Florida Geographic Data Library:
    The Design and Implementation of a State-Wide 
    GIS Data Depository
     
     
      Paul D. Zwick, Ph.D. 
      Associate Scientist and Co-Director GeoPlan Center 
      Department of Urban and Regional Planning 
      University of Florida 
      Gainesville, Florida, 32611, USA 
      paul@geoplan.ufl.edu 

      J. David Lambert 
      Research Associate 
      GeoPlan Center, Department of Urban and Regional Planning 
      University of Florida 
      Gainesville, Florida, 32611, USA 
      dlambert@geoplan.ufl.edu 

      Margaret H. Carr 
      Associate Professor 
      Department of Landscape Architecture 
      University of Florida 
      Gainesville, Fl. 32611, USA 
      mcarr@nervm.nerdc.ufl.edu 
       

     

    Abstract 

    Many government agencies have developed extensive Geographic Information Systems (GIS) databases, at considerable expense in support of their specific programs and mandates.  Other government agencies, educational institutions, consultants, and the general public could benefit from improved access to this valuable information.  In an effort to improve public access to this information, the GeoPlan Center at the University of Florida recently completed a project to develop the 'Florida Geographic Data Library' (FGDL).  The purpose of the project was to facilitate the gathering, organization, and distribution to the public of GIS databases that have been created, with public funds, by government agencies. The databases will be distributed initially on CD-ROM(s) and eventually on the Internet and with Digital Video Disks (DVD).  The project utilized "focus groups" and "pilot projects" to help determine the design criteria for both database distribution methods and end-user applications.   This paper presents results of the FGDL project. 

     

    INTRODUCTION
     

    The GeoPlan Center, a research center of the Department of Urban and Regional Planning at the University of Florida (UF), Gainesville, Florida, recently completed a project titled  'Development of the Florida Geographic Data Library'.  The Florida Department of Environmental Protection (FDEP) and the Florida Department of Transportation (FDOT) funded the Florida Geographic Data Library (FGDL).  The focus of the FDGL project is to facilitate the gathering, organization, and distribution to the public of GIS data that have been created, with public funds.  The FGDL project assists the FDEP, FDOT and other state agencies in fulfilling their commitment to provide the citizens of the State of Florida with new environmental, cultural and public education opportunities.  The FGDL project, while on-going, is already improving access to GIS data within Florida. 

    Background 

    Currently, Florida's state agencies expend a significant amount of human and computer resources fulfilling requests for GIS data.  As new and potential users of GIS technology become more aware of the availability of these GIS data, the pressure on state agencies to fulfill data requests is likely to increase, and the resources required to fill these requests will also increase.  As a result, several state agencies have sponsored the GeoPlan Center's FGDL method for the distribution of publicly funded GIS data.  The FGDL distribution method provides greater efficiency, organization and accountability for the distribution of these GIS data. 

    As a matter of policy,  most state agencies have distributed their data sets in the digital formats and map projections in which their GIS data are maintained.   Unless the agencies are motivated, through special arrangements, to provide the additional data processing required to convert the data into other formats, users are often required to provide significant additional data processing to use these data.  This method of data distribution is uncoordinated and requires that potential end-users have a relatively high degree of  technical proficiency, and the time and resources to accomplish any/all additional processing for these data sets. Unfortunately, many potential, but less technical, users can not get past this basic (but necessary) step because they don't have the training, time, or computer hardware required. 

    Since the authors believe there are many more potential users of publicly-funded GIS databases than there are current users, there is an opportunity for new GIS user groups to more easily develop if these data were provided in a 'pre-processed', 'user-friendly', and inexpensive form.  For instance, these data could be used to support and improve the planning efforts of many local and regional government agencies that presently lack affordable access.  Many non-governmental organizations could participate more fully in, and understand more clearly, public policy development, land use and environmental planning, and historical development patterns.  Students in public schools, colleges, and universities could use the information in classes and research. Private planning and development companies, business consultants, and Realtors have all shown great interest in this technology, and the many GIS databases developed by the State of Florida could provide valuable assistance to these user groups. Furthermore, as a result of FGDL, the general public could gain access to the data through local libraries. 
     
    With GIS data publishing and distribution by CD-ROM becoming a standard practice in the United States, government agencies have an opportunity to transfer data more effectively than has been previously accomplished.  Today, several state and federal agencies have published their databases in CD-ROM format and made them widely available at reasonable costs.  Many private companies have been formed to develop and distribute GIS data in proprietary formats, and have targeted business users as their primary market.  However the products offered are often to expensive for many small organizations, and the general citizenry. 

     
    TECHNOLOGICAL OBJECTIVES
     
     A single CD-ROM will store approximately 640 megabytes of data.  This storage capacity is sufficient for the current and near future  needs of the project based on tests and predictions by the research team.  This conclusion is based on a determination of the amount of Florida GIS data that is now, or will be, available for distribution.  It is also based on the assumption that the data will be distributed by the geographic extent of the 67 individual Florida Counties.  In the future, new 'DVD' media may offer much larger storage capacities (perhaps 14-16 gigabytes per disk) at affordable costs.  As a result, the authors believe the new DVD media will eventually be needed as a transfer media, especially as an ever increasing amount of GIS data, aerial photography and satellite imagery become available. 
     
    The UF team successfully created 'beta-versions' of FGDL CD-ROM(s) for a representative set of Florida counties.  These beta  CD-ROM(s) were tested on PC, Macintosh, and UNIX computing platforms and operating systems, and were found to meet performance expectations as judged by the researchers and the sample of new, inexperienced, users. 
     
    The team also evaluated the feasibility of distributing FGDL databases via the Internet/World Wide Web.  For the present, Internet transmission speeds experienced by the typical user are too slow to be practical for downloading data sets as large as many of the FGDL databases.  This does not preclude however, the hybrid use of the Internet.  Utilization of the Internet for 'home pages' to 'advertise' and 'take orders' for the data set, post new data availability, and to allow the downloading of educational materials, lesson plans, and software extensions is a wise use of web resources.  The FGDL team has completed the design and begun implementation for a hybrid Internet distribution system for the data library implementation. 
     
    CD-ROM(s) offer an advantage over Internet downloading for the FGDL data because they do not require the user to maintain large amounts of data on relatively expensive hard-disk storage devices.  It is likely that many users will want to have the entire state-wide database available to them occasionally, but would not want to keep the complete database 'on-line'.  FGDL would require too much disk storage for most users groups if downloaded (approximately 60 Gigabytes at present).  The FGDL CD-ROM(s) provide a method of storing these data in compartmentalized county modules for use as needed. 
     
    Unfortunately, GIS databases are often altered and passed on to other users without quality control or documentation.  Since CD-ROM(s) are a 'read only' media, this characteristic ensures that original release versions of the database will always be available to users.  Once received the user has a write-protected copy of FGDL and can begin to build a time series database by receiving data sets annually.  The CD-ROM media provide a simple means for data protection, data can be cataloged by year for time series analysis, and the FGDL version release of the data is maintained for reference. FGDL CD-ROM(s) also contain an on-line hyper-linked metadata, a user tutorial, and software documentation. 
     
    MANAGEMENT OBJECTIVES 
     
    A major  objective of the project was to determine the requirements for the FGDL data management and distribution system.  The digital format used for vector data in the FGDL database is ArcView 'shapefile' format. This openly-published digital format is efficient, supported on all major computing platforms, and translators have been, or are being, developed for many of the other popular GIS software.  Future versions of the FGDL database may be distributed in the 'Spatial Data Transfer Standard' (STDS) digital format being promoted by the United States government.  Imagery and raster data are distributed in ERDAS '.lan', geotiff format, and ArcInfo GRID format. 
     
    The map projection currently being used for the FGDL database is: 
     
      ALBERS 
      CLARKE 1866 
      NAD27 
      METERS 
      1st Standard Parallel:   24  0 0.000 
      2nd Standard Parallel:   31     30 0.000 
      Central Meridian:        -84   0 0.000 
      Latitude of Origin         24  0 0.000 
      False Easting:     400000.000 
      False Northing:    0.000
     

    This projection is not, however, the projection being used by all state agencies.  Consequently, the research team has strongly recommended that a standard map projection for state-wide GIS databases, suitable to the needs of all concerned state agencies, be adopted now, rather than later.  The fiscal impact of changing standard map projections will grow dramatically in the future as the user community grows and new databases are developed. 
     
    The research team has developed a data management system based on the 'ArcInfo Data Library' model using ArcInfo GIS software (by Esri, Redlands, Ca, USA). Based on this model, each state-wide data layer is 'cut' into 67 pieces, or tiles, corresponding to the geographic extent of each Florida county boundary.  The database is maintained in ArcInfo 'coverage' format to allow for efficient data processing activities.  Custom software programs have been developed to convert the tiled database into the shapefile distribution format, and to compile the individual county data sets according to a standardized organizational scheme.  Additionally, the team is testing, with assistance from Esri, the Spatial Database Engine (SDE) for Internet access and database management. 
     
    The FGDL organizational scheme for its many data layers is illustrated in Table 1.  In this scheme, each data layer is organized into a subject area directory.  This categorization of the data layers is intended to help the end-user 'navigate' more efficiently through the database to find the data required.  This will be a particularly important feature of the FGDL data products as more and more data layers are added in the future. 
     

         Table 1.  FGDL Subject Matter Directories 
 
        Directories 
        Example Data Layers
        Agriculture 
        General  and Specific Soils
        Boundary
        Agency Districts , Counties, etc. 
        Cultural 
        Land Use Classification, Historic Sites etc. 
        Habitat 
        Wildlife Habitat, etc.
        Hazards 
        Hazardous Sites, Flood Zones, etc.
        Imagery 
        Satellite Images 
        Documentation 
        Documentation for Data Layers
        Physical
        Geology, topography, hydrography, etc. 
        Planning
        Conservation Lands, Strategic Habitats, etc. 
        Property 
        Public Land Survey, Dept. of Revenue tax records, etc.
        Transportation
        Roads, Railroads 
        Quads USGS 
        Quadrangle boundaries and digital raster graph
 
     
    IMPLEMENTATION 
     
    The FGDL team has examined the issue of full-scale implementation and long-term support of the project in this first phase of the project.  Some general recommendations and alternatives have been outlined by the researchers.  For instance, full-scale implementation  of the FGDL should be accompanied by the organization of an executive, or advisory, board.  The executive board should be selected from agency personnel, librarians, and other interested parties. 
     
    Institutionalization will also require that memoranda of agreement  with cooperating agencies  be established.  Table 2 lists the state and federal agencies that have contributed data to the FGDL pilot project. 

    Table 2.  Government agencies that are contributing data members of FGDL. 
     

      Florida Department of Environmental Protection 
      Florida Department of Transportation 
      Florida Department of Community Affairs 
      St. Johns River Water Management District 
      South Florida Water Management District 
      Suwannee River Water Management District 
      Southwest Florida Water Management District 
      Northwest Florida Water Management District 
      The Florida Game and Freshwater Fish Commission 
      The United State's Geological Survey 
      The United State's Census Bureau 
      The Florida Department of State - Division of Archeological Research 
      The United States Department of Agriculture 
      The Florida Department of Geology 
      The Florida Department of Revenue 
      The Florida Natural Areas Inventory 
         
    Long-term financial support could be provided through a cooperative funding arrangement made between several government agencies;  and/or by an annual appropriation of funding from the Florida State Legislature. 
     
    The authors have recommended that in addition to these subsidies, that the FGDL charge end-users for the production costs associated with each CD-ROM.  It is the opinion of the project team that the taxpayers of Florida should not be required to pay more than the cost of producing duplicates of each of the county CD-ROM(s).  A current estimate of the production cost per CD-ROM is approximately $20.00. 

    User Community  

    User community analysis for the FGDL project was conducted to develop appropriate community identification strategies and to perform an initial market study using those strategies.  The research team determined that a four-step market analysis approach was appropriate.  The driving concept behind that approach is that the potential market should help shape and direct the product content and price.  To that end, the team's market strategy involved: 
     

       a)  Individual focus interviews with potential user representatives, 
       b)  Small-medium sized group interviews and presentations, 
       c)  Broad-scale survey research among potential representatives, and 
       d)  In-depth 'pilot projects' with at least two potential user groups. 
     
    Furthermore, the team determined that the study would target both private and public user communities since it is believed that the FGDL product will have appeal to both communities.   To that end, a range of individuals and groups were identified to query.  These included representatives from the following potential markets:  Farm and Agricultural Agencies; Realtors; Developers, Bankers and Business People; Police; State and Federal Agencies; Planners and Planning Officials; Educational Institutions;  and Librarians. 
     
    Several issues and concerns were raised by the 'focus group' participants.  For instance, most of those interviewed were concerned about having access to the appropriate hardware and software to effectively utilize the data in applications which are appropriate to their agency missions.  Another barrier to adoption identified by  the focus groups was the need for training, on-line help, and tutorials for beginners. The cost of the CD-ROM(s) was an issue for most groups (all groups believed that $20.00 was an extremely reasonable price for such valuable information).  The frequency of database updates was also a common issue. 
     
    Despite these obstacles, the results of the individual and group focus interviews, presentations, and broad-scale survey research indicate a large potential demand for the FGDL products.  The input received from the interview and survey participants has been incorporated into the recommended design and implementation criteria. 
     
    Two in-depth pilot projects with libraries and public schools were also conducted. The University of Florida Map and Imagery Library was chosen as a test site to determine the practical issues and implementation strategies for libraries, and to help identify potential user communities.  The project installed public-access terminals in the library and began development of appropriate user-interfaces to the data based on criteria developed by the librarians.  The project also surveyed a sample of librarians in Florida and found a high level of interest in the FGDL data.  However, there is concern among the librarians about offering a new, potentially very popular, service for which they are not fully prepared to respond.  Their most immediate concerns are that many libraries do not have adequate computer resources and training to respond to GIS questions.  Consequently, a multi-level service scheme has been proposed whereby most libraries would only offer 'check out' services, a smaller number would offer basic on-site computer access supported by a user-friendly interface, and a only a few libraries would offer more advanced service with a trained GIS specialist. 
     
    The P.K. Yonge Developmental Research School (PKY) is a laboratory school operated by the University of Florida Department of Education.  This school was the site of the project's second pilot project which was aimed at determining the potential for the FGDL data to enrich existing middle and high school courses.  Several PKY faculty were involved in this project and developed GIS activities that were used in their current courses.  The activities covered middle school social studies, high school social studies, and high school environmental science and technology.  Each of the teachers indicated that the GIS activity was a success.  Comments included the need for more time for future activities and the need for more teacher training.  The teachers also commented on the need for adequate computer hardware.  Customized software applications would also help get the students get going faster.  The PKY faculty involved in this project believe that GIS activities can fit well into the public school curriculum if the computer resources are adequate. 

    FGDL Application Development  

    Based on the feedback from the focus groups and pilot projects, the team established that a primary objective of the FGDL project should be the development of 'end-user-friendly' software applications and tutorials that would make the FGDL data easy to use.  Consequently, the project team initiated the development of several FGDL applications during the Fall of 1996.  During Spring 1996, the first version of these  applications were completed  and a 'beta test' program was conducted to get feedback from both novice and experienced end-users.  Development of software applications  was based upon the utilization of ArcView and its application programming language 'Avenue'.  The FGDL team also utilized an application development software called 'MapObjects' Lite for the creation of a data viewer.  Both ArcView and MapObjects Lite are software products of Environmental Systems Research Institute, Redlands, CA. 

    As a result of input from 'beta-test'  and interview participants, the team identified several general requirements for FGDL applications.  A few of the less technical requirements were: 
     

      1)  Software applications should be developed for the PC, Macintosh, and UNIX computing platforms,  and the user-interface should have a common 'look and feel' across  all platforms where possible. 
      2)  Applications  focused toward helping new, inexperienced users get started using the data quickly should be the first development priority.   To this end, these applications should include  context-sensitive 'on-line' hypertext help features and an optional 'on-line' tutorial. 
      3)  The applications should be designed to provide experienced users with features that help them start using the FGDL quickly and efficiently.  For instance, pre-defined map legends should be provided for each data layer.  Several of the experienced GIS  users that we interviewed  requested that efficient map projection capabilities be included in the applications  to facilitate  the use of other GIS data sets that are not included in the FGDL data set. 
      4)  The applications should include  a context-sensitive, hypertext -style 'metadata' feature to provide quick and efficient documentation for each data layer.  Documentation for each data layer should be optionally available to the user in three different levels of detail.  First, a 'quick-look' version of the metadata would only provide a short description  of the data layer.  Second, the user should have the option of accessing  a more detailed, but still 'basic' version of the metadata in which only the most essential documentation for  the data set would be provided. At the highest level of detail and content, comprehensive documentation for each data layer, if available, should be accessible to the user in a format complying to the digital geospatial metadata where possible. 
      5)  Many  potential users (mostly individuals) expressed  a desire to have an inexpensive , or even free, software application option available to them  that would provide basic viewing and query functions.  In other words, many individual users believed they would probably never be able to purchase full-function GIS software, however, they would like to have access to the database. 
     
    Based on these general requirements for FGDL applications,  the team has identified three alternative approaches to providing end-user applications.  The team believes that all three approaches should  be developed and implemented ,  because  each approach is targeted at a different user group with different requirements.  These applications fall into three general categories: 

    Applications designed for use with existing GIS software packages.  Aimed primarily at experienced users (or those wishing to become experienced users), this category of applications requires that users have licenses for GIS software packages.  Initially, the project team has developed applications for ArcView GIS software. These applications are distributed as 'freeware'.  This implies that the 'intellectual rights' to the application software is preserved by UF, but the application software is provided 'free of charge'.  Eventually, we would also like to develop applications for other GIS software that can 'read' ArcView shapefiles.  Users could 'download' the applications from the project's World Wide Web 'home page'. 

    The primary development goal  for freeware applications  is to provide both an easy start up for beginners, and a quick starting point for customization by  experienced users of the GIS software.  The application design methodology  for this category of application is based on the concept of developing a robust 'core' FGDL application  that fulfills all of the basic FGDL application requirements, such as basic data and metadata viewing .  The design also allows customized 'extension' applications to be 'plugged in' to the core FGDL application. For instance,  customized extensions could provide discipline-specific analytical capabilities , or data views customized to a target audience. 
     
    Plug and Play' software applications.  Many  potential, inexperienced users have suggested that an inexpensive software application option is a necessary element of a successful wide-scale implementation of FGDL. This category of applications has been designed to be included on the FGDL CD-ROM(s), so that users won't have to install or purchase software.  They simply load up the CD-ROM, and click on an icon to start viewing and querying the data.  These applications presently provide basic data and metadata viewing and query capabilities, but could be customized to the needs of specific user groups, such as libraries, schools, etc. Additionally, on-line help and tutorials are provided for the viewing package. 

    Internet  'Map Viewing' applications.  There is increasing interest in Internet/World Wide Web data viewing and query access to the FGDL data.  This type of service would provide the general public , the casual  agency user, and students with a valuable resource.  Internet 'Map Viewing' service will probably provide data viewing and query capabilities to the end-user  that are similar, if not identical, to those offered by the 'plug and play' applications.  The primary difference may be the time delays  for getting results from a query or viewing option selection caused by low Internet data transmission speeds.  The authors believe that Internet access will not replace the need for CD-ROM distribution methods in the near future because of the size of the FGDL database.  Even when some of the technical problems of transmission speed are solved, their will still be advantages to the CD-ROM methodology for many user groups. 

    The GeoPlan Center has been researching this technology to determine the feasibility of implementing  a FGDL application, using the ArcView Internet Map Server software.  Our initial findings have been very encouraging, however, there are some obvious barriers to immediate implementation that have been identified.  For instance, Internet 'Map Viewing' technology is very new, standards are evolving, and changing, very rapidly, therefore, demonstration and pilot projects  need to be executed before a commitment to full implementation is made.  Additionally, although the service would be 'free' to the public, there are some significant  costs involved in setting up the 'Internet Server' hardware and software that will be required to offer reliable service to the large number of anticipated users. 

    More information about FGDL products is available at the GeoPlan web-site www.geoplan.ufl.edu. 

    Acknowledgments  

    Special Thanks are due: 

    Ilir Bejleri, FGDL Project Manager 
    Alexis Thomas, Imagery Manager 
    Wesley Harrell, Systems and Database Administrator 
    Andrea Goethels Home Page Manager 
    Juna Papajorgji Applications Specialist (ArcView) 
    Priscila Zardo Application Specialist (Viewer) 

    All the graduate students who have worked long hours on the project over the past two years.