"Geographic information science should no longer be funded solely from the periphery of other sciences." (Onsrud, 1998)
Geographic space [1] is the arena within which social and environmental processes act. Location, therefore, is a fundamental dimension of the phenomena examined by the environmental and social sciences. In this context, geographic information must be a key component of any information system designed to support social or environmental modeling or decision-making. Geographic information systems (GISs) are designed to store, retrieve, analyze, and display geo-referenced data of various types. Computer-based GISs developed over the last 30 years now form the core of a $2 billion industry in the United States. GIS technology is used to route delivery vehicles or hazardous materials, to locate health care facilities or stores, and to plan parks and housing developments. GISs also are used to study distributions of ill health, crime, and endangered species, to assess environmental risks, and to support geodemographic marketing.
Geographic Information Science (GI Science) [2] is the basic research field that seeks to redefine geographic concepts and their use in the context of geographic information systems (GIS). GI Science also examines the impacts of GIS on individuals and society, and the influences of society on GIS. GI Science re-examines some of the most fundamental themes in traditional spatially-oriented fields such as geography, cartography, and geodesy, while incorporating more recent developments in cognitive and information science. GI Science also overlaps with and draws from more specialized research fields such as computer science, statistics, mathematics, and psychology, and contributes to progress and expansion of those fields. It supports research in political science, public health, and anthropology, and draws on those fields in studies of geographic information and society. Geographic Information Science is quintessentially multidisciplinary in every sense.
Unfortunately, as a critical and emerging area of multidisciplinary science and engineering, GI Science is in danger of being lost in the interstices of the conventional infrastructure set in place by established disciplines and funding programs. It seems clear that the basic research aspects of GI Science are most appropriately funded by an organization, like NSF, that funds basic research. However, as a relatively new and multidisciplinary topic, GI Science finds no specific home at the Foundation that accommodates all its needs—although embraced by many, GI Science is nevertheless marginalized.
Because of the importance of Geographic Information Science to basic science and the National interest, we propose a workshop to critically examine the relationship of GI Science to existing programs and initiatives at NSF, and to examine the prospects for a new Initiative (or theme) at NSF in the area of GI Science and Geospatial activities.
The Disciplinary Mix of Geographic Information Science Research
It is difficult to assess the relative strengths and interests of various disciplines in Geographic Information Science. One way to estimate the distribution of interest is through analysis of scholars who associate themselves with GI Science through the University Consortium for GI Science (UCGIS) [3]. As of July of 1998, 30 of the 43 academic members of UCGIS listed participating scholars by discipline on their UCGIS web sites. Of 659 researchers listed, 459 (70 %) were associated with fields normally funded primarily by NSF, with the remainder in more applied fields funded by NASA, EPA, NIH, and other such agencies. Of the 450+ scholars associated with NSF-funded disciplines, 141 (about 30 %) were from geography departments, with another 44 from related fields such as planning, bringing the G&RS-affiliated total [4] to about 40 %. Another 17 % of the researchers were from other SBE-associated departments. If the home departments of the remaining researchers are counted by the NSF Directorate most likely to fund them, the numbers would be: 64 (14 %) from GEO, 56 (12 %) from ENG, 45 (10 %) from CISE; 22 (5 %) from BIO, and 9 (2 %) from MPS fields.
A somewhat different perspective is gained from NCGIA's Project Varenius, [5] which articulates the basic research agenda for advancing GI Science around three foci: cognitive foundations, computational implementations, and society-technology interactions. Implicitly, this suggests that there is relative equality among the behavioral, computational, and social sciences, and that CISE, for example, may have a much larger potential role to play in GI Science than is indicated by the fact that only about 10 % of the UCGIS scholars are from CISE disciplines. On the other hand, the important and practical applications in the areas of medicine, planning, spatial analysis, statistics, and the physical sciences, for example, are only peripherally acknowledged.
Given the wide distribution of GI Science research across disciplines spanning all NSF Directorates, a workshop to assess priorities and mechanisms for supporting this multidisciplinary, cross-directorate field at the Foundation should prove extremely valuable for NSF's planning and budget process. Such a review also is critical for the future health of basic GI Science research in the United States, critical to Federal initiatives such as the National Spatial Data Infrastructure (NSDI) and to the future of the GIS industry.
Specific Objectives of the Workshop
The specific objectives of the workshop are as follows:
• to assess recent and current levels of activity at NSF in the areas of GIS, GI Science, and Geospatial research (including infrastructure development broadly defined to include educational activities);
• to 'map' the dominant concepts, principles, and activities of GI Science and Geospatial research onto traditional academic disciplines and then onto NSF programs;
• to examine the explicit and implicit interrelationships among GI Science, Geospatial activities, and various existing special NSF initiatives, themes, and thrusts (e.g., KDI, LEE, STC, Digital Libraries, Digital Government, etc.)
• to determine the extent to which existing activities at NSF are able to meet the demands for basic GI Science research support as articulated by the research community via such vehicles as recent reports from UCGIS and GISDATA on research priorities
• to identify the prospects for and the hurdles preventing cross-Directorate collaboration in the areas of GI Science and Geospatial activities at NSF, with an assessment of how improvements might be effected, and develop mechanisms consistent with NSF operating procedures to manage and distribute appropriate levels of basic research funding to the GI Science research community
• to define a five-year vision for GI Science basic research topics and needs, and develop mechanisms to link that with NSF's strategic plan and future program and initiative development.
Workshop Particulars
The workshop will be held at NSF in the Fall or Winter of 1998-99. It will involve approximately 30 participants, roughly half of them from NSF itself, and the other half from the US GI Science research community outside NSF. All non-NSF workshop participants will be invited by the workshop co-Chairs in consultation with appropriate people at NSF. Although several NSF program officers and other staff have a strong grasp of GI Science and its constituents, participation from non-NSF personnel is required to provide balance and expertise across the full range of key disciplines that are fundamental to GI Science. We believe that a two-day workshop is most appropriate, with an alternation of plenary discussions and working breakouts. At least one of the break-outs would likely divide along the lines of Directorate-specific activities and needs.
The workshop would have three non-NSF researchers as co-Chairs. David M. Mark (University at Buffalo), currently President of the University Consortium for Geographic Information Science, is proposed as the co-Chair representing Geography and the Social Sciences. Co-Chairs representing the GEO/BIO perspectives and the CISE/ENG perspectives will be selected in consultation with NSF program officers and officials in those Directorates. Lastly, since GI science knows no international boundaries, we propose to include two foreign scholars as participants. Ideally, these would be scholars who have been strongly involved in GI Science activities such as the European Science Foundation's GISDATA project (e.g., Ian Masser (formerly Sheffield, now ITC Enschede, Netherlands) or Andrew Frank (Vienna)).
Participation from all the related NSF Directorates will be solicited early in the process, and we expect that people from these Directorates will be involved in the planning process. Key contacts will include, but not be limited to: Bernard Bauer (SBE), Alan Gaines (GEO), Maria Zemankova (CISE), Mike Lesk (CISE), Cliff Jacobs (GEO), Scott Collins (BIO), Tom Baerwald (GEO), Larry Brandt (CISE), Bill Bainbridge (SBE), Cheryl Eavey (SBE), Mike McCloskey (SBE), Gary Strong (CISE), and others yet to be determined.
Once the workshop agenda is set, it will be useful to invite some participants or observers from other federal agencies that fund GI Science research, especially NIMA, NASA, and DARPA, as well as representatives of OMB, OSTP, and similar parts of the Executive branch. Input from the NRC's Mapping Science Committee also would be solicited.
The primary product of the workshop will be a 10-20 page white paper, to be written by the non-NSF workshop co-chairs with input from other workshop participants. The white paper will document recent patterns of NSF funding for GI science, outline the relations of GI Science to current and planned NSF programs and cross-cutting activities, and make recommendations to NSF regarding appropriate mechanisms for minimizing the chance that traditional disciplinary boundaries inside and outside NSF will impede funding for this multidisciplinary field. The white paper will be published on the World Wide Web. A glossy brochure could also be produced if NSF wishes and if supplementary funds for printing are made available.
Depending on the success of this workshop, it may be appropriate to refine the agenda and plans through subsequent workshops, each seeded with a core of participants from the initial cross-cutting workshop proposed in this document.
References
Goodchild, Michael F., 1992. Geographical Information Science. International Journal of Geographical Information Science. v. 6, pp. 31-45.
Onsrud, Harlan J., 1998. UCGIS Congressional Breakfast presentation, April 1998. (http://urban.rutgers.edu/Lyna/powerpoints/PresPP97-H/sld006.htm)
Notes:
1. Geographic space here refers to outdoor space at scales from several meters to the size of the planet. Spatial information or data for phenomena at geographic scales is sometimes referred to as 'geospatial' data or information.
2. GI Science (as distinct from GIS) was first identified by name in 1990 by M. F. Goodchild in a keynote address at the International Symposium on Spatial Data Handling in Zurich. The term is used as a label for the basic research field underpinning geographic information systems (GIS) (Goodchild, 1992). GI Science was included in the name of the University Consortium for Geographic Information Science (UCGIS) when it was founded in 1994/5, and in NCGIA's NSF-funded agenda-setting Project Varenius, "Advancing Geographic Information Science" (SBR-9600465).
3. The University Consortium for Geographic Information Science (UCGIS; see http://www.usgis.org) was formed in 1994/5 to serve as a unified voice for the GI Science research community. As of June 1998, UCGIS had 54 members and Affiliate Members, of which 42 were Universities. UCGIS includes the great majority of the US research universities with strength in GI Science.
4. Note: By no means have all of these researchers been funded by NSF, nor would they necessarily come only to G&RS at NSF to seek research support. They merely are from disciplines that most commonly are considered for NSF funding by the G&RS program.
5. See note 2, above.