ANNEX C

Introduction

On the occasion of this third international conference on the Global Spatial Data Infrastructure (GSDI) , it is appropriate that we review some strategic SDI (Spatial Data Infrastructure) initiatives occurring across the globe, to better understand how the GSDI can be formulated, funded and implemented. At the same time, it is important that we do not continually "reinvent the wheel" with each new GSDI conference. We should keep firmly in mind the "Findings and Resolutions" [1] of the 2nd GSDI Conference held in the USA in October 1997.

Among the key points contained in the " Findings " were:

• the GSDI " is critical to the attainment of substantial and sustainable development in both the developed and developing countries of the world ",
• the GSDI is of " vital importance " in implementing global environmental conventions and should be recognized as such at the 2001 meeting of the UN Commission on Sustainable Development,
• support for GSDI must come from the " highest levels of business, government and academia " at " local, national, regional and international levels ", including the UN, World Bank and G7 countries,
• many international organizations already work in areas related to the GSDI, at global level, and they should all be encouraged to participate in and support the GSDI's goals and activities, and
• education (about the importance and best use of spatial data) and research activities (which will enhance access to and use of GI) must be strongly supported.

Among the " Resolutions " pertinent to this paper were:

• formation of an initial organizational nucleus to create the administrative infrastructure for the GSDI initiative,
• encourage formation and participation of regional GI infrastructure initiatives,
• recognize the importance of standards, as tools for international cooperation, and support current initiatives in regard to internationally recognized standards for GI/GIS and ICT generally,
• explore ways to harmonize data definitions " without losing their primary relevance or compromising the political and legal diversity amongst nations " and ways to turn " local, national and regional data sets ... into international ones " while taking into account the needs of " different sectors of the GI community for different applications ", supporting capacity building and data collection in emerging nations, fostering the collection and use of metadata to agreed standards, finding ways to improve access to data (including clearing houses), examining ways and cost/benefit for converting legacy data and encouraging provision of services,
• acknowledge the importance of applied research and best practice (government and private sector) and the value of demonstrators and pilot projects.

This is quite a comprehensive list of issues and goals! Examining the characteristics of the more advanced national and regional SDIs may provide valuable lessons for determining the form and format for a GSDI.

The participants of GSDI-2 in Chapel Hill, NC, USA, 1997, generally agreed the following definition for the GSDI:

This definition, then, sets out what we want a GSDI to be or encompass. Michael Brand proposes a main goal for GSDI as:

If we accept this simply stated and quite sensible goal, then whatever specific elements of a GSDI are actually put in place must be able to support attainment of this goal, in a global context.

In prior GSDI discussions, we have accepted that the GSDI will build upon regional SDIs, which themselves will comprise interconnections of national SDIs. Yet few national SDIs are anywhere near completion and no regional SDI is much beyond the initial planning stages. The question is - does the GSDI need to mirror the structure or format of a national or regional SDI - and must it wait until these SDIs are already in place? When we reach global level, with the number of actors inherent at that scale, across scores of nations, sectors and disciplines, with a multitude of linguistic, cultural, economic and legal backgrounds - is it not perhaps worth considering that the implementation format for a GSDI may need to be quite different from that of national and regional SDIs?

Could a workable GSDI be put in place along simpler lines and still be able to achieve its main goal(s), while fulfilling the needs expressed in the " Findings and Resolutions "? What level of "interconnectivity" between all actors in the GI field (data owners, users, service providers), across all disciplines, nations and regions, will be required - and what can be achieved, in practice (and in fact!)? To attempt to achieve all the implied goals in the accepted GSDI definition above would require a global coordinating body more complex than any existing today, including the main agencies now existing within the UN, each of which focuses on only one major area of global interest, whether that be trade, development, health, agriculture, statistics (and cartography), etc.

Nothing succeeds like success. Failure to reach a minimum set of practically achievable objectives within a reasonable time frame (maximum of one decade?) may even harm the long-term prospects for achieving greater global cohesion at economic, political and cultural levels through wiser use of quality, readily available GI. People - and more importantly, politicians! - will lose interest. While accepting that GSDI must move forward on a range of diverse fronts, simultaneously - political, technological, legal, economic and scientific - perhaps it is wise at this stage to set limited, achievable, "doable" goals in each of these areas.

In a paper of this length, it is simply not possible to review, let alone analyze, the wide range of SDI initiatives underway at national and regional level. Therefore, the comments below are based on published reports and personal experiences of the author relating to three SDI initiatives, with a view to extracting any practical lessons to keep in mind when discussing the GSDI. At the outset, it is safe to say that no single national or regional SDI initiative has yet achieved its stated goals in all areas which it purports to cover, i.e. legal aspects, commercialization, accessibility to data, protecting rights of data owners and users (or in some cases, even fully defining these rights, legally!), best use of existing data, and more.

The US NSDI began with a very strong political mandate based on President Bill Clinton's Executive Order in April 1994. Such outright support from a head of state has not been achieved in any other national or regional SDI initiative. President Clinton's initial lead has been strongly reinforced by further activities of Vice-President Al Gore in the important areas of raising awareness at public level and developing visionary (and eye-catching!) projects, e.g. the "Digital Earth". Yet all these activities have been aimed mainly at federal level spatial data.

According to reports from many corners, discussions at conferences and private communications, the FGDC and NSDI is not having the expected impact at local government level or with legacy data (which is not affected by the Executive Order). In April 1997, the FGDC Steering Committee met with representatives of 18 US states, the National States Geographic Council and the National Association of Counties to find ways to better involve non-federal organizations in the NSDI [3] . In that same month, the FGDC stated its "vision" for the NSDI as " Current and accurate geospatial data will be readily available to contribute locally, nationally, and globally to economic growth, environmental quality and stability, and social progress ". [4]

As well as the FGDC itself, major players in the US NSDI include the Mapping Sciences Committee (MSC) of the National Research Council and the National Imagery and Mapping Agency (NIMA, formerly the Defense Mapping Agency). NIMA, as the key US mapping agency, sets the agenda for geospatial data in the US via its Geospatial Information Infrastructure workplan 1997. Interestingly, this plan seems to endorse the very practical concept of "build a little at a time, from the bottom up, but in a coordinated manner", rather than wait for some all encompassing infrastructure to be defined, mandated and implemented.

A recent report of the National Academy of Public Administration (NAPA) [5] provides an important update on the US NSDI initiative. This report, commissioned by the Bureau of Land Management, USGS, Forest Service and National Ocean Service, published in January 1998, examined "trends over the larger spatial arena" during a study which began in October 1996. The principal recommendation is that more legislation, specific to the NSDI, is needed, specifically " Draft a new statute in cooperation with state and local governments and other organizations to create an NSDI , establish a National Spatial Data Council , and better define federal agency roles and responsibilities for NSDI so as to meet the participating organizations' programmatic needs. " (underscores added by this author) This does not sound to me as though the US NSDI is nearing completion, rather it is only just beginning.

1. Complicated metadata requirements are leading to slow uptake of the FGDC standards. Costs of implementing data and/or metadata clearing houses are collectively quite large and often prevent participation of local government agencies. There is no requirement for private GI data providers to participate in the standardization initiatives.
2. The US adopted a "top down" approach whereby FGDC set standards and Presidential decree required compliance of federal agencies. But where was (and is) the grass roots support? How many local governments and private organizations have adopted metadata standards and/or are creating clearing houses? NIMA, which must respond to defense needs, proposes a "bottom up" approach, building on existing elements, but always striving for future harmonization and reduced cost. This is a "do-able today" scenario for the GSDI.
3. The research agenda is important. The University Consortium for Geographic Information Science (UCGIS) and the Mapping Science Committee of NRC draw up national priorities for spatial research. Government support, via the National Science Foundation, thus has a coordination mechanism by which it can be kept informed of the needs of the research community.
4. The main lesson? Its going to take a long time! The US is an advanced, wealthy, technologically oriented society with strong traditions in freedom of information (when collected via public funding), the value and importance of research, commercialization, free market attitudes and the bulk of the GIS vendors of the world. The 1998 NAPA report [5] indicates that the US is still far from achieving its NSDI goals. On past performance, it may be a decade (or more) before the true impact of the 1994 Executive Order is felt nationally.

In April 1987, a UK Parliamentary Committee of Enquiry into The Handling of Geographic Information, chaired by Lord Chorley, published its report [6] (now often called the "Chorley Report") which offered a wide range of recommendations for increased efficiency in collecting, handling and disseminating GI throughout government and beyond government. This enquiry predates President Clinton's Executive Order by 7 years. The report made numerous recommendations, including creation of a national, non-governmental GI association or council. In 1988, the AGI was born, which today boasts over a 1000 members - more than any similar organization in Europe.

Yet serious (public) discussion for what would become the UK's National Geospatial Data Framework (NGDF) began only 9 years later (mid-1996) and the first actual "product" of that initiative only appeared in June 1998 (the NGDF "Discovery Metadata Guidelines" [7] setting out metadata guidelines, which adhere to ISO's draft GI metadata standards 15046-15). The NGDF provides a forum for discussion and coordinated movement towards creating a range of national level spatial products, e.g. national land information registries, national level gazetteers, etc. The UK Standard Geographic Base (UKSGB) [8] provides a standard infrastructure (supported by the NGDF), a published set of commonly used core spatial units and an Internet service describing these units and their associated geographic databases. To date, only administrative and postal units and data are available via the UKSGB. The NGDF provides a formalized way to ensure that all members of the GI community can more fully participate in the development of the UK's "SDI".

The UK has one of the most advanced digital mapping capabilities in Europe, with all Ordnance Survey maps, at all scales, available digitally - for a price. The UK also has strong capability in regard to remote sensing and participates in numerous regional and global environmental monitoring, geophysical, hydrographic, marine research, coastal management and related programmes, all requiring significant amounts of spatial data. The 450+ local government authorities in the UK are represented by a single Local Government Management Board (LGMB) in regard to common problems and needs, including informatics and especially spatial data. The UK has had a deregulated telecommunications structure for some years (leading to some of the lowest telecommunications costs in Europe) and penetration of IT in business, government and the home is exceptionally high, including access to the Internet. All major US GIS vendors have offices in the UK. Yet even this highly advanced, wealthy, spatially aware nation is having problems coming to grips with the many diverse issues surrounding wider and more efficient use of spatial data at national level.

Key data owners at local and national government level still appear to be very much in the driving seat, enjoying near monopolistic control of basic spatial data and taking an ever stronger "market" approach to sales and distribution of spatial data. Many user groups have great difficulty in understanding and accepting the current trend in the UK for "cost recovery" from government departments, which is strongly enforced at national level and is now becoming more apparent at local government level as well. As the value of information, and the cost of its collection, processing and updating, becomes more widely known, recognized and accepted throughout government, we can expect a future trend towards a more commercial approach in spatial data dissemination, even in government. A recently signed National Interest in Mapping Service Level Agreement (NIMSA) between Ordnance Survey (OS) and the UK government (DETR) sets goals and performance targets to be met by OS in exchange for additional funding (£60 million over three years) to help cover the cost of improving its services generally, and specifically for survey costs in geographical areas for which cost recovery principles do not work because of low demand for the data.

To quote from the conclusions in a recent report [9] issued on the 10th anniversary of the 'Chorley Report' - " The pricing of data is still the major problem for the uptake of GI in the UK. ... Access to geographic information is still one of the greatest barriers to the uptake of GI technology by many organizations in the UK, Europe and beyond. ... (there is) an implied need for standards and interoperability. ... Copyright remains a major problem ... (there are significant) real needs for awareness, education and training. "

1. As in the case of the USA, defining and implementing an all encompassing SDI at national level takes time, money, commitment and strong political will power, regardless of how advanced technologically and "spatially" the country is and how forward looking its SDI proponents may be. If the UK (and US) have taken so long to achieve their first steps towards NSDIs, let us not be too hasty or impatient in regard to development of regional and (especially) global SDI.
2. The UK's "cost recovery" model provides a well-developed counter-point to the USA "Freedom of Information" model for distributing spatial data in the future. How these two models would operate at global level is yet to be seen, although much remote sensing data which will provide the main input to future global projects is now coming from commercial remote sensing platforms, from which data is sold at market prices.
3. Practical methods or models must be found to balance government "cost recovery" or commercial profit oriented goals with the needs of community and citizen, especially where access to relevant and appropriate spatial data is mandatory but simply not economical at full "cost recovery" or market rates.

The first main pan-European association concerning GI/spatial data was CERCO [10] - Comité Européen des Responsables de la Cartographie Officielle (Association of Europe's National Mapping Agencies), founded in 1979 as a "club" for heads of national mapping agencies to meet regularly and discuss matters of common interest. CERCO now has more than 30 members and thus represents far more than just the 15 European Union (EU) Member States. However, as its name implies, CERCO represents mainly the interests of mapping agencies and is not mandated to produce a regional SDI which would encompass the interests of numerous non-cartographic spatial data suppliers.

CERCO does, however, have a significant impact on standards development and access to cartographic data across borders, via its commercial marketing division MEGRIN - the Multipurpose European Ground Related Information Network) [11] created in June 1993. MEGRIN has been instrumental in gaining first-hand experience of processing and marketing national cartographic products across European borders, as well as participating in standardization and interoperability initiatives at European and global level. MEGRIN created one of the few truly regional, cross-border spatial data products in Europe - SABE, the Seamless Administrative Boundaries for Europe geographic database, on behalf of the EC's Eurostat (Statistical Office of the European Community). This was valuable experience for everyone concerned, not least the European Commission, as significant resources were required to create this product, which is only available at what many consider to be "high" cost.

Following the US NSDI initiative in April 1994, several EU Member State ministers, at the urging of the European GI community, wrote to the European Commission requesting that steps be taken to investigate and initiate a European "geographic information infrastructure" (EGII). Some time previously, a pan-European umbrella association for GI was already being discussed among key players in the European GI community and with the Commission's DG XIII (Information Market Directorate) in Luxembourg. From these initial talks, with full support of the Commission, including initial "seed" funding, was born EUROGI- the European Umbrella Organisation for GI [ 12 ]. EUROGI is "an association of associations", both national and pan-European (CERCO is a member). It has been instrumental in helping create national GI associations in those EU and CEEC countries where none existed previously. Today it has 17 national association members and one pan-European association member. EUROGI has been involved in the GSDI initiative from the outset and strongly supports its goals and objectives. Unfortunately, EUROGI suffers from limited funding and therefore has few resources, widely spread, to cover its pan-European remit, let alone assume new responsibilities with regard to GSDI implementation.

However, neither EUROGI nor CERCO nor any other pan-European GI-related organization has a mandate to create a "European GI Infrastructure" (or ESDI). Under the EU institutional structure, only the European Commission (the Executive Branch of "European" government) can propose "legislation" (via Directives), which are then agreed (or not!) by the European Parliament and one of various Councils of Ministers (each Council comprises the key Ministers for that discipline from each EU Member State) [ 13 ]. EC Directives which require legislation must then be enacted into law across the EU, by all Member States at national level, usually within three years (although there are often exceptions). Thus, it fell to the EC to act upon the Member States' Ministers requests for an EGII and the task of investigating and proposing the structure for this EGII was assigned to DG XIII under their Information Market Directorate based in Luxembourg.

Early in 1995 a wide pan-European consultation process began, to define the EGII, which later resulted in a draft Communication to European Parliament and Council of Ministers (the Council of Industry Ministers, in this case), which has taken more than 3 years to reach the stage where it will shortly be "adopted" by the Commission, then forwarded (early 1999) to Parliament and Council. The rationale and proposals for action to create the EGII are embodied in this Communication "GI2000: Towards a European Policy Framework for Geographic Information" [ 14 ]. The main recommendation of the Communication is creation of a "GI2000 High Level Working Party" (HLWP), comprising senior representatives of Member State governments (from those ministries responsible for spatial data at national level), heads of major pan-European GI-related organizations, users associations and other members, yet to be defined. The HLWP will be chaired by a senior Commission official.

The main recommendations for action are the identification, collection and wide dissemination of pan-European base data (topographic), strengthening the emerging national and pan-European metadata services and directories, removing such barriers to wider access to GI as have already been identified during the three-year consultation process, encouraging market growth for spatial data and the application of standards and interoperability specifications.

The GI-related research agenda in Europe was set partly by the GISDATA - Geographic Information Systems Data Integration and Data Base Design programme of the European Science Foundation's (ESF) Social Science Programme, from 1992 - 1997. When GISDATA funding ended in 1997, an independent organization was created by the GI/GIS research community itself, called AGILE - the Association for Geographic Information Laboratories in Europe. AGILE held its inaugural symposium in April 1998 and plans to hold conferences annually hereafter. While a GI/GIS research agenda is being prepared, it has yet to be published. However, like EUROGI, AGILE too suffers from far too little funding, certainly compared to the funds previously available under the ESF programme. Such lack of funding appears to be endemic in the GI community across Europe.

1. While one could argue that certain EU countries (e.g. UK, France, Finland) were actually focusing on NSDI needs even before the USA, this is certainly not true at European regional level, where the initial spark to do something was very definitely the April 1994 Executive Order for a US NSDI. Since late 1994, the EC has taken four years simply to draft and present a Communication to EU governments and the EU GI community - and the main recommendation of that Communication is to create a High Level Working Party. Do not expect anything to move more quickly at global level!
2. Lack of funding at regional level (for cross-border initiatives) seems to be a perennial problem, solved normally only by commercial intervention or under regional structural programmes, including the R&D Framework Programmes of the EU. This makes long-term planning difficult or impossible, can lead to duplicated work or lack of work in key areas of technology or science.
3. The EU has yet to secure the full and whole-hearted support of one high EU official, whether at Commission, Parliament or Council level (even though Commissioner Bangemann was a supporter of the GSDI initiative from its inception). Can we expect there to be a single "champion" for the GSDI? This author thinks not and, unless that champion held the post of "President of the Wold", it makes little difference whether he/she surfaces or not. GSDI will by its nature and scope be a "committee" animal. This is one way that it differs from national and even regional SDIs, where appearance of a "champion" can help tremendously. (Perhaps we need a "Committee of GI Champions" comprising the leading proponents of spatial information from each region and/or nation where SDIs are well underway?)
4. Is the "umbrella" organization the best model for implementing regional (and global) SDIs? This is yet to be proven, as all regional initiatives are still too new to offer proof of effectiveness of any model they may have adopted.
5. Determining who holds an official "mandate" for SDI actions, including the authority to fund (or request funding for) mutually agreed activities, is exceptionally difficult. While the EC is the main instigator of the EGII (Euro SDI), the officials in charge of this initiative show great reluctance to offer any separate funding for the work needed to better define, create and maintain the EGII resulting from the recommendations of the HLWP. The most we can expect are further small sums of "facilitating" funds, similar to those originally expended in creating EUROGI. Access to regional funds via the RTD Framework Programme or other EC programmes will be in strict competition with numerous competing subject areas and themes.
6. Under the "subsidiarity" principle of the EU, European level action (by EU institutions) is only to be initiated (and paid for!) in cases where Member States cannot undertake the required activities themselves. It is easy to see that, if one Member State operates a virtual "freedom of information" regime while another wants "full cost recovery" in regard to its spatial data, this can lead to tremendous impasses - as has already been seen in regard to certain pan-European level GI projects. Similarly, a grossly uneven playing field regarding preparedness to participate in a regional SDI can slow down the entire process for all participants.

This paper reviewed the SDI initiatives of only two nations, the USA and UK, and one region, Europe, which are all three best known to the author (an American, resident in UK and Luxembourg, working as a GI advisor to the European Commission). Yet a common thread runs through all three presentations regarding lessons we might learn in regard to our efforts to define and implement a global SDI. These can best be summarized as:

1. Everything we do in regard to GSDI is going to take far longer than any one of us expects today - but that is no reason not to proceed!
2. Getting real and practical high-level government support is imperative - but difficult enough at national level, exceptionally difficult at regional level and perhaps not even sensible at global level (at least, no single "spatial data champion" will emerge or would have the power to greatly assist us in achieving the GSDI at global level).
3. Funding will seldom, if ever, come from a single, easily identifiable source (unless someone wants to talk to Mr. Bill Gates, who is already quite interested in GI - look at TerraServer and Microsoft's investment in MapInfo and other GI-based information products!). Rather, GSDI will emerge from scores of national SDI initiatives, already being "coordinated and harmonized" at regional level (in some cases, or about to be). Thus, GSDI implementation teams will be in the unenviable position of (perhaps) being "coordinators of coordinators".
4. We cannot leave everything to "market forces", nor forget that much research and development is still needed in regard to collecting, processing, using and disseminating spatial data. Most of this research will take place at national level, with some multi-national projects being funded by regional programmes, such as the EC's Framework RTD Programmes.

It is not the purpose of this paper to probe the detailed organization structure needed to implement the GSDI. It is worth saying, however, that we should separate in our minds the two concepts of a "structure for the GSDI" and "a structure for defining and implementing the GSDI". The former may not exist, in practice, for many years to come, yet the latter must exist if we are to move forward. The distinction between the two terms is not always made plain, even in English language documents - once again highlighting the supreme importance of developing an agreed, multi-lingual glossary of terms as a priority project.

The comments below relate only to ancillary working groups that may be needed:

• Because "GI is everywhere", it is not possible for any one group to delve in any detail into the myriad problems faced by multiple disciplines. Therefore, we need to create multiple discipline-oriented working groups, from which leaders convene and report regularly (annually?) to share results of studies and/or pilot projects and publish the identified needs of their users, real experiences and barriers. These "annual statements of progress" could then be discussed at a higher (political) level "round table" where potential solutions can be proposed and examined as to feasibility and cost/benefit.

• Create a "Research & Development" or "Science" Working Group, as well as the proposed Technical Working Group. "Technology" and "science" are not synonymous, although they may be closely linked, and the participants and goals will be quite different between two such groups.

• Create a "Legal Affairs" Working Group to monitor the existing and evolving regulatory regimes under which spatial data will be collected, traded and used. Much of the work already underway in regard to EDI (electronic data interchange), e-commerce (electronic commerce), protecting copyright/IPR (intellectual property rights), "freedom of information", "fair use" of protected information, misuse of GI (to invade personal privacy) and similar issues needs to be monitored across nations and regions and reported on regularly.

The format and/or structure of a "management committee" is discussed in other contributions to GSDI-3. Conference participants are invited to remember that the current trend in the USA, Europe (and Canada and Australia?) is to devolve responsibility away from central government to local government (called "subsidiarity" in the European context). Any overall management or coordination structure proposed for the GSDI must take this trend into consideration.

In reviewing the history, development and partial results achieved to date by some major national SDI initiatives and the European SDI formulation process, it is apparent that very long lead times are needed to accomplish anything significant. In an endeavor such as the GSDI, where high-level political backing will be needed at national, regional and global level, there is always the fear that lack of demonstrable progress will lead to loss of interest - and more importantly, to loss of much needed funding, however it is offered to those trying to implement GSDI.

Therefore, we should focus on a few achievable goals for which practical activities can be identified, funded and completed successfully in as few years as possible.

Roger Longhorn is an independent consultant, working since 1992 as an expert in Geographic Information policy issues for the European Commission's DG XIII/E in Luxembourg on the IMPACT and INFO2000 information market programmes, and as consultant and technical author on the draft Communication "GI2000: Towards a European Policy Framework for GI". He has BSc and MSc degrees in engineering and management science from MIT, Cambridge, MA, USA, and has worked in the ICT area for DG III, DG XIII, DG I and UNCTAD since 1989. Prior to that he managed a software development company operating on a global basis for the maritime transport industry. Roger is a member of the UK's Association for Geographic Information, the British Association for IT in Agriculture and the Irish Society for IT in Agriculture, for which he publishes periodic news on GI and GIS relevant to the agriculture industry. He is a Fellow of the British Interplanetary Society (UK), where his main interest is use of space technology in remote sensing. Roger can be contacted at e-mail: ral@alum.mit.edu .