Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 GIS/GIScience Research Needs N ational needs in GIS/GIScience span a wide spectrum of research topics, ranging from applications through abstract theoretical considerations in cartography, computer science, geography, military science, philosophy, psychology, intelligence and security stud- ies, and sociology, among other disciplines. Some of the dilemmas evoked by GIS technology and its existing and proposed applications arouse seri- ous public policy questions and debate. The suitability of existing organi- zational arrangements to meet the demands inspired by GIS/GIScience can be questioned with specific reference to college and university struc- ture, research funding mechanisms, and research agenda setting. PRACTICAL AND THEORETICAL CHALLENGES Geographic information systems have revolutionized the ways soci- ety handles geospatial information, allowing the automation of what were previously tedious and inaccurate methods of map analysis, the construc- tion of sophisticated simulations of real systems, and the visualization of geospatial information in new and exciting ways. Much research is still needed, however, to handle new and potentially powerful datasets, to exploit research advances that have been made but not implemented in ways that are easy to use, and to provide effective tools in support of spatial decisions. In addition to enjoying new ways of collecting, viewing, and manipulating more accurate and more precise geographic data, geo- graphic information scientists continue to improve their capacities to cre- ate higher-dimensional datasets by integrating data from multiple 43
44 BEYOND MAPPING sources. Maintaining or accelerating current rates of improvement in data integration constitutes a major challenge, as does grappling with the re- lated questions of access to data (who can have what, when, and where), data preservation (what and how much to keep and how to organize what is kept) and providing access to stored information. The Mapping Science Committee's report Weaving a National Map (NRC, 2003b) commented on several of the data characteristics envisioned in the USGS plan for the National Map. The committee used the metaphor of blankets and quilts to describe the complex nature of multiresolution geospatial information that is inevitable in today's world. While organi- zations such as the Bureau of the Census must work with uniform blan- kets of a specified scale for the entire nation, local governments continu- ously create quilts of extremely high resolution for property records and infrastructure management. The committee identified this as a major is- sue, noting that "the edge-matching problems caused by variable resolu- tion will be severe and not always solvable" (NRC, 2003b, p.51). This is the type of practical technical issue that is likely to be addressed by the new Center of Excellence in GIScience that is being created by the Na- tional Geospatial Programs Office (USGS, 2005). The research agenda for this center is being created with assistance from the NRC's Mapping Sci- ence Committee. While GIS software has demonstrated its importance and functional- ity in business and government applications that rely on the creation, maintenance, and retrieval of spatially referenced information, applica- tions in some scientific domains have been more challenging. For example, atmospheric, hydrologic, and environmental scientists often need to model dynamic processes that occur in three dimensions above or below Earth's surface. The first "Environmental Modeling with GIS" meeting was organized in 1993, and a series of scientific conferences have followed. In order to meet the specialized needs of researchers in these communi- ties, software vendors have expanded their tools for analysis, interpola- tion, and flow modeling. They have also developed interactive tools that enable a researcher to visually develop process models that link data in- puts to procedures and output. These tools enable a hydrologist to create a surface flow model that can run with standard GIS software and also be shared with other colleagues for further refinement and evaluation. Clearly, three-dimensional virtual globes such as Google Earth assist with the visualization of these model results. They also provide an excel- lent way to discover real-time data sources such as USGS stream gauges. The challenge of extending existing GIS tools into complex dynamic en- vironments is being addressed by a new GIS initiative at the Center for Capacity Building in the University Corporation for Atmospheric Re- search (UCAR) within the National Center for Atmospheric Research
GIS/GISCIENCE RESEARCH NEEDS 45 (NCAR) in Boulder, Colorado. The main objective of this initiative is to promote and support the use of GIS as both an analytic and an infrastruc- ture tool in atmospheric research in order "to foster collaborative science, spatial data interoperability, and knowledge sharing with GIS." A work- shop in 2002 explored crucial issues related to integrating weather and climate data with complementary information from the physical sciences, social sciences, and related areas of the geosciences. A group of UCAR and NCAR scientists and engineers have been exploring opportunities for using GIS to enhance knowledge sharing and integration for research, applications, and education (http://www.gis.ucar.edu/initiative.html [accessed May 24, 2006]). Some additional problems and questions for which more ideas and more trained people are needed are · Ways to map and analyze such dynamic phenomena as new road construction, thunderstorm and hurricane development, or animal, hu- man, and inventory movements in near real time; · Developing real-time maps for handheld devices, including audio capabilities; · Testing possible relationships between disease outcomes and envi- ronmental, demographic, and social indicators to predict the spread of disease through human, animal, or plant populations; · Refining navigational information for the sight impaired; · Analyzing networks to identify choke points and critical nodes, and the potential effects of removing or blocking selected links in a road network or for evacuation during emergencies; · Investigating why past data-sharing efforts failed and identifying the types of institutions and mechanisms (e.g., mandates, incentives, regu- lations) most likely to succeed given current interaction among federal, state, local, and private organizations; · Methods for analyzing vast repositories of geospatial data in search of patterns and anomalies to uncover unknown associations between and among attributes; and · Portraying greater detail in response to (1) using voice or gesture activation, (2) pointing and clicking (linking and brushing) on features of an image that are linked to supplemental information (e.g., audio files), or (3) zooming to a higher-resolution image. Privacy Issues Many geospatial datasets contain sensitive information and conse- quently are restricted or licensed. For example, an insurance company's database of property information is one of its most valuable and protected
46 BEYOND MAPPING assets, yet those data would be invaluable to emergency response organi- zations during a fire or natural disaster. Other data, while extremely use- ful to fire and police organizations, could be dangerous in the wrong hands. Research is needed to identify both the technical advances required to create systems for sharing restricted data and to formulate business models that simultaneously support and constrict access to sensitive datasets. In fact, the emergence of GIS/GIScience could be taken as a stimulus for taking a fresh look at the entire concept of privacy relative to geospatial data, rather than creating a set of ad hoc solutions. Interoperability Even though geospatial data are created for numerous purposes by a variety of public and private organizations using different standards and software, it is important to maximize the interoperability of data and soft- ware. Many types of geographic data are created by a variety of public and private sector organizations. For example, when anthrax spores were detected in New Jersey postal facilities, state health officials scrambled to identify building locations, potentially affected neighborhoods, and build- ing floor plans for inclusion in the state GIS in order to facilitate analyses. Databases used by state officials were often incompatible with local infor- mation sources. The result was masses of datasets with no way to inte- grate them quickly, accurately, or efficiently. The optimal benefits of these data will be realized and the risks they pose for society will become more evident when they can be accessed, integrated, and manipulated simulta- neously by diverse organizations coordinating their activities. GIS/GISCIENCE AND SOCIETY1 As with any new technology, one of the challenges facing societies that employ geographic information systems is that of ensuring maximum benefits while minimizing the risks of misuse. Geographic data can be misused, either deliberately or through inadvertence, as often occurs when individuals innocent of the principles of mapping portray data in- appropriately. The collection of geographic data and the ability easily to copy digital data can raise questions about who owns what data, for how 1The term "GIStudies" has been proposed to designate research on the uses to which GIS and GIScience are put and investigations of the interactions between GIS/GIScience and societies. Though not widely adopted, the distinction does identify a critical realm for scien- tific and policy research, either independently or as a focus within the general topic of sci- ence, technology, and society analyses (Forer and Unwin, 1999).
GIS/GISCIENCE RESEARCH NEEDS 47 long, and for what purposes. Numerous questions regarding intellectual property with respect to geographic data are being contested, and many more will arise in the future as the use and value of geographic data in- crease (NRC, 2004). Geographic information systems and geographic information science appear to be benign technologies, but some of their applications have been questioned; as is true of any technology, GIS, though neutral in and of itself, can be used for pernicious ends. GIS makes it easier for marketers, for example, to pinpoint likely customers for purposes of advertising, but the same data and techniques can be used to profile individuals and groups for surveillance or for robbery. Because individuals can be pro- filed by linking data to their places of residence, the collection and com- parison of geographic information can result in the erosion of individual privacy (Monmonier, 2002). At what appears now to be an extreme, GPS chips can be embedded in animals, children, parolees, and rental cars, making it possible to track their movements continuously, leading to the possibility of forms of geoslavery, if some individuals acquire the capabil- ity to control the movements or locations of others using monitoring tech- nologies (Dobson and Fisher, 2003). Mapping viewsheds (the places that can and cannot be seen from a specific point) on a battlefield or for sur- veillance may be good or bad, depending which side of the war a person is fighting on or a person's attitude toward general-purpose surveillance. Applications that are highly beneficial in many situations could be omi- nous in the hands of a totalitarian government (NRC, 1997). The country needs GIScience professionals who have a sufficient background in the policy and social sciences to be sensitive to the full array of positive and negative applications their new technologies enable. In the final analysis, the mapping sciences exist to provide society with geospatial information, and their success in doing so must be the basis of any measure of their value to society. The world of geographic information creation has changed dramatically over the past few decades, as a result of new technologies for sensing, acquiring, assembling, vali- dating, disseminating, and using geospatial information. Many problems and questions remain, however, and there are many ways the supply of society's geographic information could be improved. RESEARCH INFRASTRUCTURE The country's research infrastructure for GIS/GIScience is poorly de- veloped. There exists no core outside the few federal agencies that tradi- tionally had national mapping responsibilities. Within the National Sci- ence Foundation (NSF), the major source for basic science funding has come from the Geography and Regional Science program in the Social,
48 BEYOND MAPPING Behavioral, and Economic Sciences Directorate, from various programs in the Computer and Information Sciences Directorate, and from the Educa- tion and Human Resources Directorate. No single NSF program has pri- mary responsibility for geographic information systems and geographic information science. The NSF implicitly recognized the need for coordi- nation of GIS/GIScience in the appointment in late 1998 of a GIS coordi- nator, but the position was not filled when the incumbent left the agency shortly thereafter. In 1999, the NSF sponsored a workshop on emerging research themes in GIScience (Mark, 1999). The workshop clearly differentiated two types of research associated with GIS: (1) scientific research using GIS, and (2) research that advances GIScience. Both kinds of research are often intertwined, as they frequently are in the discussions of the UCGIS. In practice one might add a third kind of research: using GIS to imple- ment the results of science in the formulation of policy, in spatial decision support systems, for example. The workshop report identified four re- search issues facing GIScience: (1) the integration of data, (2) analysis of the relationship between data and scale, (3) the implementation of models of process, and (4) usability. The workshop report also recommended that: · NSF recognize GIScience as a coherent research specialty and es- tablish a funding center for it as soon as possible; · Both basic GIScience and research using GIS be supported by the new unit to promote integration of these related research efforts; · NSF establish an internal task force consisting of representatives from all its directorates and the Office of Polar Programs, charged to meet regularly and ensure that GIScience links to all relevant parts of the foun- dation and benefits from their operations; and · NSF appoint a multidisciplinary advisory panel of non-NSF per- sonnel to assist in defining, implementing, and evaluating the new unit's effectiveness. Other federal agencies with potential interest in funding research in the mapping sciences include the National Geospatial-Intelligence Agency, the U.S. Geological Survey, the Bureau of the Census, and the National Aeronautics and Space Administration. Many other agencies have supported specific GIS applications, including the National Insti- tutes of Health and the Department of Justice. Yet, no central, coordi- nated office for funding the mapping sciences exists within the federal government. Nevertheless, funding opportunities for GIS/GIScience exist in nu- merous government agencies. Mapping scientists have successfully col- laborated with scientists from other disciplines to compete for major fund-
GIS/GISCIENCE RESEARCH NEEDS 49 ing on projects in which geospatial data, technologies, and principles are indispensable. Almost no funding for basic research has been forthcom- ing from private corporations. The research conducted by private firms has for the most part been kept as privileged data, shared only when eco- nomic advantages are perceived. Examples of private/public funding ex- ist in allied fields in Canada and Europe. ERTICO (Intelligent Transporta- tion Systems and Services--Europe) is a not-for-profit, public-private partnership whose mission is to implement intelligent transport systems and services in Europe. ERTICO's projects are financed by annual fees from its partners and by project funding from organizations such as the European Commission. In the United States, a prominent example of pub- lic-private funding is the collaboration between private companies and the U.S. Food and Drug Administration (FDA). The 1992 Prescription Drug User Fee Act (PDUFA) authorizes FDA to collect fees from compa- nies that produce human drug and biological products that FDA reviews. Perhaps a similar scheme would work in the United States for fund- ing GIScience research, with a nonprofit entity such as the Open Geo- spatial Organization disbursing funds in support of a consensus research agenda. The organization would need base funding from the federal gov- ernment and from agencies and private firms interested in the results of specific research agenda topics. One could even imagine that the non- profit entity might be supported by modest taxes on location-based ser- vices, such as one cent per month on every mobile telephone or a flat tax on each GPS device. In the absence of this or some other form of aug- mented funding, GIS/GIScience will continue to be more fragmented than necessary. EXECUTING RESEARCH AGENDAS Since 1990, several groups have proposed no fewer than 11 overarch- ing research agendas for GIS/GIScience (Table 3-1). The lists of commis- sions of the International Cartographic Association (http://www.icaci. org/en/commissions.html [accessed 24 May 2006]) and the International Society for Photogrammetry and Remote Sensing (http://www.isprs. org/tcwg.html [accessed 24 May2006]) provide insights into current re- search practice in these specialties within the mapping sciences. The most recent agenda prepared by the UCGIS lists a comprehensive set of long- and short-term major topics (Sidebar 3-1); the National Research Council's Mapping Sciences Committee has also put forth a list of re- search priorities for the National Geospatial-Intelligence Agency (NGA) (Sidebar 3-2).
50 BEYOND MAPPING TABLE 3-1 Proposed GIS/GIScience Research Agendas, 1988-2005 Date Proposed By 1988 David Rhind (1988) 1989 National Center for Geographic Information and Analysis (1989) 1992 National Center for Geographic Information and Analysis (1992) 1996 University Consortium for Geographic Information Science (1996) 1999 National Computational Science Alliance (1999) 2000 International Cartographic Association (2000) 2002 University Consortium for Geographic Information Science (2004) 2002 Ohio State University (2002) 2003 National Research Council (2003a) 2004 Robert McMaster and Mark Monmonier (2004) 2006 National Geospatial-Intelligence Agency (NRC, 2006b) The numerous efforts to devise an overarching research agenda par- allel new developments in technology and its applications. At the same time, they reflect an increased interest in investigating interactions be- tween GIS technology and society, each in its own way resting on varying conceptions of individuals, computers, mapping, and society, with GIScience at their common core (Egenhofer et al., 1999; Goodchild et al., 1999; Sheppard et al., 1999). Research about the individual is dominated by cognitive science and focuses on understanding spatial concepts, learn- ing and reasoning about geographic data, and interactions between hu- mans and computers. Research about computers is dominated by repre- sentation, adoption of new technologies, computation, and visualization. Research about society addresses the effects of technologies and societal concerns about their use. Differences among the research agendas for GIS/GIScience that have been proposed over the last 30 years generally reflect the varying focuses of the groups that have put them forward, and many common elements transcend both the multiple disciplines engaged in GIS/GIScience and the time that has elapsed since geographic information systems moved beyond the experimental stage. In some respects, the first formal research agenda was proffered in the 1987 National Science Foundation solicita- tion for proposals for the National Center for Geographic Information and Analysis (NCGIA). Formulated on the basis of broad consultation with the contemporary GIS research and applications community, the solicita- tion listed four goals for the new center (NSF, 1987): · Advancing the theory, methods, and techniques of geographic analysis based on geographic information systems;
GIS/GISCIENCE RESEARCH NEEDS 51 SIDEBAR 3-1 UCGIS Research Agenda Long-term Research Challenges Spatial ontologies Geographic representation Spatial data acquisition and integration Remotely acquired data and information in GIScience Scale Spatial cognition Space and space/time analysis and modeling Uncertainty in geographic information Visualization GIS and society Geographic information engineering Distributed computing The future of the spatial information infrastructure Geospatial data mining and knowledge discovery Short-term Research Priorities GIS and decision making Location-based services Geoslavery Identification of spatial clusters Geospatial semantic web (a web of geospatial data that can be processed by machines) Incorporating remotely sensed data and information in GIS Geographic information resource management Emergency data acquisition and analysis Gradation and indeterminate boundaries Geographic information security Geospatial data fusion Institutional aspects of spatial data infrastructures Geographic information partnering Geocomputation Global representation and modeling Spatialization Pervasive computing Geographic data mining and knowledge discovery Dynamic modeling SOURCE: UCGIS, 2004; McMaster and Usery, 2004.
52 BEYOND MAPPING SIDEBAR 3-2 NGA GIS/GIScience Research Priorities The list of research priorities for the National Geospatial-Intelligence Agency (NGA) prepared by the NRC Mapping Sciences Committee (NRC, 2006b) lists 12 salient tasks: · Assimilation of new, numerous, and disparate sensor networks within the tasking, processing, exploitation and dissemination process; · Spatiotemporal data mining and knowledge discovery from het- erogeneous sensor data streams; · Spatiotemporal database management systems; · Process automation and human cognition; · Visualization; · High-performance grid computing for geospatial data; · Image data fusion across space, time, spectrum, and scale; · Role of text and place-name search in data integration; · Reuse and preservation of data; · Detection of moving objects from multiple heterogeneous intelli- gence sources; · Geospatial intelligence ontology; and · Multilevel security. The mission of the NGA--until 2003, the National Imagery and Map- ping Agency--is to provide timely, relevant, and accurate geospatial intel- ligence to support national security. Although some of its priorities are distinct to its specialized mission, many of them overlap to a greater or lesser degree with GIScience research priorities identified by other groups. SOURCE: NRC, 2006b. · Augmenting the nation's supply of experts in GIS and geographic analysis; · Promoting the diffusion of analysis based on GIS throughout the scientific community; and · Acting as a clearinghouse for disseminating information on re- search, teaching, and applications. These goals were to be met by addressing five priority research topics each containing a number of specific tasks (Abler, 1987):
GIS/GISCIENCE RESEARCH NEEDS 53 1. New modes and methods of spatial analysis -- Social science applications for GIS-based spatial analysis --Error and error propagation --Nontraditional statistics 2. A general theory of spatial relationships --Theory of spatial relationships --Nonplanar relationships among multiple objects --Efficient data storage structures --Structures for volumetric data --Structures for time-dependent data --Methods for integrating heterogeneous data --Techniques for redefining data --Translations among different locational schemes 3. Artificial intelligence and expert systems in GIS --Automated data entry --Database summaries and indexes --Map evaluation and interpretation --Intelligent geographical information systems 4. Visualization --New options for color and motion --Three-dimensional maps --Showing error on maps --Noncartographic means of displaying spatial relationships 5. Social, economic, and institutional issues --GIS adoption and implementation --Costs and benefits of GIS --Information access --Privacy --Legal questions Many, if not most, of the long-term research challenges and short- term research priorities listed in the most recent iteration of the UCGIS research agenda (Sidebar 3-1) persist in one form or another and fit under one or more of the rubrics contained in the 1987 NSF solicitation. For the most part, the NGA priorities (Sidebar 3-2) appear to be agency mission- specific instances of the more general priorities identified in the UCGIS tabulation. The persistence of many elements (often indeed in advanced and refined formulations) of the 1987-1988 statement of priorities is per- haps to be expected in a newly emerging specialty. Alternatively, their persistence may suggest that a fresh look at the most recent iteration is in order. The UCGIS research agenda is an attempt at a consensus program based on a generally open and broad participatory process. In the committee's judgment, it should soon be endorsed and pursued by the
54 BEYOND MAPPING U.S. GIS/GIScience community or modified as needed to make it accept- able to that community. Broad support throughout the GIS/GIScience re- search and applications community would be a major step toward gar- nering increased support for GIS/GIScience research and toward producing the well-qualified GIS/GIScience professionals the country needs.
