----------------------------Original message---------------------------- 32nd Annual Clinic on Library Applications of Data Processing GIS and Libraries: Patrons, Maps and Spatial Information Report prepared by Alberta Auringer Wood, Joan Winearls, Janet Kaufman, Grace Welch, Richard Pinnell, Trudy Bodak, Yves Tessier, and Cheryl Woods Introduction From Sunday morning, April 2, through Tuesday afternoon, April 4, a group of about 100 librarians, academics, computer specialists and others attended this conference at the University of Illinois in Urbana-Champaign. Not surprisingly, librarians were in the majority. According to the list distributed at the conference, there were 59 librarians, 17 teaching faculty, four computer specialists, six students, two consultants, two state government employees, two federal government employees, two academic support staff, one government cartographer, one local government employee and one person from the library press. In terms of the librarians, there were 23 map librarians (either that alone or in combination with other subjects or media), 11 librarians, six reference librarians, six government documents librarians, four science librarians, one library director, three technical services/cataloguing librarians, one humanities and social sciences librarian, one geography librarian, one geology librarian, one media/microcomputer librarian and one city planning and landscape architecture librarian. Among the librarians and computer specialists, 47 were from university libraries; four each from federal government and public libraries; two each from state, public or science center libraries; and one each from a college library and a state historical society. They also came from both sides of the continent and twenty-seven states or provinces. The largest number of participants out of the 64 total were from Illinois (21), but the next largest group was from Ontario (six). There were four participants from California; three each from Michigan, Wisconsin and Nebraska; two each from Pennyslvania, New York, Louisiana, D.C., Indiana and Missouri; and one each from Iowa, Tennessee, Maine, Massachusetts, Virginia, Vermont, Ohio, Washington, Colorado, Oklahoma, Quebec and Newfoundland. There were 14 session chairs, support staff, student volunteers and faculty from Illinois. In terms of speakers, there were four from Illinois and Florida, three from New York, two each from Iowa and California, and one each from Minnesota, Virginia, Maryland and Missouri. (AAW) Highlights GIS in libraries is MUSHROOMING! Hardware and software are dropping in cost, but data remains high, aside from U.S. government data. Numerous World Wide Web pages are appearing. Social scientists, emergency preparedness centers, and others use GIS. Cataloguing of data presents problems in physical description, bounding coordinates, time and edition or version, and includes metadata, some of which must be provided by the producer. Librarians provide the "value added" service of interpretation. A Government Information Locator Service is in process; a new government initiative to assist in the management of their resources and to assist the public to find information via Internet and according to Z39.50. Burgeoning electronic resources will affect collection policy, by eliminating the need to collect certain types of maps in paper format, e.g. non-major U.S. cities. Lotus Screen Cam can be used to prepare on screen tutorials. University of Minnesota is developing an electronic atlas of the state to be accessible via the Internet. A GIS can assist in estimating and profiling library "market" areas. It is essential to provide time and training for librarians. There is a public access electronic atlas in the St. Louis Public Library that has been a huge success. GIS is a natural extension of information services in the library. Use of user interfaces can be monitored using a "mouse trap". Tremendous amount of global change data and access by geographic area is needed. Illinois Department of Natural Resources has produced a CD-ROM for the state that includes map data at scales of 1:24 000, 1:100 000 and 1:250 000. How can we apply "farming by the foot" to fishing? There will be problems with data volume. For example, the Alexandria Digital Library Project will need 100 MB on a screen all at once. In summary, there is a clear need for digital libraries. (AAW) Comments and Observations From the Conference Attendees The general feeling about the conference was that it was well presented with a good variety of speakers and sessions. It was not too technical and the thought of GIS in libraries does not seem as scary and unmanageable. There are challenges and opportunities to be gained, but the process should be well planned and unhurried. (CW) There are a number of standard creating initiatives in the U.S. at the federal level. What is happening in Canada? We know far more about what is happening in the U.S. than in Canada, and it seems that U.S. libraries are very involved in the process. This involvement does not seem to be the case in Canada. As an association we need to find out more about what is happening and whether we can play a role. Also, are U.S. librarians developing their own rules for cataloguing geomatic data sets; will they be compatible with the rules developed by the CGSB? Several interesting funding projects were discussed: the Alexandria Project to create a digital library of spatial information and the Dept. of Education grant to create the ACIC at the University of Minnesota. What grants are available for Canadian map libraries? Involvement in GIS activities in U.S. libraries varies considerably; many of the participants at the conference were only just starting out and hoping to get some answers or direction. In Canada, there seems to be less variation in GIS capabilities in map libraries. There are sophisticated set-ups, such as at the University of Minnesota, and then libraries only planning on acquiring their first workstation. Without question though, GIS is a reality that libraries of all types will have to come to terms with in the very near future. The practical discussions were the most interesting, such as the development of the St. Louis metropolitan atlas at the St. Louis Public Library. The project demonstrates how a good definition of what could be done in a public library setting and the willingness to try something new, combined with interest and support from professionals in the community, can bring about a successful implementation. (GW) This very informative and thought-provoking meeting of spatial data producers, managers (partly us) and users brings new insights to our profession, but also raises new questions regarding the increasing "spatialization" of information, and the consequences on studies and research work deriving from it. Following are some personal observations, more "impressionist" than analytical, presented "a la carte" as they came to mind. It is said that nearly 80% of all information can be attached to a spatial entity (or georeferenced). Space has significance in the distribution of phenomena: disease incidence, pollution occurrences, clientele distribution, road accidents, social behaviors, etc. More, spatial analysis is about to redefine such long-standing sciences as the social sciences, according to professor Gerard Rushton (The University of Iowa). Environmental sciences and geography have integrated spatial analysis in their paradigm (an evidence not to be questioned), but social sciences are about to do it. According to Rushton, social sciences try to answer two types of questions: how things are (applied role of the social sciences) and how the human system works (fundamental role of the social sciences). The better you know the social reality, the better you can explain it. Geographic information systems bring a new tool of analysis to the social scientist and to the science itself, in introducing spatiality in the significance of observations. For instance, as demonstrated, the distribution of a disease in a city correlated with polluted land fills years before can explain the phenomenon and indicate solutions to this health problem-solving. What are the consequences of more spatiality on information processes and uses? Certainly, there is a greater demand for georeferenced information and a greater demand for tools and assistance in using this information. How is it different from services provided around printed georeferenced information (a paper map)? GIS brings new insights to "map" librarianship, but raises as well, new questions to the profession. As we provided printed maps, tracing tables and assistance in map reading, we will have to provide georeferenced data, "tracing screens" (workstations) and assistance in correlating layers of information. A new professional paradigm is ours to embrace! We have no choice but to pave the way to electronic map librarianship, in order to provide better support to new trends (more spatiality) in research work and in spatial decision systems. It strikes me how U.S. public information sources are more and more managed all across their producing agencies, as opposed to the Canadian situation. The integration of managerial components within the American government regarding spatial data is exemplary. You begin with the National Information Infrastructure (NII) concept. Then, as one of the possible applications, you have the National Spatial Data Infrastructure(NSDI), meaning the technology, policies, standards, and human resources to manage spatial data, as created by President Clinton's Executive Order 12906. To communicate and exchange, the National Geospatial Data Clearinghouse (NGDC) exists as a distributed electronic network of geospatial data producers, managers and users. To coordinate the development of the NSDI, there is the Federal Geographic Data Committee (FGDC) that has already adopted standardized documentation of geospatial data (metadata) and promoted plans by agencies to provide public access to geospatial data, among other responsibilities. In order to support the decennial census of 2000, a National Digital Geospatial Data Framework (NDGDF) has been created. Not insignificant to mention is a liberal attitude for public access to geospatial data, with the Government Information Locator Service (GILS) being created to identify, locate and access geographic data. When you have a sense of vision for change, higher political commitment, coordinating bodies, means for exchange, standards for production, partnerships between actors, consideration for clienteles, you have a winning mix to introduce change fruitfully. All these are ingredients that are lacking badly in Canada, where no appearance of integrated management seems to exist for spatial data. Haven't we been, as potential spatial data managers (and promoters), contributing to this stagnation? We tend to consider technological advances in a horizontal view, introducing an equipment or a software, without relating sufficiently these components to a larger environment, without reassessing and questioning existing practices. In many of the developments or projects presented during the clinic, reference was often made to the relations between data, technology, human resources and problems to be solved. These represented a sort of vertical integration of components, a holistic approach in designing solutions. Technology is not neutral in conditioning human operations. Components within a system interact on others. If spatial data is considered a new input for a researcher, new demands will be put before the information infrastructure (the map library, hopefully) for access and service, which will require new training for staff, new equipment and new interactions within the infrastructure (data librarian, computer people, etc). Interdependency is another word for vertical integration (more than ever, and outside our small world). Are we sufficiently open minded in that direction? On-line public access catalogues (OPACs) replicate a conceptual framework established originally by Charles Cutter in his Rules for a Dictionary Catalogue... at the begining of the present century. We approach the end of the same century (!) without significant progress in retrieval. Nearly not quite so, hopefully. We are more aware of providing user support in retrieval systems, of offering richer querying mechanisms. The Alexandria Project will make room for tutorials in retrieval. We speak of semantic networks and hypertext to expand the capacity of querying devices. We develop non verbal area searching using graphics, polygons. Will we continue to abide by Cutter's rules in forcing new information units into alien categories? Will we continue to produce thickening guides to search OPACs? Will we force users to adapt themselves to our systems, or will we borrow users search patterns to design systems accordingly? New insights? New questions? New challenges! Stimulating! (YT) Workshops and Keynote On Sunday, there were two optional "workshops" and a keynote address. The first workshop was an "Overview of GIS Concepts" presented by Marc P. Armstrong, Associate Professor, Department of Geography and Program in Applied Mathematical and Computational Sciences, The University of Iowa. He used software called "GIS Tutor" to present this. The major books, journals and trade magazines in the field were listed, such as S. Aronoff's Geographic Information Systems: A Management Perspective (Ottawa: WDL Publications, 1989), Cartography and Geographic Information Systems, and GIS World. He noted that a bibliography of videos on the topic was available from the National Center for Geographic Information and Analysis. He pointed out that a number of professional organizations provide a forum for learning more about GIS through their publications and conferences, for example, the American Congress on Surveying and Mapping and the American Society for Photogrammetry and Remote Sensing with their annual conventions, the GIS/LIS meetings in the fall, and the Auto Carto symposiums every two years. For keeping up to date electronically, he gave lists of Web sites with GIS related content and also a list of sites available using MOSAIC. There is also a listserver called GIS-L which is available as a newsgroup, too (comp.infosystem.GIS). Using the GIS tutor, Dr. Armstrong defined what a GIS is, how it works, what hardware is required, something of its development, what it is made up of and some of the applications for it. He also provided some information on GPS (global positioning systems). In terms of GIS software, he pointed out that MapInfo grew out of a PC environment and is not topologically correct, while ArcView and Intergraph came from a mainframe environment and are topologically correct. One software, GRASS, was developed by a government agency. The second "workshop" was on "Spatial Analysis in the Social Sciences" presented by Gerard Rushton, Associate Professor, Department of Geography and Program in Applied Mathematical and Computational Sciences, The University of Iowa. He defined spatial analysis as the ability to cross-correlate two or more overlays of information, such as merging a zoning map with soils and land use. By using overlaying, he felt that users could start to make information driven decisions. He pointed out the linkages between social science and spatial analysis and stated that in the 1960s geography was viewed as a spatial science while in the 1970s and 1980s it was pushed out and in the 1990s it has been rediscovered. Spatial analysis is now integrated with GIS and involves getting information about a point. To illustrate this he described a study that was the spatial analysis of the relationship between infant mortality rates and birth defect rates in Des Moines, Iowa, 1989-1992. He also discussed the use of the U.S. Bureau of the Census TIGER data and its limitations. In addition, he noted the application of interactive computer-based systems which help decision makers use data and models to solve unstructured problems about the spatial organization of activities. A benefit of the latter is providing improved information for decisions and a more rational process of location decision making. He illustrated this with an Iowa study on restructuring of school districts (not yet implemented by the state). After a reception and dinner, there was "keynote address" by Mark Monmonier, Professor, Department of Geography, Maxwell School, Syracuse University entitled "Technological Transition in Cartography". This was a "look back" on a book that he did in 1985, and how it related to today. He gave a detailed, chapter by chapter, description of the contents and especially noted the problems of collecting and archiving digital data. Dr. Monmonier discussed how the book was received by noting the nine reviews and their ratings; it was said to be available yet for purchase! He emphasized the increased diversity of the integration of maps and closed with demonstrations of several "dynamic" (computer produced) maps that he had prepared some time ago. It was not exactly what had been expected in terms of a keynote. (AAW) Describing and Cataloging GIS Monday morning, Mary Larsgaard, Assistant Head, Map and Imagery Lab, Davidson Library, University of California, Santa Barbara spoke on "Cataloging Planetospatial Data in Digital Form". She indicated that at UC Santa Barbara they have a policy for cataloguing all data being acquired through Project Alexandria, one of six large National Science Foundation grants for digital libraries which started in June 1994. One of the problems has been in finding a USMARC equivalency for every FGDC field. Briefly, Larsgaard indicated that cataloguing problems for data fall into four main areas. The first is physical description and tag 256 - file characteristics from chapter nine computer files. Here the dilema is whether you define the item as a computer file or map or multi-media etc. For 256 you can only enter computer program or a similar term and the number of programmes or bytes. But what is needed is other information such as colour and other graphic characteristics. The first piece of information that is requested is the size of the file and you cannot count maps as these are infinite. For tag 300 colour becomes a question of the gray scale and a palette of 1-256 colours and is often dependant on software. Under $a the number of megabytes is added. The second problem area is the question of production versus publication. A good example here is to ask if an Internet f ile is published. Presumably, there is also a problem here with edition or version particularly if it is always being updated. The third problem area is that of merging with another standard such as the Federal Geographic Data Committee (FGDC) standards. These are inherently two different things. FGDC is mainly for use by data producers. Some information will only be known by the producer and not by the cataloguer. The FGDC standards are new while AACR2 and MARC come from older book-oriented traditions. Finally, Ms. Larsgaard discussed multilevel description. The problem here is providing for tiles, which are areas adjoining each other, and for layers, or coverages of data, for these or for other areas. There can also be multiple versions of the same thing - e.g. if an aerial photo is catalogued as part of a set and then one or more are scanned, do you prepare separate records for these which would make for much more work. Subject headings are not all consistent, and there is a need for a heading for world or earth with subdivisions. For 352 $i, we also need to be able to indicate raster or vector. Bounding coordinates are essential for digital files. Local time is used for satellite data and this must be recorded somewhere. Ms. Larsgaard indicated that she had looked a bit at the Canadian rules for Geomatics cataloguing, but it is obviously time for us to look at these in some detail and start to use them for our various map data. Michael Domaratz, Member, Federal Geographic Data Committee (FGDC) Secretariat and a Cartographer at the U.S.Geological Survey (USGS) reviewed "Metadata Standards and the FGDC". The FGDC of USGS is the national group mandated to support proper standards for medadata and the creation of the National Spatial Data Infrastructure (NSDI). The FGDC promotes the development, maintenance and management of distributed data base systems that are national in scope for geographic data. Domaratz noted that the future belongs to those who control the tools for filtering and making sense of data. It will be up to librarians to sort out how they will integrate with all the other groups involved with data, e.g. geographers, cartographers, computer scientists etc. Content standards for digital geospatial metadata were approved in June 1994 to provide a common set of terms that will be used in describing data. Federal agencies must use these beginning in 1995. The standards answer four major questions about data - availability of data for a specified location, the fitness for use in a specified need, access routes to the data and transfer of information needed to use a set of data. The standards are based on the premise that the person best set up to prepare medadata is the producer. However, Mr. Domaratz mentioned that recording all the changes made to data could mean very lengthy records. Mr. Domaratz noted that one of the greatest problems is knowing where to find the data - data for which the U.S. government has spent billions of dollars. To solve this the National Geospatial Data Clearinghouse (NGDC) was set up in April 1994. The records are on the Internet and the base configuration is WAIS, although some records are on gophers, WWW etc. Many federal agencies are now at work inventorying their data and mounting this on the Internet. Much of the data is also available through the Internet and vast numbers of files are already being transferred even without the Clearinghouse. For instance, now that all the DLG and DEMs for the U.S. are up on the site at Sioux Falls over 700,000 files were shipped in the first year. The USGS EROS Data Center recorded nearly 49,000 downloads in its first 3 months. Further information of FGDC initiatives is available in the FGDC Newsletter and this and a list of about 20 documents can be acquired from Publications, USGS, 590 National Center, Reston, VA 22092. Although the Canadian contingent had heard earlier about FGDC , NSDI and the Clearinghouse the effect of the speech was to make us all realize what vast quantities of data there are in the U.S., mostly available free or at a nominal charge and to conclude that we are data-starved in Canada. Re the metadata: during discussions it became clear that we had no idea whether questions of standards are being worked on in Canada. (JW) Users and Interfaces After the mid-morning break, Myke Gluck, Assistant Professor, School of Library & Information Studies, Florida State University spoke on "Spatial Information: Users' Needs & Competencies of the General Public". He reported on the results of several studies he has conducted in public libraries with a focus on three key research questions: 1) user needs for spatial data - what are they? 2) formats and tasks requested by users, and 3) formats which users find most useful. The needs of users were categorized as facts (geographical data) and information ( making the data usable). Users of spatial information acquire their knowledge procedurally, by walking around and looking at an area, as well as from representations of spatial information presented on maps. Dr. Gluck identified three broad categories of map tasks: reading (what), analysis (how) and interpretation (why). Dr. Gluck has conducted a number of experiments to assess how fast and accurately individuals can complete a number of map tasks. He described a translation exercise in which one group of subjects were given a map and asked to describe what was represented on the map. A second group of subjects were given a description and asked to create a map based on the information contained in it. Subjects in the first group were generally unhappy with their descriptions of the data presented on the map because: they needed too many words to describe the data; it took too long to complete a good translation; and he map was too complex. Subjects in the second group were equally unhappy with their translations from text to map, although for different reasons: drawing ability; poor recall of standard symbols used on maps; and difficulty in showing temporal changes. Focus group discussions were held with the subjects who participated in the translation exercise. Participants were asked about: their major personal uses of maps and how they act to make sense of a map - what steps do they take first, second, etc.? In another study 82 subjects between the ages of 14 and 56 years were surveyed about the materials they read to assess their geospatial competencies. Dr. Gluck has concluded that experience with information increases competency with tasks to a greater degree than formal education. The final study described by Dr. Gluck involves determining public library users' geospatial data needs by doing a content analysis of questions asked at the Tallahasee Public Library. There was considerable variation in the needs ranging across educational (course assignments), professinal/career (job hunters), personal (hospital visit) and recreational (hikers/campers). Maps were only one of a variety of sources consulted by users. The second speaker in this session was Ray Larson, Associate Professor, School of Library and Information Studies, University of California, Berkeley who covered "Geographic Information Retrieval and Spatial Browsing". His presentation covered the following: 1) what geographic information retrieval (GIR) is, 2) geographic and spatial querying and browsing, 3) geographic and spatial indexing, and 4) a demonstration of some GIR systems and geographically indexed information. Dr. Larson described GIR as being concerned with providing access to georeferenced information sources. It includes all of the traditional areas of information retrieval research with the addition of spatially and geographically oriented indexing and retrieval. Furthermore, it combines aspects of database management systems research, user interface research, GIS research and information retrieval research. Geographic and spatial information retrieval systems are needed to access digital libraries, which are increasing at a rapid rate, and for the next generation of online catalogues. Geographic and spatial querying both imply querying relationships within a particular coordinate system. They can be defined as queries about the spatial relationships of entities geometrically defined and located in space. Geographical coordinates are geometric relationships, i.e. distance and direction can be measured on a continuous scale (100 km east of Toronto). Spatial relations may be both geometric and topological i.e. spatially related, but without measurable distance or absolute direction (inside the city limits). There are a variety of types of spatial queries: 1) point-in-polygon: what is located at this x,y coordinate? 2) region queries: what is in this region? 3) distance and buffer zone queries: what cities lie within x miles of border of Ontario and Quebec? 4) path queries: what is the shortest route from Winnipeg to Brandon? and 5) multimedia queries using non-map georeferenced information: what are the names of farmers affected by flooding in Monterey and Santa Cruz Counties, California? Spatial browsing combines ad hoc spatial querying with interactive displays. It has advantages and disadvantages. The former include: the accuracy of a full GIS may not be needed; a comrehensible searching metaphoris adequate for many materials. The latter include: c lutter and differing scales; the need for good and accurate geographical indexing. Traditional geographic indexing which uses place names from LCSH and name authorities also present problems: names are not unique; places referred to change size, shape and names over time; there are spelling variations; some places are temporary conventions (study areas, etc.). Geographic coordinates have some advantages over names: they are persistent regardless of name, political boundary or other changes; they can be simply connected to spatial browsing interfaces and GIS data; and they provide a consistent framework for GIR applications and spatial queries. One example of a geographic and spatial indexing system is GIPSY, developed by Allison Woodruff and Chritian Plaunt. It is designed to operate on the full text of documents, extracting geographic terms and attempting to identify the coordinates of the places discussed in the text using a combination of evidence. GIPSY uses the USGS Geographic Names Information System (GNIS) and Geographic Information Retrieval and Analysis System (GIRAS) to associate names with coordinates of named places, geographic features and land use characteristics. Identified places are added as "elevations" with each place adding a weight based on its frequency in the text and database characteristics. The resulting map is analysed to identify the most likely locations and coordinates for those locations are extracted. Examples of GIRs are: the UC Berkeley Digital Library Project, the NSF/NASA/ARPA Digital Library Project, GRASS Links public access GIS, and the Canadian National Atlas Information System (NAISMap). Larson summarized by pointing out that: GIR and spatial browsing can provide valuable new searching mechanisms for digital and traditional libraries with georeferenced collections; automatic georeferencing is possible and can aid in retrospective conversion; and GIR and spatial browsing should be among the components of digital libraries and next-generation online catalogues. The last speaker in the session was William Moen, Ph.D. Candidate, School of Information Studies, Syracuse University who described an innovation, "The Government Information Locator Service (GILS): Identifying, Locating, and Accessing Geographic and Spatial Data". GILS was created for a number of reasons: 1) the increase in government information in electronic format, 2) finding tools were not adequate e.g. USGPO catalogues, 3) technological advances made it possible, and 4) to enhance public access to government information. It will assist the federal government to manage its resources and users to know what is available, where it is available and how to get it. GILS is not specific to spatial data, but it will permit identification of data that is available. Mr. Moen stressed that GILS is a locator service or put another way, it is a point of entry to finding out about government information NOT the information itself. The information may be in any format - paper, electronic, map, etc. Each record consists of 20 core data elements and metadata is provided by the producers of the data. GILS provides a uniform interface to multiple servers and sets out standards for attributes of search queries. It can be searched by titles, key words and phrases, personal names and the record source. GIS In Academic Libraries I After lunch on Monday afternoon, Brent Allison, Director, John R. Borchert Map Library, University of Minnesota presented the practical side of GIS in a map library setting. With the assistance of a Department of Education grant received in 1992, the John R. Borchert Map Library established an "Automated Cartographic Information Center" (ACIC) to meet the needs of its patrons for a variety of GIS applications. Their configuration has evolved to an 11 workstation network with access to scanning and digitizing equipment and a colour printer. The ACIC is networked through the university's backbone making it the principal campus resource for GIS information. The ACIC is a DOS Windows and Mac operation providing to users a range of GIS and automated cartography products. Desktop mapping/GIS capabilities are provided by Atlas GIS (mostly used for digitizing), ArcView, and MapInfo. MapInfo is the most popular; it is easy to learn, permits layering of data and the easy production of customized maps. CD Export provides access to Tiger files in MapInfo format and Wessex permits the easy extraction of census data in a spreadsheet format. In addition, the ACIC provides a number of electronic atlases such as PC-USA, Global Explorer, MapExpert, Map 'N Go. In the first year, printed guides were developed by graduate students for the three GIS/desktop mapping programs. These guides however, have been rarely used. In their place, online tutorials are being developed, which will embed movies showing the steps involved in creating a map . LOTUS Screen Cam ($69.00), for instance, records movements on the screen which can be incorporated in word processing documents. Users of the ACIC must attend an orientation session before they are permitted access to the workstations. However, on-going assistance for users of the ACIC is still needed and extra student assistants had to be hired. Mr. Allison cautioned that often the requirement for on-going support after the period of a grant is overlooked when making grant applications. The ACIC is now involved in the development of an electronic atlas as part of ARL/GIS Literacy Project. The ARL Electronic Atlas is modelled after the urban atlas series produced by the U.S. Bureau of the Census using 1970 census data. The project, which is being developed cooperatively with a number of other libraries, will create maps of major metropolitan areas using socioeconomic and environmental data from the 1990 U.S. Census. The maps will be accessible on the Internet. Mr. Allison displayed sample maps from the project, pointing out that the atlas will be of interest to the public, schools, libraries and researchers. It is anticipated that the atlas will be particularly useful for people who do not have the resources or skills to construct the maps themselves. The maps will be in ""gif"" format and will include the statistics used to derive the maps. A prototype of the atlas is now on the Web. To close out this session, Christie Koontz, of the Florida Resources and Environmental Analysis Center, Institute of Science and Public Affairs, Florida State University described how GIS can be used by libraries, particularly public libraries, as a decision making tool to define their services and market areas. This capability is particularly useful in today's environment where budget considerations are forcing the consolidation of services. Many public libraries are hampered by less than optimum geographic locations for their facilities. Until recently, library managers have had to conduct market profile studies rather haphazardly. However, GIS is changing this, permitting managers to define more precisely the size of their current market, the composition of the market (age/sex/education, etc), and predict what the future market will be. She did point out that use of GIS is not without its pitfalls: problems of data accuracy and data availability exist; sometimes data is mislabelled; and until recently, there was a steep learning curve associated with using U.S. census data. Using an example from Evansville-Vanderburgh County Public Library, Ms. Koontz illustrated how GIS could be used to identify a new branch location. Tiger data was analyzed using GIS Plus software. She discussed a number of techniques that could be used to conduct the market analysis: assign census tracts to branches, or if more precision is required, assign block groups; zip codes can be used, or a radius of a certain value from the branch. Two miles is the radius most commonly quoted for public library services. (GW) GIS in Libraries II The second Monday afternoon session was led off by Dean K. Jue, Director, Spatial Analysis, Research, and Training Program; Florida Resources and Environmental Analysis Center, Institute of Science and Public Affairs, Florida State University who spoke on "Implementing GIS in the Public Library Arena". Mr. Jue contacted as many American libraries with GIS as he could. He determined that 64 libraries had GIS accessible to the public, including 13 public libraries (as of April 1995). His intention was to review library experiences in implementing GIS and to develop decision models for the implementation process. Mr. Jue identified a number of data use models including: 1) personal use model (librarians serve only as custodians of data), 2) chauffeur-driven model (librarians are GIS users for their patrons), and 3) adaptive interpersonal use model (combination of the above two). He also identified three spatial data user types: 1) map user (makes use of existing products), 2) personalized map user (makes use of existing data to produce personalized map for his own needs), and 3) map maker (acquires, manipulates, and analyzes data in a sophisticated manner). Based on the comments and responses he received from the 64 libraries, Jue developed a decision flow chart for implementing GIS in libraries. Key decision boxes are: adequate staffing levels? adequate equipment for GIS? are equipment, software, data secure? easy access to constituency data? (i.e., constituency identified?) are staff computer-literate? For example, libraries with inadequate staffing levels and inadequate GIS equipment are advised to implement the map user and personal use model. At the other extreme, if a library has adequate staffing and GIS equipment, has security measures in place and easy access to constituency data, then it can implement the map maker/adaptive interpersonal use (or chauffeur-driven) model. The responding libraries offered the following advice to libraries wishing to implement GIS. Provide the supporting librarians with enough time and training to learn the software as well as to provide the proper end-user support (from 48% of the 64 responding libraries); implement the GIS in a controlled environment with controlled access and go slow and define the role of the library and librarians carefully (28%); obtain state of the art equipment, including good printer and software protection (20%); make sure staffing level is adequate to support the project (17%); develop pre-canned and relevant data sets that are easy for the casual user to access (13%). Closing out the afternoon was "St. Louis Public Library's Electronic Atlas; Elements of a Successful GIS Application in the Public Library Environment" by Ann Watts, Coordinator, Information and Technology Services, St. Louis Public Library. The Electronic Atlas project began three years ago, using exhibit money with assistance from URISA (Urban and Regional Information Systems Association). Support for the project includes one 486 PC with 8 MB RAM and ArcView software, but no additional staff. The computer sits unattended in a public area and anyone can sit down and use it. Library staff have prepared a finite number of pre-planned "views" using census summary tapes; Ms. Watts commented that the 80/20 rule is very relevant in this situation. Staff are willing to do referenceinterviews over the telephone and to prepare maps for theseclients; copies of these maps are not kept by the library. Nor are patrons allowed to download information or copy files. She made the following comments about the impact of thisproject upon staff. Staff get new users started and teach themhow to solve problems with GIS; alternatively staff will operatethe system for patrons who don't want to become directlyinvolved. Staff do know their own limitations. Interestingly, the library has found that crib sheets do not work well; patronsprefer to ask staff for assistance rather than consulting userguides. Ms. Watts offered the following reasons for the success of theproject. First, the scope of the project is limited to the local county; the library does not collect data for adjacent counties. There has been good support from "outside" people including URISAmembers as volunteers and staff at the Illinois State Data Center at Southern IllinoisUniversity. Cost of the project has been minimal (less than $5,000) and yet, because of leverage, the library has been able to provide a colour printer. The service has been cost effectiveeven though the library provides colour printing at no charge to patrons. She did emphasis that such a project requires an"internal champion," someone who can work the political system from within the organization. (RHP) Specialized GIS Applications On Tuesday morning "The Role of Special Libraries in Emergency Preparedness and Response" was the subject of Robert Lee Chartrand's presentation which focused on the role of the library in emergency response, particularly. Mr. Chartrand is a consultant and was formerly with the Library of Congress. He stressed that libraries should plan to be ready. He mentioned several 'local areas of action'. One of these was the concern that official power centers have full government support. There sould be period reviews of guidelines and a monitoring of emergency management. The role of advisiory groups should be emphasized and federal/state interaction should be strengthened. It was important to make use of groups such as those involved in hotlines and services. He noted that it was important to incorporate information technology tools and techniques and to adopt recommendations of studies in the field. The various ways in which GIS can play a role were pointed out, such as the preparation of emergency scenarios and ensuring the availability of maps, charts, and air photos. Other ways that he noted were the selective storage of local information for emergency services (e.g. fire departments) and the provision of graphic information for public presentation. Mr. Chartrand also stated that corporations in a community have an obligation to help the local area. There are some deficiencies that exist, however: lack of access, lack of standardization, and a need to improve hardware and software compatibility. In his closing remarks, Mr. Chartrand stressed that we can help each other, and we can learn from the emergency management profession. We must ask ourselves: "Are we doing as much for our community within our profession as is possible? This address was followed by "Spatial Access to and Display of Global Change Data: Avenues for Libraries" by Linda L. Hill, Senior Research Scientist, Center of Excellence in Space Data and Information Science, Universities Space Research Association at the Goddard Space Flight Center. She talked primarily about data that is in federal agencies, the ideal spatial retrieval system, and systems that are available that have geographic search components. In her presentation, Ms. Hill mentioned the Global Change Data and Information System (GCDIS), a multi-agency project of the Global Change Data Management Working Group within the U.S. Global Change ResearchProgram. She distributed a handout on the current status of GCDIS and the most active departments/agencies in GCDIS. She outlined the five responsibilities of its Library Information Subgroup as building an infrastructure of libraries and librarians for GCDIS implementation, evaluating GCDIS from a library user's perspective and providing user needs analysis, linking data resources to information resources for knowledge management, promoting GCDIS to libraries and developing approaches to user education, and advising on data and information processing standards and systems from the library perpective. Ms. Hill went on to sketch her view of what an ideal spatial access information system should look like. The ideal system should include all types of data and information, specific area representations, spatial query capabilities, and translation of place names into spatial definitions. Other attributes it must have would include retrieval mechanisms that return a ranked list based on a metric of geographic similarity, ways of limiting or filtering, browsing snapshots of images, full documentation, and mechanisms for ordering data. She ended her presentation by noting and briefly describing five example systems: GC-ASK, EOSDIS, Global Change Master Directory, NOAA Marine Geology & Geophysics On-Line Searches, and the National Geospatial Data Clearinghouse. (TB) State and Local Initiatives in Spatial Data and GIS This second Tuesday morning session was devoted to state and local initiatives for addressing issues in community-based spatial information uses. Two cases were presented: publication of Illinois spatial data on CD-ROM and Champaign County's electronic network. The Illinois Department of Energy and Natural Resources with its divisions has developed over 10 years very comprehensive spatial digital databases at the state level. In order to make this information more accessible, many of these databases were published onC D-ROM in April 1994. This initiative reflects an attitude towards better access and sharing of spatial data by citizens and groups having an interest in Illinois people and resources. This electronic publication is intended for state and local agencies, libraries, schools, public interest groups, and the private sector. It is available free of charge (yes) except for a fee of $100 for private enterprises or individuals. Copies sold covered disk production costs. The CD-ROM contains layers of information for different topics such as hydrography, flood zones, roads, railroads, utilities, natural areas and preserves, archaeologic probability, boundaries, landfills, wells, public water supply intakes, towns, etc. Data come from local, state or federal sources. Most of the data is at scale of 1:100,000 with different scales depending on the map layer, and all are in ARC/INFO format. The data are not explicitly copyrighted, but prior notice to the Department is required before redistribution. The other local initiative presented pertains to thedevelopment of the Champaign County Network (CCNET) by theChamber of Commerce in co-operation with the National Center for Supercomputing Applications of the University of Illinois at Champaign-Urbana. The intent was to identify applications for a county-wide network and to design a network to connect the entire county and provide a model that can be adopted in other communities. The benefits are for business and industrial firms, schools and colleges, farming, medical care, and county planning. The listing of application task forces is revealing in itself: agribusiness, community and government resources and libraries, education, health care, small business, and geographic information networks. Current network access tools include Prairienet, a dial-up modem access via telephone lines and the Internet, and the Mosaic browser. For information about CCNET, the URL is: http://www.prairienet.org/SiliconPrairie/ccnet.html. An application for agricultural planning of GPS and GIS was brilliantly illustrated in a not so futuristic implementation of information technologies. A farmer would sample the soil for georeferenced analysis using GPS, passing this data to aGIS-based fertilization plan, and then to a GPS-governed automatic-adjusting fertilizer machine. The crop yields would be recorded, georeferenced through GPS, by the harvesting machine, and the plan for seeding the next year could be tabulated spatially! (YT) Generalization and Summary The Generalization Problem presented by Barbara Buttenfield, Associate Professor, Department of Geography, State University of New York at Buffalo focused on the nature of GIS data and the concepts of scale. Scale change, growth and changes of scale, adding and eliminating information due to scale, and models of growth and scale changewere discussed. GIS data is customarily troublesome and the concept of scale is complicated. She spoke of three main problems: data volume, indexing and metadata content and browsing. Data volume presents the problem of a lot of memory being used for one screen of information, so you need to compressthe data. Indexing focuses on data tiling and the need for better interface design. What map series and scale are appropriate? Metadata problems pertains to knowing that the data exists, determining its fitness for use and browsing capabilities. Dr. Buttenfield spoke about allometry and the scalability of phenomena and processes, not just the mathematical ratio. Her simple example was of a braided stream's detail being lost at 1:100,000 and 1:250,000 maps when compared to 1:24,000 maps. She concluded by stressing that there is a need for digital libraries and access to spatial information, a need to prepare now for growth and complexity that will surely happen and the important role that librarians and information scientists will play. Linda C. Smith, Conference Co-Chair and Professor, Graduate School of Library and Information Sciene, University of Illinois at Urbana-Champaign, closed the conference with a few brief words. She noted that the talks and other presentations illustrated the challenges and opportunities offered by GIS in libraries. There will be a proceedings volume for the conference. (CW)