Implementation Approach

Many success stories can be reported that are encouraging to those just starting out on SDI development. However, it may be equally helpful to know that they are not alone in encountering difficulties. It may take some time until efforts bear fruit and different strategies and approaches may need to be considered to get people on board (see Example 5).

Delays in success: As the GI community in South Africa has frequently requested, the technology for capturing and publishing metadata has been put in place by the National Spatial Information Framework (NSIF) directorate in charge of implementing the national SDI. For the users, there are no costs associated with this clearinghouse (the Spatial Data Discovery Facility). However, despite the best efforts by the NSIF, the fact that the clearinghouse is available does not seem to be in people's heads yet and they still come out with statements like "what we really need is ...". Moreover, people do not contribute metadata to be included in the system.

Yet this lack of awareness and participation is likely to be temporary. In a recent survey of the South African GI community, about 70% of the participating organisations considered the clearinghouse provided by the NSIF a very important facility but only a small percentage indicated that they possessed the necessary metadata skills (Wehn de Montalvo 1999). Once these skills are in place, the use of, and contributions of metadata to, the Spatial Data Discovery Facility are likely to increase.

While there is no prescriptive recipe for building a SDI, the following aspects have emerged as 'lessons learned' from the international arena of SDI developments. They may need to be adapted to the specific political system and social context within which a SDI is being developed.

• Build a consensus process: build on common interests and create a common vision
• Clarify the scope and status of the SDI
• Exchange best practices locally, regionally and globally
• Consider the role of management in capacity development
• Consider funding and donor involvement
• Establish broad and pervasive partnerships across private and public sectors
• Develop clearinghouses and use open international standards for data and technology

Even in contexts where the community of technicians involved in GIS development is small enough to allow all the members to know each other, there often is no apparent willingness at the institutional level to co-ordinate and harmonise the development of the systems. The development of a SDI will require cultural and organisational changes so as to manage the whole shifting process. This entails mobilising resources so that people in different organisations can adjust.

Creating a common vision: The Australian experience in establishing a national SDI shows that getting people on board has been a long process and has been driven by ANZLIC in terms of awareness raising and making the major components of the ASDI more tangible. Informal collaboration is fairly smooth. As the number of Australia's spatial data stakeholders is quite small, most people know one another, so ideas and knowledge get exchanged quite easily. Formally, ANZLIC is the formal process to endorse collaboration activities, but in reality people just go to the individuals or agencies who have worked in specialist areas to get advice and help. The ASDI is therefore not too regimented. The States, Territories and the Commonwealth are all working together on most national implementation projects such as the Australian Spatial Data Directory (ASDD), Australia's fully distributed metadata directory.

Masser (1999) has summarised the objectives of most national SDI's. These are intending to promote economic development, to stimulate better government and to foster environmental sustainability. A selection of SDI vision statements is provided below.

Colombia (ICDE): http://www.igac.gov.co/indice.html
Europe (EUROGI): http://www.eurogi.org/objectives/
Finland (NGII): http://www.nls.fi/ptk/infrastructure/vision.html
United Kingdom (NGDF): http://www.ngdf.org.uk/
United States (NSDI): http://www.fgdc.gov/nsdi/strategy/goals.html

But a common vision for a SDI may be missing or hindered by reasons such as culturally based resistance. In many instances, information is linked to personal power and tends to be strictly controlled in a top-down manner. This "personalised" approach to information may be one important reason for a lack of a shared SDI approach and also hindering the various stakeholders to produce a shared common vision of a national SDI. High-level commitment and support may be crucial in implementing a change in culturally-bound attitudes.

A common and shared vision about spatial data collaboration and co-operation may fundamentally change the landscape for a nation wide exchange of data and information. In order to get the various stakeholders on board, it may be essential to insist on joint development of a common vision. This may entail a cultural change in the attitude towards information and the exchange of information, a new approach how to manage and share information. The process of getting the concerned parties involved to accept and to actively support the idea of a SDI will need both a strong lead and a lot of creativity in order to minimise unnecessary resistance and not to de-motivate or suffocate creative initiatives.

The vision needs to be developed jointly and shared with the conceivable stakeholders and indicate the incentives for developing a SDI so that people are mobilised to change their behaviour in accordance with the shared vision.

A participative approach to co-operation and co-ordination should be considered in order to build on common interests. This also entails initiating a participative process among the representatives of the already existing database systems. It would make sense to bring the up to now independent system owners, stakeholders, donors, representatives of international organisations active in the field of GIS, soft- and hardware suppliers, and database managers, including their technical staff, to sit together at a round table in order to develop a common concept of a nation wide SDI.

The common standards and procedures the stakeholders will have to agree on will not necessarily fit into their actual database set up but a participative approach and a transparent decision making process will help them to understand the basic questions and to accept the resulting needs for change. Participative processes and transparent decision making are strong arguments to motivate the independent parties to invest their resources in a common project.

The vision needs to be communicated widely using various media to reach all stakeholders. Plans should be developed and implemented regarding the dissemination of information on SDI activities that are under way, including the information about the SDI components, available technological best practices, and the promotion of the use of existing technologies and standards to support the development of a SDI, for example by establishing WWW pages on the Internet or using printed media or CD-ROM where Internet connections are limited.

SDI Scope and Status Clarification: Two broad categories with respect to the status of a national SDI can be identified (Masser 1999), i.e. a SDI resulting from a formal mandate (as was the case in the US, for example) and a SDI growing out of existing spatial data co-ordination activities (as was the case in Australia). While a formal mandate benefits from the provision of funds, existing co-ordination activities provide a base for collaboration. The scope of a SDI may be all-inclusive or focusing on a subset of stakeholders, such as public sector, private sector, or NGOs, with voluntary or mandatory participation. Regardless of which category a SDI falls into and regardless of the breadth of its scope, both should be clarified as early as possible.

An active co-ordination body (committee or commission) to co-ordinate tasks and provide leadership during the process of creating a national SDI should be considered. This would need to be sufficiently empowered to carry out the co-ordination task. In order to implement a SDI, it may not be necessary to establish new organisations and institutions. Instead, existing ones could be strengthened. This would require a revision of the mandates of that institution to ensure that it is well equipped to deliver.

However, the promotion of an existing institution to the co-ordination body for a SDI needs to be carefully considered. The institution needs to be chosen carefully so as to be aware of potential conflicts of interest that may be perceived between the institution's existing mandate and the additional SDI-related activities. For example, a National Mapping Organisation may end up carrying out the SDI co-ordination task and policy development while also acting as a major data producer. This may hamper the support for the SDI initiative from potential participants that could perceive it as biased. Example 8 demonstrates that although it may take some time for the co-ordinating body to gain support, a crucial element to success is how its mandate is perceived.

Perceived mandate: In Portugal, the national SDI (SNIG) is co-ordinated by the National Center for Geographic Information (CNIG). CNIG is not a major data producer, like many agencies in other countries that are responsible for co-ordinating a national SDI. Development of the SNIG was slower than expected mainly due to lack of available digital GI and the incipient computer technologies used by most GI producers. The fact that the CNIG is not a major data producer facilitated the interactions with the GI producers, as they recognised the role of the CNIG as being a complementary one that did not harm their own mission.

The task of promoting and developing a SDI is not restricted to the public sector. In Japan, for example, the private sector is a major driver behind the establishment of a national SDI (see Example 9).

Private sector involvement: In 1995, the Government of Japan established a Liaison Committee among Ministries and Agencies on GIS that is to provide SDI-like functions in the Government in implementing a national SDI in Japan. Private companies in Japan set up the National Spatial Data Infrastructure Promoting Association (NSDIPA), a non-profit organisation to promote the concept of national SDI in Japan. The activities of NSDIPA are aimed at gaining wide awareness of the necessity of the National Spatial Data Infrastructure. It is a group that strives towards the benefits of society and fosters a new information services industry by demanding activities of the government, municipalities and other organisations and by sharing this information with both the public and private sectors.

The representatives of all major sectors or interest groups should be involved. The co-ordinating body, once nominated and appropriately mandated, can then produce a series of activities which need to be accomplished with deadlines and output milestones. The implementation process should be approached in a multi-disciplinary and multi-sectoral way. All related organisations will have their role to play in the SDI development process.

Working groups constitute the platforms for more collaboration among stakeholders by pooling resources and harmonising initiatives to avoid duplication. The involvement of stakeholders is a key issue for the future development of a SDI.

Exchanges of Best Practice and Awareness Creation: Lessons in awareness creation about SDI can be drawn upon from various countries. These suggest that presentations and publications are just some of the activities that can be pursued to advocate and advance SDI development. Networks of communication (see Example 10) can also play an important role. A list of activities includes:

• Outreach through support for SDI from high-profile individuals.
• Promotion of SDI principles through presentations.
• Education through workshops, training courses and material.
• Provide "train-the-trainer" technical workshops to explain the origins, purpose, and strategies for implementation of the SDI-endorsed standards.
• Use pilot projects to demonstrate the value of spatial data and a SDI to improve decision making in communities.
• Establish networks of communication to enable participants to exchange experiences with SDI implementation.
• Facilitate information sharing through newsletters, web pages, and publications: regularly inform interested parties of SDI-sponsored activities and initiatives.
• Provide a forum for debate, analysis and the identification of issues relevant to SDI development.
• Help interested parties or groups to use the spatial data clearinghouse to locate sources of data, training and expertise.
• Offer interested parties the opportunity to participate in Working Groups and Subcommittees, as appropriate.

Networks of communication: EUROGI, the European Umbrella Organisation for Geographic Information, seeks to raise awareness of the value of GI and improve the sharing of knowledge between members themselves and between EUROGI and the European Commission. Communication is facilitated through on-line discussion forums and EUROGI directories where people are able to tell others about their activities by completing a form to add information to a directory or search a directory to read about other peoples' activities.

Examples of how demonstration projects can be used to create awareness of the usefulness of a SDI are detailed in Example 11.

Community Demonstration Projects: The FGDC has worked with the Administration and Federal Agencies to promote several Community Demonstration Projects (http://www.fgdc.gov/nsdi/docs/cdp.html) across the country. These NSDI based pilots are designed to demonstrate the value of spatial data and the NSDI to improve decision making in communities. The Demonstration Projects address a number of issues including flood management, local/regional crime management, Citizen-based land use analysis, environmental restoration. The NPR and the FGDC jointly queried FGDC membership to seek interested communities, offering only in-kind federal help (federal staff, training, etc but no dollars) to maturate the projects. Shortly after the selections, the six selected communities joined together to apply for a grant under the Government Information Technology Services Board (GITS). They were awarded over $600,000 dollars as part of their grant request. These projects are expected to report back in May 2000, with the detail of each effort utilised to help articulate the value of NSDI to enhance place-based decision-making, and to help communities understand the costs and processes associated with establishing NSDI operations.

Community-based capacity building: In 1998, the FGDC working with OMB and its federal agency representatives began a $40M multi-agency budget initiative to accelerate the application of the NSDI to improve the decision-making power of communities in addressing livability issues. The Community Federal Information Partnership (CFIP) (http://www.fgdc.gov/nsdi/docs/schaeferbrief/index.htm) was first announced by the Vice President as part of a Sustainable Communities Speech delivered in 1998 at the Brookings Institute. The C/FIP would provide grants to communities to activate capability and tools for place-based decision-making, and would provide federal agencies with additional funding to help make their spatial data more available for public access. The outcome of CFIP for fiscal year 2000 is still being worked through the congressional budgeting process at the time of this writing.

Exchange of best practice: The FGDC has developed metadata, Clearinghouse, data Standards training, and has developed and offers metadata tools. Assistance is provided by FGDC and FGDC trained partners to local, state, federal, tribal, and international organisations seeking to establish or improve their SDI.

The Role of Management in Capacity Development: An important barrier to change is an organisation's capacity to adopt new standards and technologies. While the introduction of specialised software, for example for the creation of a geospatial catalogue, is relatively easy, its effective use depends on the technical capabilities as well as organisational support. Awareness creation of SDI components should be considered down to the lowest level and with strong management support and leadership. Capacity development should be a prime concern of senior management. It includes the theoretical issues and the practical hands-on capabilities to implement the SDI components.

This issue of building local capacity will to be a major constraint to the success of a SDI in many developing countries. As job specific technical competencies will be stipulated, it will be necessary to review positional titles, remuneration packages and salaries. The staff rotation system in the "Department of Geological Surveys" in Zimbabwe is a case of "best practice" in how "brain drain" can be avoided and serves as an example of how staff can be motivated within a "Learning Organisation". This system is designed to enhance the capacity of personnel within the department, therefore reducing the need for external recruitment of technical staff.

The personnel resources for SDI in many countries are very limited since most of the GIS implementations being built up are understaffed. A pool of qualified staff has to be created if the projects are to become sustainable. What makes it difficult for countries such as Zimbabwe, for example, is not only the number of specialists required, but also the working conditions offered. "Brain-drain" is a serious problem: the fact that skilled personnel are leaving their jobs too often, too soon. Human capacity development and long term career planning should be of prime concern to senior management. It includes the training, theoretical issues and practical hands-on capabilities to implement projects and programs, as well as the working conditions. Working conditions need to be considered not only with respect to salary, but even more importantly with respect to the work climate, motivation and professional perspectives.

Compensating for high staff turnover: One of the US NSDI Community Demonstration Projects is taking place in the Police Department in Baltimore, Maryland. The Police Department has come to realise that a SDI is good for managing vital crime data in addition to the classic mapping data that many rely on for the base mapping. The Baltimore Police budget is tight, they have a high turnover in staff. By capturing metadata and using clearinghouse capabilities, they can better assure the proper management of critical crime data used by the department and throughout the region as part of a regional crime management collaboration between community police organisations.

Senior management of all concerned institutions should consider the development of standards a priority. They should closely supervise technical work groups and assure that the desired results will be produced. Matters like the standardisation of data and the harmonisation of classification schemes cannot be left to technicians alone because they entail political decisions. Senior management should be acknowledged as a driving force behind the build-up of a SDI.

Funding and Donor Involvement: Funding and adequate resources can present a major constraint to SDI development when awareness of the importance of SDI is lacking at the local, national or regional level and there is no existing SDI-like initiative or a mandate to develop a SDI to which sufficient funds have been assigned.

Nevertheless, in order to ensure funding, it may more persuasive to potential funders to have something to show already (for example, a clearinghouse system) rather than a concept document alone. This does not have to involve huge costs since clearinghouse components are available free over the Internet (link to Chapter 4). In addition, justification for the limited cost of this initial development may well be found within existing projects or initiatives (for example, documenting data holdings is a part of sound information management).

Innovative use of resources can ensure that funds stretch a long way. For example, with a 'carrot and stick approach', incentives can be created for the adoption of SDI principles. Using small, non-repeating grants to stimulate the development of the application layer of the SDI can work well where there is broad base of existing expertise that can be encouraged (see Example 13).

Program of Grants: The FGDC in the US has maintained a relatively small but persistent Cooperative Agreement Program of Grants (CAP) to help communities validate and initiate NSDI concepts (http://www.fgdc.gov/publications/publications.html). The FGDC initiated the CAP program to provide seed money to stimulate co-operative activities among organisations to begin implementing the NSDI. Rooted in the premise that building the NSDI is a shared responsibility and collaborative efforts are essential for its success, the CAP program has worked to seed 270 NSDI resource sharing projects across the county involving more than 1300 organisations. These projects have resulted in helping state governments, libraries, universities, local government organisations, and private sector entities to become stable contributing sources on the NSDI. While the FGDC funding level for the CAP has been somewhat limited ($1 - $2M yearly), annual funding has been persistent since 1994, and recently the number of grants awarded has been increasing - communities appear to be doing more with less.

Reports of different SDI funding mechanisms from Australia and Portugal suggest that the provision of central funds is an important contributor to accelerated SDI development (see Example 14 and 15).

Decentralised funding: In Australia, there is no major national funding allocation for the ASDI (unlike the US and Canada). Each jurisdiction (States, Territories and Commonwealth) are all financing their own programs. The States and Territories in Australia are each developing their SDI's, so in fact the ASDI is all the individual jurisdictional SDI's puzzled together. This approach has some drawbacks. It would be more coherent if a national SDI pool of funds was available to draw on and bring influence to bear. Industry groups are not really engaged in Australia yet to the same degree as in the US or Canada. Making the ASDI politically attractive for large scale national funding has not occurred yet but continuing efforts are under way. One notable success was the establishment of the Australian WWW Mapping Consortium as a full member of the OpenGIS Consortium (OGC). 23 Australian industry, R&D and Govt Groups have all come together to share ideas and work on an "Australian Web Mapping Testbed" which making good progress, allowing them to provide some new input into the OGC process.

Centralised funding: The creation of the Portuguese national SDI, the SNIG, was endorsed by public funds. The approval by the Portuguese Government and by the European Commission (at the end of 1994) of a program integrated in the Regional Development Plan 1994-1999 did cover a specific budget assigned to the support of the SNIG development. Part of the funding was used to speed up the creation of digital geographic information, namely for conversions of existent geographic information into digital formats, and for the purchase of satellite data and existent digital topographical data for local GIS implementation in municipalities. Another part of these funds was used to provide major public data producers with Internet servers, routers and communication infrastructures. A small fraction of the funds is still being used to build WWW interfaces and applications to facilitate the access to geographic information available at the different institutions integrated in the SNIG network. In the Portuguese case, funding was a major factor that helped the fast development of SNIG since 1995. It did accelerate a process that would have taken years to grow up. At the present moment, a total of 117 public institutions, including almost all GI producers, have joined the Portuguese spatial data infrastructure.

GIS implementations in developing countries are often functioning under special conditions that need to be considered during the initiation of a SDI at national or regional level. In many countries the lack of local financial resources means that GIS implementations are not financially sustainable and therefore depend primarily on donor funds. Usually donor support for these projects is provided under certain conditions such as a time limit for implementation after which there are no further disbursements of funds. The future of many of these systems is uncertain beyond the end of international assistance.

Another aspect of donor-funded GIS implementations is that often the projects have been initiated by donors according to their own objectives and little attention has been paid to the requirements and capacities of the host organisations. The result is that there is insufficient co-ordination of the technical support and funding activities of different donors. In some cases donors may not be willing to work with each other and this can impose limits on the co-operation or data exchange between projects that are funded by different donors. A lack in capacity to co-ordinate donor activities coupled with competition among the donors themselves can hamper a SDI initiative.

Under these conditions, the co-operation with donors is a critical aspect of the development of a national SDI. While the existing co-operation should not be exposed to strain, a co-ordinated SDI-based approach would change the priorities for GIS implementations. This potential conflict could be avoided if donors would be invited as partners to take part in the participative process defining the components of a nation wide SDI.

In order to develop (or renew) a national SDI in a (multi) donor-funded GIS context, a useful approach has been developed by Ryerson and Batterham (2000). This approach entails an evaluation of GIS projects with regard to:

• clients' needs and desires,
• an assessment of the recipient country's capabilities in terms of meeting those needs,
• an assessment of related activities by other donors,
• an assessment of current technology and its direction,
• donor country capabilities and capacity if the aid is tied, and
• costs.

Initially funded by Donors, the SADC Regional Remote Sensing Unit (RRSU) in Harare, Zimbabwe, has been integrated in the organisational structure of the Southern African Development Community (SADC) since 1998. The Unit is funded by the 14 Member-states (Angola, Botswana, Democratic Republic of Congo, Lesotho, Malawi, Mauritius, Mozambique, Namibia, Secheylles, South Africa, Swaziland, Tanzania, Zambia, and Zimbabwe), and still receives some additional donor contributions.

The SDI work implemented by the RRSU was never part of the original work plan. It was identified when GIS technology was to be used for basic analytical procedures. This could not be done because the data sets were incomplete or incompatible. At the time the SDI work started, the Unit was still depending on donor assistance and technical assistance from the FAO. Consequently, changes had to be made to the work programme, something which had to be discussed with the donor and technical assistance partner.

From the regional and international partners in the development (the data suppliers) of the RRSU spatial data sets, no financial contributions were required.

The RRSU spatial data sets were originally developed for GIS-based applications in support of early warning for food security. However, the data sets are being recognised as one of the major spatial data base developments in the SADC region, and for this reason, the RRSU continues to attract donor funding. The spatial data base activities were originally not foreseen as a major task - but this has changed considerably over the years. (http://www.zimbabwe.net/sadc-fanr/intro.htm)

Broad and pervasive partnerships across public and private sectors: Co-operation and partnerships across different levels of the public sector and with the private sector can be helpful at every stage of SDI development to collect, build, share, and maintain spatial data.

Since no one organisation can build a SDI, collaborative efforts are essential for its success. The FGDC in the US encourages federal, state, local, and tribal governments, academia, the private sector, and non-profit organisations to work together within a geographic area to make geospatial data available to all. So-called 'cooperation groups' are formed that enable all parties to participate in, and contribute to, the national SDI in the areas of their strength and expertise. Guiding policies and procedures for these cooperation groups have been developed (http://www.fgdc.org/funding.html). Co-operation among Federal, State, local, private, and academic sectors is expected to be based on shared responsibilities, shared commitment, shared benefits, and shared control aiming at improving the spatial data delivery system (see also Example 17).

The task of building relationships to further the implementation of the NSDI in the US has been a large and continuing effort. The effort has been made difficult by the fact that organisations, functions, and responsibilities are diverse and spread out across the country. Initial efforts concentrated on FGDC initiatives to build relationships with Coordinating Groups that have formed to represent issues within the States, and with Organisations and Associations that represent levels of government of key interest groups nationally. This has helped focus the work of the different groups and has established strong linkages with some of the key elements needed for a long-term national network of partnerships. The FGDC's efforts has also been helped by the fact that many in the United States recognise the value of geographic information to the decision making needs of communities. Geographic information is collected at all levels. Most of the data originates at the local level, but very important types of data come from other levels, including complete information of an issue or topic that transcends jurisdictional boundaries (region or state). Thus there is a growing level of support for policies, interfaces, standards, and relationships that enable government, companies, organisation and citizens to interact and share in the collection and dissemination of geographic information across jurisdictions.

In the Canadian context, public and private-sector partnerships focus on partnering and leveraging the resources of the private sector to accelerate access to spatial data and technology development. GeoConnections, the program responsible for implementing the Canadian Geospatial Data Infrastructure (CGDI) has placed particular emphasis on partnerships between the federal and provincial and territorial governments and the private sector and academia. Programs focus on working across governments, and with stakeholders and the private sector to advance the amount of information accessible through 'clearinghouse' systems, the development of data frameworks to ease data integration, fostering advanced technology and application development, and building supportive policies to speed industry growth. To this end, guiding principles for the provincial and territorial government agencies involved in geomatics have been agreed upon.

1. Data should be collected once, closest to the source and in the most efficient way possible, with a view towards increasing the vertical integration of the data.
2. Geo-info data should be as seamless as possible, with co-ordination across jurisdictions and boundaries when possible.
3. Data should be collected, processed and maintained according to international standards to maintain data integrity across databases, and to enable the addition of value, further enhancement and easy access and use.
4. Upon agreement, partners should contribute equitably to the costs of collecting and managing the data, and should be allowed to integrate the resulting information into their own databases, for their own use and for further distribution to their stakeholders.
5. There should be an attempt to harmonise terms and conditions for use where practical. In the absence of such agreement, each agency should be free to establish its own terms and conditions for such information.
6. Agreements between agencies will normally be negotiated on a case-by-case bilateral or multi-lateral basis, according to these principles of partnership.
7. Partnerships between agencies should be simple and support the principles of the CGDI, open to the participation of interested stakeholders within any level of government, the education communities or the private sector.
8. A group or agency within each province and within the federal government should be designated to promote and co-ordinate the development of a common geospatial data infrastructure, both within its jurisdiction and between jurisdictions.
9. CGDI is national in scope, and must meet the needs of a wide range of geospatial user communities, data producers and different areas of the private sector.
10. CGDI must consist of a set of co-ordinated and interrelated policies, practices and possibilities that build on the vision.

Develop clearinghouses and use common standards for data and technology: The technical underpinning of a SDI is a common framework of standards, tools and services based on these standards. In this three-tier model, applications work with metadata and data content and services that exist on the enabling infrastructure. The following technical elements are important components of a SDI:

• quality metadata,
• residence of metadata in on-line directories,
• good data management,
• access to services on-line,
• their documentation in directories and
• reference implementations of software to demonstrate capabilities.

The development of the Portuguese Spatial Data Infrastructure serves as an example of the importance of outreach activities that parallel the implementation of the technical elements of a SDI (see Example 18). The Portuguese SDI differs from other SDI's by having a centralised metadata catalogue. Usually metadata is organised in a distributed way. Nevertheless, the example demonstrates that in order to gain support for the system (i.e. increasing the number of users of the system), new interfaces were developed according to feedback from the users themselves and through the development of tools that are more devoted to the needs of citizens. The Portuguese experience also shows that a SDI can be developed incrementally with improvements implemented step by step.

User involvement in the technical implementation: In 1990 the Portuguese government created SNIG, the Portuguese GI infrastructure, as a national public service (http://snig.cnig.pt). The main goal of SNIG was to ensure the connection of Portuguese users and producers of digital geographic information through a network. This goal implied the development of catalogues describing the available geographic information. By that time, the majority of public agencies were more concerned with the production and organisation of digital geographic information than with the dissemination process. It was felt that data producers were not prepared to manage their own metadata registers. Thus, the creation and maintenance of the metadata that currently supported the SNIG was organised in a centralised way by its co-ordinator, the National Center for Geographic Information (CNIG). By the end of 1994, taking advantage of the multiple opportunities for publishing data offered by the World Wide Web, the Portuguese GI metada ta catalogues were implemented in a Relational Database Management System and CNIG started to build an HTML interface to allow the query of the metadata and the retrieval of the available data sets. The SNIG network was finally launched on the Internet on May 3, 1995. The main concern was to connect users to the available digital data sets, creating an operational system that could be improved in the following years. Therefore, the metadata catalogues were not based on any metadata standards. Common sense, some guidelines provided by the CORINE Catalogue of Data Sources Project and the identification of the main geographic information sources were used to design the database. During this stage, the system structure and design was mainly oriented towards the professional user.

Subsequently, the creation of new metadata catalogues obliged to rebuild a new WWW interface. While the first SNIG interface was developed without the implementation of formal usability studies, usability testing was required to support further developments of the SNIG. In order to rebuild the SNIG site, qualitative research involving users was carried out for the first time. The research was designed to answer to the following questions:

• Which would be SNIG potential users groups?
• What geographic information options would users need?
• What would the users be looking for in an infrastructure such as SNIG?

The main results of this research pointed out that it would be necessary to develop a friendlier interface. The new interface should adopt informal and non-technical terminology and include search engines by terms and geographic location. The need of more geographic information for non-professional use and the adoption of more common data formats were also stated. It would also be important to include raster images to illustrate the available information. In July 1999, an alternative SNIG user interface was launched (GEOCID). GEOCID is more appealing and information oriented, avoiding complex tasks and navigating routes to access to the data. In addition, new applications were developed based on the information citizens are interested in. An application that allows the user to navigate through Continental Portugal, select specific locations and download the part of the orthophoto he is visualising on screen was developed. The launch of GEOCID was a big success (http://ortos.cnig.pt/ortofotos/ingles/).

Recommendations: Outreach and Capacity Building Options for Implementing a SDI

By overcoming inefficiencies, a coherent and consistent SDI can ensure that geographic information may be used to address complex social, environmental, and economic issues. The following guidelines indicate some of the outreach and capacity building activities that can be used to foster the implementation of a SDI:

• A practical step to take in the development of a national SDI is the development of a vision, detailing a vision of the desired future and a clear sense of how SDI components could serve that future and help to realise it. This also involves setting clear priorities and defining a strategy or policy to accomplish the vision.

• A workshop organised with the stakeholders to define and create a national co-ordinating body, considering its structure in terms of an existing or newly created institution, working groups and/or committees. In countries where GIS implementations are highly dependent on donor involvement in terms of funding and technical expertise, donor representatives should be considered as stakeholders and included in the process of building a SDI. The co-ordinating body needs to be mandated to manage the required activities and devise an action plan to co-ordinate the activities.

• Consideration needs to be given to the necessary resources for implementing strategy, policy or plans and activities, considering staff, technical know-how, material, and funding opportunities such as innovative partnerships.

• Formal working groups should be organised around well defined objectives, strategies, plans, programmes, and actions, and not simply for informal and limited consultations. These working groups would be made up of interested parties and experts to deal with specific aspects of SDI such standards (metadata, exchange), national data sets, policy, clearinghouse and how to assimilate existing technological solutions into the local context.

• Awareness creation of SDI components should be considered down to the lowest level and with strong management support and leadership.

• Plans should be developed and implemented for the dissemination of information on SDI activities under way, including the information about the SDI components, available technical best practices, and the promotion of the use of existing technologies and standards to support the development of a SDI, for example by establishing WWW pages on the Internet or using printed media or CD-ROM where Internet connections are limited.

• Measures should be taken to monitor, analyse and participate in, developments at international levels that affect the use of standards and supporting technologies in the national context. This entails assigning clear administrative responsibility for tracking key developments are the international level and within the GSDI community.

• Within SDI development, the role of donors should be clarified to support activities by way of following local priorities such as interoperability of different GIS implementations rather than wishing to be associated with a particular type of activity irrespective of cost effectiveness or fit with the broader institutional or national objectives.

GSDI (1998) "Conference Resolutions, Recommendations and Findings", 3rd Global Spatial Data Infrastructure (GSDI) Conference, Canberra, A.C.T., Australia, 17-19 November.

Gouveia, C., Abreu, J., Neves, N., Henriques, R. G. (1997) "The Portuguese National Infrastructure for Geographical Information: General Description and Challenges for the Future", GISDATA Conference Proceedings.

Henriques R. G., Fonseca, A. et al. (1999) "National System for Geographic Information (SNIG): The Portuguese National Infrastructure for Geographic Information", Madame Project: 1st Progress Report.

Mapping Science Committee, Board on Earth Sciences and Resources, Commission on Geosciences Environment and Resources, and National Research Council (1994) "Promoting the National Data Infrastructure Through Partnerships", Washington, D.C.: National Academy Press.

Mapping Science Committee (1993) Toward a Coordinated Spatial Data Infrastructure for the Nation, Washington, DC: National Academy Press. Execrative Summary available at http://38.217.229.6/NAPA/NAPAPubs.nsf/00a36275d19681118525651d00620a03/229b79ae768d77e48525658c0061a3bd?OpenDocument).

Masser, I. (1999) "All Shapes and Sizes: The First Generation of National Spatial Data Infrastructures", International Journal of Geographical Information Science, Vol. 13 (1), pp. 67-84.

Mbudzi, M., Jairosi, Y., Vogel, D. and Bohnet, D. (1997) "Best Practices on Environmental Information Systems (EIS): The Case of Zimbabwe", Program on Environmental Information Systems in Sub-Saharan Africa, May.

Mendes, M. T., Joaquim, S.P., Hengue, P. and Gerbe, P. (1998) "Best Practices on Environmental Information Systems (EIS): The Case of Mozambique", Program on Environmental Information Systems in Sub-Saharan Africa, May.

Nicolau, R. (1998) "Adoption of the Metadata Standards within SNIG", workshop on "Challenges and Future Developments of GI Infrastructures: The Portuguese Experience", GIS PlaNET'98 Conference, Lisbon, FIL, 7-11 September.

Ryerson, R.A. and Batterham, R.J. (2000) 'An Approach to the Development of a Sustainable National Geomatics Infrastructure', Photogrammatic Engineering and Remote Sensing, January, pp 17-28.

United States National Academy of Public Administration (1998) "Geographic Information for the 21st Century: Building a Strategy for the Nation", Executive Summary, January, http://www.napawash.org

Wehn de Montalvo (1999) "Survey of Spatial Data Sharing Perspectives in South Africa - Views on the Exchange of Spatial Data Across Organisational Boundaries", Summary Report, SPRU - Science and Technology Research, University of Sussex, December.

Table of Contents