Alison Davey and Keith Murray, Ordnance Survey
1 What is the National Geospatial Database
2 The Three Pillars of the NGD: Collaboration, Standards
and Data Access
3. Results of the NGD Seminar: A Cooperative and
a Proactive Approach
4. Next Steps
At the AGI '95 conference the proposal by Ordnance Survey for the creation of a National Geospatial Database (NGD) won the prize for the best paper[1]. During 1996 Ordnance Survey has, in discussion with many other parties, drawn up an outline plan of action to provide a framework for linking together the many different geospatial data sets held by government departments in a National Geospatial Database. This paper provides an update of what has happened since AGI '95.
Geospatial data is broadly defined as any data that describes some phenomenon and has a geographical location reference such as a National Grid reference, a postcode or a latitude and a longitude. When implemented, the NGD will be the over-arching framework to facilitate the linking, combining and widespread use of geospatial data collected, maintained and held by different organisations, particularly (but not exclusively) where data gathering uses public funds.
The NGD will provide the data baseline that will enable many applications and services, including the National Land Information Service (NLIS), ScotLIS and other existing initiatives, to identify and integrate underlying data sets with greater ease. Thus the NGD is a framework for 'adding value' to the many existing geospatial data holdings that will improve the quality of the data sets and minimise duplication of effort. The NGD will not be a single, massive, physical database but a set of standards, procedures and processes to link together existing data, to formalise their availability, describe their quality in a consistent way (and hence the uses to which they can be put), and provide tools for exploring the data. In this way the NGD will support the realisation of a conceptual 'one stop shop' for geospatial data.
To achieve the NGD there must be easily accessible information describing the geospatial data and how the data can be accessed. In addition, geospatial data should be compiled with the aim of making it easier to integrate the data with other data sets.
A culture of collaboration and partnership between organisations in the geospatial data community will need to further develop to realise the NGD benefits. Many hundreds of geospatial data sets exist which are owned or held by many different organisations. Each organisation has geospatial data to satisfy its own mandate. For this data to be used effectively more than once people are going to have to collaborate. Working together is not new, already there are examples of groups or projects which are in sympathy with the aims of the NGD.
Many initiatives are already up and running which demonstrate that organisations do collaborate and work together. Examples include:
The UK Standard Geographic Base
A consortium of government organisations funded a study of what would constitute a standard geographic base for the UK. The final report in April 1996 contained recommendations for
The core spatial units include basic land and property units, unit postcodes, wards and streets. The study also concluded that there would be significant benefits for information handling if administrative and postal boundaries could be brought into alignment.
An implementation steering group and a project group have been established to develop a business plan and an implementation plan by March 1997. The National Land Information Service (NLIS) and ScotLIS
The NLIS project in England and Wales and the ScotLIS project in Scotland each involves a consortium of organisations investigating the technical and business issues surrounding on-line access to information on land and property. Each of these initiatives aim to bring together data from a number of different organisations to provide services to customers. The Bristol pilot study in the NLIS project has resulted in a land and property gazetteer for Bristol that is used to link data sets and the use of a central hub for accessing disparate datasets. A consultancy study will report on possible business models for NLIS.
In addition to specific projects, several organisations or groups are working toward wider and more intelligent use of geospatial information. Examples include:
The Association for Geographic Information (AGI)
Part of the mission of the AGI is "to spread the benefits of geographic information", which means that AGI is the natural home for taking forward some of the necessary activities in the NGD process by drawing on the expertise of its membership and encouraging people to work together. Members of the AGI Council have been in discussion with Ordnance Survey on how AGI can best assist the development of the NGD. AGI has an important role to play in promoting awareness of the NGD and as a voice for the interests of the geospatial data community.
The Inter-departmental Group on Geographic Information (IGGI)
IGGI has a stated aim of identifying barriers to effective use of geographic information and taking steps to remove the barriers. The group was established in 1993 to provide coordination between government departments in the development and effective use of geographic information. The UK Standard Geographic Base study was commissioned by IGGI.
The Local Government Management Board (LGMB)
The local government community will be making a very active and important contribution to the NGD process through the LGMB and through the creation of relationships for the release of local government data. For example, the National Street Gazetteer concession will ensure the creation of a national gazetteer from local gazetteers in local government and access to the gazetteer for the wider community. LGMB are working on the implementation of BS 7666 on spatial data sets for geographical referencing.
Many existing initiatives are inter-linked: NLIS and ScotLIS require a BS 7666 compliant land and property gazetteer, the UK Standard Geographic Base proposals include a land and property gazetteer and a street gazetteer. The NGD could provide maintained information on geospatial initiatives to enable easier understanding of the links between initiatives, their timescales and their main areas of focus.
The wider geospatial data community, through the NGD, needs to draw benefits from the results of the many initiatives to ensure there is no duplication of effort and that future work builds on their success.
For the NGD to be a success the geospatial data community needs to develop and implement standards and best practice for the data itself and for processes and procedures to ensure that data sets are accessible, well-described and exchangeable. A minimal set of standards is needed that are simple enough to win acceptance by data providers, users and GIS vendors.
The British standards for addresses, and for street and land and property gazetteers (BS 7666) are starting to be implemented by Local Authorities and others. Use of unique identifiers and these standards will support data linking and combination. International standards that are under development will also support the NGD (once the standards are implemented). The European standard on metadata from CEN/TC 287 (which is expected to be available from the British Standards Institute in November 1996) is just one example of a standard that could assist the NGD process.
A successful NGD will need good metadata services. Already in Britain there is the Spatial Information Enquiry Service (SINES), which is run by Ordnance Survey on behalf of IGGI. This metadata service for geospatial data held by government could be expanded to create a readily available and comprehensive register of any geospatial data in the UK that people wished to describe (including data held by private organisations). Any national index will need to involve both data suppliers and users. Work is in-hand to improve access to SINES, data currency and the directory content. The UK Standard Geographic Base report also contains proposals for a metadata directory of spatial units. There is likely to be more than one geospatial data metadata service in the UK. Through the NGD it should be possible to work towards linking metadata services to transparently serve metadata to users. Sample data could also be included in the metadata service (possibly geographically limited).
A project bid has been submitted for an EC INFO2000 project on the European Spatial Metadata Infrastructure. The project aims to create an environment in which metadata systems can communicate, thereby linking metadata services. The bidding consortium, in addition to Ordnance Survey, includes organisations in the Netherlands, France, Portugal, Spain and Italy.
Many applications for which there is market demand should develop from the NGD. For these applications to succeed more work is needed on the legal and commercial issues surrounding geospatial information. Data access models and methods need to be developed that take account of Intellectual Property Rights, onward use of data, privacy, charging, dissemination, liability and responsibility for data quality while encouraging the widest possible use of the data. Good practice guides could document different access models applied by different organisations as a guide to other organisations wishing to make their data more accessible. How organisations should market their data will not be dictated through the NGD but we should be able to draw together and publicise experience to date.
Services, such as NLIS-type developments, will always be market-driven. Owners of the 'content' embedded in the services must be free to decide on the market channels, access methods, charging policies and partnership arrangements. NGD-compliant applications will be taken forward in a variety of different, and possibly competing, ways. These commercial decisions will be distinct from the NGD infrastructure. The NGD-compliance will facilitate other data sets being easily and sensibly linked to each service at a later date (where there is market demand).
A seminar on the NGD was held in London on 5 June 1996 to identify what needs to happen to develop the NGD. This was attended by over 60 people, including data holders, software suppliers and the academic community. Four scenarios were discussed: market forces, cooperative, proactive and central control. Most of the audience favoured an approach that was both cooperative and proactive.
These approaches highlighted that:
Such a cooperative and proactive approach should help improve knowledge on what geospatial data already exists, ensure that data can be more easily combined, ensure people are better able to combine data sensibly, and ultimately make it easier for services and applications that use geospatial information to be developed.
Necessary activities to implement the NGD include:
Ultimately the more services and applications that are developed, the greater the value to the economy of geospatial data.
During and subsequent to the seminar, there has been a widely expressed view that it is time for Ordnance Survey to stop just talking about the NGD and start putting in place mechanisms to enable the whole geospatial data community to make it a reality. As a result of bilateral discussions between Ordnance Survey and others on the realisation of NGD, it is proposed that management of the implementation be taken forward by establishing: