1. INTRODUCTION

The proposed development of national geospatial data infrastructures has received considerable attention from institutional suppliers, the private sector and user communities in North America, Europe and many Asia-Pacific countries. Evolving from earlier data sharing and program coordination efforts, the term encompasses the sources, systems, network linkages, standards and institutional issues involved in delivering spatially-related information from many different sources to the widest possible group of potential users.

These components of geospatial data infrastructure are being implemented in different forms and at different rates around the world. A few datasets (e.g., Digital Chart of the World) are available internationally. However, it has been the particular local and national mix of available datasets, access mechanisms, tools, policies and institutions which have defined the spatial information "marketplace" in any given country.

Technological developments and existing efforts in more developed nations are already playing a major role in transforming traditional approaches to data collection, packaging, distribution and dissemination programs. However, cultivation of a global geospatial data infrastructure (or "GGDI") will first require its definition (however broad), followed by recognition of the diversity and similarities, convergence and conflicts, requirements and constraints characterizing users and producers of geospatial data around the world.

A clearer understanding of the term depends in large measure on the respective interests, requirements and constraints of the small collection of key "stakeholder communities" influencing GGDI developments today. Beyond examining how the characteristics of global efforts differ from those of their local, regional and national counterparts, it is also important to examine the "local to global transition" - the direct or generalised contributions of these more focused programs to the wider GGDI.

This paper is one of two discussion documents prepared in advance of an international seminar on Global Geospatial Data Infrastructure to be held in October, 1997 at the University of North Carolina Alumni Center in Chapel Hill, North Carolina, U.S.A. The objectives of this document are:

to provide a working definition of "global geospatial data infrastructure" (GGDI) and describe in some detail its potential requirements, components and functions;

to identify regional and global initiatives which could contribute to building a GGDI;

to provide input to the preparation of a GGDI implementation plan.

The second discussion document - dealing with GGDI Implementation issues and constraints - is being prepared by a team led by Dr. David Rhind of Ordnance Survey.

2. BACKGROUND

2.1 GII: The Communications Component

The vision of a global information infrastructure (GII) is very much wrapped up with issues surrounding the development of telecommunications infrastructure. In particular, as the world becomes more interconnected and interdependent, the world of telecommunications policy has been increasing in importance.

Such policy is currently driven by the belief that communications and information infrastructures collectively represent a principal "transforming technology" serving important social goals, and that development of these infrastructures is fundamental to economic growth. As well, it reflects a strong movement towards privatisation of services and increased competition in the telecommunications sector.

These driving factors go back at least to Marshall McLuhan's concept of a "global village", and reflect both the beliefs developed and lessons learned from programs and reviews undertaken over the past 25 years. For example:

In the mid 1970s, the French government released a report describing the imminent emergence of an "information society" [Simon and Minc, 1978]. Among many other issues, the report addressed questions of national technological sovereignty and introduced the term "telematique." At about the same time, Canadians were promoting the Telidon technology as the foundation for the "electronic highway." [Godfrey and Parkhill, 1979].

In 1983, Japan's Ministry of Posts and Telecommunications unveiled its "Teletopia Plan" to support the spread of advanced telecommunications across the country and to assure high-level information and communications functions for regional cities.

In 1987, the European Commission released a Green Paper on telecommunications which talked about modern communications networks constituting the nervous system of modern social and economic life; the paper went on to argue that European unity would depend in large measure on a unified, ubiquitous and advanced European communications infrastructure. [Commission of the European Communities, 1987].

By 1993, U.S. Vice President Al Gore was promoting the concept of an "information superhighway" and advancing the idea that an advanced communications and information infrastructure should be a national priority. In support of this idea, Gore stressed five key policy elements: (i) maximum possible reliance on private enterprise; (ii) promoting and protecting competition; (iii) ensuring open networks and universal access; (iv) avoiding information haves and have-nots; (v) maintaining flexible, adaptable rules and regulations. At the first World Telecommunication Development Conference, held in Buenos Aires, Argentina in March 1994, he presented a similar list of principals as the foundation for a Global Information Infrastructure (or "GII").

In May 1994 the European Commission published Europe and the Global Information Society. Commonly referred to as the "Bangemann Report", it presented the findings of a high level group of business people and community leaders brought together under the chairmanship of European Commissioner Dr. Martin Bangemann. This report formed the basis for much of the subsequent work of the European Commission on strategic planning for the Information society. It also influenced subsequent work of the G7 (including a February 1995 meeting of G-7 officials at which a broad consensus emerged for fostering a "Global Information Society") and national projects such as the U.K. Government's Information Society Initiative and government.direct.

Against this backdrop, the business community - whose commercial activities are becoming increasingly global - has become the principal force for the pro-competitive restructuring of telecommunications and information markets.

Governments, on the other hand, are increasingly being urged to facilitate rather than control these industry-driven activities by creating a legal and regulatory environment that supports efficient investment and innovation, and promotes full and fair competition. In many situations, politicians and public-sector executives are being expected to drive the technology into government by supporting testbeds for new technologies, fostering the transfer of resulting technologies to the private sector, promoting assimilation and use of applications and technology through government procurement, and developing applications that support government operations and dissemination of government information [Brown, 1995].

While numerous discussions of information concerns have been raised (See [Kahin, 1993], for example), there is no question that the telecommunications sector was largely responsible for framing the initial definition and discussion of global information infrastructure. This is reflected in the focus on the conduit rather than content defining the infrastructure, and the apparent "technology-push" rather than "demand-pull" emphasis in many early discussions of applications and target markets.

The concept of geospatial data infrastructure (GDI) was promoted from very different roots, with a much greater focus on content and concerns over the communication media treated as being largely beyond the community's control. The next section explores some of these roots and values underlying the vision of spatial data infrastructure as it is now being refined.

2.2 Managing Geospatial Information: From Integrated Mapping to Geospatial Data Infrastructure

The data-sharing paradigms underpinning the development of geospatial information networks today are also not new:

In the 1960's, proponents of integrated mapping practices advocated the registration, overlay, interpretation and analysis of different "layers" or themes of spatially-related datasets- possibly originating in different organisations - to the practical solution of important problems in land use planning and resource inventory (e.g., [Tomlinson, 1967]; [McHarg, 1969]).

Through the 1970's, the multipurpose cadastre concept launched major topographic and cadastral "base-mapping" mega-programs to support land administration at the local, state and federal levels across North America, Australasia and in emerging nations (e.g., [Roberts, 1968]; McLaughlin/NRC, 1974]). By putting in place a reliable and comprehensive basic map series, these mega-programs were intended to reduce perceived duplications of effort in basic mapping by end-users and encourage greater focus on creation and maintenance of special-purpose "thematic" layers.

By the early 1980's, the notion of "information as a corporate resource" [Diebold, 1979] and the information resources management movement encouraged individual organisations to implement collective approaches to the collection, management and sharing of designated hardcopy and computer-based data holdings of "corporate-wide" interest.

The manifestation of these data sharing precepts evolved from early dreams of centralised "land information databanks" through the 1960's and 1970's (e.g., [Cook et al., 1967]; [Roberts, 1968]) into the vision of more complex distributed land information networks in the 1980's. This vision conveyed the idea of linking together organisations responsible for the management of land-related information in a jurisdiction into a network to form a "virtual" geographic information system which could be queried in a manner similar to a single database. Hearle [1962] suggested such a concept at the state- and local government level almost thirty-five years ago, and researchers in the land administration, geography and geomatics communities have since examined the institutional and technological issues involved in considerable depth (e.g., [Palmer, 1984]; Sedunary, 1984]; [Onsrud and Rushton, 1995]; [Rhind, 1996] and many others).

Anne Branscombe [1982] introduced the term "information infrastructure" to refer collectively to the various media, carriers and even physical infrastructure used for information delivery. The term was being used in a much broader context by the late 1980's, and the notion of infrastructure as an enabling agent (i.e., enabling users to "plug in" to independent databases) was being adopted by the larger information processing community. Neil Anderson [1990] extended the definition by suggesting that information infrastructure should possess the following three important characteristics:

(1) the contents (data), conduit (telecommunications network) and flow-control procedures should be standardised;

(2) the major sources and users must be networked together; and

(3) the network must be customised for easy third-party access.

Through the 1980's and early 1990's, technological limitations and long, expensive database-loading programs served to limit the number of jurisdiction-wide, multi-participant land information projects that would qualify as "geospatial information infrastructure" under Anderson's definition. Even so, critical mass has now been reached in a number of more recent enterprise- or jurisdiction-wide efforts. At least five important reasons account for this acceleration:

(1) Increasing prominence of spatial data handling within organisations: There now exists a proven track record of intelligent application of GIS and desktop mapping-based processes to an increasing number of strategic and operational problems in business and government. Spatial data management - while not always optimally funded - is nevertheless considered an integral component of most enterprise-wide information management strategies by senior management, end-users and system integrators.

(2) Robust, easy-to-use and relatively inexpensive tools: Desktop mapping and spatial data viewing software packages (e.g., ArcView, MapInfo, GeoMedia and Maptitude) are now widely available at relatively low cost, and Internet-based spatial data viewing and map-making tools (e.g., CARIS Internet Server, Autodesk Map Guide, GeoMedia Web Browser and ESRI Internet Browser ) are also beginning to emerge.

(3) Ubiquitous data: The necessary digital spatial datasets covering entire areas of interest have now become more accessible, widely available and better packaged than ever before. Government departments in developed countries are under increasing pressure to provide robust, consistent, well-documented and competitively-priced digital data products to end-users and/or value-added repackagers. Especially in the United States, strong competition among value-added resellers has fed the "buyer's market" for spatial data - at least in the larger centres.

(4) Ubiquitous Communications: A critical mass of organisations and individuals have now worked through the technical and cultural transformations inherent in providing employees with desktop access to corporate networks and, more recently, the Internet. As well recent advances in communications and database technology are finally resolving what were serious technical limitations in the management and transfer of large, distributed datasets [Coleman, 1994].

(5) Greater availability of experienced people: Most important, a large and increasing number of specialists and informed end-users are now familiar with the capabilities and limitations of both the tools and datasets and with how they relate to the requirements of their particular business.

Finally, although it had a relatively small role in the first few rounds, the ubiquitous and inexpensive positioning, tracking and navigation capabilities of GPS are already pushing the next wave of market developments and having an effect on the manner in which end-users view and employ existing spatial datasets. The contribution of GPS (and subsequent positioning technologies) to future spatial data infrastructure efforts cannot be under-estimated or simply dismissed as a transitory "technology-push" phenomenon. Perhaps even more than the mapping and GIS products with which we are familiar, positioning "appliances" will drive the requirements, demands and practices of spatial information users over the next decade.

Clearly, Anderson's second two criteria - dealing with interconnection and easy third-party access - are being met on a daily basis already. With a few promising developments excepted, the first criteria (dealing with standardisation) is still being addressed - albeit perhaps less successfully - through a series of ad-hoc policies, one-on-one agreements and proprietary product development or data integration strategies.

By the early 1990's, the concept of spatial data infrastructure (SDI) development was being proposed in support of accelerating geographic information exchange standards efforts, selected national mapping programs and the establishment of nation-wide spatial information networks in the United States [Mapping Sciences Committee, 1993], the United Kingdom [Rhind, 1992], Canada [McLaughlin, 1991] and the European Community [Brand, 1995].

Interest in extending these early efforts to the global level has been evident from several different communities:

First, military organisations from a number of the NATO countries have been cooperating in the preparation of VMap data products and DIGEST Tools to support the so-called "Global Geospatial Information and Services Initiative" [Lenczowski, 1997]; [McKellar, 1996];

Similarly, the International Hydrographic Organization has made major efforts to cooperatively develop global standards defining the nature and content of electronic nautical charts produced by the hydrographic organisations of different countries [Anderson, 1996]

Third, the first Conference on Emerging Global Spatial Data Infrastructure - held near Bonn, Germany in September, 1996 - brought together key public-sector data producers and members of selected special interest groups from a number of the more developed nations around the world [EUROGI, 1996b].

Finally, the Santa Barbara Statement prepared from the Interregional Seminar on Global Mapping for Implementation of Multi-National Environmental Agreements (held Santa Barbara, California, USA, in November 1996) made a strong plea for the accelerated collection, promotion and use of the output from national and global mapping programs and the coordinated development of a global spatial data infrastructure [ISCGM, 1996].

Some of the key activities and requirements of these interest groups will be discussed in greater detail in Section 4.

3.0 DEFINING GLOBAL GEOSPATIAL DATA INFRASTRUCTURE (GGDI)

There is understandably no clear agreement on what "(geo-) spatial data infrastructure" and "geographic information infrastructure" efforts should or should not include: both the existing situation and most appropriate implementation strategies will necessarily vary from country to country. However, since development of a global geospatial data infrastructure does involve a common understanding of at least a fundamental set of concepts and terms, it is important to at least introduce for discussion the components and competing visions driving the various infrastructure efforts at this time.

3.1 First Approximations

A first approximation of the term Global Geospatial Data Infrastructure can be achieved by defining its components:

"Global" implies that anyone, anywhere will eventually be technically able to access to non-confidential data about the whole planet and its constituent parts.

McKee [1996] defined "geographic" data as those data describing phenomena directly or indirectly associated with a location and time relative to the surface of the Earth. The authors have opted to use the term "Geospatial" rather than "spatial" or "geographical" in order to be as inclusive as possible.

The word "Data" was chosen in lieu of "information", since the authors envision an environment in which much of the decision-making with respect to packaging data into information will reside with the user.

Webster defines "Infrastructure" as "...the underlying foundation or basic framework of a system or organisation."

The challenge is to come up with a definition which is not too restrictive and does not artificially limit thinking. This is especially critical in a GGDI which reflects the convergence of telecommunications, information services and information technology sectors, but yet is more than just the physical facilities used to transmit, store, process, and display voice, spatial data, and images.

In a broader context, Robert Pepper of the U.S. Federal Communications Commission expresses this challenge in the following manner:

"When we talk about infrastructure, we tend to think about wires - hardware. Infrastructure is far more than that. It is people, it is laws, it is the education to be able to use systems. If you think about the highway system, we tend to think about bridges and interstates, but the infrastructure also includes the highway laws, drivers' licenses, McDonalds along the roadside, gas stations, the people who cut the grass along the highways, and all of those support systems. You cannot talk about infrastructure in the telecom-information sector without also talking about the human support systems."

Beyond these components, Kelley [1993] believes "infrastructure" shares the following characteristics with data and information: (a) it exists to support other economic or social activities, not as an end in itself; (b) it incurs a relatively high initial capital cost; and (c) it has a relatively long life. Therefore it requires long-term management and commitment of funds.

3.2 GGDI Components

McLaughlin and Nichols [1990] suggested the components of a spatial data infrastructure should includesources of spatial data, databases and metadata, data networks, technology (dealing with data collection, management and representation), institutional arrangements, policies and standards, and end-users. (See Figure 1.) While the model they employed for this definition (actually extended from [Ford, 1990]) was useful in terms of trying to visualise the linkages between the components, the model had difficulty rationalizing the genuine overlaps between components (e.g., "Technology" vs. "Networks, "Institutional Arrangements vs. "Policies", etc.)

By comparison, Kelley [1993] likens spatial data infrastructure to the "skeleton" supporting the generation, flow and use of data. He sees it as including:

fundamental spatially-related data sets which can be used in a wide range of applications;

a pool of planning, management and technical expertise supporting cost-effective use of spatial information; and lastly

systems, standards and protocols to allow timely and efficient access to the high volume of spatially-related data within the fundamental data sets.

Kelley further suggests that these elements collectively serve to increase the sharing of data across electronic networks. This sharing, in turn, should minimise duplication of collection efforts, fosters new partnerships and enable more active civic participation in community planning.

Again along similar lines, the proposed "Victorian Geographic Data Infrastructure" (VGDI) in the State of Victoria, Australia is defined as the people, technology (software, hardware, data, telecommunications), and policies necessary to share geographic data across all levels of government, the private and non-profit sectors, academia and the community [Smith and Thomas, 1996] (with italics added by the author). Again, the driving vision is implementation of a distributed network of compatible data bases linked by common standards and protocols, each managed by appropriate custodians nominated in each jurisdiction.

3.3 Building a Working Definition

Summarizing the above considerations and terminology, then, the authors suggest the following working definition of GGDI:

"A Global Geospatial Data Infrastructure encompasses the policies, technologies, standards and human resources necessary for the effective collection, management, access, delivery and utilisation of geospatial data in a global community."

It would be diplomatically naive, operationally problematic and economically - well, just plain impossible to assume that each of these four elements should be or could ever be the same in every country. For this reason, the clear and growing need for global data coverage to satisfy specific environmental science initiatives will have to be satisfied for the foreseeable future by a patchwork quilt of selected global, national and project-specific datasets.

While truly global infrastructure remains a longer-term goal, Section 3 of this paper has already presented clear examples of transnational geospatial data infrastructure (or "TGDI") programs already underway by government, military and commercial interests. In these cases, the emphasis is placed on establishing application-specific information infrastructure in place on a cross-border, continental or perhaps even inter-continental basis. Taking these programs into account, it begs the question whether or not it is possible and worthwhile to collect and integrate the output from these individual TGDI efforts into some larger GGDI.

This opens up other questions for discussion. For example:

What are the respective driving applications, characteristics and requirements inherent in transnational and global geospatial data infrastructure efforts?

Where is the overlap? What are the steps to developing some minimum level of practical compatibility between national, transnational and global efforts?

Where are the gaps? What transnational efforts are we not taking full advantage of right now in trying to improve data coverage (especially in emerging nations)?

Do the routine applications and economic benefits exist to formalise a "vertical highway" between local, regional, national, transnational and global initiatives? Is such an option even practical?

Key interest groups promoting their respective visions of GGDI will be introduced in the next section, while Section 5 will be discuss alternative perspectives and social policy concerns .

3. GGDI STAKEHOLDER COMMUNITIES

3.1 The Military

Military needs for surveying and mapping internationally date back centuries, with the garrison painters being an interesting 19th-century example. While global mapping efforts were undertaken as far back as 1880, it was the military communities of NATO, the Warsaw Pact and their respective allies which began addressing the global mapping challenge after the Second World War. The creation and sustained funding support of the U.S. Defence Mapping Agency (now NIMA) and other military mapping organisations world-wide - as well as the efforts of NATO's Digital Geographic Information Working Group (DGIWG) to produce common military spatial data standards - are just two examples of the military commitment to global mapping.

The military (especially U.S. military) has played a significant role in developing and eventually spinning off the technology components which support global geospatial data infrastructure today. Some of the best-known examples of this - GPS, remote sensing, special-purpose computer hardware and software, and even the Internet itself, were developed to support basic mapping, surveillance, and command & control systems. Defence budgets have played a major role in supporting leading-edge research and development in industry, government and academic laboratories across the developed world.

Beyond the tools, however, military organisations have played a major role in spatial data collection and/or integration at a wide variety of scales. This work - albeit largely unavailable to civilians yet - has been carried out at in support of (e.g.,) battlefield command & control, emergency relief assignments, peacekeeping duties, general intelligence, cruise missile testing and even bilateral arrangements for drug enforcement. Since even the digital products have been produced over the past twenty-five years, their nature, content and format may vary widely.

Especially in the post cold-war period, the shift from battlefield-oriented warfare to more of a peacekeeping and anti-terrorist orientation has placed new demands on military's geospatial information infrastructure. Today, the application of information technology to weapons has involved "gathering huge amounts of data; processing them so that relevant information is displayed on a screen; and then destroying targets, at much greater distance and with much greater accuracy than was previously possible." [The Economist, March 8, 1997; p. 15]

Future warfare will be "multi-dimensional", requiring the integration of information to support land, air, sea and even space-based operations in the process. This revolution in military affairs revolves around advances in three types of systems:

(1) Intelligence gathering - through sensors in satellites, aircraft, unmanned aircraft;

(2) Intelligence processing - advanced C4 systems (for command, control, communication and computing) make sense of the data, display it on screen and assign weapons to particular targets); and

(3) Response - acting on all this intelligence (e.g., by using long-range precision strikes to destroy targets).

More advanced systems now in development are being designed to integrate these three components into a single system which "would allow a commander to watch a screen displaying everything going on in a battle, select targets and destroy them by pressing a button."

4.2 Science and the Environment

The history of systematic surveying and mapping efforts in support of scientific purposes goes back to at least the early 15th century with Infante Don Henrique of Portugal, Henry the Navigator. Project-oriented experiments (e.g., Eratosthenes calculation of the earth's circumference in the 3rd Century BCE ) go back centuries earlier.

Mapping the world's resources on a global basis has long been recognised as an important but problematic task by a variety of different disciplines. For example:

Between 1923 and 1985 there have been at least 26 calculations of closed forest land. The only long time series data for forest area come from the FAO returns of land cover for the 35 years between 1950 and 1985.

A study known as the Agro-ecological Zones Project began in 1976 to obtain a first-order estimate of the production potential of the world's land resources. Building on the 1:5,000,000 FAO/UNESCO Soil Map of the World, project scientists superimposed a climatic inventory characterizing temperature and moisture regimes matched to crop requirements.

Past problems with such inventories have included, for example, differences in scale and map detail of source materials, differences within single component classification systems, inconsistencies of survey techniques and even incompatibility of inventory objectives. (See [Townshend, 1991] for a more detailed discussion of these problems.)

To some extent, the influence of Our Common Future, the report of the World Commission on Environment and Development by Gro Harlem Brundtland (Chair) in 198_ highlighted once again the need of reliable geographic information and, to some extent, re-energised the international scientific community. A number of global change initiatives have been implemented to improve quality of observations and interpretation, manage large quantities of global change data, and communicate the results of global change research to the international community. Examples of key international programs include:

USGCRP Global Change Data and Information System (GCDIS)

the World Climate Research Programme (WCRP), which is part of the World Climate Programme

the International Geosphere-Biosphere Program (IGBP);

the Human Dimensions of Global Environmental Change Programme (HDP).;

Committee on Earth Observation Satellites (CEOS);

Global Mapping Initiative

These initiatives have not proceeded without some criticism or concerns from inside and outside the scientific community. Higher-profile concerns focus on such issues as the role of "big science" and state-based planning, limited funding resources and attendant difficulties in defining priorities, perceived American domination of the research agenda and, perhaps most important, continuing debates over what this work is really telling us with respect to (e.g.,) desertification and global warming. Cost and availability of appropriate spatial data coverage - either on a global basis or over selected areas - also remain key concerns of scientists in these programs.

4.3 The International Maritime Community

(from [Anderson, 1997])

The commercial shipping industry world-wide depends upon up-to-date information for safe and efficient navigation, and the International Hydrographic Organization (IHO) has been successful in developing international standards to support hydrographic data and electronic charts.

The European Countries have the most advanced number of Hydrographic Offices with the capability for ENC production. An initiative under the European Community called ECHO 926 is bringing these countries together to develop a common, networked electronic chart infrastructure. All of the countries have adopted IHO's S57 Standard and are moving into full ENC production of their territorial waters. A similar cooperative initiative to create ENC's of the St. Lawrence Seaway system is now underway between the United States and Canada.

The real challenge continues to be financing the construction of the marine geospatial data infrastructure. Since infrastructure is very costly to build, it is always difficult to raise the initial financing. Experiences in Canada have demonstrated that the use of Electronic Chart Systems and precise GPS dramatically reduces the risk of groundings, improves operational efficiency and reduces insurance costs [citation]. In that country, it was determined that the initial design and application of the marine geospatial data infrastructure could best be financed through electronic charting efforts where: (1) there was a clearly identified client base willing to pay for services; and (2) there exists a global requirement for standards-based products.

The military, global science and maritime communities are but three stakeholders in global - or at least transnational - GDI that have already developed cogent arguments and/or business cases for its extension. Other stakeholders - in particular, the increasing number of organisations involved in locating and tracking people or things and directing them from location to another - will emerge within the next five years.

5. ALTERNATIVE PERSPECTIVES AND ISSUES

This section of the paper attempts to promote further discussion on the topic by doing three things. After trying to briefly summarise the alternative perspectives which shape the debates over GDI implementation today, a new framework for discussion of the issues is presented. The section concludes with a brief listing of some of the social questions to be considered in GDI implementation.

5.1 Accommodating Alternative Perspectives

There are drawbacks to defining a concept in terms of its components rather than, for example, its appearance and function. The gap between perception and reality now characterizing many GDI efforts reflects the tensions that exist between competing visions of both the final destination and the routes to get there. Depending on their respective mandates and objectives, individuals or interest groups within stakeholder organisations may justify, design, implement and evaluate infrastructure-building efforts from one or several of five different perspectives:

(1) A data-driven perspective

(2) A technology-driven perspective

(3) An institutional perspective

(4) A market-driven perspective

(5) An application-driven perspective

The "Data-Driven" Perspective

This perspective reflects the values and aspirations of those involved in the creation, maintenance and dissemination of selected geospatial "foundation datasets" designed to be shared by a wide variety of users on a jurisdiction-wide basis. Much of the early impetus for both spatial data distribution and the sharing of spatial data between organisations has been management-driven, with a few visionaries seeing the economic and operational benefits of early mission-driven, multipurpose mapping, charting and land records programs.

These same players have more recently given top-down support to national and international spatial data standards efforts and, most recently, land information networking and infrastructure initiatives. Their interests and concerns are among the most widely-documented because, in part, they represent a natural extension of the values and efforts involved in both building and maintaining major mapping and land information programs and then working to ensure the resulting data is used as widely as practicable.

This data-centric viewpoint is illustrated most effectively by the Australia-New Zealand Land Information Council (ANZLIC) in their four-part model for NSDI, incorporating:

(1) an institutional framework which defines the policy and administrative arrangements for building, maintaining, accessing and applying the standards and data sets;

(2) technical standards defining the technical characteristics of the fundamental data sets;

(3) fundamental data sets (ANZLIC proposed a list of twenty-eight) which are produced within the institutional framework and fully comply with the technical standards; and

(4) a distribution network which is the means by which the fundamental data sets are made accessible to the community, in accordance with policy determined within the institutional framework, and to the technical standards agreed.

[ANZLIC,1994]

Again, note the emphasis on fundamental data sets in each part of the model, giving the infrastructure the appearance of a "virtual spatial data library" more than anything else. This "library" paradigm is shared by other GDI initiatives as well.

The Technology-Driven Perspective

Increasingly influenced by developments in personal computing, communications, positioning and database management technology, less constrained by institutional barriers, and more comfortable operating in a global marketplace, technology-centered organisations have a somewhat simpler vision of GGDI. According to McKee [1996], for example, a global spatial data infrastructure is like a wheel with technology as its hub, each spoke a different country. (See Figure 2.) Each country has SDI components or levels dealing with legacy data, culture, academic resources, professional organisations, governmental agencies, and legal and regulatory structures (for land tenure, privacy, intellectual property, environment, census, etc.)

While much of the interoperability focus has appeared to rest on promoting tighter interaction between GIS-related components, its real paybacks may emerge through facilitating smoother integration between Global Positioning System (GPS) units and a variety of other application software packages - applications which may or may not have a "spatial tradition".

This perspective dwells on the contribution of technology to the development of information infrastructure. At the risk of gross over-simplification, it is grounded in the beliefs (backed up by history) that: (1) the impact of "technology-push" on the design and evolution of infrastructure initiatives cannot be underestimated; (2) some minimum level of compatibility through standardisation actually increases choice in the marketplace; and (3) the ultimate choice of tools and standards will shape the ultimate form and content of the spatial information marketplace.

The Institutional Perspective

The institutional perspective looks at the issue of GDI through the lens of inherent mandates, responsibilities, limitations, conflicts and financial constraints of the respective organisations and constituencies involved. By necessity, it is coloured by concerns over the administration and financing of the programs and other initiatives supported by the data and the technology. Refined and communicated largely by senior government program managers, this perspective has been shaped by the organisational realities of internal and inter-organisational cooperation, relations with other sectors and, more recently, by "new public management" pressures towards greater financial accountability, more tightly-focussed program delivery, cost recovery and higher return on investment.

Members of the FGDC, IACG, EUROGI and ANZLIC have all made substantial contributions to the processes and formal mechanisms involved in the collection, maintenance and distribution of (largely) government datasets. They understandably take a "supply-side" view tempered by the realities of having to obtain results and progress in large, complex bureaucracies with limited funding, decentralised control and sometimes-conflicting information policies.

The Market-Driven Perspective

Advocates of this perspective emphasise the importance of market demand and customer satisfaction in defining and prioritizing products, product coverage, on-going maintenance and access to products and services. They are more concerned with nearer-term return on investment and are less burdened by public-sector concerns over jurisdiction-wide coverage or mission-driven program support unless there is a satisfactory guaranteed income stream associated with the venture.

Their concerns center on adding value to datasets in key "high-traffic" markets (e.g., demographic data / market research firms), expanding and diversifying the market for spatial data products (e.g., the high-resolution satellite imaging companies) and/or providing improved on-line access to existing datasets (e.g., Microsoft, Teranet and others). They are wary of cooperative efforts which seem to show little progress and promise only longer-term returns, and often want to "get on with the job" - with either proprietary or open solutions as the need warrants and the customer dictates.

Due to this (understandable) attitude, the inherent fragmentation of the industry and - in some cases - heavy reliance on mission-driven government programs, the various sectors of the mapping industry have left matters of national and international policy to their counterparts in government. However, given the increasing prevalence of: (1) strategic partnerships between industry and government; (2) outsourcing of system management and data distribution arrangements; (3) the dramatic growth of a competitive spatial data industry in the United States and elsewhere; and, most recently, (4) the commercialisation of high-resolution space-based remote sensing, there is a clear need to take into account the values and expectations inherent in this perspective.

The Application-Driven Perspective

By definition, most users view spatial data as a means to an end. While they may be fully familiar with their own data which they add to an application (e.g., tree types for forest management, property values and demographics for school site location, meteorological observations for weather prediction, etc.), they may possess only a peripheral interest in the underlying data collection methodology or original intended use of a road network, DTM or electronic nautical chart dataset. The basic "foundation datasets" are considered a given - the interest is in the products, services and new knowledge which may be developed atop this base.

The belief that end-users form the largest single "interest group" in GGDI is misleading. In fact, they form many different groups whose perspectives are sometimes at odds with one another. While all desire reliable, up-to-date data as quickly and as inexpensively as possible, their underlying requirements and traditions with respect to practical data usage may vary widely.

Pushed ahead by Internet developments, expectations regarding rapid access to data perhaps best characterise this perspective at this time. An early examination of selected Internet news groups and discussion lists indicated that the monthly number of inquiries regarding availability of on-line spatial data covering a particular area of interest increased by more than 400% between June 1991 and June 1993 [Coleman and McLaughlin, 1995]. The volume of such requests has increased substantially since that survey was taken. Further, with the growing interest since 1994 in on-line spatial data distribution, the World-Wide Web has clearly come into its own as a leading environment for data searching, browsing and retrieval.

This growth in spatial data usage and interest has also resulted in growing levels of expectation concerning the availability and easy usage of on-line resources. Most desktop mapping users are accustomed to working with well-documented, often easy-to-use software packages with pre-packaged datasets. Those familiar with desktop publishing have come to expect reliable transfer of graphics and text from one application to another in almost real-time, and have neither the expertise nor the desire to consider the special problems involved in spatial data exchange. From the viewpoint of such consumers, the data is no more than a commodity which - like software - should be competitively priced, readily acquired, robust and reliable within given bounds.

Where the data is publicly available and free of copyright restrictions, private companies have responded to this group by offering pre-packaged datasets optimised to work with a given software package and available for next-day delivery via courier or, more recently, for immediate delivery via the Internet.

Where copyright and cost-recovery issues do apply, government agencies have responded by enhancing their products and/or improving their distribution capabilities through agreements with strategic partners, value-added repackagers or designated sales agents. In some cases, slow initial rates of product acceptance outside government departments have led suppliers to re-examine their policies with respect to data pricing, usage royalties and/or product packaging.

Discussion

The authors do not suggest that the perspective of every individual and interest group involved in GDI development neatly fits exclusively into one of these categories. Quite the contrary, most such individuals interact widely with others, and - over time - have modified their concerns and priorities in light of this interaction.

Even so, participants in collaborative infrastructure-building programs have experienced confusion and frustration when trying to reconcile the respective values and priorities involved. Indeed, very few individuals in any jurisdiction may be able to appreciate the respective values involved in any two or three of the above perspectives, much less all five. Moreover, the more one can appreciate the bigger picture, the less one typically knows about the technical details or political implications of any one subject area. Progress is slow, and "mis-connections" are frequent.

Questions arising from this include:

"Top-Down", "Bottom-Up", "Technology-Push", "Laissez-faire": Are these approaches mutually exclusive? If not, how can they be combined over the next ten years to cultivate the growth of geospatial data infrastructure(s) within and between countries?

"It is now possible, therefore I need it": Can we predict the medium-term impact of technology-convergence (communications, positioning and computing) and interoperability on user "requirements" and - by extension - on the nature and form of GDI components?

5.2 Defining the Study of GDI: Issues at the Interface

Clearly, GGDI has the potential to be greater than the sum of its parts. While many groups already address the four components mentioned in the working definition (i.e., technology, policies, standards and human resources), it may be valuable to focus on the interfaces and relationships between these components [Schell, 1997].

The cells in Table 1 identify examples of the various issues associated with the relationships between these components. Note that, while we are already seeing some convergence of opinions among the more technical components (i.e., toward the lower-right corner of the Table), there is still considerable divergence of opinion in the people- and policy-oriented relationships described in the upper left quadrant.

5.3 Questions for Discussion

Criticisms of information infrastructure - in fact, the overall global information society vision have emerged from a wide variety of quarters, and from both developed and developing communities. Rather than discuss these issues, it may be more useful to leave them as questions for consideration by Conference participants. These questions may include:

Is a single definition of "basic" infrastructure even appropriate, or is it too ambitious or naive an idea at this time?

Should our attention be directed to a more limited geographic focus? For example, much of what has been brought under the GIS vision may really only be appropriate at this time for those OECD countries characterised by democracy and well-developed market economies.

What can we recommend to less developed countries? What are the alternatives? Are their opportunities for "leapfrogging" developments in North America, Europe and Australasia?

Many from the developing world originally criticised the emphasis on advanced communications at a time when too many of their citizens still lack access to the most basic telephone services. However, the subsequent rapid uptake in wireless communication in the Far East and the Caribbean may enable users there to move into more advanced remote applications faster than users in more developed countries.

What is the real influence of pricing on the nature and direction of GDI development?

Is there any common agreement on what customers are prepared to pay for? How can transnational or global information infrastructure efforts be designed to accommodate different attitudes and policies with respect to pricing of information?

...and perhaps two of the most important questions or issues to be considered in the discussions:

How much of this information infrastructure effort will grow "organically" regardless of our plans, policies or actions?

The tension over finding the appropriate balance between "prescriptive" and "enabling" policies is beginning to become as pronounced in the geospatial world as it is in the wider information infrastructure debates. Where do we draw the line?

Can we recommend appropriate combinations of prescriptive and enabling policies to promote effective application of geospatial data infrastructure in support of substantial and sustainable economic development?

This brings us back to the Conference theme. Are there any "basic values" which forward planning efforts or some form of regulatory regime should be designed to support? What are the most important benefits to be sought from additional public investment in information infrastructure?

CONCLUDING REMARKS

Geospatial data infrastructure is seen as a new and exciting environment accompanied by remarkable technological developments, full of promise and opportunity for research, education, commerce and even the promotion of democracy. However, it is also an "electronic frontier" in which laws and values and standards remain unwritten, and where individuals and institutions will be working in pioneering conditions.

It was not the intent of the authors to fully define GGDI or to provide a ready set of answers and strategies to conference participants. Quite the contrary. If our working definition of GGDI and the introduction of the notion of transnational infrastructures result in disagreement and constructive debate, that is good. If our interpretation of the various interest groups and interface issues - however flawed - provides some framework for further discussion, then so much the better.

Finally, it is hoped that this paper will serve to introduce more application- or issue-specific presentations from others on the first day of the Conference - and adequately provide a lead-in for the second Theme Paper and more heated discussions on GGDI implementation to be held on Day #2.

REFERENCES

Andersen, N.M. [1990]. "ICOIN Infrastructure". Proceedings of a Forum on the Inland Waters, Coastal and Ocean Information Network (ICOIN), pp. 4-17. The Champlain Institute, Fredericton, New Brunswick, Canada. June.

ANZLIC. [1994]. Strategic Plan 1994-97. Australia New Zealand Land Information Council, Canverra, A.C.T., Australia.

Brand, M.J.D. 1995. "The European Geographic Information Infrastructure (EGII)". Paper presented at the ESIG Conference - 27-29 September 1995, Lisbon, Portugal. URL Reference http://www.frw.ruu.nl/eurogi/forum/esig.html.

Brand, M.J.D. [1995]. THE EUROPEAN GEOGRAPHIC INFORMATION INFRASTRUCTURE (EGII), Remarks by the President of EUROGI, ESIG Conference - 27-29 September 1995, Lisbon, Portugal. URL http://www.frw.ruu.nl/eurogi/forum/esig.html

Branscomb, A. [1982] "Beyond Deregulation: Designing the Information Infrastructure." The Information Society Journal, Vol. 1, No. 3, pp. 167-190.

Coleman, D.J. and McLaughlin, J.D. [1994]. "Building a Global Spatial Data Infrastructure: Usage Paradigms and Market Influences". Proceedings of the 1994 Canadian Conference on Geographic Information Systems, Ottawa, Ontario. Vol. 1, pp. 31 - 45.

Coleman, D.J. [1994]. "Geographic Information Systems Performance in a Broadband Communications Environment". Unpublished PhD. Dissertation. Department of Surveying and Spatial Information Science, University of Tasmania, Hobart, Tasmania, Australia. February.

Commission of the European Communities. [1987] Towards a Dynamic European Economy: Green paper on the Development of the Common Market for Telecommunications Services and Equipment.

Dawe, P. [1996] "An Investigation of the Internet for Spatial Data Distribution". Unpublished Master of Engineering Report, Department of Geodesy and Geomatics Engineering, University of New Brunswick, Fredericton, New brunwick, Canada. March.

EUROGI [1996a]. "Towards a European Policy Framework for Geographic Information: A Working Document", Working Group report prepared on behalf of the European Organization for Geographic Information. November. URL http://www2.echo.lu/gi/gi2000/en/egii/gi2000dd.html.

EUROGI, DDGI, AI, ILI/ILA, OGC, FGDC and FIG [1996]. Proceedings of the Conference on Emerging Global Spatial Data Infrastructure, Dr. Martin Bangemann, Patron. Königswinter, Bundesrepublik, Deutschland (Germany). . Distributed through the European Umbrella Organization for Geographic Information (EUROGI), Mr. Michael Brand, President. September.

Godfrey, David and Douglas Parkhill. [1979]. Gutenberg Two. Toronto: Press Porcepic Ltd.

ISCGM [1996]. "Santa Barbara Statement on Global Mapping for Implementation of Agenda 21." Prepared at the Interregional Seminar on Global Mapping for Implementation of Multi-National Environmental Agreements, Santa Barbara, California, USA, 13-16 November 1996. International Steering Committee for Global Mapping (ISCGM). URL http://www1.gsi-mc.go.jp/iscgm-sec/sbs/SBS.html.

Kelley, P.C. [1993]. "A National Spatial Information Infrastructure." Proceedings of the 1993 Conference of the Australasian Urban and Regional Information Systsms Association (AURISA), Adelaide, South Australia, Australia.]

Lenczowski, Roberta E. [1995]. "The Global Geospatial Information and Services Initiative." GIS in Government, The Federal Perspective, 1994. Fort Collins: GIS World Books, 1995.

Mapping Science Committee [1993]. Towards a Coordinated Spatial Data Infrastructure for the Nation. Report of the Mapping Sciences Committee, Board on Earth Sciences and Resources, Commission on Geosciences, Environment and Resources, National Research Council. National Academy Press, Washington, D.C., U.S.A.

McHarg, I.L. [1969]. Design with Nature. Doubleday Books, New York, New York, U.S.A.

McKee, L. [1996]. "Building the GSDI - Discussion paper for the September 1996 Emerging Global Spatial Data Infrastructure Conference". Proceedings of the 1996 Conference on Emerging Global Spatial Data Infrastructure, Königswinter, Bundesrepublik, Deutschland (Germany). European Umbrella Organization for Geographical Information (EUROGI), September.

McKellar, D.G. [1996]. "The Global Geospatial Information and Services Inititative." J2 Geomatics, National Defence Canada. WWW URL http://www.j2geo.ndhq.dnd.ca/documents/GGIS/GGISdoc.html. March.

McLaughlin, J.D. [1991]. "Towards National Spatial Data Infrastructure." Proceedings of the 1991 Canadian Conference on GIS, Ottawa, Canada, pp. 1-5. Canadian Institute of Geomatics, Ottawa, Canada. March.

Onsrud, H.J. and G. Rushton [1995]. Sharing Geographic Information. Center for Urban Policy Research, Rutgers University Press, New Brunswick, New Jersey, U.S.A.

Palmer, D.W. [1984]. "A Land Information Network for New Brunswick". MScE. thesis, Department of Surveying Engineering, University of New Brunswick, Fredericton, N.B., Canada.

Rhind, D. [1992] "The Information Infrastructure of GIS." Proceedings of the 5th International Symposium on Spatial Data Handling, Vol. 1, pp. 1-19. Humanities and Social Sciences Computing Lab, University of South Carolina, Columbia, S.C., U.S.A.

Sedunary, M.E. [1984]. "LOTS and the Nodal Approach to a Total Land Information System". Proceedings of "The Decision Maker and Land Information Systems"--FIG Commission III International Symposium, ed. A.C. Hamilton and J.D. McLaughlin. Edmonton, Alberta, Canada. October.

Simon, Nora and Alain Minc. [1978]. L'information de la societe, Rapport M. le President de la Republique. Paris: La Documentation Francaise.

Smith, M. and E. Thomas [1996]. "Lessons Learned for the Leading Edge - Emerging Global Spatial Data Infrastructure". Proceedings of the 1996 Conference on Emerging Global Spatial Data Infrastructure, Königswinter, Bundesrepublik, Deutschland (Germany). European Umbrella Organization for Geographical Information (EUROGI), September.