physical environmental quality, andsocio-political change. The authors suggest that the degree ofsuccess inter-jurisdictional cooperative groups attain is linked tothe objectives to be achieved (indicating the level of expectedpolitical resistance), the institutional format to be used (indicat-ing local autonomy), and the underlying social paradigm to beused (indicating the way issues will be approached, or even ac-knowledged, by the group).

By itself, Nunn and Rosentraub’s argument still begs thequestion why cooperate using a GIS? The second motivation fora multi-jurisdictional GIS is the interest recently expressed bystate and federal organizations in GIS technology and policy. Forexample, during the current and past two legislative sessions, theCalifornia Assembly has repeatedly expressed interest in devel-oping GIS capabilities at the local level through the proposal of aboard to manage grant applications written for this purpose. Pas-sage of such a bill into law could represent a new source of fund-

ing for additional cities, counties, and special districts to partici-pate in a large multi-jurisdictional GIS. At the federal level, thereare two efforts under way to enhance the detail, accuracy andaccessibility of spatial data: the National Spatial Data Infrastruc-ture (NSDI) and the Local Update of Census Addresses (LUCA)for the U.S. Census 2000.

NSDI is a project by the Federal Geographic Data Commit-tee (FGDC) to consolidate and assign various tasks of geographicdata collection under the jurisdiction of the federal government.The members of the FGDC are the U.S. Departments of Inte-rior, Commerce, Agriculture, and Transportation, respectively.The departments exchange data on their respective jurisdictionsconsistent with FGDC standards for quality, accuracy, and trans-mission. As a side task, the FGDC also is intended to “provideguidance and promote cooperation” between the federal govern-ment and the state and local governments regarding collection,presentation, and dissemination of spatial data (Office of Man-agement and Budget 1990). As of March 1999, this side task hasreceived greater attention by way of the Community/Federal In-formation Partnership, a budget item proposed by the Clintonadministration in order to expand NSDI’s reach to the local level(National States Geographic Information Council 1999).

Under LUCA, the Census Bureau hopes to enlist local juris-dictions in assisting with the year 2000 census by having themcorrect for geocoding and address database mistakes using theBureau’s Master Address File and Topographically IntegratedGeographic Encoded Record (TIGER) files (Williamson 1998).Areas where local jurisdictions may be able to enhance censusresults include review of subdivisions that should be incorpo-rated in the census tally to allow checking accuracy of municipalboundaries, amending TIGER files with street name or addresschanges or planned changes, and analysis of multi-family dwell-ing units and unusual addresses or “atypical” (i.e., illegal) hous-ing units that meet the Bureau’s definition of a household.Although LUCA is targeted at municipalities, there is still a rolefor larger jurisdictions to play. Counties and local area formationcommissions can provide the Census Bureau with informationon unincorporated county territory not yet under a specific mu-nicipal jurisdiction but which will be incorporated before 2010.Both NSDI and LUCA show how a multi-jurisdictional GISwould serve such inter-jurisdictional efforts as a common researchresource with information beyond the immediate confines of theparticipating members, expanding their analytic capabilities with-out surrendering sovereignty.

Examples of Multi-Jurisdictional GIS

Local ExamplesTo better understand how multi-jurisdictional systems wouldwork in practice, it helps to examine the organization of existingdistributed GIS operations. Because the practice of distributedGIS development is relatively new, much of the following case-specific discussion comes from non-academic resources. Threeexamples of a multi-jurisdictional system currently in use are the


Indianapolis Mapping and Geographic Infrastructure System(IMAGIS) in Indiana, the Cincinnati Area Geographic Informa-tion Systems Consortium (CAGIS) in Ohio, and the WinnebagoGeographic System (WINGS) in Winnebago County, Wiscon-sin. Of these, CAGIS and WINGS have the best historical docu-mentation available, so they are the focus of this section. All ofthese endeavors have undergone major software and system up-grades within the past seven years specifically for the purpose ofsecuring reliable relationships between project members: WINGSin 1992, CAGIS in 1994, and IMAGIS in 1996 (Elliot 1996;Quinn et al, 1999; Marion County 1999). In each case, the indi-vidual participants of each multi-jurisdictional system were us-ing their own GIS since the early to mid-1980s but, seeing theneed to develop consistent systems, cooperative forums devel-oped by mutual consent for discussion of issues and opportuni-ties beyond any single level of government. Benefits were realizedin reduced staff time to complete tasks and lower maintenancecosts due to maintaining only one set of records.

The WINGS and CAGIS examples demonstrate importantpoints. The spark for developing WINGS was the need of boththe county and its municipalities to keep pace with its share ofincreasing regional growth in a cost-effective manner. Designedaround a cooperative agreement between the county and six cit-ies within its jurisdiction, the project eventually expanded to all21 municipalities in the region (Elliot 1996). The operationalgoal of the new system was to cut costs for municipal and countyservices by reducing duplicate filings procedures for development,sharing common property databases and computer-generatedmaps between municipal departments with similar goals, andproviding a secure, comprehensive system for maintaining mu-nicipal records (American City & County 1995). Analysis of theeconomic and housing trends of the ten-year period prior to theinception of WINGS bears out this point. Winnebago lost agri-culture as an employment base at a proportion 50 percent greaterthan the rest of the state, while experiencing faster than averagegrowth in managerial and technical service employment (9% and15% above state norms, respectively). Concurrently, the numberof housing units classified as urban grew at a rate 38 percentfaster than the state as a whole, while housing units classified asrural increased at a rate less than half (48%) that of the state(U.S. Dept. of Commerce 1980a, U.S. Dept. of Commerce1980b, U.S. Dept. of Commerce 1990). The case was similar forthe CAGIS group: inconsistencies between various municipal andcounty maps and databases lead to increased delays in processingbuilding permits, infrastructure repair, and public service requests.After CAGIS reached full operational status, the developmentprocess was shortened five to seven months on average in theCincinnati area, while the local sewer district gained six milliondollars in reduced maintenance costs and increased revenues.Infrastructure changes made in one department were reflectedacross the system in near real time, regardless of the jurisdiction(Quinn et al 1999).

In both cases, the driving force for developing the coopera-tive arrangement was similar: the need for consistency of infor-

mation and enhanced predictive capabilities across several inter-dependent organizations. WINGS was built around 20 commoncritical layers of information for all participants, such as roadnetworks, infrastructure, and parcel data. Individual participantscould supplement the system with their own information, andby 1995 the system contained 65 layers of information on all 21cities and over 85,000 land parcels. Special GIS applications andinterfaces were designed for individual departments as needed(American City & County 1995). As the project has grown, ithas come to include 200 layers of information and now is beingused as a forecasting tool for such tasks as population projection,urban sprawl management, and farmland preservation in acounty-wide context. Thus, Winnebago’s experience has gonebeyond simply enhancing reactive capabilities to the develop-ment of new proactive tools. Although primarily designed forinformation dissemination, analysis of archived CAGIS data couldserve the same role for both the city of Cincinnati and HamiltonCounty, because the system was designed with similar goals inmind.

The expansion the WINGS project experienced is a logicaloutcome of successful cooperative GIS development. Yet, bothWINGS and CAGIS are geographically constrained; neither ex-amines implications beyond the borders of the county. Arguably,concentrated growth has implications beyond the county level.How should other jurisdictions react? A detailed examination ofattempts at developing a larger-scale, multi-jurisdictional com-puting environment, contrasted with a current, similar multi-jurisdictional GIS operations would show how such a systemsconcept could be improved in this aspect. The USAC programand the SCAG ACCESS project provide those examples.

USAC: A Multi-jurisdictional GISCollapsesProject Goals. The Urban Information Systems Inter-AgencyCommittee (USAC) is one of the first coordinated attempts todevelop a broad based cooperative prototype municipal infor-mation management system. (Eichelberger 1992; Tosta andCroswell 1992). The ultimate goal was to design an integratedsystem to be used by more than one level of government in theirrespective decision-making processes (National Academy of Sci-ences 1976). This was to be achieved through the prototypedevelopment of a comprehensive database known as the Inte-grated Municipal Information System (IMIS), the premise ofwhich was to build “a total municipal information system basedon four functional subsystems, representing Public Safety, PublicFinance, Physical and Economic Development, and Human Re-sources Development” (Kraemer and King 1977). USAC identi-fied six related operational goals, including integrated dataprocessing, operations-based automation, prototype development(including system analysis, conceptualization, design, development,implementation, and evaluation), complete project documenta-tion (for project replication), and transfer of developed systemsand subsystems.


Project Design and Organization. As described by Kraemerand King, the federal government, through the Department ofHousing and Urban Development (HUD), selected six cities(Wichita Falls, TX; Charlotte, NC; Dayton, OH; Long Beach,CA; Reading, PA; and St. Paul, MN) for initial participation.Universities in the participating jurisdictions were included forthe purposes of monitoring, technical advising, and evaluatingIMIS development. These six sites were chosen from a pool of 79applicant cities with populations between 50,000 and 500,000.Between 1970 and 1977 (when USAC was formally disbanded)$26 million was spent in pursuit of developing an IMIS for eachcity: $20 million spent by the federal government, $6 millioncontributed by the local jurisdictions.

USAC implemented a “consortium approach,” which placedmost of the responsibility and power for project completion inthe hands of municipalities, who were expected to enlist the helpof private enterprise for technical expertise and university assetsfor project evaluation. It was anticipated once IMIS was devel-oped in each city, it subsequently could be connected to stateand federal agencies. While initial stages of the project were pur-sued enthusiastically, it was soon realized that technical complexi-ties were severely underestimated, due at least in part to the factthat IMIS was a prototype. As a result of these technical com-plexities and associated budget shortages, it was realized that de-velopment of IMIS on the desired scale was not possible. Wherefederal and local goals conflicted, local goals received priority fromthe primary project implementers. The prime example Kraemerand King give is the desire of the participating municipalities todevelop a working system for their own jurisdiction, while fed-eral officials wanted to develop a research tool. This conflictprompted federal review and eventually the imposition of strictadherence to USAC goals. Cities responded by meeting only theirofficial obligations, anticipating the eventual end of the project(Kraemer and King 1977).

Analysis. Kraemer and King note that if the USAC did notdevelop a fully transferable IMIS, then it at least encouraged thedevelopment of large-scale computing at the municipal level. Thenumber of data processing employees in USAC cities increasedby 74 percent as compared to 47 percent in the other applicantcities between 1970 and 1975. Computing capacity increased by2,500 percent and the number of computer terminals increasedby 550 percent in USAC project cities over the same time pe-riod. For applicant jurisdictions that were not included in USAC,computing capacity jumped 690 percent and the number of com-puter terminals increased 1,000 percent. In addition to this in-creased processing power, local jurisdictions took the USACexperience as indicating that large-scale municipal computingprojects designed to handle municipal records could at least beconceived, and that smaller subsystems could in fact be built.From the collapse of USAC, then, some of the fundamental re-quirements for future multi-jurisdictional cooperation in data-base development were identified. These included the need forgreater processing power and computer talent at the local level,the need for project participants to be able to see their work in

relation to a larger system, and a realization of both fiscal andstaff resource costs involved.

SCAG ACCESS: A Multi-Jurisdictional GISUnder ReviewProject Goals. The two major goals of ACCESS are the improve-ment of communications and coordination between the South-ern California Association of Governments (SCAG) and itsconstituent jurisdictions, and the creation of cooperative subre-gional planning institutions. The project was envisioned to “ . . .give the subregional Associations [i.e., Councils/Associations ofGovernments] and the constituent local governments the toolsthey need to engage in a coordinated planning process, and whichwill allow for all local governments to become more fully in-volved in subregional and regional planning decision making.”To that end, SCAG provided GIS software and computer hard-ware free of charge to participating member cities within the coun-ties under its jurisdiction (Figure 1).

Project Design and Organization. Currently, OrangeCounty participation in ACCESS is facilitated by the OrangeCounty Council of Governments (OCCOG), created to reviewand comment on regional and sub-regional planning activitiesand their impacts on such issues as air quality, demographics,transportation modeling, and implementation of new technolo-gies. At the time of this writing, OCCOG membership is com-prised of 36 Orange County cities and selected special districtsincluding various water districts, the Orange County SanitationDistrict, and the Transportation Corridor Agencies (responsiblefor toll road project management).

Orange County participation in ACCESS is being handledby the OCCOG-Technical Advisory Committee (OCCOG-TAC), a collection of staff from the various OCCOG participat-ing agencies who analyze the opportunities and impacts presented

Figure 1: Jurisdiction of Southern California Association ofGovernments

Source: http://www.scag.org/bin/scag_map


by state, federal, and SCAG programs. The OCCOG-TAC co-ordinates ACCESS implementation between SCAG staff, par-ticipating Orange County cities, and the Center for DemographicResearch (CDR, a social science and statistics research organiza-tion attached to the California State University at Fullerton)through monthly meetings.

The organizational structure of ACCESS is designed to besimultaneously “top-down” and “bottom-up” (Figure 2). Localjurisdictions and special districts receive initial subregional fore-casts from SCAG, acting as the MPO, through the sub-regionalcouncils of governments (e.g., OCCOG for Orange County).Once distributed at the local level, municipalities and specialdistricts can distribute ACCESS information to individual di-visions within their organization for review. When this reviewis complete, the corrections and comments are aggregated andgiven to the sub-regional Council of Governments and the Sub-Regional Coordinator for the purpose of providing a unifiedresponse to SCAG. SCAG then makes what they believe arethe appropriate corrections, and the process begins again. Afterseveral iterations of this process, the final sub-regional forecastsare distributed, and individual project participants can then usethe sub-regional forecasts for their own planning needs. Throughthis process, counties and cities can ensure that their voice isheard in regional planning decisions, and SCAG has a means ofchecking and refining their projections, which serve as the ba-sis for meeting their obligations under such federal environ-mental statutes as the regional MPO. In addition, local userswho have common or overlapping borders can use ACCESS asa frame of reference when attempting to resolve conflicts. Aca-

demic institutions are considered potential participants becausealthough they participate in OCCOG discussions, they haveno voting power and as such cannot directly affect the outcomeof the sub-regional review process. Academic institutions can,however, use ACCESS to conduct research, and therefore ben-efit from project participation.

The technological core of ACCESS is a set of customizedGIS project files based around the regional analysis functions forwhich SCAG holds responsibility, such as the MPO for the South-ern California region. Regional project files include growth fore-casting for employment and general population growth;demographic analysis at the census tract and census block level;regional employment analysis using Standard Industrial Classifi-cation codes; and transportation analysis based on the RegionalTransportation Plan developed by SCAG. Local use of ACCESSis supported by: 1) property identification capabilities (useful onlywhen municipalities include their own property records; 2) land-use analysis based on the Anderson Land Use Classification Sys-tem; 3) a damage assessment routine that allows local jurisdictionsto tie their property files to an electronic form which can be sub-mitted electronically to the Federal Emergency ManagementAgency or the California Governor’s Office of Emergency Ser-vices; and 4) a general data viewer, containing information onecologically sensitive areas, political districts, and municipal gen-eral plans. In addition, local jurisdictions can supplement AC-CESS with any of their own information and combine it withACCESS project files.

Figure 2: SCAG ACCESS Organizational Flow Chart


By April 1997, 18 of the original 32 OCCOG member agen-cies chose to participate in ACCESS (Figure 3).

Participation in ACCESS is free to SCAG member agen-cies, although there is a $4,000 biannual subscription fee for theproject package for non-members. The software component canbe leased by itself from SCAG for $2,000. In addition to thesefees, there is a lease agreement with Thomas Bros. Maps, Inc. for$500 per year, regardless of whether a project participant is aSCAG member agency. Financing for local participation in AC-CESS was derived from three sources in the 1997/98 OCCOGbudget. The first two are related to California Assembly Bill 2766(AB 2766), managed by the Mobile Sources Air Pollution Re-duction Review Committee (MSRC). AB 2766 was implementedfor the purpose of improving regional air quality through trafficmanagement and the reduction of vehicle trips and harmful ve-hicle emissions. ACCESS can be used to review how local roadprojects and land uses tie into regional transportation plans,making AB 2766 discretionary funds available. Second, citiesreceived AB 2766 subvention funding to improve local air qual-ity and participate in projects of regional benefit. It was antici-pated that member cities participating in ACCESS would usesome of these funds to support the project. Third, funding forACCESS participation was specifically requested by the OCCOGin their SCAG Overall Work Program (SCAG-OWP) for 1997/98, totaling $50,000/$20,000 for application of GIS studies tothe Regional Transportation Plan (RTP) and $30,000 for a dem-onstration project analyzing building permit data in conjunctionwith subregional demographic data provided by the CDR.

Analysis. While ACCESS provides local jurisdictions with anew level of analysis, it has substantial problems. The largest dif-ficulties are in the areas of data accuracy and supplementation,

basic systems training and management, restrictive licensing agree-ments, and insufficient exploration of advanced spatial analysiscapabilities. These issues are examples of Evans and Ferreira’s dis-cussion of the interaction between technological and organiza-tional issues mentioned earlier; they are interconnected in such away that isolating them is not easily accomplished. First, SCAGdoes not provide property files for local jurisdictions, meaningthat cities have to supply this information on their own. This canbe an expensive obstacle to local system implementation. Pricesfor property information files can range between two and fivedollars per parcel, depending on the provider. In cities with thou-sands of parcels, individual acquisition may become prohibitivelyexpensive.

Currently, OCCOG is negotiating to acquire property filesfrom the private provider authorized by Orange County. Costsof $2 per parcel for initial distribution and $.60 per parcel foryearly updates were originally discussed with the data provider,although using the OCCOG as a collective negotiating forumresulted in a lower unit price ($1 per parcel, $.10 per parcel up-date fee). Regardless of who eventually does the format changes,a system will have to be developed to ensure that a usable endproduct is delivered. If there are inconsistencies in database for-mats across different GIS programs used to convert the propertyfiles, costs for corrections will be incurred (both in terms of imple-mentation delays and monetary considerations). The OCCOGis attempting to solve this difficulty of property supplementationby working with the CDR on a pilot project with a subgroup ofACCESS participants. The pilot project would convert and in-stall County parcel files on ACCESS machines in participatingcities, for the purpose of determining how such procedures canbe streamlined in the future.

Second, for the GIS information provided there is a seriousdisjunction between what ACCESS displays and what local par-ticipants need. Part of this is due to the differences in levels ofanalysis for the various ACCESS participants. For example,growth forecasts and environmental map coverages have beendeveloped at scales that are adequate for SCAG but useless to thelocal participants. In some cases, this has led to high levels ofspatial error (upwards of 200 feet) that are insignificant to a re-gional planning organization but of great concern to smaller ju-risdictions. In theory, the organizational structure of ACCESSshould correct for these errors of scale through the communica-tions process, making SCAG aware of the technical difficultiesthrough the sub-regional coordinator. More fundamentally dam-aging are the inadequacies of and the restrictions placed on theunderlying data. Based on 1990 U.S. Census data, ACCESS filesno longer reflect current trends in demographics or employment.Even if this were not the case, ACCESS coverages are not ar-ranged in such a way as to be user friendly. Components of AC-CESS cannot be extracted easily for use in other GIS projects, orwithin sub-projects of ACCESS itself. This inability to reorga-nize and update data is due to specialized programming codeassociated with ACCESS, restricting modification or movementof the master data files. In addition, because of the licensing agree-

Figure 3: Municipalities Participating in SCAG ACCESSProject as of April 1997


ments with software and data vendors, ACCESS cannot be placedon a municipal network without renegotiating with the respec-tive companies for a site license. Thus, there is only one terminalper city with full ACCESS capabilities, limiting the number oftasks it can serve at one time. This directly limits the utility ofACCESS as an interdepartmental GIS at the local level, and infact these data restrictions have driven away many potential par-ticipants from this project.

Third, the full abilities of ACCESS to conduct advancedanalysis as yet have not been clearly demonstrated to local par-ticipants. Training has been sparse and generally off topic, focus-ing on advanced systems well beyond the logistical capability ofproject participants or the ACCESS software itself. A few par-ticipants have used ACCESS in specific planning functions (e.g.,public works management) or as a training device for their staff,but none have fully integrated the system into their planningprocess. In addition to training difficulties, spatial analysis capa-bilities of ACCESS are weak and the upgrade capability unknown.ACCESS cannot run spatial statistics or provide three-dimen-sional analysis; the extensions that exist for this purpose have notbeen tested for their compatibility with the existing customizedprograms. It is entirely possible that the specialized code protect-ing the master data files will conflict with these spatial analysisextensions, yet local jurisdictions will need this ability if they areto participate in future regional modeling endeavors.

Contrast: USAC versus SCAG ACCESSIn order to compare the USAC and ACCESS case studies, a nec-essary first step is to discuss the evolution of computing environ-ments and examine how that has changed the organizationalcontext of distributed computing efforts. From a feasibility stand-point, technological change has made large-scale computing ingeneral cheaper and easier to comprehend. Kraemer and King(1977) noted that for full implementation of IMIS, 500 kilobyteswould be necessary for the individual terminal and eight mega-bytes of total computer storage capacity would be necessary. Suchfigures now are inadequate for running desktop systems evenwithout GIS. In addition, the command structures for using GIShave shifted from lines of code to graphical interfaces using on-screen “buttons,” divorcing the need to have detailed knowledgeabout computer systems and programming languages from theimplementation of GIS. This makes the learning of the day today operational tasks easier, increasing the likelihood that thesystem will be adopted at least in part. Relational database soft-ware now makes it relatively easier to maintain municipal recordsof greater detail and number, meaning more task-specific infor-mation can be collected, separated, and reported as needed. Desk-top modem technology, not even available at the time USACwas implemented, facilitates the transfer of information betweenjurisdictions. Such advances in capability between the time USACwas concluded and ACCESS was introduced to Orange Countysubstantially lessens the difficulties of participation and coordi-nation burdens placed on those who would design a multi-juris-dictional GIS today.

Still, even if the technology has been made drastically moreaffordable and understandable to the lay person, learning how touse it remains an obstacle. It was stated in the previous paragraphthat relational database software has made it relatively easier tomanage larger and more detailed municipal records. The natureof this simplification is that database information no longer needsto be part of the original GIS application program itself, mean-ing that changing municipal records does not require changingthe GIS that uses it. This should not be taken to mean that work-ing with a relational database application program is easy. Visu-alization of the logical relationships between records is difficult,and though all commercial database programs today adhere to acommon form of expressing database inquiries, each relationaldatabase program in existence can have its own special user inter-faces. Each interface can require a new session of training. Simi-larly, although all GIS can perform many of the same underlyingtasks (e.g., property identification, “within distance” location,boundary determination, image overlays), each can have a differ-ent user interface, requiring the user to be retrained. More fun-damentally, conceptualization of how a GIS works cannot beachieved without at least a concurrent (and preferably prior) ex-posure to fundamental GIS issues such as data management,graphics, and computer hardware.

The need for both technical and organizational assistancehas been a major stumbling block for both USAC and ACCESS.For USAC, part of these problems may have stemmed from thefact that group meetings were scheduled on a semi-annual basis.Infrequency of direct contact may have contributed to a lack ofconsistent communication and levels of training between thevarious branches of government and their respective goals (al-though the criticism of USAC’s lack of systems training is some-what tempered by the fact it was an attempt to develop aprototype). Additionally, Kraemer suggested that the role of theuniversities in USAC was “a double bind,” stating they were sup-posed to be simultaneously “friendly inside advisors and objec-tive outside critics.” The role of critic often prevented the flow ofinformation necessary for a complete analysis of the system, andthe role of friendly advisor prevented honest identification ofproject faults. (Kraemer and King 1977)

By comparison, ACCESS implementation in Orange Countyhas attempted to head off this problem through the recruitmentof a technical coordinator already within the project, throughthe CDR. Because CDR has been involved in providing supple-mentary data, it is well acquainted with the technical aspects ofACCESS and the problems described in the previous section.Monthly meetings provide a forum in which regular project up-dates and training issues can be addressed. But without moreassistance from SCAG, this input cannot be translated into solu-tions. Though the CDR technical coordinator may be able toidentify problems and perhaps develop temporary solutions,SCAG is responsible for making changes to the underlying sys-tem. Realizing this, SCAG is developing new versions of AC-CESS in conjunction with sub-regional coordinators to identifyand correct problems encountered in previous releases.


This development of ACCESS revisions in response to sub-regional comments demonstrates a difference in organizationalstyles between the USAC and ACCESS. HUD’s authoritarianmanagement structure and strict adherence to regulations as itbecame obvious that federal and local goals were diverging gen-erated an indifference, if not outright hostility, toward federalgoals at the local level. In contrast, although SCAG did not origi-nally deliver certain necessary components for local implementa-tion of the ACCESS project (namely, the local propertyidentification files), the project was intended to at least accom-modate, and hopefully integrate, with local analytic needs. Inaddition, the more frequent meeting of ACCESS participantsprovides greater opportunity to address differences in projectperspective.

It would appear the ends pursued by USAC and ACCESSare subtly different. The overall goals of IMIS development andACCESS are identical: Use common information displayed inthe same frame of reference for the purposes making local, re-gional, and eventually federal policy based on increased efficiencyof data gathering, greater accuracy of underlying data, and lowerredundancy. However, whereas USAC was implemented for thespecific purpose of determining whether a prototype system couldbe built and replicated with the current technology available,ACCESS is designed to involve local jurisdictions more directlyinto an existing regional planning process using established tech-nologies. Because ACCESS has the benefit of a stable technologybase and a broader forum for organizational input, this increasesthe likelihood that it could succeed where USAC fell short. Criti-cally, though, increased likelihood of success is not an ironcladguarantee. Identification of specific local goals to which ACCESScould be applied had to be conducted by the local jurisdictions;otherwise ACCESS would become just another good idea thatfailed in the implementation phase.

To that end, a survey was conducted by the OCCOG toidentify the general readiness of Orange County cities to partici-pate in a multi-jurisdictional GIS. The results are moderatelyencouraging. Out of 31 surveys distributed, 23 were returned invarious usable forms. ACCESS presents an initial exposure tothe benefits of a GIS for 12 of the 23 respondents, 8 of which arebelow the median size of Orange County cities ranked by popu-lation. Although two of those 12 cities are not SCAG members,and thus must pay the project subscription fees, these costs arestill substantially less than those incurred if these cities were toattempt to create their own GIS system independently. It is un-likely any of these cities would have a tax base broad enough todevelop GIS capabilities without the ACCESS program. In ad-dition to the reduced cost aspect compared to creating a newsystem, the areas of inquiry proposed as the basis for ACCESShave been well received by the participants. Of five project cat-egories users expressed having included in a regional GIS data-base (mapping and zoning, infrastructure planning, facilitiessiting, demographic and transportation analysis, and housingforecasts) all received more than 50 percent popular support, andfour received more than 75 percent. Recommendations for fu-

ture ACCESS projects include regional economic development,business license and crime trends, aerial photographs, and anInternet site where questions regarding ACCESS can be postedand answered.

Where Does ACCESS Go from Here?Enthusiasm for a cooperative effort is a necessary, but not a suf-ficient, condition for success. Using Budic’s organizational ac-ceptance criteria (1994) as a benchmark, some norms fordetermining success of GIS environments can be developed:breadth of user base, versatility of data and applications, ease ofuse, and project tenure. Using these criteria for judging ACCESS,there are serious obstacles to overcome.

Despite the expressed interest in the project, ACCESS execu-tion has not been widespread in Orange County. Training on howto use the system and incorporate it into local jurisdictions hasbeen inadequate and, as a result, interest at the local level has waned.In addition, component failures and data sharing restrictions ham-pered delivery of initial hardware and software systems to localparticipants; those machines that have been delivered have beeneither sidelined or converted to other uses within the cities. Com-ponent failures were attributed to reduced hardware quality im-posed by budget constraints, while data sharing restrictions wereput in place because of agreements with private data providers.Because ACCESS participation is voluntary and executed in a co-operative forum, these issues are known to non-participatingOCCOG members, hindering expansion of the system.

In spite of these failures, there is still interest on the part ofSCAG to pursue ACCESS in Orange County. The underlyinginterest in multi-jurisdictional GIS development has been ex-pressed, so the task at hand is to determine whether ACCESScan be salvaged. A good first step is a proposed implementationof pilot projects for ACCESS system upgrades in a few selectcities. This gives local participants a chance to see what problemsexist with current approaches, test new methods, and resolve dif-ficulties prior to implementing new procedures and system up-grades throughout the sub-region. Technological issues can inpart be solved by discussion between new ACCESS participantsand the technical coordinator for the OCCOG, but there arecertain standardization aspects that can be addressed now thatwill lessen these problems, which should not necessarily be leftup to a committee structure. These should be left to engineersand technical experts to ensure compatibility, accuracy, and effi-ciency in data collection and use throughout the system, both athigher levels of government as well as the local jurisdiction. In-consistencies in data can become a costly source of error, and areavoidable if a constant, logical set of procedures is established asa reference point.

Assuming it continues, ACCESS also will face organizationalchallenges in the coming years. ACCESS can act as the core fromwhich smaller cities build their own GIS, but in a larger sense,the ability to share and revise technical information at severaldifferent levels of government simultaneously will influence theway in which all forms of local and regional planning will take


place. In order to adapt, ACCESS implementation must be or-ganized in such a way as to facilitate new members beyond therelationship between the cities and SCAG. There already are signsthat higher levels of government are interested in pursuing sucha relationship in ACCESS specifically and distributed GIS ingeneral, which would greatly expand the project beyond its cur-rent relationship. Alternatively, the funds also could be used toupgrade the capabilities of existing multi-jurisdictional GIS par-ticipants. To be prepared for this type of policy change, SCAGand OCCOG should anticipate, and in fact promote, new rela-tionships between local, regional, state, and special jurisdictions,so that original goals are not inconsistent with new project re-quirements.

ConclusionsAs recently as 1991, multi-jurisdictional systems could not haveexisted; the technological base would have been insufficient, thecosts astronomical, and the conceptualization of the general ca-pabilities by project participants inadequate. Since then, com-puter processing power has grown exponentially, GIS have becomemore refined and user friendly, new sources of funding have beenidentified, and the literature on the types of problems GIS havebeen used to solve has grown. The necessary technological andfinancial support for such a system appear to have caught upwith what is a fundamentally sound strategy to further coopera-tive governance.

Multi-jurisdictional GIS applications also provide directionfor academic instruction and research. The pallet of case studiesinvolving GIS must now be broadened to include examinationsof how differences of local and regional interests (based in largepart on geographic jurisdiction and legal responsibilities) can bereconciled. This will require greater clarification of the role theGIS can play in topics traditionally related to planning such aspolitics, finance, and law. For those who would teach the appli-cation of GIS, instruction must not only include the creation ofuseful projects, but also how those projects relate to organiza-tions beyond the single client and how analysis must change withthe increase of interested and technologically capable parties.

At this point, a warning regarding the investigation and fu-ture application of multi-jurisdictional GIS is in order. Levineand Landis (1989) warned against what they called “the false godof comprehensiveness,” stating that no single GIS software pack-age can accomplish every conceivable task. Huxhold (1995) fur-ther stated that “a successful GIS is built, not bought,” arguingan organization should not build a GIS around the specific ap-plications software they choose. Rather, they should choose thesoftware and create necessary database expansions according tothe goals they wish to achieve and examine how those choicesinteract with other organizations with whom they interact on aconsistent basis. Those who would attempt to build a multi-ju-risdictional GIS must recognize that it cannot be designed at theoutset to solve every conceivable multi-jurisdictional planningissue. Instead, the system must be designed to facilitate expan-sion as new tasks arise.

Aside from these pieces of technical advice, there must bemutual respect between participants of a multi-jurisdictional sys-tem and a realization that individual members will use the sys-tem to further their own goals as well as group goals. This is thepolitical aspect of any type of cooperative planning. GIS applica-tions will not change that, although they may change the way inwhich these political issues are framed and resolved. To use atechnological analogy, the advent of the pocket calculator in en-gineering fields was a major advancement in terms of the speedat which analysis could be done. But the calculator did not changeany of the underlying principles on which engineering is based.The same is true of GIS. Just because the capability exists to runanalysis faster and in greater detail does not substitute for thefundamental physical, architectural, environmental, economic,or social principles of the planning process.

Even though USAC collapsed and ACCESS is in need ofserious revision, there is still reason to be optimistic about thepossibility of using GIS in a federal, state, and local multi-juris-dictional contexts. The first step in successful cooperative plan-ning and conflict mitigation is to get all relevant participantsreading from the same page. Multi-jurisdictional systems pro-vide that frame of reference, giving regional jurisdictions the abilityto see in greater detail how their plans affect smaller entities, andgiving municipalities the ability to use their data in planning pro-cesses beyond city limits.

Notes

1. Kaiser, Gerry, Traffic Engineer, City of Neenah, Wisconsin(Personal Communication, 6 July, 1999).

2. Levine, David. Database Administrator for Winnebago Geo-graphic System Project, County of Winnebago, Wisconsin(Personal Communication, 6 July 1999).

3. Bills, T. Principal of Data and GIS, Southern California As-sociation of Governments (Personal Communication, 6 May1997). The discussion of SCAG ACCESS entails moreproject specific analysis than the USAC example, due in nosmall part to the involvement of the author in implementa-tion of ACCESS in Orange County, California from Sep-tember 1996 to June 1997. The greater detail is in no waymeant to minimize the contributions of other research pre-sented here, but rather give as detailed a description as pos-sible to a case study which the author has personal knowledge.

4. Southern California Association of Governments, About Ac-cess, http://www.scag.org/public_docs/d62.htm

5. SCAG includes the area between Ventura County and Im-perial County as one geographic region. Orange County isone of seven subregions under SCAG jurisdiction.

6. Halls, D.K. Manager of Policy and Legislation, League ofCalifornia Cities, Orange County Division (Personal Com-munication, 19 June, 1998).


7. Walsh, D.J, GIS Specialist/Demographer, Center for De-mographic Research, California State University at Fuller-ton (Personal Communication, 14 July 1998).

8. A supplemental source of local information was disaggre-gated data from Orange County Projections, 1996 (OCP-96) under a Memorandum of Understanding between theCDR, the League of California Cities, the County of Or-ange, the Orange County Transportation and Fire Authori-ties, the Orange County Sanitation Districts and theTransportation Corridor Agencies. These data sets were pro-vided at the partial Census Tract level in an ArcView-readableformat. Among the deliverable products related to ACCESS,CDR developed the following; 1.) satellite imagery based landuse inventory update capabilities; 2.) master polygon files; 3.)multi-stage geocoding, and; 4.) interactive demographics abili-ties. These CDR project deliverables were to be used in con-junction with the street map database leased from ThomasBros. Maps to the participating agencies and distributedthrough SCAG. The Thomas Bros. data sets provide streetlevel and zoning information to be used as reference pointsfor the distribution of OCP-96 data within each city.

9. For example, the city of Huntington Beach was used as amodel during an April 1997 training session as an exampleof how a GIS could be used to connect building permit ap-plications across all the necessary departments simulta-neously, reducing the amount of time needed to completepermit reviews. This was a poor example for two reasons.First, this type of system could probably not be designedwith ACCESS because of the data and system constraintsplaced on the project. Second, the Huntington Beach ex-ample was developed entirely outside of ACCESS projectparticipation. The Huntington Beach example demonstrateda system which, while technically sophisticated and extremelyuseful, was beyond the scope of implementation.

10. The survey instrument can be obtained by contacting theauthor through the Department of Urban and Regional Plan-ning - School of Social Ecology, University of California,Irvine.

Acknowledgments

The author wishes to thank Scott Bollens, Dan Walsh, EricHeikkila, Marlon Boarnet, and three anonymous referees for theirtime and comments related to this work. Their assistance helpedto improve the clarity and comprehensiveness of this discussion.The author is further indebted to Janet Huston and Daryl Hallsat the League of California Cities, Orange County Division, andTerry Bills at the Southern California Association of Governmentsfor providing the research opportunity and access to the docu-mentation that was the genesis for this research. Any errors oromissions are the sole responsibility of the author.

References

Anthes, G.H., 1993, Phoenix Maps Out GIS Plan.Computerworld, 27(24), 81-84.

Bollens, S.A. and D.R. Godschalck, 1987, Tracking Land Sup-ply for Growth Management. Journal of the American Plan-ning Association, 53(4), 315-327.

Budic, Z.D., 1994, Effectiveness of Geographic InformationSystems in Local Planning. Journal of the American PlanningAssociation, 60(2), 244-263.

California Department of Conservation, Division of Mines andGeology, Index to Seismic Hazard Zone Maps (April 1998),http://www.consrv.ca.gov/dmg/shezp/maps.htm.

California Legislative Counsel, Search Bills: 1997-98 Session,Assembly Bill 1293 (September 1997), http://www.leginfo.ca.gov/bilinfo.html.

Calkins, H.W. and R. Weatherbe, 1995, Taxonomy of SpatialData Sharing. In Sharing Geographic Information, Onsrud,Harlan and Gerard Rushton (Eds.) (New Brunswick: Cen-ter for Urban Policy Research), 4, 65-75.

Cincinnati Area Geographic Information System, The CAGISOrganization (February 1997), http://www.cagis.hamilton-co.org/CAGIS/DataDictionary/Cagorganization.html

City Expands GIS Use, Explores Data Enhancements, 1997.American City & County, 112(12), 14-19.

Denning, J., 1993, Small-Government GIS. Civil Engineering,63(6), 52-54.

Eichelberger, P., 1992, Building the IS/GIS Capability to Sup-port Policies, Plans, and Programs: The Local GovernmentPerspective. In Wellar, B. and Parr, D. (Eds.), URISA 1992Annual Conference Proceedings, Washington, D.C., July 1992,5, 262-271.

Elliot, B., March 1996, A Successful GIS Project ManagementModel,. Government Technology, http://govt-t e c h . g ov t e c h . n e t : 8 0 / 1 9 9 6 / g t / m a r / p r o j m a n 2 /projman2.shtm.

Evans, J. and J. Ferreira Jr., 1995, Sharing Spatial Information inan Imperfect World: Interactions between Technical andOrganizational Issues. In Onsrud, H. and, G. Rushton (Eds.),Sharing Geographic Information. (New Brunswick: Centerfor Urban Policy Research), 26, 448-460.

Federal Geographic Data Committee, Resolution of the Mem-bership of the National States Geographic InformationCouncil (March 1999), http://www.fgdc.gov/nsdi/docs/nsgic.htm.

Heikkila, E.J., 1998, GIS Is Dead; Long Live GIS! Journal of theAmerican Planning Association, 64(2), 350-360.

Howland, M. and F.E. Lindsay, 1997, Where Do Tenants ComeFrom? Using a Geographic Information System to Study theDemand for New Office Space. Journal of the American Plan-ning Association, 63(3), 356-364.

Huxhold, W., 1991, An Introduction to Urban Geographic Infor-mation Systems (New York: Oxford University Press).


Huxhold, W., 1993, The Application of Research and Develop-ment from the Information Systems Field to GIS Imple-mentation in Local Government: Some Theories onSuccessful Adoption and Use of GIS Technology. In Masser,I. and, H. Onsrud (Eds.), Diffusion and Use of GeographicInformation Technologies (Boston: Kluwer Academic Publish-ers), 1.4, 61-75.

Huxhold, W.E. and A.G. Levinsohn, 1995, Managing GeographicInformation Systems (New York: Oxford University Press).

Indianapolis/Marion County GIS Team, Indianapolis/MarionCounty GIS History (December 1998), http://www.indygov.org/gis/history.htm.

Innes, J.E. and D.E. Simpson, 1993, Implementing GIS for Plan-ning: Lessons from the History of Technological Innova-tion. Journal of the American Planning Association, 59(2),230-236.

Intergraph Corporation, Intergraph Adopts Oracle 8 SpatialCartridge (June 1997), http://www.intergraph.com/venezuela/49.htm.

Kraemer, K.L. and J.L. King, 1977, A Requiem for USAC(University of California, Irvine: Urbis Group).

Leno, T., 1989, Is GIS Good for Local Government? PublicManagement, 71(6), 15-16.

Levine, J. and J.D. Landis, 1989, Geographic InformationSystems for Local Planning. Journal of the American PlanningAssociation, 55(2), 209-220.

Levine, N., 1996, Spatial Statistics and GIS: Software Tools toQuantify Spatial Patterns. Journal of the American PlanningAssociation, 62(3), 381-391.

Lober, D.J., 1995, Resolving the Siting Impasse: ModelingSocial and Environmental Locational Criteria with a Geo-graphic Information System. Journal of the American Plan-ning Association, 61(4), 482-495.

Maganinin, S.M., 1998, The Big Count. Planning, 64(5), 13-16.

Miller, L., 1995, GIS Branches into Urban Forest Management.American City & County, 110(2), 35-45.

Nale, D., 1995, Digital Orthophotography: The Foundation ofGIS. American City & County, 110(8): 26-39.

National Research Council, Assembly of Engineering, 1976, Lo-cal Government Information Systems/A Study of USAC andthe Future Applications of Computer Technology (Washington,D.C.: National Academy of Sciences).

Nedoviç-Budic, Z. and D. Godschalk, 1996, Human Factors inAdoption of Geographic Information Systems: A Local Gov-ernment Case Study. Public Administration Review, 56(6),554-567.

Nunn, S. and M.S. Rosentraub, 1997, Dimensions ofInterjurisdicitonal Cooperation. Journal of the American Plan-ning Association, 63(4), 205-219.

Office of Management and Budget, Circular No. A-16 Revised(October 1990), http://www.whitehouse.gov/OMB/circulars/a016/a016.html

Orange County Regional Advisory and Planning Council, 1996,Regional Advisory and Planning Council Agenda, ConsentItem I. A: Minutes from April 18, 1996, Meeting. SantaAna, California. 3.

Quinn, B., R. Allen and M. Sweeney, Moving Maintenance intothe Mainstream: The CAGIS Integrated GIS and Permit Man-agement System (April, 1999), http://www.cagis.hamilton-co.org/CAGIS/whitepaper/p444.htm

Shamsi, U.M., 1996, Storm Water Management Implementa-tion Through Modeling and GIS. Journal of Water ResourcesPlanning and Management, 122(2), 114-127.

Southern California Association of Governments, About AC-CESS, http://www.scag.org/public_docs/d62.htm, OverviewMap of SCAG, http://www.scag.org/bin/scag_map

Tosta, N. and P. Croswell, 1992, The Effects of Policies on Imple-mentation and Use of Information Technology. In Wellar,B. and, D Parr. (Eds.), URISA 1992 Annual Conference Pro-ceedings, Washington, D.C., July 1992, 5, 272-289.

Unwin, D., 1983, Introductory Spatial Analysis. In Huxhold,William, An Introduction to Urban Geographic InformationSystems (New York: Oxford University Press), 2, 57.

U.S. Department of Commerce (July 1983), 1980 Census ofHousing: Detailed Housing Characteristics, Wisconsin Part 51,1. Washington, D.C.: U.S. Department of Commerce, Bu-reau of the Census.

U.S. Department of Commerce (August 1983), 1980 Census ofPopulation: General Social and Economic Characteristics, Wis-consin Part 51, 1. Washington, D.C.: U.S. Department ofCommerce, Bureau of the Census.

U.S. Department of Commerce, Bureau of the Census, 1990Census Lookup (1.4a) http://venus.census.gov/cdrom/lookup/CMD=LIST/DB=C90STF3A/LEV=STATE.

Ventura, S.J. 1995. The Use of Geographic Information Systemsin Local Government. Public Administration Review, 55(5),461-467.

Williamson, C., 1998, All Counting Is Local. Planning, 64(5),14-15.

Winnebago System Reengineers Government, 1995. AmericanCity & County, 110(2), 38.

Think about your coworkers and peers.

Think about your IT vendors.

Think about students who aspire to work inthe field. And tell them about URISA!

NewKnowledgeDatabase

Visit the searchable knowledge database of URISA materials on theURISA website, http://www.urisa.org/topics.htm. The database cur-rently includes abstracts and papers (when available) from the threemost recent URISA Annual Conferences. URISA is working to add thepresentations from other conferences, as well as those from the URISAJournal and other educational publications owned by URISA. The Knowl-edge Database is available to URISA Members only and is password-protected (user = URISA2000, password = ORLANDO).

Did you know that most people join URISAbecause a member like you encouraged themto join?

URISA Journal ■ Special Report 45

Executive SummaryScenarios for geographic information use in the year 2010 suggestgreat potential to extend the capabilities of scientific researchers,decision-makers, and the public. This potential, however, will onlybe realized if there are substantial advances in Geographic Infor-mation Science, enhancing knowledge of geographic concepts andtheir computational implementations. To assess the needs for basicresearch in this emerging science and technology field, a workshopwas held at the National Science Foundation January 14-15, 1999.Workshop participants represented a broad range of the disciplinesinvolved in Geographic Information Science and technology. Theworkshop identified two important research streams: research inbasic Geographic Information Science (hereafter, GIScience), andresearch using geographic information systems (hereafter, GIS). Itis imperative that research in these two areas be integrated, as ap-plications motivate the science, and awareness of theory improvesapplications. Basic research in GIScience has several compellingcomponents. First is software integration, a general problem thatneeds specific research to solve its geospatial dimensions. Second,scale and resolution are spatial problems that interact with the scales(characteristic lengths) of environmental and social processes andwith data quality. Third, process models are a general computingproblem, but again geographic applications will require uniquelygeographic solutions. And fourth, usability of systems and tech-nologies is also a major component in need of research. In addi-tion, uncertainty and spatial dependence were recognized asimportant crosscutting research themes. GIScience is clearly a co-herent research field of strategic importance.

Workshop participants agreed that there is an urgent needfor a focused investment in GIScience, and that the NationalScience Foundation is the most appropriate U.S. agency to dothis. Such an investment is consistent with several importantnational trends, represented by the President’s Information Tech-nology Advisory Committee (PITAC) report, the Administration’s

A SPECIAL REPORT

Geographic Information Science:Critical Issues in an Emerging

Cross-Disciplinary Research Domain

Editor: David M. MarkNational Center for Geographic Information and Analysis

Department of GeographyState University of New York at Buffalo

Buffalo, New York 14261 USA

FY 2000 Information Technology for the Twenty-First Century(IT2) initiative, and the National Spatial Data Infrastructure. Theworkshop found that there is a coherent research communitypoised to make advances in GIScience if sufficient research sup-port is made available.

The workshop participants made the following recommen-dations to the National Science Foundation:

1. The National Science Foundation should recognize the im-portance of GIScience as a coherent research field, and shouldfocus a funding activity in this area as soon as possible.

2. Both basic GIScience, and research using GIS, should besupported from the new activity, to promote the integrationof these research areas.

3. The Foundation should establish an internal task force, withrepresentatives from all the Directorates and the Office ofPolar Programs, that would meet regularly to ensure thatthe new GIScience activity includes and benefits all relevantparts of the Foundation and their constituents.

4. The Foundation should appoint a multidisciplinary advi-sory panel of non-NSF personnel to assist in defining, imple-menting, and evaluating the effectiveness of this activity.

The University Consortium endorsed these recommenda-tions for GIScience in June 1999.

Visions from 2010Technological trends suggest that the world of the scientist willbe very different a decade from now. Information technology,communications infrastructure, microelectronics, and relatedtechnologies could enable unprecedented opportunities for dis-covery, and new ways to do research. To make this more con-crete, here are some visions of some aspects of the practice ofgeospatial1 research in the year 2010.


■ A geomorphologist doing fieldwork at a Long Term Eco-logical Research site in New Mexico slips on a headset as sheleaves the site office. The headset combines glasses, earphones,and a tiny microphone, and weighs little more than a pair ofsunglasses did in the 1990s. When she reaches her studyarea, she issues a simple voice command, and a red wire-frame display of the microtopography of the hill slope as itwas surveyed by a graduate student who worked here in 1997is superimposed on the landscape in front of her. Her en-hanced reality system lets her see areas where there has beensignificant erosion over the last decade, since there, the oldersurface appears like a net stretched above the ground. Afterexamining this simulated surface for a while, the researchertakes out a hand-held pointing device, and begins to pointat the current surface in various places and click on them. Asshe works, a green mesh appears, connecting the points shehas collected, and in places where it appears too far above orbelow the land surface, she collects additional points to makethe digitized data fit the real micro-topography. Over thenext several hours, she asks for a report on the total volumeof material removed in one area since the 1997 study, andviews a simulation of the runoff and erosion that might re-sult from a 5 cm per hour rainstorm. As she works, her dataappear simultaneously in the LTER data office and in thelab at the east-coast university where she is based, allowingher colleague to ask her for more surface height data in anearby area. The next day, she wears a slightly heavier head-set that incorporates a digital camera, so that an 8th gradescience class in Oklahoma City can join her for a half hourto see how fieldwork happens, asking her questions in realtime...

■ A sociologist is studying crime in a city in the northeasternUnited States, trying to understand the pattern of assaults.Sitting in front of a multimedia system, he requests that allassaults in the past year be shown on a map of the city. Whilehe is looking at the map, the system computes correlationswith other available data, and notes several phenomena thathave spatial associations with the crime data. One of these isan association between the pattern of assaults and the den-sity of bars (drinking establishments); the researcher acceptsthis particular suggestion, and the system adds the bar loca-tions to the map. Next, the researcher opens a modelingwindow and composes a rule: a “bar assault” is any assaultwithin 100 meters of a bar, between the hours of 5 p.m. and2 a.m. local time. He then asks for all bar assaults to beshown as yellow dots, and to display assaults that are not barassaults in red. Next, he has the system show area lighting,traffic patterns, and police patrol patterns. The system auto-matically runs standard correlations and plots summary as-sociations so that individual events can be examined. Thevariables with the highest correlations appear in a window,ordered from strongest to weakest correlation, and the sys-tem asks if he would like to see correlations between similarvariables that have been published in similar studies of other

U.S. cities. The system is providing this spatially enabledscientist with tools and methods to facilitate spatial think-ing and inference, spatial analysis, and spatial statistics. Thesystem automatically finds background data he requests,based on either coordinates or place names, checks that thedata are compatible in terms of scale, accuracy, and mapprojection, and integrates data from different sources auto-matically, leaving the researcher to concentrate on thinkingabout the crime patterns themselves and their possiblecauses...

■ Members of the general public, including school children,are obtaining detailed information about any place on Earththrough an intuitive interface that looks like a large manipu-lable globe. They rotate the globe to put any region in theforefront, or simply speak to the system to ask it to show aparticular place or region. As they zoom in, they see ever-increasing detail. The default view shows what the planetlooks like at the current moment from the chosen perspec-tive, but the user can ask for clouds to be removed, for theentire planet to be illuminated, or for thematic informationsuch as political boundaries, population densities, endan-gered species, or land values to be shown. One person usesthe system to travel back in time to look at agricultural pat-terns in southern Mexico in 1450. Another turns the timeback half a billion years, and then watches continents formand move into their present positions. Yet another travelsinto a possible future world in a global warming scenario; toproduce the images, the system invokes a Global ClimateModel developed several years earlier in a research centerthat has been made available to the public through this digi-tal earth. Although people without technical training easilyuse the digital earth, scientists and policy makers also usedata from digital earth as input to their models...

These scenarios may seem like science fiction, but much ofthe technology to support them is already available in prototypeform or is being developed: high-speed wireless information links,real-time multimedia satellite transmission, high-performancecomputing, global positioning systems (GPS) chips, content-basedretrieval from digital libraries. The development and dissemina-tion of such systems requires substantial advances in our knowl-edge of GIScience, associated knowledge of human-computerinteraction, and models of environmental and social processesthat shape our geographic world.

A Need for ResearchDramatic developments in communication, information andcomputational technologies alone promise to revolutionize ourlives even further. Advances in these fields will change the wayscience is performed and expand its capabilities dramatically. Theywill influence the ways we teach and learn—perhaps even theway we think. Our scientific adventures are far from over.2

GIS and spatial analysis methods are powerful tools for theanalysis and synthesis of geographically distributed phenomena,


and form a critical component of the information infrastructurefor science. Such systems are applicable to a wide variety of prob-lems, including many areas of basic and applied research. GIScienceis an inherently multidisciplinary field that underpins GIS.3 TheGIS software, data, and services industry is estimated at $4.2 bil-lion in the United States alone, and appears to be growing at around20 percent per year.4 GIScience research, and research using GIS,have been funded by the National Science Foundation through awide range of programs and other activities, including every Di-rectorate to some degree. However, except for a special solicitationissued in 1987 to establish the National Center for GeographicInformation Analysis, no program or special activity has focusedexplicitly and directly on GIScience and GIS.

In order to explore this situation, a workshop was held atthe National Science Foundation on January 14-15, 1999.5 Thegoals of the workshop were to explore the relationship ofGIScience to existing programs and initiatives at NSF, and toexamine the prospects for new initiatives or other activities in thearea of GIScience and geospatial information. Twenty research-ers from outside the Foundation, representing many of the disci-plines and fields active in and dependent on GIScience,participated in the workshop, along with many members of theNSF staff. The participants are listed in the Appendix to thisreport.

In the remainder of this report, we first distinguish two dis-tinct but deeply interconnected research areas: basic research inGIScience; and research using GIS. We present arguments re-garding why research in both areas is critical to the advancementof our knowledge of geographic information. They should bothreceive special funding emphasis at NSF, and the funding shouldbe awarded to catalyze multidisciplinary research that integratestwo research domains where possible. After reviewing four majorcomputing problems that have a role in GIScience research, weoutline several trends and opportunities in science, technology,and policy that would make an immediate NSF response verytimely. After providing a summary of the community that wouldbe likely to respond to an NSF activity in GIScience andGeospatial Activities, we close with specific recommendations tothe Foundation.

Research Using GISThe early development of GIS was led by applications in landmanagement and record keeping in government. GIS also havebecome powerful tools for researchers in the environmental andsocial sciences. GIS can support both exploratory and confirma-tory analysis, provide tools for both inductive and deductive ap-proaches, and support both scientific research and theimplementation of public policy based on GIS models. How-ever, GIS and geospatial technologies are not used in research aswidely as they could be or should be. One major barrier is thelack of interoperability among GIS and geographic informationtechnologies themselves, and between GIS and other informa-tion technology. Research communities often have their own soft-ware for pre-processing of sensor data or for analysis; however,

due to the nature of commercial GIS, it may be difficult or im-possible to integrate such software with components of com-mercial GIS, or to rewrite scientific models in the macro- ormodeling languages of GIS. Problems arise due to inadequatedocumentation of data quality, and the propagation of errorthrough GIS or other analysis. Differences due to scale and reso-lution also impede GIS adoption in some sciences. Second is abarrier related to dimensionality and temporality of geographicphenomena–current commercial GIS are essentially 2-dimen-sional and static. The scientific focus on processes and explana-tion of environmental phenomena may require three spatialdimensions, or time, or both, and this requires extensions to geo-graphic representations available in GIS. Third is an ease of usebarrier. Most GIS software today is not easy to use, but requiresextensive training. Issues of human-computer interaction impedescientific adoption of GIS, especially in fields where computerliteracy is not high. Across many scientific domains, researchersusing geographic information struggle when they attempt to com-pare or integrate their data with data collected and processed byothers. Important insights can be lost due to this impediment,which arises because of a lack of theory and methods to performintegration of geospatial information across different data mod-els, scales, and phenomena. Identifying and tracking derived orprocessed information relative to primary information also is acritical issue.

There is wide variability in the levels of adoption of GIS,spatial analysis, and related tools and methods across the envi-ronmental and social sciences. Scientific advances often are drivenby the availability of both analytical tools for analysis and datarequired by those tools, and some sciences are missing out on theinsights that could be provided through spatially-explicit prob-lem solving enabled by GIS. For example, several impedimentsto greater use of GIS by geologists have been identified, and aretypical of many other fields as well:

■ the lack of 3- and 4-D oriented spatial analysis tools; theseinclude mathematical, cognitive, and statistical tools;

■ the inability of generally available systems to accurately de-pict the natural variability of geologic features, or to repre-sent associated uncertainties;

■ the lack of access to subdiscipline-specific tools for explor-ing and modeling geologic systems– for example, many suchtools have been developed for the oil industry but are tooexpensive for all but a few universities to obtain; and

■ the lack of well done examples to help break the inertia ofscience. Most geoscientists know about GIS, but do not em-brace it because they wonder if the investment in time, ef-fort, and research dollars to deal with GIS will result in betterscience.

These or similar impediments likely are applicable in mostof the environmental and social sciences. But if and when suchimpediments are overcome, GIS can have a significant role inaccelerating diffusion of ideas across the disciplines. And since


untrained viewers often can understand maps and other graphicdisplays, GIS can serve an important role in communicating sci-ence to the public.

Research in GIScienceGIScience is the basic research field that seeks to redefine geo-graphic concepts and their use in the context of GIS. GISciencealso examines the impacts of GIS on individuals and society, andthe influences of society on GIS. GIScience re-examines some ofthe most fundamental themes in traditional spatially orientedfields such as geography, cartography, and geodesy, while incor-porating more recent developments in cognitive and informa-tion science. It also overlaps with and draws from more specializedresearch fields such as computer science, statistics, mathematics,and psychology, and contributes to progress in those fields. Itsupports research in political science and anthropology, and drawson those fields in studies of geographic information and society.

GIS are similar to many statistical packages, in that they arecommercial software systems widely used both within and out-side the research community. However, there is an important dif-ference between the two areas: statistics has a more universallyagreed-upon foundation, whereas there is not yet an equivalentmature foundation for GIS software. GIScience seeks to providethe theoretical foundation for GIS, just as the discipline of statis-tics provides foundations for statistical software. Again as in thecase of statistics, basic research in GIScience is a legitimate (thoughyoung) scholarly enterprise in its own right. But the positive ef-fects of GIScience on the GIS software industry, and on basicand applied research using GIS, are inescapable. And GIS areused in some of our most pressing societal problems, such ascrime, health, and disaster response.

Basic research in GIScience addresses complex problems thatrequire multidisciplinary solutions. Fundamental problems insuch a field are at risk of falling through the cracks between tra-ditional disciplines, especially in the absence of targeted fundingto support it. GIScience may begin with deep ontological ques-tions regarding the nature of space and phenomena in space. Isthe concept of space itself different among different fields of study?If there are differences, what are the common elements? Resultsfrom studies of spatial cognition and spatial language have rarelybeen used to build spatial query languages, and work in roboticson objects moving in space over time cannot easily be integratedwith work on spatio-temporal databases for moving objects. Evenmethods of spatial analysis developed in geography and regionalscience are often difficult to integrate into a GIS framework.

The Need to Integrate Theoretical and AppliedGIS ResearchThe workshop participants strongly endorsed an integrated ap-proach to both kinds of GIS-related research. This research fieldis clearly an area in which applications motivate the science. Dif-ficulties encountered in applying GIS to spatial problems andphenomena can expose interesting and significant problems re-

quiring basic GIScience research for their solution. Likewise,awareness of theory can improve applications, putting them on asolid conceptual foundation.

Enabling GIS Use through GIScienceIntegration, scale, process models, and usability are major researchissues facing GIScience. These research issues apply to a widerange of domains where digital computers are employed, but manyof the specific answers that GIS needs are unique to geographicinformation, and require GIScience research for their solution

IntegrationMarket forces that promote software integration andinteroperability in business have not had as much effect on sci-entific software, which often has been developed with the nar-row needs of a specific research community in mind. Existingtool systems do not always make it easy to respond to new tech-nologies for data collection or processing, and the problem iscomplicated by uneven levels of technical abilities and trainingacross the disciplines. GIS can serve as a frame for scientific dataintegration, but there are conceptual impediments to the inte-gration of some scientific models with software. Data fusion isan integration problem, and conflation, the process of combin-ing spatial data from different sources, is also critical. Whengeospatial data from different sources are combined, it is a chal-lenge to preserve the semantics inherent in the component datasets, unless each was prepared strictly according to a commonstandard. Data quality for the results of data fusion or conflationmay be difficult to characterize, especially with regard to posi-tional accuracy.

ScaleEven without formal training in cartography, most people realizethat the scale of a paper map influences the amount of detail thatcan be portrayed. But they may not realize how pervasive theinfluences of scale and resolution are on the analysis and otheruse of geospatial information in computers. Some of this is alegacy effect, since much geospatial information today was de-rived from maps. And for remotely sensed imagery, spatial reso-lution is a characteristic of the design of sensing instruments.Different measurement and positioning technologies will usu-ally produce data with different positional accuracy, capable ofresolving different levels of detail in geographic phenomena. Agreat deal of other geographic information is available not forpoints but for zones. For example, census data, a cornerstone ofmuch social science research, are spatially aggregated in order toprotect the confidentiality of individual records. The aggrega-tion rules are based on a minimum population, and thus are largerin low-density areas and smaller where population density is high.The zones may also change from one census to the next. Simula-tions have shown that correlations between variables may varyconsiderably under different aggregation scenarios, calling intosome question findings that are based only on analysis of data for


the zones that happen to have been selected by the census bu-reau. Such scale or aggregation effects are not restricted to thesocial sciences but have been addressed in ecology as well. If basicGIScience can produce new methods of analysis that can mini-mize effects of scale and resolution of new and legacy geospatialdata, this would have significant benefits for fields conductingresearch that uses geospatial information.

Process Models6

There is much knowledge embedded in the processes carried outto solve problems. Exploitation of geographical information notonly requires having the right maps, but also requires knowinghow to use them. The science of describing process knowledge ismuch less advanced than the science of describing data. For in-stance, even a simple task such as a search for the nearest restau-rant cannot now be expressed effectively in any of the toolsavailable on the Web. For spatial searches, these algorithms areclosely linked to geographic representations, and there is a recog-nized interplay between process models and information repre-sentation that is poorly understood. Research, experiments, anddevelopment are required to make the wealth of the data that isbecoming available fit the tools of the researchers and the public.Some of the key research topics here are spatial dynamic model-ing, non-linear transformations, linkage and integration of pro-cess models with information systems, computability, andvalidation. The computability dimension may include heuristicsto determine which solution method to use, based on the sizeand difficulty of the problem.

UsabilityUsability often refers to issues of human-computer interaction(HCI), user interface design, and training. Indeed, these issuesare of central importance, since if the user does not understandthe system and its user interface, the system will at best be usedinefficiently, and at worst will be used incorrectly and produceinvalid results. At the workshop, though, participants put usabil-ity in a much broader context. Access is an important factor inthe ability of a particular person to use GIS: access to data, accessto processing power, access to technology, access to training—allof these influence system use. More broadly yet, are people awarethat systems and data exist at all? If so, how can they find outwhat data are available, whether they are fit for use? Issues of howusers communicate their needs to the system get us back intoHCI. Typing and mouse clicks are not the only means of interac-tion, but the potential of other interaction methods such as sketch-ing, touch screens, and voice have hardly been examined for thegeospatial context. The usability issue should also examine moredeeply the potential value of collaborative decision-making acrossdistances for spatial decision support. Usability also has a societalcontext: a single system, based on the same data, should be ableto adapt to serve the needs of different sub-populations with dif-ferent backgrounds and needs.

Research ChallengesA different way to motivate basic research is through grand re-search challenges. The workshop program did not allow timefor consensus building regarding such challenges, but in theopening session of the workshop, one of the participants pre-sented four grand challenges for GIScience. Although they re-flect the particular priorities and curiosities of one workshopparticipant, they are representative of the fundamental scien-tific questions that will drive GIScience in the next decade.

Challenge 1: RepresentationThe central idea here is the challenge of representing the infinitecomplexity of the real world within the digital computer. Thereal world is usually thought of as a spatio-temporal continuum,whereas the digital computer has finite capacity, and representsconcepts and values in a discrete code. To meet this challenge,GIScience must examine the geographic concepts that are usedby environmental and social scientists in their research, includ-ing the ontology of reality at geographic scales. GIScience re-search in this area will be conducted by experts in geographictheory and geographic representations, by domain scientists whostudy geographically distributed environmental and social phe-nomena, by knowledge engineers and information scientists, andby philosophers.

The challenge: To find ways to express the infinite complex-ity of the geographical world in the binary al-phabet and limited capacity of a digitalcomputer.

Challenge 2: UncertaintyIf the representational challenge cannot fully be met, we mustaccept that geospatial data include uncertainty. This uncertaintycan include measurement error, error due to imperfect interpola-tion between measurements, gaps (incomplete data), artifacts ofgraphic or digital processing, and occasional blunders. Or it maybe due to the nature of the phenomena themselves, such as theextents of objects with indistinct or graded boundaries. The sci-entific measurement model is available for some aspects of un-certainty, but strong spatial dependencies complicate the situationconsiderably. And some spatial processes are essentially stochas-tic, and thus have an inherent uncertainty component.

The challenge: To find ways of summarizing, modeling, andvisualizing the differences between a digitalrepresentation and real phenomena.

Challenge 3: CognitionAlthough some information comes directly from sensor into spa-tial databases, human operators who use human judgment intheir work have developed much geospatial information. Datafrom maps have been through processes of symbolization, ab-


straction, and generalization. Thus inclusion of map-based in-formation in GIS means that cognitive concepts are already in-corporated into spatial databases. Many spatial or geographicabilities are part of common knowledge or common sense andare characteristics of most people by the time they reach the ageof 12. Use of GIS and spatial analysis tools may depend on theseabilities, but also on other concepts not learned until graduateschool. In some ways, this is similar to the challenge of represen-tation, except here it is the correspondence between binary rep-resentations, computational methods, and cognitive concepts thatis the focus.

The challenge: To achieve better transitions between cogni-tive and computational representations andmanipulations of geographic information.

Challenge 4: SimulationIt could be argued that one cannot claim to have understood aprocess if one cannot build a computer simulation of that pro-cess whose output cannot be distinguished from data about real-world instances of the process. Part of the popularity and impactof fractal mathematics has been the degree to which graphic ren-derings of fractal functions simulate, to some degree, landscapesor other natural phenomena. In detail, however, fractal math-ematics has little in common with geomorphic processes, and anexpert could quickly distinguish fractal terrains from real ones.Successful simulation of geographic phenomena would not onlyprovide a way to confirm process geographic processes, but italso can provide generic data for testing algorithm performance,graphical procedures, and other GIS methods.

The challenge: To create simulations of geographic phenom-ena in a digital computer that are indistin-guishable from their real counterparts.

The Data ChallengeOne additional technological trend that requires a response isthe increasing quantity of data being collected and archived. Thefact that very large volumes of scientific data were becoming avail-able was already evident in the late 1980s, but a dozen years later,data volumes are drastically higher again. Commercial remotelysensed data will soon be available at 1-meter resolution. Locally-produced data is being registered on the Internet in the clearing-houses of the National Spatial Data Infrastructure, and with GPSand wireless technologies making even more data available, theflood of geospatial data will be spectacular. However, the exist-ence of unimaginable quantities of data does not guarantee thatresearchers will find needed data more quickly and easily. Just asColeridge’s ancient mariner was thirsty when becalmed in a vastocean, the current scientists cry could be “Data, data everywhere,but not the information I need.” This will be a problem for allkinds of data of interest to scientists, but solutions for geographicinformation will need focused research efforts. Even if the right

data can be found, complex problems need sophisticated toolsthat may fail to be useful or usable, for reasons noted above.

The visions for 2010 presented in the first part of this reportwill be difficult to realize if the major research issues and grandresearch challenges presented in this section are not met head onby a concerted, coordinated program of multidisciplinary research.Such a research program will required the participation of na-tional funding agencies.

Trends and Opportunities

Why now?Research using GIS has been happening for decades, as has re-search into basic theories and concepts of computing about geo-graphic space. But, as noted above, current technologies andsocietal trends are producing great increases in availability of anddemand for GIS and services by all sectors of society. The tech-nology and systems will continue to be pushed by military, com-mercial, and administrative applications, and the research sectoris small by comparison. Without investment on behalf of theresearch community, the full potential of these systems and tech-nologies for scientific use is unlikely to be realized. Funding fromthe margins of other disciplines and programs is unlikely to pro-vide the kind of base funding that this emerging multidisciplinaryfield of study requires.

The President’s Information Technology Advisory Commit-tee7 was established to provide the government with guidanceand advice on all areas of high performance computing, commu-nications, and information technologies. The Interim PITACReport submitted in August 1998 noted that under funding ofresearch related to information technologies is a threat to “U.S.leadership in the emerging 21st-century information-basedeconomy.” The worldwide market for GIS software and serviceshas been projected to approach $4 billion in 1999, and U.S. com-panies appear to have more than half of the world market forGIS software. Including services as well as software and data,1998 market in the U.S. alone is estimated at $4.2 billion.8 TheU.S. GIS industry is very healthy at present, but could stall outin the future without sufficient government support of basic re-search into the theoretical foundations of GIScience. Many as-pects of current commercial systems are based in innovationsdeveloped in the public or academic sectors in the 1960s and1970s. The PITAC report specifically mentions GIS in a sectionon socio-economic impacts of an investment in information tech-nology research.

If the PITAC report leads to increased funding for basic re-search in information science and technology, a non-trivial pro-portion of those funds should be specifically directed towardGIScience research. The general public easily understands manyaspects of GIS and related technologies, and an investment inthis area is likely to produce tangible benefits obvious to manysectors of society. Such investment could also help assure thatU.S. industry will continue to lead in this area. GIS software also


serves other industries, and is used in data management by alllevels of government. The Next Generation Internet (NGI) andthe development of a Global Disaster Information Network(GDIN) are other trends that are consistent with the need forfurther advances in GIScience and related technologies. Mobileinformation systems with positioning systems will provide newopportunities and challenges for the telecommunications indus-try, and have positive implications for the broader field of infor-mation technology.

Why NSF?GIScience needs a broad, cross-disciplinary coverage and activeinvolvement of researchers in order to make necessary advances.GIScience needs research in theoretical geography, mathematics,cognitive science, and basic computer and information science,as well as in areas of science and engineering where GIS is ap-plied to scientific problems, such as ecology, earth science, socialsciences, and other areas. Clearly, GIS and GIScience researchshould be supported by many federal agencies. However, theNational Science Foundation is the only agency in the UnitedStates that supports all aspects of the GIScience and GIS, fromtheoretical topics to scientific applications. The National ScienceFoundation is well placed to make a difference to scientific andtechnical progress in GIScience.

There are parallel initiatives already being planned withinthe Foundation as well. NSF’s Directorate for Geosciences (GEO)is undertaking a major long-range planning effort to develop avision of the cutting-edge issues for the geosciences during thefirst decade of the 21st century.9 One of the driving forces be-hind this effort, code-named “GEO Beyond 2000,” is improvedscientific models, and the other is technology.

The second development is the revolutionary increase in thecapability of computer information, and sensor technologies. Ourcurrent ability to monitor and observe the Earth system on allspatial and temporal scales is unprecedented, and, when coupledwith our ever-increasing ability to store and retrieve vast quanti-ties of archived information for detailed examination, providesfor much more rapid knowledge generation and dissemination.These rapidly improving technologies will advance the scientificresearch to provide the sophisticated tools and monitoring sys-tems that policy makers will need to make informed decisions.

If GEO Beyond 2000 goes forward, it could contribute sig-nificantly to needs for ‘Research using GIS’ in the geosciences.However, it is unlikely that it would provide the broadmultidisciplinary support for basic GIScience and for researchusing GIS in the biological and social sciences.

“This Just In...”In the months following the workshop, potential support forcomputer and information science and technology continued toimprove. The U.S. Administration’s FY 2000 budget includes$366 million of new money for computing and communicationstechnology, to be implemented through several Federal agencies,

of which NSF’s $146 million is the largest component.10 NSF’sFY 2000 budget11 states:

NSF has been asked to serve as lead agency for theAdministration’s FY 2000 Information Technology for theTwenty-First Century (IT2) initiative. IT2 grew from the ef-forts of several agencies and responds to recommendationsmade by the President’s Information Technology AdvisoryCommittee (PITAC), which termed federal support for in-formation technology “dangerously inadequate”. Partneragencies include the Departments of Defense and Energy,the National Aeronautics and Space Administration, theNational Institutes of Health, and the National Oceanic andAtmospheric Administration. IT2 involves a total federalinvestment of approximately $366 million in FY 2000.NSF’s FY 2000 investment in IT2 totals $146 million. Thisincludes $110 million funded through NSF’s Computer andInformation Science and Engineering Activity for researchin software systems, scaleable information infrastructure, andhigh-end computing.

The participants in the present GIScience workshop felt thatspecific GIScience research is needed to solve geospatial dimen-sions of information technology, we recommend that a portionof any increased research money for information technology bedirected specifically toward GIScience.

A Research Community Ready to RespondFunding initiatives such as the one we are recommending areunlikely to be effective in the absence of a community of scholarsthat provides both potential applicants as well as norms againstwhich proposals would be evaluated. GIScience in the UnitedStates has such a community. Although the term Geographic In-formation Science was coined only in the early 1990s, it labeleda science and engineering field that had been emerging duringthe previous two decades as a consequence of technological tran-sitions in the mapping sciences, convergence of spatial analysismethods, and the development of new technologies for collec-tion and processing of geospatial data. In 1994, representativesof 34 U.S. universities and other research organizations met anddecided to establish an organization “dedicated to the develop-ment and use of theories, methods, technology, and data for un-derstanding geographic processes, relationships, and pattern.”Named the University Consortium for Geographic InformationScience (UCGIS), the organization has grown to include 50 uni-versities and four other organizations as full members, and someof the leading U.S. IT and GIS firms as affiliate members.12 TheUniversity members are estimated to have more than a thousandindividual GIS-related researchers and educators, and these areonly part of the U.S. academic research community. In 1996, theUCGIS established 10 “National Research Priorities” that havebeen used to promote research in the field,13 and in 1997, theydetermined eight “National Education Priorities”.14 Taken to-gether, the UCGIS research priorities cover a rather broad range


of specific research problems in GIScience, and thus these priori-ties have a different granularity from the research areas discussedin this report. Other resources for the field include the NationalCenter for Geographic Information and Analysis and its currentProject Varenius (“Advancing Geographic Information Science”),the Open GIS Consortium, the Federal Geographic Data Com-mittee, and a good number of refereed journals15 and scholarlyconferences. The OpenGIS Consortium16, established in 1994to provide a formal structure and process for developing a speci-fication for interoperable geoprocessing, is another indication ofa research and development community. Primarily composed ofprivate sector organizations, OpenGIS also includes governmentand academic participants. Thirdly, in the public sector, the Fed-eral Geographic Data Committee (FGDC)17 was established bya Presidential Order, also in 1994, to support public and privatesector applications of geospatial data in such areas as transporta-tion, community development, agriculture, emergency response,environmental management, and information technology; stateand local governments are involved through the National StatesGeographic Information Council (NSGIC) and the NationalAssociation of Counties (NACo).

Methods and Levels of FundingIn a climate where research funding is widely acknowledged tobe tight, it is difficult to know what level should be devoted tonew activities in GIScience. There are, however, several recent orfuture tendencies that may provide some basis for such decisions.For example, recently, NSF devoted $50 million per year to theKnowledge and Distributed Intelligence (KDI) initiative. For thefuture, the PITAC report recommended a diverse portfolio offunding for information technology, including single-investiga-tor efforts, multi-investigator projects, and centers. One of themore novel ideas presented in the PITAC report is “Expeditionsinto the 21st Century,” described as follows:

“Expeditions into the 21st Century” will be virtual centersthat bring together scientists, engineers, and computer sci-entists from academia, government, and industry to “live inthe technological future.” The mission of these expeditionswill be to report back to the Nation what could be accom-plished by using technologies that are quantitatively andqualitatively more powerful than those available today.

They recommend that such Expeditions be very well funded:

The full term of an Expedition would be ten years. To en-courage truly aggressive efforts, very high annual fundinglevels should be possible, say up to $40 million per center.

PITAC also recommended “Enabling Technology Centers”at up to $10 million per year.

The workshop participants believe that investment on thatorder of magnitude is needed to have a real influence on amountof basic GIScience research conducted and the number of disci-plines now using GIS and geospatial methods in their research.

Considering the size of the U.S. GIS industry and of govern-ment annual expenditures on GIS, GIScience research fundingon the order of $40 million per year is readily justified. NSF hasavailable many models for administering and funding research,ranging from individual proposals and programs to cross-Direc-torate initiatives and Centers. Educational initiatives should notbe ignored, nor should physical infrastructure such as comput-ing hardware or buildings to house GIScience research activities.Geographic information pervades, or should pervade, all com-puting about the environment and society, for administration,management, planning, disaster response, and business as well asapplied research. GIScience provides the basic intellectual un-derpinnings for geographic information technologies, andGIScience research should be supported at levels appropriate tothe importance of these technologies and their application.

RecommendationsThe workshop participants make the following recommendationsto the National Science Foundation:

1. The National Science Foundation should recognize the im-portance of GIScience as a coherent research field, and shouldestablish a funding activity in this area as soon as possible.

2. Both basic GIScience, and research using GIS, should besupported from the new activity, to promote the integrationof these research areas.

3. The Foundation should establish an internal task force, withrepresentatives from all the Directorates and the Office ofPolar Programs that would meet regularly to ensure that thenew GIScience activity includes and benefits all relevant partsof the Foundation and their constituents.

4. The Foundation should appoint a multidisciplinary advi-sory panel of non-NSF personnel to assist in defining, imple-menting, and evaluating the effectiveness of this activity.

At its Council meeting on June 26 1999, the UniversityConsortium for Geographic Information Science passed a reso-lution that “UCGIS strongly supports the recommendations toNSF set forth in the NSF-funded workshop report titled ‘Geo-graphic Information Science: Critical Issues in an Emerging Cross-Disciplinary Research Domain’.”18


Appendix A: Workshop Participants

Workshop Co-Chairs:David M. Mark, Geography, State University of New York at BuffaloLeal A. K. Mertes19, Geography, UC Santa BarbaraRichard R. Muntz, Computer Science, UCLA

Steering Committee:Max J. Egenhofer, Spatial Information Science and Engineering,

University of MaineMichael F. Goodchild, Department of Geography, University of

California-Santa BarbaraCharles M. (Chuck) Meertens20, UNAVCOBarbara Tversky, Psychology, Stanford University

Other Workshop Participants:Lawrence E. Band, Geography, University of North CarolinaRoy K. Dokka, Geology and Geophysics, Louisiana State Uni-

versitySusan L. Epstein, Computer Science, Hunter College, City Uni-

versity of New YorkStephen C. Hirtle, Information Sciences, University of PittsburghDavid R. Janecky, Geochemistry, Los Alamos National Labora-

toryCarol A. Johnston, Natural Resources Research Institute, Uni-

versity of Minnesota, DuluthStephanie King, John A. Blume Earthquake Engineering Center,

Stanford UniversityWerner Kuhn, Geoinformation, University of Muenster, Ger-

manyHarvey J. Miller, Geography, University of UtahDonna J. Peuquet, Geography, Pennsylvania State UniversityHanan Samet, Computer Science, University of MarylandEric S, Sheppard, Geography, University of MinnesotaMichael Stein, Statistics, University of ChicagoThomas M. Usselman, Earth Sciences, National Academy of

ScienceGio Wiederhold, Computer Science, Stanford University

NSF Observers:Frank D. Anger, Computer-Communications Research, CISEThomas J. Baerwald, Deputy Assistant Director, GEOBernard O. Bauer, Geography and Regional Science, SBEBennett I. Bertenthal, Assistant Director for SBEScott G. Borg, Office of Polar ProgramsLawrence E. Brandt, Experimental and Integrative Activities,

CISEJohn M. Briggs, Ecology, BIOJohn C. Cherniavsky, Senior Advisor for Research, EHRAlan M. Gaines, Senior Science Associate for Spatial Data and

Information, GEOJulie Palais, Office of Polar Programs

Rita V. Rodrigues, Experimental and Integrative Activities, CISEJames L. Rosenberger, Statistics and Probability, MPSMichael H. Steuerwalt, Applied Mathematics, MPSMaria Zemankova, Information and Data Management, CISE

Notes

1. The term geospatial is used in this report to refer to spatialinformation (positions, sizes, shapes, orientations, relations)for phenomena at geographic scales. In contrast, spatial in-formation could refer to the same characteristics at any scale,from sub-molecular to intergalactic. Geographic informa-tion is used to refer to both geospatial and non-spatial at-tributes of geographic phenomena.

2. House Committee on Science, 1998. Unlocking Our Fu-ture: Toward a New National Science Policy: A Report toCongress. http://www.house.gov/science/science_policy_report.htm

3. In this report, we take a broad view of GIS as any softwarefor handling geographic information, and do not limit theterm to current commercial off-the-shelf software.

4. National Academy of Public Administration, 1998. Geo-graphic Information for the 21st Century: Building a Strat-egy for the Nation. Washington, DC: National Academy ofPublic Administration, Report 98-01, p. 298.

5. Funding for this workshop was provided by the NationalScience Foundation, through a supplement to award SBE-9600465. Support from the Foundation is gratefully ac-knowledged. In addition to the members of the SteeringCommittee, Bernard Bauer, Alan Gaines, and MariaZemankova (all of NSF) and Patricia Shyhalla (Univer-sity at Buffalo) were especially helpful in the planning ofthe workshop.

6. The term process model is used here to refer to computa-tional processes, and not to models of physical or social pro-cesses.

7. http://www.ccic.gov/ac/

8. NAPA report, 1998, op cit., p. 298.

9. http://www.geo.nsf.gov/adgeo/geo2000/

10. http://www.access.gpo.gov/usbudget/fy2000/pdf/budget.pdf, page 107.


11. National Science Foundation, Fiscal Year 2000 Budget Re-quest, Overview. http://www.nsf.gov/bfa/bud/fy2000/overview.htm.

12. http://www.ucgis.org/

13. UCGIS, 1996. Research Priorities for Geographic Infor-mation Science. Cartography and Geographic InformationSystems 23(3). http://www.ncgia.ucsb.edu/other/ucgis/CAGIS.html

14. http://www.ncgia.ucsb.edu/other/ucgis/ed_priorities/contents.html

15. International Journal of Geographical Information Science,Cartography and Geographic Information Systems,Geoinformatica, Transactions in GIS, Geographical Systems,Spatial Cognition and Computation, and others.

16. http://www.opengis.org/

17. http://www.fgdc.gov/

18. http://www.ucgis.org/CM-MN.html

19. Leal Mertes and Chuck Meertens are members of the work-shop Steering Committee but were unable to attend theworkshop.

20. See previous note.

URISA Journal ■ Review 55

Review of Current Journal Literature

Zorica Nedovic

Editorial Intent

For the Review of Current Journal Literature, we have selected 20 journals of related interest to URISA members (see next page forlist of journals). The Journal Literature editor scans these journals for articles she feels are relevant to the URISA audience. Selectedarticles are then assigned to one of nine categories. The nine categories are:

■ Analysis, Modeling, and Simulation■ Applications■ Cartography■ Data■ Hardware/Software Technology■ Implementation and Management■ Remote Sensing and GPS■ System Concepts and Theory■ Other Issues and Topics

Please note that there are no clear-cut boundaries between categories, and some articles may qualify for entry in more than one.However, we decided against repetitive entries for a single article, so the reader should be advised to look in more than one categoryfor a particular entry. Also, note that some entries retrieved online lack full information on page numbers.

If you have suggestions or comments about our procedures or about the section, please contact URISA.

Zorica Nedovic

In this issue…

Editor Zorica Nedovic continues to be the driving force behind this section. Thanks also to Ted Koch, of the Wisconsin StateCartographer’s Office, who reviewed the cartographic journals; and Ken Dueker, of Portland State University, who reviewedTransportation Research Record.


Review of Current Journal LiteratureZorica Nedovic, Editor

Selected Journals

IJGIS International Journal of Geographical InformationScience (monthly)Taylor and Francis Ltd., 1. Gunpowder Square,London EC4A 3DE, UK

SLIS Surveying and Land Information Systems(quarterly)American Congress on Surveying and Mapping,5410 Grosvenor Lane,Suite100, Bethesda, MD 20814-2122

CGIS Cartography and Geographic InformationSystems/Science (quarterly)American Congress on Surveying and Mapping,5410 Grosvenor Lane,Suite 100, Bethesda, MD 20814-2122

CA Cartographica (quarterly)University of Toronto Press, 5201 Dufferin Street,Downsview, OntarioM3H 5T8, Canada

PERS Photogrammetric Engineering and RemoteSensing (monthly)American Society of Photogrammetry and RemoteSensing, 5410 Grosvenor Lane, Suite 210,Bethesda, MD 20814-2160

IJRS International Journal of Remote Sensing(monthly)Taylor and Francis Ltd., 1. Gunpowder Square,London EC4A 3DE, UK

CG Computers and Geosciences (monthly)Pergamon Press Ltd., Linacre House, Jordan Hill,Oxford OX2 8DP, UK

MISQ Management Information Systems Quarterly(quarterly)Carlson School of Management, University ofMinnesota, 271 19th Ave.South, Minneapolis, MN 55455

JAPA Journal of the American Planning Association(quarterly)American Planning Association, 122 S. MichiganAve., Suite 1600, Chicago, IL 60603

EPB Environment and Planning B (bimonthly)Pion Limited, 207 Brondesbury Park, LondonNW2 5JN, UK

CEUS Computers, Environment and Urban Systems(bimonthly)Pergamon Press, Inc., Fairview Park, Elmsford,New York 10523

LUP Landscape and Urban Planning (bimonthly)Elsevier Science B.V., Journal Department, P.O.Box 211, 1000 A.E. Amsterdam, The Netherlands

EM Environmental Management (bimonthly)Springer-Verlag New York Inc., 175, Fifth Avenue,New York, NY 10010

JEM Journal of Environmental Management (monthly)Academic Press Ltd., 6277 SeaHarbor Drive,Orlando FL 32887-4900

PAR Public Administration Review (bimonthly)American Society for Public Administration(ASPA), 1120 G Street NW, Suite 700,Washington, DC 20005-3885

JUA Journal of Urban Affairs (quarterly)JAI Press, Inc., 55 Old Post Road No.2, Box 1678,Greenwich, CT 06836-1678

TRR Transportation Research RecordTransportation Research Board, National ResearchCouncil, 2101 Constitution Avenue, Washington,DC 20418

EDQ Economic Development Quarterly (quarterly)Sage Publications, 2455 Teller Road, ThousandOaks, California 91320

JRS Journal of Regional Science (quarterly)Regional Science Department, University ofPennsylvania, 3718 Locust Walk, Philadelphia,Pennsylvania 191014-6209

1. Analysis, Modeling, and Simulation

Gustafson, EJ, and Crow, TR. 1998. Simulating Spatial and Tem-poral Context of Forest Management Using HypotheticalLandscapes. EM 22(5): 777-787.

Hossain, M, and McDonald, M. 1998. Modelling of Traffic Op-erations in Urban Networks of Developing Countries: A Com-puter Aided Simulation Approach. CEUS 22(5): 465-484.

Hwang, DM, Karimi HA, and Byun, DW. 1998. UncertaintyAnalysis of Environmental Models Within GIS Environ-ments. CG 24(2): 119-130.

Kernohan, BJ, Millspaugh, JJ, Jenks, JA, and Naugle, DE. 1998.Use of an Adaptive Kernel Home-Range Estimator in AGIS Environment To Calculate Habitat Use. JEM 53(1):83-89.

Landis, J, and Zhang, M. 1998. The Second Generation of theCalifornia Urban Futures Model - Part 1 – Model Logicand Theory. EPB 25(5): 657-666.

Landis, J, and Zhang, M. 1998. The Second Generation of theCalifornia Urban Futures Model. Part 2: Specification andCalibration Results of The Land-Use Change Submodel.EPB 25(6): 795-824.

Lee, J, and Stucky, D. 1998. On Applying Viewshed Analysisfor Determining Least-Cost Paths on Digital ElevationModels. IJGIS 12(8): 891-905.


Marker, JT Jr., and Konstadinos, GG. 1998. Truck Traffic Pre-diction Using Quick Response Freight Model Under Dif-ferent Degrees of Geographic Resolution: GIS Applicationin Pennsylvania. TRR 1625: 118-123.

Marr, AJ, Pascoe, RT, Benwell, GL, and Mann, S. 1998. Devel-opment of a Generic System for Modelling Spatial Pro-cesses. CEUS 22: (1) 57-70.

Murray, AT, and Estivillcastro, V. 1998. Cluster Discovery Tech-niques for Exploratory Spatial Data Analysis. IJGIS 12(5):431-443.

Obee, AJ, Griffin, EC, and Wright, RD. 1998. Using A GIS toOvercome Data Adversity — Industrial Air Pollution RiskModeling in Tijuana, Mexico. PERS 64(11): 1089-1096.

Semboloni, F. 1997. An Urban and Regional Model Based onCellular Automata. EPB 24(4): 589-612.

Sui, DZ. 1998. GIS-Based Urban Modelling — Practices, Prob-lems, and Prospects. IJGIS 12(7): 651-671.

Swetnam, RD, Ragou, P, Firbank, LG, Hinsley, SA, and Bellamy,PE. 1998. Applying Ecological Models to Altered Land-scapes — Scenario-Testing With GIS. LUP 41(1): 3-18.

Veregin, H. 1999. Error Propagation in Environmental Model-ing with GIS. IJGIS 13(2): 186-188.

Vieux, BE, Mubaraki, MA, and Brown, D. 1998. Wellhead Pro-tection Area Delineation Using a Coupled GIS andGroundwater Model. JEM 54(3): 205-214.

Wang, B, and Manning, RE. 1999. Computer Simulation Modelingfor Recreation Management: A Study on Carriage Road Use inAcadia National Park, Maine, USA. EM 23(2): 193-203.

Westervelt, JD, and Hopkins, LD. 1999. Modeling Mobile In-dividuals in Dynamic Landscapes. ISGIS 13(3): 191-208.

Wu, F, and Webster, CJ. 1998. Simulation of Land Develop-ment Through the Integration of Cellular Automata andMulticriteria Evaluation. EPB 25(1): 103-126.

Wu, F. 1998. Simulating Urban Encroachment on Rural LandWith Fuzzy-Logic-Controlled Cellular Automata in a Geo-graphical Information System. JEM 53(4): 293-308.

Wu, F, and Webster, CJ. 1998. Simulation of Natural Land UseZoning Under Free-market and Incremental DevelopmentControl Regimes. CEUS 22: (3) 241-256.

Yates, PM, and Bishop, ID. 1998. The Integration of ExistingGIS and Modelling Systems: With Urban Applications.CEUS 22: (1) 71-80.

2. Applications

Alam, M, and Fekpe, E. 1998. Application of GIS Technologyin Freight Data Analysis: Case Study of I-90/I-94 Corri-dor Analysis. TRR 1625: 173-183.

Ambrosia, VG, Buechel, SW, Brass, JA, Peterson, JR, Davies, RH,Kane, RJ, and Spain, S. 1998. An Integration of RemoteSensing, GIS, and Information Distribution for Wildfire De-tection and Management. PERS 64(10): 977-985.

Bishop, ID. 1998. Planning Support: Hardware and Softwarein Search of a System. CEUS 22(3): 189-202.

Bresnahan, PJ. 1998. Identification of Potential Hazardous WasteUnits Using Aerial Radiological Measurements. PERS64(10): 995-1001.

Cousins, SAO,and Ihse, M. 1998. A Methodological Study forBiotope and Landscape Mapping Based on CIR Aerial Pho-tographs. LUP 41(3-4): 183-192.

Craig, WJ. 1998. The Internet Aids Community Participationin the Planning Process. CEUS 22(4): 393-404.

Deppe, F. 1998. Forest Area Estimation Using Sample Surveysand Landsat MSS and TM Data. PERS 64(4): 285-292.

Ding, C. 1998. The GIS-Based Human-Interactive TAZ DesignAlgorithm -Examining the Impacts of Data Aggregation onTransportation-Planning Analysis. EPB 25(4): 601-616.

Edamura, T, and Tsuchida, T. 1999. Planning SupportSystem for an Urban Environment Improvement Project.EPB 26(3):381-392.

Elwood, S, and Leitner, H. 1998. GIS and Community-basedPlanning: Exploring the Diversity of Neighborhood Per-spectives and Needs. CGIS 25(2): 77-88.

Gamba, P, and Casciati, F. 1998. GIS and Image Understandingfor Near-Real-Time Earthquake Damage Assessment.PERS 64(10): 987-994.

Grabaum, R, and Meyer, BC. 1998. Multicriteria Optimizationof Landscapes Using GIS-Based Functional Assessment.LUP 43(1-3): 21-34.

Harris, B. 1999. Computing in Planning: Professional and In-stitutional Requirements. EPB 26(3): 321-332.

Hopkins, LD. 1999. Structure of a Planning Support Systemfor Urban Development. EPB 26(3): 333-343.

Hsiao, N, Sato, S, Arima, T, Hagishima, S, Kim, K, and Kameno,T. 1998. Using Computer Graphics to Compare the VisualEnvironments of Urban Streets in Japan and Taiwan. CEUS22(3): 277-298.

Jensen JR, Halls, JN, and Michel, J. 1998. A Systems Approachto Environmental Sensitivity Index (ESI) Mapping for OilSpill Contingency Planning and Response. PERS 64(10):1003-1014.

Kammeier, HD. 1999. New Tools for Spatial Analysis and Plan-ning as Components of an Incremental Planning-SupportSystem. EPB 26(3): 365-380.

Kim, T, and Muller, JP. 1998. A Technique for 3D BuildingReconstruction. PERS 64(9): 923-930.

Klosterman, RE. 1999. The What if? Collaborative PlanningSupport System. EPB 26(3): 393-408.

Klosterman, RE, and Sikdar, PK. 1998. Computer Support forUrban Planning and Management. CEUS 22(3): 185-188.

Lathrop, RG, and Bognar, JA. 1998. Applying GIS and Land-scape Ecological Principles to Evaluate Land ConservationAlternatives. LUP 41(1): 27-41.

Lee, J, Tian, L, Erickson, LJ, and Kulikowski, TD. 1998. Ana-lyzing Growth-Management Policies with Geographical In-formation Systems. EPB 25(6). 865-880.


Levine, N, and Kim, KE. 1998. The Location of Motor VehicleCrashes in Honolulu: A Methodology for Geocoding In-tersections.” CEUS 22(6): 557-576.

Li, X. 1998. Measurement of Rapid Agricultural Land Loss inthe Pearl River Delta With the Integration of Remote Sens-ing and GIS. EPB 25(3): 447-461.

Mainguenaud, M. 1997. Constraint-Based Queries in a Geo-graphical Database for Network Facilities. CEUS 20(2):139-151.

Mandelbaum, SJ. 1997. Making and Breaking Planning Tools.CEUS 20(2): 71-84.

Martin, D. 1998. Automatic Neighbourhood Identification fromPopulation Surfaces. CEUS 22(2): 107-120.

Mondello, C, and Schneider, F. 1998. Disaster Response andRecovery Support from Airborne Remote Sensing. PERS64(10): 958-960.

Molnar, DK, and Julien, PY. 1998. Estimation of Upland Ero-sion Using GIS. CG 24(2): 183-192.

Nichol, JE. 1998. Visualisation of Urban Surface TemperaturesDerived from Satellite Images. IJRS 19(9): 1639-1649.

Wende, A, O’Neill, and Harper, E. 1997. Location Translationwithin GIS. TRR 1593: 55-63.

Obermeyer, N. 1998 The Evolution of Public ParticipationGIS. CGIS 25: (2) 65-66.

Podger, NE, Gage, JD, Teeter, R, and Lillesand, TM. 1998. Useof Image Data to Facilitate Navigation of an Airport Emer-gency Response System. PERS 64(10): 963-968.

Ramsey, EW, Chappell, DK, Jacobs, DM, Sapkota, SK, andBaldwin, DG. 1998. Resource Management of ForestedWetlands — Hurricane Impact and Recovery Mapped byCombining Landsat TM and NOAA AVHRR Data. PERS64(7): 733-738.

Sarjakoski, T. 1998. Networked GIS for Public Participation—Emphasis on Utilizing Image Data. CEUS 22(4): 381-392.

Sexton, WT, Dull, CW, and Szaro, RC. 1998. Implementing Eco-system Management — A Framework for Remotely SensedInformation At Multiple Scales. LUP 40(1-3): 173-184.

Shi, X, and Yeh, AGO. 1999. The Integration of Case-BasedSystems and GIS in Development Control. EPB 26(3): 345-364.

Shiffer, Michael J. 1998. Multimedia GIS for Planning Supportand Public Discourse. CGIS 25(2): 89-94.

Sifakis, NI, Soulakellis, NA, and Paronis, DK. 1998. Quantita-tive Mapping of Air Pollution Density Using Earth Obser-vations: A New Processing Method and Application to anUrban Area. IJRS 19(17): 3289-3300.

Singh, RR. 1999. Sketching The City: A GIS-Based Approach.EPB 26(3): 455-468.

Srinivasan, S, George, V, and Aten, B. 1997. A Computer-BasedTool For Defining Regions Of Similar Characteristics.CEUS 20(2): 111-137.

Talen, E. 1998. Visualizing Fairness - Equity Maps For Plan-ners. JAPA 64(1): 22-38.

Talen, E. 1999. Constructing Neighborhoods From The Bot-tom Up: The Case For Resident-Generated GIS. EPB26(4): 533-554.

Walsham, G, and Sahay, S. 1999. GIS For District-Level Ad-ministration In India: Problems And Opportunities. MISQ23(1): 39-65.

Wong, D. 1997. Enhancing Segregation Studies Using GIS.CEUS 20(2): 99-109.

Yeh, AGO, and Shi, X. 1999. Applying Case-Based ReasoningTo Urban Planning: A New Planning-Support System Tool.EPB 26(1): 101-116.

Zhao, F, Wang, L, Elbadrawi, H, and Shen, LD. 1997. Tempo-ral GIS And Its Application To Transportation. TRR 1593:47-54.

3. Cartography

Andrienko, GL, and Andrienko, NV. 1999. Interactive MapsFor Visual Data Exploration. IJGIS 13(4): 355-374.

Brewer, CA, and McMaster, RB. 1999. The State of AcademicCartography. CGIS 26(3): 215-234.

Cartwright, W. 1999. Extending The Map Metaphor Using WebDelivered Multimedia. IJGIS 13(4): 335-353.

Chrisman, NR. 1998. Rethinking Levels of Measurement forCartography. CGIS 25(4): 231-242.

Filin, S, and Doytsher, Y. 1999. A Linear Mapping Approach to MapConflation: Matching of Polylines. SLIS 59(2): 107-114.

Fremlin, G, and Robinson, AH. 1998. Maps as Mediated Seeing. CA35(1,2): 1-139.

Harrie, LE. 1999. The Constraint Method for Solving Spatial Con-flicts in Cartographic Generalization. CGIS 26(1): 55-69.

Lloyd, R. 1997. Visual Search Process Used in Map Reading.CA 34(1): 11-32.

Morrison, JL. 1999. The State of Government Cartography in1998. CGIS 26(3): 167-200.

Peterson MP. 1999. Active Legends For Interactive CartographicAnimation. IJGIS 13(4): 375-383.

Richardson, DE, and Mackaness, WA. 1999. ComputationalProcesses for Map Generalization. CGIS 26(1): 3-6.

Ricotta, C, and Avena, GC. 1999. The Influence Of Fuzzy SetTheory On The Areal Extent Of Thematic Map Classes.IJRS 20(1): 201-205.

Ruas, A. 1998. A Method For Building Displacement In Auto-mated Map Generalisation. IJGIS 12(8): 789-803.

Ruggles, AL, and Armstrong, MP. 1997. Toward A ConceptualFramework for the Cartographic Visualization of NetworkInformation. CA 34(1): 33-48.

Stehman, SV. 1999. Comparing Thematic Maps Based On MapValue. IJRS 20(12):2347-2366.

Stehman, SV. 1999. Basic Probability Sampling Designs For The-matic Map Accuracy Assessment. IJRS 20(12): 2423-2441.

Vasiliev, IR. 1997. Mapping Time. CA 34(2): 1-47.


4. Data

Arctur D, Hair, D, Timson, G, Martin, EP, and Fegeas, R. 1998.Issues And Prospects For The Next Generation Of The Spa-tial Data Transfer Standard (SDTS). IJGIS 12(4): 403-425.

Boxall, J. 1998. Spatial Data Infrastructures: Developments,Trends, and Perspectives form Converging Viewpoints—Introductory Remarks. CGIS 25(3): 129-132.

Brich, SC, and Fitch, GM. 1997. Opportunities For CollectingHighway Inventory Data With The GPS. TRR 1593: 64-71.

Brydia, RE, Turner, SM, Eisele, WL, and Liu, JC. 1998. Devel-opment of Intelligent System Data Management. TRR1625:124-130.

Burkholder, EF. 1999. Spatial Data Accuracy as Defined by theGSDM. SLIS 59(1): 26-30.

Buttenfield, BP. 1998. Looking Forward: Geographical Infor-mation Services and Libraries in the Future. CGIS 25(3):161-172.

Coleman, DJ, and Nebert, D. 1998. Building a North Ameri-can Spatial Data Infrastructure. CGIS 25(3): 151-160.

Dukewilliams, O, and Rees, P. 1998. Can Census Offices Pub-lish Statistics for More Than One Small Area Geography -An Analysis of The Differencing Problem in Statistical Dis-closure. IJGIS 12(6): 579-605.

Devogele, T, Parent, C, and Spaccapietra, S. 1998. On SpatialDatabase Integration. IJGIS 12(4): 335-352.

Flewelling, DM, and Egenhofer, MJ. 1999. Using Digital Spa-tial Archives Effectively. IJGIS 13(1): 1-8.

Gao, J. 1998. Impact of Sampling Intervals on The Reliabilityof Topographic Variables Mapped from Grid DEMs at aMicro-Scale. IJGIS 12(8): 875-890.

Hendriks, PHJ. 1998. Information Strategies For GeographicalInformation Systems. IJGIS 12(6): 621-639.

Jaakkola, O. 1998. Multi-scale Categorical Data Bases withAutomatic Generalization Transformations Based on MapAlgebra. CGIS 25(4): 195-208.

Lamont, M, and Marley, C. 1998. Spatial Data and the DigitalLibrary. CGIS 25(3): 143-150.

Laurini, R. 1998. Spatial Multi-Database Topological Conti-nuity and Indexing — A Step Towards Seamless GIS DataInteroperability. IJGIS 12(4): 373-402.

Lembo, AJ, Jr., and Hopkins, PF. 1998. The Use of AdjustmentComputations in Geographic Information Systems forImproving the Positional Accuracy of Vector Data. SLIS58(4): 195-204.

Lembo, AJ, Powers, C, and Gorin, ES. 1998. The Use of Inno-vative Data Collection Techniques in Support of Enter-prise Wide GIS Development. PERS 64(9): 861-890.

Leung, Y, and Yan, JP. 1998. A Locational Error Model ForSpatial Features. IJGIS 12(6): 607-620.

Lopez, XR, and Larsgaard, M. 1998. Towards a CaliforniaGeospatial Digital Library: A Strategy for NetworkedKnowledge. CGIS 25(3): 133-142.

Lunetta RS, Lyon, JG, Guindon, B, and Elvidge, CD. 1998.North American Landscape Characterization Dataset De-velopment and Data Fusion Issues. PERS 64(8): 821-829.

Maceachren, AM, Wachowicz, M, Edsall, R., Haug, D, andMasters, R. 1999. Constructing Knowledge From Multi-variate Spatiotemporal Data: Integrating GeographicalVisualization With Knowledge Discovery in DatabaseMethods. IJGIS 13(4): 311-334.

Martin, D. 1998. Optimizing Census Geography — The Sepa-ration of Collection and Output Geographies. IJGIS 12(7):673-685.

Masser, I. 1999. All Shapes And Sizes: The First Generation ofNational Spatial Data Infrastructures [Review]. IJGIS13(1): 67-84.

Murphey, DA. 1999. Presenting Community-Level Data in an“Outcomes and Indicators” Framework: Lessons fromVermont’s Experience. PAR 59(1): 76-82.

Nickerson, BG, and Gao, F. 1998. Spatial Indexing of LargeVolume Swath Data Sets. IJGIS 12(6): 537-559.

Olivera, F, and Maidment, D. 1998. GIS Use for HydrologicData Development for Design of Highway Drainage Fa-cilities. TRR 1625: 131-138.

Osborn, KJ. 1998. United States Mexico Transboundary AerialPhotography And Mapping Initiative. PERS 64(11): 1085-1088.

Russomanno, DJ. 1998. Utility Network Derivation from LegacySource Data for Feature-Based AM/FM Systems. IJGIS12(5): 445-463.

Sexton, WT, and Szaro, RC. 1998. Implementing EcosystemManagement — Using Multiple Boundaries for Organiz-ing Information. LUP 40(1-3): 167-171.

Stefanakis, E, and Sellis, T. 1998. Enhancing Operations withSpatial Access Methods in a Database Management Sys-tem for GIS. CGIS 25(1): 16-32.

Tao, C, Li, RX, and Chapman, MA.1998. Automatic Recon-struction of Road Centerlines from Mobile Mapping Im-age Sequences. PERS 64(7): 709-716.

Thomas, F. 1998. Generating Street Center-Lines form Inaccu-rate Vector City Maps. CGIS 25(4): 221-230.

Tveite, H, and Langaas, S. 1999. An Accuracy AssessmentMethod for Geographical Line Data Sets Based on Buffer-ing. IJGIS 13(1): 27-47.

Van Beurden, AUCJ, and Douven, WJAM. 1999. AggregationIssues of Spatial Information in Environmental Research.IJGIS 13(5): 513-527.

Walter, V, and Fritsch, D. 1999. Matching Spatial Data Sets: AStatistical Approach. IJGIS 13(5): 445-473.

Worboys, M. 1998. Computation with Imprecise GeospatialData. CEUS 22(2): 85-106.


5. Hardware/Software and Technology

Bergen, SD, Mcgaughey, RJ, and Fridley, JL. 1998. Data-DrivenSimulation, Dimensional Accuracy and Realism in a Land-scape Visualization Tool. LUP 40(4): 283-293.

Carr, JR, and Mela, K. 1998. Visual Basic Programs For One,Two or Three-Dimensional Geostatistical Analysis. CG24(6): 531-536.

Carr, JR. 1998. A Visual Basic Program for Principal ComponentsTransformation of Digital Images. CG 24(3): 209-218.

Chainey, S. 1999. Mapping With Microsoft Office. IJGIS 13(1):93-94.

Cole, S. 1998. Of Maps and Macros — Object-Oriented Spread-sheet GIS. EPB 25(2): 227-243.

Deutsch, CV. 1998. Fortran Programs For Calculating Con-nectivity of Three-Dimensional Numerical Models and forRanking Multiple Realizations. CG 24(1): 69-76.

Estalrich, J, and Trilla, J. 1998. Gatagrass — A Graphical UserInterface for Using With Grass GIS. CG 24(5): 501-506.

Gahegan, M. 1999. Four Barriers To the Development of Ef-fective Exploratory Visualisation Tools for the Geosciences.IJGIS 13(4): 289-309.

Hipple, JD. 1998. Autodesk Autocad Map Release 2.0. PERS64(2): 97-99, 106.

Jones, RM. 1998. An Analysis of Computer-Supported Co-op-erative Work Systems to Support Decision-making in Re-gional Planning. CEUS 22(4): 335-350.

Kahkonen, J, Lehto, L, Kilpelainen, T, and Sarjakoski, T. 1999.Interactive Visualisation of Geographical Objects on theInternet. IJGIS 13(4): 429-438.

Knoxrobinson, CM, and Gardoll, SJ. 1998. GIS-Stereoplot —An Interactive Stereonet Plotting Module for Arcview 3.0Geographic Information System. CG 24(3): 243-250.

Kraak, MJ, and Maceachren, A. 1999. Visualization for Explo-ration of Spatial Data. IJGIS 13(4): 285-287.

Laurini, R. 1998. Groupware for Urban Planning: An Intro-duction. CEUS 22(4): 317-334.

Murnion, S. 1999. GIS Online. IJGIS 13(5): 529-530.Pebesma, EJ, and Wesseling, CG. 1998. GSTAT — A Program

for Geostatistical Modelling, Prediction and Simulation.CG 24(1): 17-31.

Peng, ZR. 1999. An Assessment Framework for the Develop-ment of Internet GIS. EPB 26(1): 117-132.

Prisloe, S. 1998. The Geographic Transformer Version 3.08 forWindows95 or Windows NT. PERS 64(8): 777-783.

Shen, Q. 1998. Spatial Technologies, Accessibility, and the So-cial Construction of Urban Space. CEUS 22(5): 447-464.

Rhyne, TM. 1999. A Commentary On GeoVRML: A Tool for3D Representation of Georeferenced Data on the Web.IJGIS 13(4): 439-443.

Royle, JA, and Nychka, D. 1998. An Algorithm for the Con-struction of Spatial Coverage Designs With Implementa-tion in SPLUS. CG 24(5): 479-488.

Simmonds, DC. 1999. The Design of the DELTA Land-UseModelling Package. EPB 26(5): 665-684.

Snay, RA. 1998. Using the HTDP Software to Transform Spa-tial Coordinates Across Time and Between ReferenceFrames. SLIS 58(4): 235-246.

Stein A, Bastiaanssen, WGM, Debruin, S, Cracknell, AP, Curran,PJ, Fabbri, AG, Gorte, BGH, Vangroenigen, JW,Vandermeer, FD, and Saldana, A. 1998. Integrating SpatialStatistics and Remote Sensing. IJRS 19(9): 1793-1814.

Verbree, E, Van Maren, G, Germs, R, Jansen, F, and Kraak, MJ.1999. Interaction in Virtual World Views — Linking 3DGIS With VR. IJGIS 13(4): 385-396.

Wright, D, Wood, R, and Sylvander, B. 1998. ARCGMT - ASuite Of Tools For Conversion Between Arc/Info(R) AndGeneric Mapping Tools (GMT). CG 24(8): 737-744.

6. Implementation and Management

Ang, S, and Straub, DW. 1998. Production And TransactionEconomies And IS Outsourcing: A Study Of The US Bank-ing Industry. MISQ 22(4): 535-552.

Barndt, M. 1998. GIS Public Participation GIS—Barriers toImplementation. CGIS 25(2): 105-112.

Berry, FS, Berry, WD, and Foster, SK. 1998. The DeterminantsOf Success In Implementing An Expert System In StateGovernment. PAR 58(4): 293-305.

Broadbent, M, Weill, P, and St Clair, D. 1999. The Implicationsof Information Technology Infrastructure For BusinessProcess Redesign. MISQ 23(2): 159-182.

Brown, MM, and Brudney, JL. 1998. A Smarter, Better, Faster,And Cheaper Government - Contracting And GeographicInformation Systems. PAR 58(4): 335-345.

Buchanan, R, and Carr, T. 1998. The Impact Of Electronic Dis-semination - What Do Users Think. CG 24(6): 595-596.

Butler, JC, and Trippel, D. 1998. Another Node On The Inter-net - Bridging The Communication Gaps Between TheUser And The Information Technology Support Staff. CG24(2): 201-202.

Chan, TO, and Williamson, IP. 1999. The Different Identitiesof GIS and GIS Diffusion. IJGIS 13(3): 267-281.

Compeau, D, Higgins, CA, and Huff, S. 1999. Social CognitiveTheory And Individual Reactions To Computing Tech-nology: A Longitudinal Study. MISQ 23(2): 145-158.

Craig, WJ, and Elwood, SA. 1998. How and Why CommunityGroups Use Maps and Geographic Information. CGIS25(2): 95-104.

Edelson, DC, and Gordin, D. 1998. Visualization For Learners- A Framework For Adapting Scientists Tools. CG 24(7):607-616.

Goodman, PS, and Darr, ED. 1998. Computer-Aided SystemsAnd Communities: Mechanisms For Organizational Learn-ing In Distributed Environments. MISQ 22(4): 417-440.


Jain, H, Ramamurthy, K, Ryu, HS, and Yasaiardekani, M. 1998.Success Of Data Resource Management In DistributedEnvironments - An Empirical Investigation. MISQ 22(1):1-29.

Jennings, ET, and Ewalt, JAG. 1998. Interorganizational Coor-dination, Administrative Consolidation, And Policy Per-formance. PAR 58(5): 417-428.

Karahanna, E, Straub, DW, and Chervany, NL. 1999. Informa-tion Technology Adoption Across Time: A Cross-SectionalComparison Of Pre-Adoption And Post-Adoption Beliefs.MISQ 23(2): 183-213.

Kopczynski, M, and Lombardo, M. 1999. Comparative Perfor-mance Measurement: Insights And Lessons Learned FromA Consortium Effort. PAR 59(2): 124-134.

Kumar, K, Vandissel, HG, and Bielli, P. 1998. The Merchant OfPrato - Revisited - Toward A Third Rationality Of Infor-mation Systems. MISQ 22(2): 199-226.

Lacity, MC, and Willcocks, LP. 1998. An Empirical Investiga-tion Of Information Technology Sourcing Practices -Les-sons From Experience. MISQ 22(3): 363-408.

Nedovic-Budic, Z. 1998. The Impact Of GIS Technology. EPB25(5): 681-692.

Pinsonneault, A, and Rivard, S. 1998. Information TechnologyAnd The Nature Of Managerial Work - From The Pro-ductivity Paradox To The Icarus Paradox. MISQ 22(3):287-311.

Ross JW, Vitale, MR, and Beath, CM. 1999. The UntappedPotential of IT Chargeback. MISQ 23(2): 215-237.

Sambamurthy, V, and Zmud, RW. 1999. Arrangements For In-formation Technology Governance: A Theory Of MultipleContingencies. MISQ 23(2): 261-290.

Segars, AH, and Grover, V. 1998. Strategic Information Sys-tems Planning Success - An Investigation Of The Con-struct And Its Measurement. MISQ 22(2): 139-163.

Straub, DW, and Welke, RJ. 1998. Coping With Systems Risk:Security Planning Models For Management Decision Mak-ing. MISQ 22(4): 441-469.

Tulloch, DL, Barnes, D, Bartholomew, D, Danielsen, D, andvon Meyer, N. 1997. The Wisconsin Land InformationProgram: Supporting Community Land Information Sys-tem Development. SLIS 57(4): 241-248.

Venkatesh, V. 1999. Creation of Favorable User Perceptions:Exploring the Role of Intrinsic Motivation. MISQ 23(2):239-260.

Watson, RT, Pitt, LF, and Kavan, CB. 1998. Measuring Infor-mation Systems Service Quality — Lessons from Two Lon-gitudinal Case Studies. MISQ 22(1): 61-79.

Zigurs, I, and Buckland, BK. 1998. A Theory of Task/Technol-ogy Fit and Group Support Systems Effectiveness. MISQ22(3): 313-334.

7. Remote Sensing and GPS

Adams, JM. 1999. A Suggestion for an Improved VegetationScheme for Local and Global Mapping and Monitoring.EM 23(1): 1-13.

Afek, Y. and Brand, A. 1998. Mosaicking of Orthorectified AerialImages. PERS 64(2): 115-125.

Alsalman, ASA. 1999. Evaluating the Accuracy of Differential,Trigonometric and GPS Leveling. SLIS 59(1): 47-52.

Arnaud, M, and Flori, A. 1998. Bias and Precision of DifferentSampling Methods for GPS Positions. PERS 64(6): 597-600.

Cihlar, J. 1999. A New Methodology for Land Cover Mapping.IJRS 20(8): 1457-1459.

Cihlar, J, Xia, QH, Chen, J, Beaubien, J, Fung, K, and Latifovic,R. 1998. Classification By Progressive Generalization —A New Automated Methodology for Remote Sensing Mul-tichannel Data. IJRS 19(14): 2685-2704.

Cortijo, FJ, and Delablanca, NP. 1998. Improving ClassicalContextual Classifications. IJRS 19(8): 1591-1613.

Couloigner, I, Ranchin, T, Valtonen, VP, and Wald, L. 1998.Benefit of the Future Spot-5 and of Data Fusion to UrbanRoads Mapping. IJRS 19(8): 1519-1532.

Cracknell, AP. 1998. Synergy In Remote Sensing – What’s In APixel [Review]. IJRS 19(11): 2025-2047.

Cunningham, K. 1998. A Perspective on GPS Use With Imag-ery. PERS 64(7): 661-662.

Cunningham, K. 1998. Resolution — A Shared Concept Be-tween GPS and Digital Imagery. PERS 64(9): 873, 878.

Ebadi, H, and Chapman, MA. 1998. GPS-Controlled StripTriangulation Using Geometric Constraints of Man-MadeStructures. PERS 64(4): 329-333.

Foody, GM. 1998. Sharpening Fuzzy Classification Output toRefine the Representation of Sub-Pixel Land Cover Dis-tribution. IJRS 19(13): 2593-2599.

Gao, J, and Skillcorn, D. 1998. Capability of Spot Xs Data InProducing Detailed Land Cover Maps at the Urban-RuralPeriphery. IJRS 19(15): 2877-2891.

Hepner, GF, Houshmand, B, Kulikov, I, and Bryant, N. 1998.Investigation of the Integration of Aviris And IFSAR ForUrban Analysis. PERS 64(8): 813-820.

Hodgson, ME. 1998. What Size Window For Image Classifica-tion — A Cognitive Perspective. PERS 64(8): 797-807.

Igbokwe, JI. 1999. Geometrical Processing of Multi-SensoralMulti-Temporal Satellite Images for Change DetectionStudies. IJRS 20(6): 1141-1148.

Kalkhan, MA, Reich, RM, and Stohlgren, TJ. 1998. AssessingThe Accuracy of Landsat Thematic Mapper ClassificationUsing Double Sampling. IJRS 19(11): 2049-2060.

Kartikeyan, B, Sarkar, A, and Majumder, KL. 1998. A Segmen-tation Approach to Classification of Remote Sensing Im-agery. IJRS 19(9): 1695-1709.


Martin, DJ. 1998. An Evaluation of GPS-derived OrthometricHeights for First-Order Horizontal Control Surveys. SLIS58(2): 67-82.

Mas, JF. 1999. Monitoring Land-Cover Changes: A Compari-son of Change Detection Techniques. IJRS 20(1): 139-152.

Mesev, V. 1998. The Use of Census Data in Urban Image Clas-sification. PERS 64(5): 431-438.

Motrena, P, and Rebordao, JM. 1998. Invariant Models forGround Control Points in High Resolution Images. IJRS19(7): 1359-1375.

Parks, W. 1998. Accuracy of GPS-derived Orthometric Heightin San Diego County, California. SLIS 58(1): 31-46.

Phinn, SR. 1998. A Framework for Selecting Appropriate Re-motely Sensed Data Dimensions for Environmental Moni-toring and Management. IJRS 19(17): 3457-3463.

Rollet, R., Benie, GB, Li, W, Wang, S, and Boucher, JM. 1998.Image Classification Algorithm Based on the RBF NeuralNetwork and K-Means. IJRS 19(15): 3003-3009.

Sansosti, E, Lanari, R, Fornaro, G, Franceschetti, G, Tesauro, M,Puglisi, G, and Coltelli, M. 1999. Digital Elevation ModelGeneration Using Ascending and Descending ERS-1/ERS-2 Tandem Data. IJRS 20(8): 1527-1547.

Sharma, KMS, and Sarkar, A. 1998. Modified Contextual Clas-sification Technique for Remote Sensing Data. PERS 64(4):273-280.

Shrestha, RL, Carter, WE, Lee, M, Finer, P, and Sartori, M. 1999.Airborne Laser Swath Mapping: Accuracy Assessment forSurveying and Mapping Applications. SLIS 59(2): 83-94.

Smits, PC, Dellepiane, SG, and Schowengerdt, RA. 1999. Qual-ity Assessment of Image Classification Algorithms ForLand-Cover Mapping: A Review and A Proposal for A Cost-Based Approach. IJRS 20(8): 1461-1486.

Soler, T, Hall, LW, and. Reed, CK. 1998. Establishment of GPSHigh Accuracy Reference Geodetic Network in the Carib-bean. SLIS 58(1): 13-24.

Stoms, DM, Bueno, MJ, Davis, FW, Cassidy, KM, Driese, KL,and Kagan, JS. 1998. Map-Guided Classification of Re-gional Land Cover With Multi-Temporal AVHRR Data.PERS 64(8): 831-838.

Stow, DA. 1999. Reducing the Effects of Misregistration onPixel-Level Change Detection. IJRS 20(12): 2477-2483.

Verstraete, MM, Pinty, B, and Curran, PJ. 1999. MERIS Poten-tial for Land Applications. IJRS 20(9): 1747-1756.

Wulder, M. and Boots, B. 1998. Local Spatial AutocorrelationCharacteristics of Remotely Sensed Imagery Assessed withThe GETIS Statistic. IJRS 19(11): 2223-2231.

Zhang, J, and Foody, GM. 1998. A Fuzzy Classification of Sub-Urban Land Cover from Remotely Sensed Imagery. IJRS19(14): 2721-2738.

8. System Concepts and Theory

Abdelguerfi, M, Wynne, C, Cooper, E, Roy, L, and Shaw, K.1998. Representation of 3-D Elevation in Terrain Data-bases Using Hierarchical Triangulated Irregular Networks— A Comparative Analysis. IJGIS 12(8): 853-873.

Abel, DJ, Ooi, BC, Tan, KL, and Tan, SH. 1998 Towards Inte-grated Geographical Information Processing. IJGIS 12(4):353-371.

Abel, DJ, Taylor, K, Ackland, R, and Hungerford, S. 1998. AnExploration of GIS Architectures for Internet Environ-ments. CEUS 22(1): 7-24.

Alsalman, Abdullah SA. 1999. Evaluating the Accuracy of Differ-ential, Trigonometric and GPS Leveling. SLIS 59(1): 47-52.

Chase, SC. 1999. Supporting Emergence in Geographic Infor-mation Systems. EPB 26(1): 33-44.

Couclelis, H. 1998. Worlds of Information: The GeographicMetaphor in the Visualization of Complex Information.CGIS 25(4): 209-220.

Deakin, RE. 1998. 3-D Coordinate Transformations. SLIS 58(4):223-234.

De Berg, M, van Kreveld, M, and Schirra, S. 1998. Topologi-cally Correct Subdivision Simplification Using the Band-width Criterion. CGIS 25(4): 243-258.

Dutton, G. 1999. Scale, Sinuosity, and Point Selection in Digi-tal Line Generalization. CGIS 26(1): 33-54.

Earl, CF. 1997. Shape Boundaries. EPB 24(5): 669-687.Gahegan, M. 1998. Scatterplots and Scenes: Visualisation Tech-

niques for Exploratory Spatial Analysis. CEUS 22: (1) 43-56.Hassen, K, and Beard, K. 1998. Visual Evaluation of GIS Algo-

rithms Using a Reference Grid. CGIS 25(1): 42-50.Hillier, B. 1999. The Hidden Geometry of Deformed Grids:

or, Why Space Syntax Works, When It Looks as Though ItShouldn’t. EPB 26(2): 169-192.

Hodgson, ME. 1998. Comparison of Angles from Surface Slope/Aspect Algorithms. CGIS 25(3): 173-185.

Holt, A, and Benwell, GL. 1999. Applying Case-Based Reason-ing Techniques In GIS. IJGIS 13(1): 9-25.

Huang, YN. 1998. Spatial Range Queries Using Traces. IJGIS12(6): 561-577.

Karimi, HA, and Blais, JAR. 1997. Current and Future Direc-tions in GIS. CEUS 20(2): 85-97.

Lee, CM, and Culhane, DP. 1998. A Perimeter-Based Cluster-ing Index for Measuring Spatial Segregation – A CognitiveGIS Approach. EPB 25(3): 327-343.

Leung, Y, Leung, KS, and He, JZ. 1999. A Generic Concept-Based Object-Oriented Geographical Information System.IJGIS 13(5): 475-498.

Liang, EH, and Lin, SG. 1998. A Hierarchical Approach to Dis-tance Calculation Using the Spread Function. IJGIS 12(6):515-535.

Lin, FT. 1998. Many Sorted Algebraic Data Models For GIS.IJGIS 12(8): 765-788.


Mower, J. 1999. Parallel Processing Algorithms For GIS. IJGIS13(2): 185-186.

Ormsby, D, and Mackaness, W. 1999. The Development ofPhenomenological Generalization within an Object-ori-ented Paradigm. CGIS 26(1): 70-80.

Papadias, D, Karacapilidis, N, and Arkoumanis, D. 1999. Pro-cessing Fuzzy Spatial Queries: A Configuration SimilarityApproach. IJGIS 13(2): 93-118.

Raper, J, McCarthy, T, and Williams, N. 1998. GeoreferencedFour-Dimensional Virtual Environments: Principals andApplications. CEUS 22(6): 529-540.

Saalfeld, A. 1999. Topologically Consistent Line Simplificationwith the Douglas-Peucker Algorithm. CGIS 26(1): 7-18.

Wang, Z, and Muller, JC. 1998. Line Generalization Based onAnalysis of Shape Characteristics. CGIS 25(1): 3-15.

9. Other Issues and Topics

Allan, S. 1999. The Digital New World Order: A View fromthe Private Sector. CGIS 26(3): 201-214.

Banerjee, D, Cronan, TP, and Jones, TW. 1998. Modeling ItEthics — A Study In Situational Ethics. MISQ 22(1): 31-60.

Benbasat, I. and Zmud, RW. 1999. Empirical Research in In-formation Systems: The Practice of Relevance. MISQ 23(1):3-16.

Bishr, Y. 1998. Overcoming The Semantic and Other Barriersto GIS Interoperability. IJGIS 12(4): 299-314.

Butler, JC. 1998. The Internet — A Catalyst for Change. CG24(7): 617-621.

Dykes, J, Moore, K, and Wood, J. 1999. Virtual Environmentsfor Student Fieldwork Using Networked Components.IJGIS 13(4): 397-416.

Epstein, EF, Hunter, GJ, and Agumya, A. 1998. Liability Insur-ance and the Use of Geographical Information. IJGIS 12(3):203-214.

Ghilani, C. 1997. College and University Programs in Survey-ing in the United States. Appendix. Special issue U.S.National Report to the International Federation of Sur-veyors (FIG) 1998. SLIS 57(4): 237-240.

Gibson, DW. 1999. Conversion Is Out, Measurement Is In—Are We Beginning the Surveying and Mapping Era of GIS?SLIS. 59(1): 69-72.

Harris, T, and Weiner, D. 1998. Empowerment,Marginalization, and “Community-integrated” GIS. CGIS25(2): 67-76.

Hepner, G, and Palmerlee, T. 1998. Understanding UC GIS.PERS 64(9): 869-872.

Higano, Y, and Shibusawa, H. 1999. AgglomerationDiseconomies of Traffic Congestion and AgglomerationEconomies of Interaction in the Information-Oriented City.JRS 39(1): 21-49.

Heikkila, EJ. 1998. GIS Is Dead - Long Live GIS. JAPA 64(3):350-360.

National Academy of Public Administration. 1997. FederalGovernment’s Needs and Programs for Geographic Infor-mation—Excerpted from Appendix E of Geographic In-formation for the 21st Century. SLIS 57(4): 256-260.

Slonecker, ET, Shaw, DM, and Lillesand, TM. 1998. EmergingLegal and Ethical Issues in Advanced Remote Sensing Tech-nology. PERS 64(6): 589-595.

Smith, HJ, and Hasnas, J. 1999. Ethics and Information Sys-tems: The Corporate Domain. MISQ 23(1): 109-127.

Szaro, RC, Berc, J, Cameron, S, Cordle, S, Crosby, M, Martin,L, Norton, D, Omalley, R, and Ruark, G. 1998. The Eco-system Approach - Science and Information ManagementIssues, Gaps and Needs. LUP 40(1-3): 89-101.

Voisard, A, and Schweppe, H. 1998. Abstraction and Decom-position in Interoperable GIS. IJGIS 12(4): 315-333.

Wikle, TA. 1998. Continuing Education and CompetencyProgrammes in GIS. IJGIS 12(5): 491-507.

Yonder, A. 1999. Local and Global: Management of Cities inthe Information Age. JAPA 65(2): 232-233.

Special Journal Issues

Special Issue on GIS, Journal of Housing Research 9 (1), 1998.Special Issue on GIS and Redistricting Social Science Computer

Review, 17 (1), 1999.Special Issue on Geospatial Information Sharing and Standard-

ization, Annals of Regional Science 33, 1999.

Tell Someone About URISA Today!

NOTICE IS HEREBY GIVEN thatproposals for LAWA UTILITY SURVEYAND DOCUMENTATION (LUSAD)PROJECT for THE LOS ANGELESWORLD AIRPORTS in the CITY OFLOS ANGELES, CALIFORNIA, are to bereceived by the Executive Director of theDepartment of Airports (DOA) of the Cityof Los Angeles at the Project ManagementOffice (PMO), Los Angeles WorldAirports, City of Los Angeles, Administra-tion Building, #1 World Way Room 225,Attn. Homero Bosso, Los Angeles,California, 90045 prior to 2:00 PM onMonday April 17, 2000. All Proposals areto be delivered in person by the bidder, thebidder’s representative, or by a deliveryservice.

The project includes all work covered by,and incidentals, to completely gather andverify existing information, survey airportgrounds to accurately map all active and/orabandoned utilities and systems infrastruc-tures, using the latest and best technologyavailable, produce four (4) electronicmaster drawings (one master drawing foreach airport) of all utility infrastructure,including a data information storing/sorting scheme and other airport opera-tional infrastructure systems, and toproduce soil stratification report for airportareas (geological reports). It is part of theproject to provide an informationmanagement system application (preferablya geographic information system typeapplication) that will access the informa-tion surveyed and documented LAWA-wide. Along with this management systema bar code label scheme to tag each cable,splice, connector, termination, vaults,manholes, closets, junction boxes, and anyother cable related issues must be proposed.The successful Consultant will beresponsible for hiring sub-consultants andsub-contractors, providing projectmanagement, planning, coordination,surveying, field checking, design, develop-ment, implementation, testing, hold weeklyprogress meetings, make meeting minutes,and other related services as required forfully implementing the referred tasks,including working closely and interfacing

their efforts with other LAWA projects, andcontractors performing Geographic Informa-tion System (GIS), and Document andDrawing Management System (DDMS)work. In addition, the Consultant is alsoresponsible for providing, but not limited to:• System Analysis and Information

Technology Application Design Expertise• Business Process Development and

Implementation for the use and mainte-nance of LUSAD deliverables

• Documentation (Including Databases andAutoCAD Drawings)The consultant shall provide all labor,material, tools, plant equipment, taxes,insurance and permits if required, andperform all operations required for theproper execution and to fully implementan automated system for maintainingLAWA airport utility infrastructure atLAX, ONT, VNY and PMD airports.All work shall be performed in strict

conformity to the Standard Specifications forPublic Works Construction, 1997 Edition, asmodified by the Los Angeles City StandardPlan S-610-17, and the RFP.

PROJECT MUST BE COMPLETEDWITHIN TWO (2) YEARS from the dateof the Notice to Proceed.

INSURANCE MUST BE IN PLACE ATCOMMENCEMENT OF PROJECT.

A proposal meeting has been scheduledfor Wednesday, March 22, 2000 at 9:30 AM

in the Samuel Greenberg Board of AirportCommissioners meeting room located at 1World Way, LAX Administration, LosAngeles, CA. All proposers are stronglyencouraged to attend.

LAWA has an Affirmative Action policy, aswell as a policy to provide Minority andWomen-Owned Business Enterprisesopportunities to compete for and participatein the performance of all airport contracts.Details of these and other policies aredescribed in the RFP and will be discussed atthe proposal meeting scheduled for Wednes-day, March 22, 2000.

Each proposal is to be submitted inaccordance with the RFP Instructions. A listof sub-contractors and sub-consultants isrequired to be submitted with the proposal.

Each proposal is to be accompanied by aproposal bond, a certified check, or acashier’s check for ten percent (10%) of theamount of the proposal. The proposalbond or check is a guarantee that theproposer to whom the contract is awardedpursuant to this Notice will execute thecontract and furnish the required insuranceand contract bonds.

At the time of execution of the contract,a bond for one hundred percent (100%) ofthe contract price is required for thefaithful performance of the contract, andan additional bond for one hundredpercent (100%) of the contract price isrequired to secure the payment of anyclaims for any materials or suppliesfurnished, or any work or labor done underthe contract. The faithful performancebond must remain in full effect for oneyear after acceptance of the work.

Prevailing wage, antitrust claimassignment, nondiscrimination, affirmativeaction provisions, Minority Business/Women Business Enterprise Program,Child support obligations are to be part ofany contract awarded pursuant to thisNotice.

The Provisions of California PublicContract Code Section 22300, if desiredby the Contractor, will also be a part of anyContract awarded pursuant to this Notice.

The RFP for LUSAD will be availableon Wednesday, March 1, 2000. A copycan be received by downloading fromLAWA’s Internet site, WWW.LAWA.ORG.More information can be obtained bycalling LAWA’s Project Management Officeat (310) 338-1075. A fee of $50.00 will becharged for all proposals mailed. The$50.00 fee can be paid by cashier check orpostal money order made out to “LosAngeles World Airports“. No cash will beaccepted.

The Board of Commissioners of the LosAngeles World Airports reserves the rightto reject any or all proposes and to waiveany informality.

January 14, 2000.Lydia H. KennardInterim Executive DirectorDepartment of Airports (DOA)

REQUEST FOR PROPOSALS FORLOS ANGELES WORLD AIRPORTS (LAWA)

UTILITY SURVEY AND DOCUMENTATION (LUSAD) PROJECT

ontents - urisa · 45 geographic information science: critical issues in an emerging...

Documents