GSDI and ESDI - Views on Interoperability and

Spatial Data Infrastructures in Europe.

PETER A. BURROUGH, Utrecht

ABSTRACT

This paper traces the history and context of international spatial data initiatives in Europe, outlining the efforts made by bodies such as Directorate General XIII/E of the European Union, EUROGI, MEGRIN, ESF-GISDATA, CEO, OGIS and others to further interoperability, the use of meta data and efficient data exchange. The internal and external social, economic, political and technical forces driving internationalism in GI in Europe are described with special attention for the aspects in which Europe differs from North America. Experience with several discipline-related GI data integration efforts is presented in terms of the ease or difficulty with which interoperability has been achieved. The role of the Open GIS Consortium and US GI Companies in determining Europe's approach to interoperability is examined and suggestions are made as to how Europe can improve coordination and cohesion to enhance its world position with respect to GI matters.

1. THE HISTORY OF INTERNATIONAL SPATIAL DATA INITIATIVES IN EUROPE.

Although the opening up of the single market in Europe means that both organisations and individuals originating in any single Member State may move freely across borders to compete for work and opportunities in other Member States there are still many situations where neither the spirit nor the letter of the various European Treaties is followed at national nor local levels. Even where agreed EU Directives have been approved by Parliament and the Council of Ministers, in a number of diverse subject areas, Member State governments are still often loath to implement them - at the end of 1996, something like 12 Member States were in default of implementing important, fully agreed EU Directives, and many were or will be taken before the European Court of Justice if the appropriate national law changes are not enacted shortly. Small wonder, therefore, that sharing of spatial data is fraught with difficulties. Until relatively recently, most spatial data were regarded as national property, or even as state secrets, that were definitely not to be shared.

In recent years, however, much has been happening to change the situation in Europe with respect to geographical information. A year ago, under the auspices of European Commissioner Dr Martin Bangemann, a meeting was held in Bonn to discuss activities concerning the Emerging Global Spatial Data Infrastructure (GSDI). This meeting was organised by the European Umbrella Organization for Geographic Information (EUROGI) together with the Deutscher Dachverband für Geoinformation (DDGI), and North American and International Institutes: the Atlantic Institute, the Institute for Land Information and its Land Information Assembly (ILI/ILA), the Open GIS Consortium, Inc. (OGC), the US Federal Geographic Data Committee (FGDC) and La Fédération Internationale des Géomètres (FIG), Commission 3 (Chenez 1996). This was perhaps the most far reaching activity in the history of the globalization of Geographical Information initiatives in Europe. A second GSDI meeting will be held in Raleigh, North Carolina on 20-21 October 1997, with Michael Brand, President of EUROGI in the Chair. Not long ago, the idea of a national spatial data infrastructure was unthinkable in many countries. Ten years ago there were few persons outside specialist areas (in which I include the armed forces) who would have envisaged the creation of shared, on-line geographical databases that are available to all persons with a personal computer linked to the World Wide Web. In Bonn, people did not just discuss if this was technically possible, but also how this development could be guided along the best possible paths, while recognizing that the huge amount of work to be done to properly identify or specify even a basic framework within which to address, discuss and resolve the many issues involved in globalization of GI.

Before going further, I must clarify the meaning of "Europe". The term "European Union (EU)" refers to the current political association of 15 states (Finland, Sweden, Denmark, Germany, UK, Ireland, Netherlands, Belgium, Luxembourg, Austria, France, Italy, Spain, Portugal, Greece), "Western Europe" includes all the foregoing plus Norway, Iceland, Switzerland and the smaller states like Andorra, Monaco, the Vatican, Liechtenstein and San Marino. "Former Eastern Europe" includes the new Baltic States of Estonia, Latvia and Lithuania, and the states of Poland, Russia, Belarus, Ukraine, Bulgaria, Romania, Macedonia, Albania, Czech Republic, Slovakia, Slovenia, Hungary, Croatia, Bosnia and Yugoslavia. Obviously, with such diversity in history and political organisation it is difficult to speak of a single "European" view of anything, let alone spatial data. Yet means must be found to improve the coordination of the collection and use of GI on a continental basis, e.g. in environmental monitoring (Chernobyl is still there - and leaking), for watershed management (the major river basins of Europe transgress political boundaries), coastal zone management or in transport planning (land routes, air routes and sea), to mention but four.

The current state of the international situation with respect to spatial data in Western Europe is the result of several actions taken by a relatively limited number of individuals in several key organisations and disciplines. At the upper political levels the stage was set by former Commission President Delors who in December 1993 presented the European Council in Brussels with a major vision for Europe. A critical element was the development of the information society, especially within the triad of the European Union, the United States and Japan. High-level, senior representatives from the industries which would implement the information society met early in 1994 under Commissioner Martin Bangemann and made recommendations which were presented to the European Council at the Corfu summit in June 1994: these were later given more substance in the Commission's action plan "Europe's Way to the Information Society" (COM(94) 347 of 19.7.94). The European Council in Essen in December 1994 further underlined the importance of the emerging information society on the role of new information services, and especially the content, or the information itself. The G7 Ministerial Conference in Brussels on 25-26 February 1995 confirmed the opportunities the information society should offer and stressed the need for global co-operation. Several of the projects defined at this summit involve significant use of geographic information, in particular projects on Environment and Natural Resources Management, Global Emergency Management and Maritime Information Systems (Burrough 1996).

At the levels of professionals, disciplinary organisations and governmental and inter-governmental bodies there have been recent moves towards a European approach to harmonising activities concerning spatial data. MEGRIN - the Multipurpose European Ground Related Information Network, set up as a collaborative project by CERCO (Comité Européen des Responsables de la Cartographie Officielle) in 1993 has created SABE (seamless dataset of the administrative units) from digital data supplied by all MEGRIN partners (Salgé 1997). EUROGI - the European Umbrella Organisation for Geographic Information, was researched by a handful of professionals in 1992-1993 before it was set up with support from an EC DG XIII/E IMPACT programme initiative in 1994. The creation of EUROGI as a single pan-European umbrella organization representing national and other pan-European GI/GIS related umbrella organizations significantly affected the development of national spatial data organisations in countries where national organisations did not yet exist. Today, most countries of the European Union and EFTA are members of EUROGI, which plays an important role in leading the discussion and communicating ideas about European GI initiatives through studies, workshops, presentations at International GI conferences, and an on-line web site (http://www.frw.ruu.nl/eurogi/eurogi.html)

Other European activities include the work of GI standards committees (specifically CEN TC287/TC278 - David et al. 1996), the establishment of international GI conferences and workshops (e.g. EGIS - 1990-1994, AM/FM 1988- present; the Joint European Conference JEC - 1995-1997, the EC-GIS Workshops held in June in each year, to name but a few), the setting up of academic research programs like GISDATA, sponsored by the European Science Foundation, and training programmes sponsored by COMETT (GISIG). The EGIS and ESF-GISDATA academic initiatives throughout Europe are being followed up by the creation of AGILE - the network for Academic Geographic Information Laboratories in Europe.

Private sector initiatives include the development of seamless road databases for all of western Europe, moves towards interoperability and the publication of several European monthly trade journals (e.g. GIS-Europe and Geomatics Info Magazine) and several book series by major publishers. The activities concerning road navigation systems have stimulated interest in de facto standards (TC278) for the GSDI discussion. DGIII (industry) has provided support to GI industrial initiatives.

European initiatives in space research and remote sensing are significant in terms of investment and the GI products they produce, but in general activities in remote sensing have proceeded independently of conventional mapping and ground-based GI initiatives. The European Space Agency (ESA) and private companies (e.g. SPOT Image and companies such as Eurosense to name but one) have made major investments in imaging technology (hardware, software, interpretation techniques, etc.) generally, in regard to GI. The EU-financed Centre for Earth Observation (CEO) located at the Joint Research Centre at Ispra, Italy (sponsored by 50 Mecu of Member State government money, matched by 50 Mecu of EC funding ) is a major player in the coordination of GI and earth observation data from remote sensor platforms.

In spite of this large and diverse activity concerning GI in Europe, the term 'interoperability' is relatively new here. Initially, the concept of 'interoperability' referred to problems in the United States of America when transferring spatial data from one commercial computer system to another. Very soon, the concept embraced the problems that different groups of users in the same government or organisation had in linking data sets for the same areas that had been created at different times, by different disciplines for different users or applications. There are analogous problems in Europe both within and between countries. At first the problems of exchanging GI were perceived as being merely technical and it was thought that they could be solved by the use of agreed data transmission and exchange standards. To facilitate the development of standards, committees were set up both in North America and Europe, and elsewhere. Perceiving that many factors could affect the quality and value of spatial data, the concept of 'Meta data' - literally 'data about data' was born to provide users with information about the methodology, lineage, provenance, ownership, copyright and pricing policies of the data suppliers. Inevitably, standards for metadata were immediately perceived as also being necessary.

An important part of the discussion on interoperability, data exchange and metadata, has been the question as to which individuals or groups in Europe should take the lead in setting up standards for GI and systems of metadata, particularly for international applications. Should Europe simply follow a lead from North America, or from the vendors of GI systems to proceed directly to a Global spatial data infrastructure, or should Europe work from national systems initiated by national agencies to a European spatial data infrastructure, perhaps via MEGRIN or through GI users working through EUROGI, or through the European remote sensing path via the Centre for Earth Observation or from a lead given by the European Commission?

There may be no simple answer to this question because the degree to which issues such as interoperability, data exchange, metadata, etc. matter to different types of GI user varies widely. A local authority/commune parks manager is generally only concerned with "his" GI (if aware of it at all!), whereas a regional planner of parks development perhaps within Dept. of Environment (or equivalent) is concerned with a much wider view, and both types of use could benefit from nation-wide surveys (land use, hydrology, geology, acid rain fallout, etc.) carried out by yet another national or even international organisation. The problem comes in motivating (by legislation, by encouraging some form of altruistic "best practise", or otherwise) purely local interests to collect, process and make available their GI on a wider basis, perhaps to users, and for purposes, not even dreamed of at the time of collection, while operating within typically tight local budget restrictions. Even in the USA the new "GI metadata standards" apply only to federal and (some) state government GI, but not to local government, which still accounts for three quarters of all GI in America

Realising that commercial operators and government agencies often have different priorities and views, during the last 30 months, the EU Directorate General XIII/E, EUROGI and other interested parties in the European GIS circuit have prepared the document GI2000 "Towards a European Policy Framework for Geographic Information" (DG XIII/E 1997a). This has been freely circulated on the Internet and by conventional means in order to open up the discussion concerning the need to establish a European policy framework for geographic information through which data can be created, marketed, used, reused and shared in a cost-effective manner for the benefit of society. The document, which has undergone much refinement, points out the potential of a good GI policy for providing better and more efficient government as well as new business opportunities for the nascent European geographic information industry. The GI2000 concept contains many ideas about how a European Geographic Information Infrastructure could arise, and identifies the strengths and weaknesses within Europe towards this end. In particular the following issues that affect decisions at global, European and national levels are noted.

Stimulating the creation of base data, bearing in mind data interoperability

Stimulating the creation of metadata services which follow/use agreed standards

Lowering legal barriers and reducing potential risks for GI collection and use

Stimulating public/private synergy, especially access to "public information"

Co-ordination at European level, while respecting the principle of subsidiarity at national level

Technological support (isn't this directly related to more help with R&D?)

Research and Development

Quality and standards

Awareness and training

Market awareness of the value of GI to both the owners of datasets and to potential users.

DG XIII/E also commissioned three supporting studies on GI-Metadata, GI-Policy, and GI-Base data; executive summaries of these studies are available on the World Wide Web (DG XIII/E 1997b). Clearly, within some EU Directorates General there is interest and activity at integrating GI over the three levels from NSDI through ESDI to GSDI, though there is still much to be achieved. GI is so heavily used within and by the European Commission itself that a separate Unit exists in the European Statistical Office (Eurostat), called GISCO - GIS at the Commission, whose mandate is to help advise other EC Directorates in regard to better coordinated use of GI. However, GISCO has no remit in regard to assisting or advising non-EC users of GI. Clearly, some body, agency, organisation must be found or formed which can take on this wider advisory and coordination role.

2. THE INTERNAL FORCES DRIVING INTERNATIONALISM IN GI IN EUROPE - WHY SHOULD WE BE WORKING TOGETHER ANYWAY?

The collapse of the Berlin wall in 1989 marked the beginning of a new era in European history. In the previous decades the main political divisions were East and West and the demarcation line ran through the European heartland. Before the collapse, Western Europe and the European Community were inward looking, focusing on economic integration, world trade and transatlantic and Mediterranean issues; the dominant model of political and economic organisation was the principle of the independent sovereign state, whose integrity is fundamentally sacrosanct. Immediately pre-1989 there was no vision of a Greater Europe as there was under Charlemagne and in the later Middle Ages when the name "Europe" meant "Unity of Christendom". The broad visions of a continental political and cultural entity faded when Europeans explored, conquered and divided the world to suit themselves; in recent centuries these visions were briefly revived by forceful figures, and repulsed. Today, partly as a result of 1989, partly as a result of globalization of the world economy, a dream of a greater Europe is being revived, but this time one in which European unity will be based on the free will of European peoples who are able to express their opinions democratically (Andriessen 1996).

The success of economic, social and political moves towards a greater Europe as an important economic and social power block in the world will depend on skills and services, not raw resources. Three quarters of European jobs are in the service sectors, including of course the provision and use of geographical information, with its importance for running and managing any corporate, spatial entity, be it nation state or local authority.

Just as nation states have used Geographical Information in their land tenure and national "ordnance" or topographical mapping programmes to give rights and status to their existence, so must a wider Europe not only be able to draw on spatial data that are relevant at European political levels, but also provide information that will be shared among the people at all levels in the member states. At the level of European government and European government agencies there are political requirements to address issues that cover the whole of the European Union. These include not only Europe-wide issues like demography, agriculture, communications or the environment, which are addressed by bodies like Eurostat and the European Environmental Agency (EEA), but also a host of separate projects that have been funded under the many EU Commission programmes of the various Directorates General, ranging from the technology driven RTD Framework programmes to monitoring adherence to the CAP (Common Agricultural Policy), transport planning (TENS), and scores of others. Many of the hundreds of the latter projects, ranging from studies of soil erosion in Mediterranean lands to cross-border projects on cadastral systems to environmental studies modelling the effects of artificial radionuclides in the food chain, use spatial data in different ways. In general, the methods used by different projects are not at all co-ordinated and it is often incidental if they use common methods, data and standards.

The challenge in Europe is not just to move from NSDI to GSDI, but rather first to move from NSDI to ESDI and then on to GSDI, focusing in the latter case on those issues or types of GI which are needed on a truly global basis. In addition to the hierarchical steps there are also discipline related activities that need to be incorporated into a properly co-o The scope of geographic information in the European context is vast and its applications include issues such as:

Mobility in the workforce

Tourism

International infrastructure, e.g. telecommunications and transport

Banking

Security - police, military, border control (Schengen)

Primary production/Agriculture/subsidies (planning and monitoring)

Business - products such as road navigation, atlases, digital databases

Business - market research and services

Studies of international flows of goods, materials, commodities, capital, waste, etc.

Today no single Institution has a mandate to oversee all these activities, though some European agencies, namely EUROSTAT, the European Environmental Agency, the CEO at Ispra, and EUROGI and its member CERCO are working in parts of the field. Improved coordination through all agencies being members of EUROGI to work jointly on ESDI could have many benefits, particularly for the international growth zones that straddle the borders of one or more countries in Europe (Table 1). All these are areas where technology is being used to provide more efficient information services for the maintenance of employment and living standards in a complex, modern society. In addition to the direct use of spatial information there is also major investment throughout Europe in research and training, and a trend towards Europeanization in scientific disciplines that use spatial data (Geology, Hydrology, Soil Science, Meteorology, Climatology).

3. THE EXTERNAL FORCES AFFECTING GI IN EUROPE

As well as the internal forces driving the need for greater harmonisation in spatial data in Europe, global market forces are forcing Europeans to compete internationally with North America and other leading trading blocks, e.g. Asia and Pacific Rim, and this also includes aspects of geographical information. Most computer hardware and software used in handling spatial data is produced in other parts of the world, namely the United States and the Far East. Most of the readily available digital data sets at world or regional level are also available from the United States, though much of the extension of GI skills to developing countries results from European aid and training.

Europeans are major participants in international science programmes that use spatially referenced data such as the International Geophysical Biophysical Programme (IGBP), the World Health Organisation (WHO), the World Meteorological Organisation (WMO), United Nations Environment Programme (UNEP), and many others. Many international businesses operating globally are European in origin or have strong European connections.

Because of the rapid technological advances in IT hardware and software (ever greater power and capability at steadily decreasing cost), any lead that non-EU trading blocks have in hardware and software is rapidly becoming less important for GI users compared to the European strengths in GI skills and innovation, advances in GI modelling, etc. Where the hardware/software platform is produced becomes far less important every year in relation to the value of the GI itself, of its accessibility to all potential users in both traditional and novel fields of research, commerce and government. Information is the key - GI data is the raw product - modelling, data interoperability and accessibility are of paramount importance for all GI users.

The dependence on external (non-EU) suppliers of GI materials on the one hand, and the reliance of others on European support and skills on the other means that Europeans cannot be inward looking when developing spatial data infrastructures. It is essential to adopt a global strategy, to speak a global language if you are going to participate properly, and I would argue strongly that Europe as a political entity should adopt a leading position in the development of organisations that are capable of enhancing the use of GI at the global level to solve global problems.

4. FACTORS DIFFERENT IN MOST EUROPEAN STATES COMPARED TO THE US.

In spite of many believing that the harmonising of GIS at the European, or even the global level is a natural good, it is important to realise that there are many factors, some rooted in history, some in culture, some in politics, that affect the ease with which technical solutions can be devised and implemented. These involve a wide range of issues from simple inertia to the adoption of different policies on data availability and pricing, questions of copyright and legal issues which differ between European states, not to mention conceptual and discipline-related differences in the actual content of the spatial data that are related to culture and language and history (Burrough and Masser 1997). The pre-1989 transatlantic focus in trade and culture continues, providing Europe and North America with a shared technology base of hardware and software tools for GI. North America is perceived as the largest world source of GI skills (GIS WORLD 1996).

Due to different levels of economic development and training facilities there is still a wide variation in levels of skills and awareness both within and between countries in Europe, partly because of the different status of academic subjects like geography in schools and universities. There are technical differences between countries such as the use of different base levels and coordinate systems, but generally, these are easier to fix than are cultural or political divisions. Surmounting such divisions requires strong political leadership in regard to GI. Many aspects inherent in using GI wisely, effectively and efficiently require a political leadership that is missing today in many EU Member States (where there is still often no national "GI champion"). The situation is similar but complicated at EU level, though certain parts of the European Commission are now looking more closely at ways to help advance GI in Europe more effectively.

5. THE RESISTANCE TO GSDI IN EUROPE

Most resistance to a European or a global approach to SDI stems from a range of relatively local issues rather than any coherent opposition philosophy. Issues causing resistance include:

National mandates for NMAs and other agencies which rely on/use GI, limiting their activities to the national boundaries.

A lack of funds to work outside the national borders

A lack of interest or awareness, jingoism and nimbyism and fear of exploitation

Political issues and subsidiarity

Too little inter-disciplinary contact

Lack of a clear lead from appropriate EU-DGs and other EU organisations

Uncertainty or difficulties with copyrighting and other legal issues.

Given good will, a strong directive from the European Commission, backed by the Council of Ministers and the European Parliament, and the patience of politicians, government officials and professionals, there seems no reason why most of these should not be overcome in time, at least to some degree. No doubt there will always be areas where it is sensible for a nation state to retain secrecy about the activities occurring on certain parts of its soil.

Table 1. Major trans-border regions in Europe.

Region Countries
1The Baltic Sea Countries with land within the Baltic Sea catchment include Denmark, Sweden, Norway, Finland, Russia, Estonia, Latvia, Lithuania, Czech Republic, Slovakia, Ukraine, Belarus, Germany, Poland.
2The Rhine Germany, Switzerland, France, Belgium, Luxembourg, Netherlands
3The Meuse France, Belgium, The Netherlands, Luxembourg
4The Ruhr complex Germany, Netherlands, Belgium
5Mediterranean Coastal strip Italy, France, Spain
6The Alps France, Germany, Austria, Switzerland, Italy, (Slovenia).
7The North Sea United Kingdom, Norway, Denmark, France, Belgium, The Netherlands, Germany, Sweden.
8The Nordkap region Norway, Sweden, Finland, Russia
9The northeast Atlantic Portugal, Spain, France, Ireland, United Kingdom,
10Paris-Brussels region France, Belgium.
11Southern Scandinavia Denmark, Sweden, (Germany, Norway).
12The Pyrenees France, Spain.
13Southern alpine industry France, Switzerland, Italy (Lyon - Geneva - Milan).

6. LESSONS TO BE LEARNED FROM ATTEMPTS AT DISCIPLINE-SPECIFIC GSDI.

Europeans actively participate in many international organizations responsible for collecting and analysing spatial data. Some groups already have their own version of a GSDI: examples are the Petroleum industry, the International Geophysical Biophysical Programme (IGBP), the US Military, the UNEP-GRID programme, the World Meteorology Organization (WMO) and the World Health Organization (WHO). Of these, some collect their own data while others rely on data from national agencies. Experience shows that collecting your own data gives more satisfactory global sharing of good data than depending on others, as in the social sciences where data collection may be constrained by imposed boundaries and inappropriate spatial units (Burrough and Masser 1997).

A brief evaluation of two international programmes, namely the IGBP and the WHO reveals some of the problems of setting up GSDI in the framework of different disciplines.

6.1 The International Geophysical-Biosphere Programme (IGBP) of the International Council of Scientific Unions - (ICSU) has seven established Core Projects dealing with the Biosphere, the Hydrological Cycle, Global Change, Terrestrial Ecosystems, Global Atmospheric Problems, Ocean Fluxes, Land-Ocean Interactions in the Coastal Zone, Past Global Changes and Land Use/Cover Change. All involve the collection, management and interpretation of data, much of it concerning spatial and temporal physical, chemical or biological phenomena, which are used for quantitative models of the short-term and long term processes affecting the Earth's Geosphere and Biosphere, including those essential for human life and activity. Much data are satellite images, but many other standard kinds of digital data such as geographical coordinates, elevation, soil data, hydrological data, ground meteorological data, land use and land cover are used. No single Core Project has specific responsibility for data acquisition, data compatibility and information exchange so IGBP has set up a Data and Information System (IGBP-DIS) and a Task Force on Global Analysis, Interpretation and Modelling (GAIM). IGBP-DIS is responsible for liaison between Core Projects and data producing agencies and international data centres while GAIM's role is to promote the development, application and application of global models of biochemistry and physical climate. The rationale for IGBP-DIS is built on several generic issues of data acquisition and management that are relevant to GSDI, namely

the increasingly integrated nature of global science, requiring efficient exchange of many kinds of data, which in turn require efficient data exchange mechanisms, meta data and quality control

the current lack of, or incompleteness of many important data sets at Regional, continental and world scales.

the huge volumes of data and their complexity create special problems of manipulation and analysis

the need to archive large amounts of data to detect significant global changes

the need for consistent and complementary data management policies that surpass national boundaries and take issues of data pricing, copyright and ownership into account.

The IGBP programmes are set up and run by international teams of scientists whose main aim is to understand physical change at the global level. They use readily available data when possible, but collect their own data in their own ways for many kinds of research. Their emphasis is largely on data collection and modelling leading to improved understanding of the physical environment. Extensions of this work to social problems are in their infancy. GI issues are not regarded as limiting so long as data are readily available and exchangeable. There is an expectation that data paid for by public research should be freely available on international networks and that commercial prices for data should not apply.

>From a purely European perspective, we should note that there are an increasing number of situations where groups of European member states need to work together on environmental problems that straddle their borders. Examples are the environmental management of the Baltic Sea (Langaas 1997), issues of water quality and quantity in the catchment of the River Rhine (Kwadijk and Sprokkereef 1997) and land degradation problems in Mediterranean areas (Cammeraat and Prinsen 1997). From these authors' reports it is clear that only when a study is set up jointly and uniformly over several countries (e.g. land degradation in Mediterranean areas) is the setting up and use of standard methodology and data sharing possible. When building joint databases from existing data in several countries many problems must be overcome ranging from geodesic registration and measurement, to methods of description and laboratory analyses, to issues of data availability, copyright and pricing. In the recent GI-Policy study, international users of environmental data feared that future pricing policies might make their work impossible, though commercial data suppliers and national mapping agencies that are forced to recover costs took an opposite stance (Burrough et al. 1997). Clearly, there is much to do to ease the integration of spatial environmental data in Europe.

6.2 The World Health Organisation (WHO) is concerned with monitoring data on people, not physical processes. They are interested in the spatial and temporal incidence of disease (epidemiology) and the quality of human health in different parts of the world. Though this is of supreme importance, much WHO research is poorly funded. Socioeconomic data are generally less well defined than data collected in the physical sciences. There is also a lack of awareness of spatial variation and its implications for data analyses and conclusions. Data are usually collected with respect to artificially crisp delineations of space such as state boundaries, census districts or postcodes that may have little relevance for the processes of disease incidence and propagation. Many data may be inadequate for many purposes and may be collected differently in different countries, making comparisons difficult.

The WHO notes that:

There is a lack of comprehensive and comparable data needed to monitor environmental conditions and their effects on health.

To achieve comparable environmental health information systems within the Region (i.e. Europe), strategies with clearly defined objectives should be adopted for comprehensive environmental health monitoring, taking into account all routes of exposure. They should be adapted to local needs but, for data to be comparable and to facilitate regional analysis, a harmonised approach must be taken to sampling, analytical and statistical procedures, quality assurance and reporting.

To facilitate the monitoring of changes in health status due to environmental factors, core environmental health indicators should be developed and applied throughout the European Region.

For environmental health information systems to be of practical use in decision- making, better data are needed on exposure-response relationships. This requires further epidemiological research.

Clearly, the WHO has problems with data, data aggregation and standards. They are dependent on national sources and national methods of spatial aggregation and classification. There is a lack of exchange standards and a paucity of metadata. Most data are collated from external sources so there is little control over data quality and resolution.

In contrast to the WHO, there are very successful commercial agencies that make serious profits from geographical data that are collected by national and international agencies outside their control, using spatial units that are not of their choice. Geodemographics (Webber 1997) is the art of combining marketing data from store cards, air miles actions and mail order firms with readily available spatially registered data from censuses, local authorities and post code districts. These commercial agencies provide retailers with spatially precise, useful data on market potential and penetration. The classification systems of socio-economic groups used have been shown to be reasonably homogeneous over many different countries, so in this area some de facto GSDI has been achieved, though possibly not of a kind that most geodesists or cartographers would accept.

7. TECHNOLOGICAL AND THEORETICAL ADVANCES IN SPATIAL DATA ANALYSIS.

Generally speaking, the technological developments in GI for the last 25 years have led to the development of a data rich but theoretically poor resource. Most effort has been spent on producing the digital equivalent of the paper map as a repository of information about objects and their attributes in a flat, Cartesian space. International (mostly US-European) research is ongoing on how to refine this, essentially reductionist paradigm, how to store and retrieve large volumes of data, how to describe it properly (metadata), how to exchange it between different hardware and software systems (interoperability), how to best accommodate the temporal aspect of GI and how to best present GI to users, including in real-time, 3D simulations. Accessibility is growing through the Internet and related technical services for information exchange. These issues are beginning to be addressed by recent publications in the GI field (e.g. Burrough and Frank 1996, Burrough and McDonnell 1997, Chrisman 1996). There is a major need for advanced R&D into spatial/temporal modelling tools which are "GI aware" or "GI enabled" in order to transform GI into useful knowledge about trends and changes of the earth. Setting up a standard GIS is easy, collecting and maintaining data is costly, yet getting something useful out of the total system often depends upon creating models of the real world, based on the GI, which is extremely difficulty to do, or to do well. Failure at the modelling level in certain application areas such as environment and socio-economic studies may negate the entire investment in the GI and GIS.

8. PERCEIVED DANGERS OF ESDI/GSDI.

If GSDI is imposed from without (i.e. outside Europe) Europeans will be presented with a fait accompli that may have serious economic and cultural disadvantages. Although American writings on GSDI pay attention to the social, political and cultural benefits of a GSDI, the view is necessarily from a North American perspective. >From a European point of view there are issues that merit more discussion, including the following:

Dependency the need to rely on organisations in other countries for the provision of data and services, as evidenced by the WHO experience. At a world level there is a concern that in spite of openess and the World Wide Web, US domination of hardware, software and data allows them to act in their own convenience and strategies, as with the accuracy of GPS signals that is controlled by the US military.

Exploitation. So long as GSDI results in improved services through land registration, reliable utilities, and a clean environment, no one will worry. But technology is only an enabler - you can use it for good or bad. In democratic lands there should be sufficient checks and balances - but often there are not. Not all GI-enabled actions will be available to everyone because not everyone has the money or the skills to carry out the tasks of bringing data together and analysing them. Specialists can enhance their economic or strategic positions through the use of GI - we are currently seeing how publicly available spatial data, linked to economic and social data can be used to enhance the activities of marketeers, commodity traders and paramilitary organisations. Experience in recent years has demonstrated how those outside the law can use technology for organised crime.

Technological prowess is also effective at enhancing differences in economic standing and therefore tensions between different political groups. One possible effect of GSDI might be to aggravate the current dependency of the less developed world on the more technically advanced nations. There is also the potential to enhance the differences between individuals having or not having access to the technology.

Lack of continuity and discipline. The hallmark of a modern, responsible state is continuity in its basic organization, including the provision of statutory information for its citizens. The problem with many developing countries is that for one or another reason they do not have this continuity or stability in public organisation that is essential for the proper use of information systems for the public good. An increasing threat to many modern states is the drive away from the provision of services paid for by government for the public good to services contracted out to individuals or small private organisations. There are needs for checks and standards.

There are needs to encourage self-discipline in the use of electronic information, including GI. Police can use GIS to determine the best places to grow cannabis and therefore increase their chances of an arrest, and so can ordinary people, because the information about how cannabis grows is based on basic ecology. The giving of recipes for bombs or providing tips on terrorism etc. on the Internet are examples of a lack of self-discipline that are potentially seriously damaging.

In the context of technological inequality there has been a call (Lummaux 1996) for a international organisation to oversee and facilitate the harmonisation of continental standards, taking account not only of the leading players in North America and Europe but also the developing world. The focus of such activities would not necessarily be on the techniques of GI but on the effective distribution of practical knowledge for resource planning, healthcare and sustainable development.

Dilution of cultural values. The concepts embodied in GIS reflect the thinking of certain classes of technically oriented people in relatively few parts of the world. The power of the ideas is evident from the speed at which these ideas are assimilated (e.g. my own book - Burrough 1986- has been translated into Japanese, Chinese and Thai, and there is an ARC-INFO in every country). Instead of enhancing cultural richness, however, bringing in a system that has been developed largely within the cultural boundaries of an alien mindset often means that the recipients treat the imported system as the only reality and apply it indiscriminately (Hall 1977). The consequent danger is a world in which the view of reality is that of the CAD-CAM computer simulation, with concrete streets and massive tower blocks, and a fast food outlet on every corner. This is a vision of a bland dystopia where nothing is bad, because it has been optimised, but nothing is good, nothing excels - everything is efficient, dull and predictable. A world in which there are no challenges to fire the young. I hope I speak for Europe when I say that such a world is not what we wish to create. But we must be aware that through a multitude of small decisions, each aimed at standardising, optimising and reducing differences between disciplines and cultures the dull, lowest common denominator might be what we end up with. Surely that is not the aim of the GSDI?

Unfortunately, there seems to be no single, strong, properly resourced "voice" with the political mandate to speak for Europe at European level. This situation must be redressed in the immediate future, or we face the prospect of continued disharmonisation in the collection and use of GI across Europe and a reduction of the impact of Europeans in world affairs. Let us now examine some recent developments in GIS interoperability that may have far reaching effects on the ability of Europeans to manage their own GI in a way that is best for them.

9 THE OPEN GIS CONSORTIUM

The Open GIS Foundation (OGF) was set up in the United States in 1992 to address the problems and expense encountered by GI users in government and industry who want to share and distribute spatial data, but for technical or organisation reasons are prevented from doing so by having databases that cannot easily be linked together. This purely technical problem of a lack of interoperability was caused by GI software vendors having protected, largely secret, internal data structures in the belief that market share could best be ensured by locking the customer into the single system that they, for other reasons, may have chosen.

The OGF, now the Open GIS Consortium (OGC), a non-profit making, but commercial organisation, introduced the idea of OGIS - Open Geodata Interoperability Specification (Hecht 1994, Schell 1995) to attack the technical problems of exchanging data between different spatial databases. In a very short time the OGC has persuaded GI software vendors and others that the ability to share and transmit spatial data in ways unhindered by proprietary data standards will provide a better commercial future for the GI industry than the previous closed approach. The goal of the OGC's Open GIS Project is to provide a comprehensive suite of open interface specifications that enable software developers to write interoperating components that provide transparent access to geodata and geoprocessing resources over networks (McKee and Kuhn 1997). The arguments for this approach have nothing to do with national or international ideals, except in so far that in the United States, the Federal Government takes the line that data collected by the State should be freely available to its citizens, and they would like to do that with as few extra costs as possible. Instead, as proclaimed in the recent Open GIS advertising supplement in the September 1997 issues of GIS World and GIS Europe, the main reasons for GI interoperability (and hence for GSDI) are purely commercial.

There's only one way to make sure we're on the true path to efficient solutions that generate the greatest wealth: put commercial interests in charge of developing the necessary technology, and let them adopt corresponding standards through a consensus process.

Cliff Kottman, Vice President, Technology Development, Open GIS Consortium, Inc.

The degree to which the OGC are persuading GI organisations around the world of the validity of their message can be seen from the impressive current list of 103 corporations, government institutes and academic institutions that have become members of the OGC. More interesting than the list is the breakdown in categories:

It should be noted that while US Government member organisations may be few in number, those involved are extremely important (eg USGS), but the 'European Government' organisations are actually two national bodies, namely RAVI from the Netherlands and the Finnish Geodetic Institute: no truly European organisations (eg the Centre for Earth Observation, the European Environment Agency nor Eurostat) are yet members. This may have something to do with the fees, and the fact that OGC is an American, and not a European business venture. The same reason, but inversely so because university membership dues are low, may explain the large number of European University members. European Companies include Laserscan, SmallWorld and Siemens Nixdorf.

Even more interesting is to note those North American Business members of OGC that are extremely active in Europe:

These companies supply most of the GI software, hardware and database technology used in Europe. They employ thousands of Europeans and clearly have a large stake in what happens here.

The possible consequences of OGC for European Spatial Data Initiatives.

Before OGIS, the path to interoperability in ESDI was seen to be organised in a top-down fashion. National GI organisations as members of EUROGI are indirectly in touch with the EU Directorates General and the European Commission. Therefore, once agreement on ESDI has been reached at the top political level, the decisions on standards, mandates and agreements would be passed down through EUROGI to the national GI organisations who would then disseminate the message to their governmental and commercial members. This conventional model does not include the European and other GI software vendors, in spite of efforts of EUROGI some years ago to obtain unanimity and support from European vendors. Then little concrete support was forthcoming, in spite of sympathetic reaction to suggestions about interoperability being sound sense.

The OGC model is essentially 'bottom up', and based on commercial interests. Because Europeans buy the products of the largely US GI vendors, any interoperability standards and procedures agreed by US commerce will very likely become the modus vivendi for Europeans, whatever the EU or national governments may wish. This means that the power of the international market place will have a very strong effect on the aims and success of the initiatives listed in the many versions of the GI-2000 document. This course of events suggests to me therefore, that interoperability is not going to be an area where European Union funding will be provided for ESDI - if US business is going to provide a workable solution for little investment on Europe's part (even if the solution is not what Europe would really want), then why invest large amounts of money? Surely, one could argue, we will get something we can use without having to be proactive.

The great danger with sitting back and waiting for someone else's solution to your problem is that their cure might be worse than your ailment. Therefore, while realising that a European alternative to US-lead interoperability is unrealistic, it is absolutely essential that Europeans are involved in all aspects of the interoperability process, from NSDI to ESDI to GSDI. Referring back to Sections 7 and 8 above, we see that interoperability is not just a question of technical data exchange, but the development of a common language, including semantics and grammar. Simple technical solutions may interfere with, or impose certain cultural values and customs. GI is a data-rich but theoretically poor resource - new systems will have to be more intelligent and new paradigms for spatial data processing may require new approaches to interoperability. EUROGI is currently taking part in OGC discussions on behalf of its members, but much more needs to be done. In particular, EUROGI needs much more active support from its member organisations and their government and commercial members and sponsors to ensure that Europe gets the solutions Europe needs.

To conclude:

10. CONCLUSIONS: WHAT CAN WE USEFULLY ACHIEVE WITH GSDI IN EUROPE?

This paper has demonstrated that there are important world developments operating towards harmonization and unity in geographic information., and that many European institutions and individuals are playing key roles in the international discussions and moves towards implementation. The perceived benefits for Europe are seen in financial, political, socio-economic, commercial and technical terms. The perceived dangers have more to do with Europe being forced to operate as a client rather than as a partner in GSDI and these dangers could be most readily circumvented by Europe having a single organisation that had the mandate, the competence and the trust to operate on behalf of the large number of interested parties who at present all too often must operate individually. I believe that it would be of great benefit to us all if the EU would provide that mandate and the funds to back it up.

In the absence of that single, mandated voice, let us not attempt to achieve too much too quickly, but go for practical, useful actions that will demonstrably remove inefficiencies and improve value and quality in the provision and use of GI.

Let us set up a dialogue in a common language to explore the benefits and disbenefits of sharing spatial data among countries and commercial organisations.

Let us consider the GSDI as a language of communication, with a foundation in the culture of all the world's peoples, not just those who happen to live in particular areas.

Let us jointly lobby the EU Institutions, i.e. the European Parliament, the European Commission and the Council of Ministers, to take the necessary lead in the legal and political issues involved and to encourage them to stimulate the proper development of a successful ESDI/GSDI that is in the interest of all people in Europe.

Let us ensure that our academic, research and training institutions can produce GI specialists, GI products and GI users of world class, through proper funding and a coordinated focus on the needs of the GI community.

Let us play an important, leading role in the GSDI process - its definition, organisation and implementation - because only by being a member of the club can we seriously have an impact on decisions.

Finally, let us take steps to mitigate against the serious misuse of GI for fraud, for taking unfair advantage or for the invasion of personal privacy.

Acknowledgements: I wish to thank Roger Longhorn, Martin Littlejohn, Christian Chenez, Michael Brand and Ian Masser for their constructive comments to earlier drafts of this paper. The opinions expressed are my own, however.

10. REFERENCES.

Andriessen, Frans H.J.J. (1996), Speech delivered on the occasion of the Opening of the 28th International Geographic Union Congress, The Hague, 4 August, 1996.

Burrough, P.A. (1986), Principles of Geographical Information Systems for Land Resources Assessment, Oxford University Press.

Burrough, P.A. (1996), A European View on the Global Spatial Data Infrastructure. Paper No. 6 in: Proceedings of a Conference on the Emerging Global spatial Data Infrastructure, organised under the Patronage of Dr Martin Bangemann, Member of the European Commission, Bonn, 4-6 September 1996.

Burrough, P.A. and Frank, A.U., (1996), Geographic Objects with Indeterminate Boundaries. Taylor and Francis, London.

Burrough, P.A. and McDonnell, R.A. (1997), Principles of Geographical Information Systems. Oxford University Press, Oxford (in press).

Burrough, P.A. and Masser, I (1997), European Geographic Information Infrastructures: Opportunities and Pitfalls. Taylor and Francis, London (in press).

Burrough, P.A., Craglia, M., Masser, I., and Rhind, D. 1997. Decision-makers perspectives on European Geographic Information Policy Issues. Transactions in GIS (in press).

Cameraat, E. and Prinsen, H., (1997), The Medalus Georeferenced Database: an application used for Land Degradation research in the European Mediterranean. In: Burrough, P.A. and Masser, I. (eds) ), European Geographic Information Infrastructures: Opportunities and Pitfalls. Taylor and Francis, London (in press).

Chenez, C.C. 1996. Proceedings of a Conference on the Emerging Global spatial Data Infrastructure, organised under the Patronage of Dr Martin Bangemann, Member of the European Commission, Bonn, 4-6 September 1996.

Chrisman, N. (1996), Exploring Geographic Information Systems, Wiley, New York.

David, B., van der Herrewegen, M., and Salgé, F., (1996), Conceptual Models for the Quality of Geographic Information. In: Burrough, P.A. and Frank, A.U. (eds) Geographic Objects with Indeterminate Boundaries. Taylor and Francis, London, 193-206.

DGXIIIe, (1997a) GI-2000: Towards a European Policy Framework for Geographic Information. DGXIIIe, Luxembourg..

DGXIIIe, (1997b) World Wide Web sites; http://www2.echo.lu/gi/en/gi-meta.html

Hall, E.T., (1977). Beyond Culture. Anchor Books, Doubleday, New York.

Hecht, L.G. (1994). Open GIS Foundation fosters industry synchronization. GIS World June 1994.

IGBP (1994). Global Change Report No. 30. IGBP Global Modelling and Data Activities 1994- 1998. International Council of Scientific Unions (ICSU), Stockholm 1994.

Kwadijk, J. and Sprokkereef, E., (1997), Development of a GIS for Hydrological Modelling of the River Rhine, in: Burrough, P.A. and Masser, I. (eds) ), European Geographic Information Infrastructures: Opportunities and Pitfalls. Taylor and Francis, London (in press).

Langaas, S., (1997), Transboundary European GIS Databases: a review of the Baltic Sea Region Experiences, In: Burrough, P.A. and Masser, I. (eds) ), European Geographic Information Infrastructures: Opportunities and Pitfalls. Taylor and Francis, London (in press).

Lummaux, J-C., 1996. Une autre vision du GSDI. AFIGEO paper, Paris.

McKee, L. and Kuhn, W. (1997). The OpenGIS Consortium's Purpose and Technical Approach. In: Fritsch/Hobbie (eds). Proceedings 46th Photogrammetric Week, Stuttgart, September 1997, pp 237-242.

Salgé, F., (1997), From an understanding of the European Activity to the reality of a European Dataset, In: Burrough, P.A. and Masser, I. (eds) ), European Geographic Information Infrastructures: Opportunities and Pitfalls. Taylor and Francis, London (in press).

Schell, D. (1995). Open GIS. Editorial. Geo Info Systems, May 1995.

Webber, R. (1997), Developments in Cross Border Standards for Geodemographic Segmentation. In: Burrough, P.A. and Masser, I. (eds) ), European Geographic Information Infrastructures: Opportunities and Pitfalls. Taylor and Francis, London (in press).

WHO (1995). Concern for Europe's Tomorrow: Health and the Environment in the WHO European Region. WHO European Centre for Environment and Health, Wissenschaftliche Verlagesellschaft mbH Stuttgart 1995, Section 19.6, pp 522-523.

______________________________________________________________________________

Peter A. Burrough

Netherlands Centre for Geoecology, Faculty of Geographical Sciences, Utrecht University,

Post box 80115, 3508 TC, Utrecht, the Netherlands.

Tel: ++31 302532766/2749/2044 email: P.Burrough@frw.ruu.nl