Abstract
The National Geospatial Data Framework (NGDF) is the UK’s spatial data infrastructure. The emphasis is on a framework of standards, metadata, and access, rather than the more centralised approach in other countries involving base datasets. It has evolved in a rather different way from those in other countries, notably the USA and Australia, due principally to the government, legal, and business environment in the UK. In recent years, the emphasis has been on practical projects and delivery of services. Two services have now been launched – the Data Locator and the Data Integrator, through the delivery mechanism of ‘askGIraffe’, a website gateway to the services and data. Further information on the NGDF in the UK, including many of the references used in this paper, is available on www.ngdf.org.uk

The UK situation
In the UK today, over 40 government departments and other organisations produce geospatial data for their own needs, spending approximately £400 million annually on data collection and assembly. Despite relatively centralised government there are serious inconsistencies between different regions of the country, and different data collectors. Much of the data is not available for re-use or is collected in ways that make it difficult to use with other datasets and most datasets remain poorly documented. In addition, there are still inconsistencies in the ways different computer systems treat sophisticated geospatial data and the geospatial information that is available is often embedded within services and is taken for granted. Many data holders, particularly those in the public sector, do not advertise the fact that they hold geospatial information. This is often because the information

• is collected for a specific purpose (e.g. research);
• is considered redundant and retained only for evidence after a report has been published; or
• is not seen as being of interest outside the department.

Furthermore, delivery of the underlying data would require an infrastructure and supporting documentation that does not exist in such organisations.

Four main obstacles could be identified which prevented better use of geospatial information. These were

• A lack of knowledge about what data exists. Metadata was poor or non-existent. Where it did exist it was collected to varying standards and so could not be easily collated.
• Poor access to information. Many data collectors do not have a remit to provide access to data. As a result there was a lack of publicity, infrastructure and documentation. Additional obstacles included legal restrictions, and pricing policies.
• Inability to integrate information due to inconsistencies. The UK has a myriad of spatial referencing systems, all developed for different applications. Integration is difficult or impossible. As a result questions asked of the data cannot be answered.
• Low levels of use of geographic information. As a result of the situation outlined above, there were low levels of use of GI, leading to duplication of effort and makeshift solutions to problems.

These problems were recognised in 1995, and it was these that the NGDF initiative sought to address.

Formation of NGDF
The NGDF was launched as an initiative at the Association of Geographic Information (AGI) conference in 1995 (Nansen, Smith & Davey, 1995). See also Rhind’s paper to this Workshop. From the outset, the participants were from government, academia and the private sector. Following an encouraging reception, work began to further refine NGDF and suggest a practical and effective method of implementation. A well attended seminar was held in June 1996 at which key players discussed and agreed a way forward and a progress report was presented at the AGI conference in 1996, which outlined the recommended way forward and raised some key issues to be addressed by NGDF. The first NGDF meetings were held in the autumn of 1996. The general approach was to move forward by involving all the major players. There was no explicit government directive as there had been in the USA, but there were some enabling government policies such as the Modernising Government and Joined Up Government policy initiatives.

Initially progress was slow, both due to differences in opinion over the strategic direction for NGDF, and a lack of funds and resources to carry out the work. A workshop was held in April 1998 to try to resolve these issues and as a result, the NGDF Strategic Plan was developed. This enabled funds to be sought from the National Interest Mapping Service Agreement (NIMSA) and the Invest to Save Programme. Funds were secured from these two in late 1998/ early 1999.

NGDF – The organisation
NGDF is a voluntary organisation made up of representatives from the government and commercial sectors. Initially it consisted of an Advisory Council, and Board and a Task Force to take forward the work. Ad-hoc groups were formed for various topics.

The structure was adjusted in 1998 with the demise of the Advisory Council; the AGI undertook this role by having a representative on the NGDF Board. The NGDF now consists of an NGDF Management Board, two Project Boards for the major projects (Metadata and the Standard Geographic Base), and working groups. An NGDF Central Management Team is employed to carry forward the work of NGDF.

NGDF has three streams of funding

• a NIMSA activity that funds the Central Management Team,
• Invest to Save budget that funded the two major projects until the end October 2000, and
• contributions made by NGDF Management Board members that fund ad-hoc projects.

The Management Board sets policy, commits resources, formally approves NGDF activities and deals with political and commercial issues as they arise. The Board contains some 15 varied but core organisations that are represented by their Chief Executive or an empowered representative. These organisations can be categorised under 4 broad headings

• ‘Umbrella’ groups: Association for Geographic Information (represents the public and private sectors), Improvement and Development Agency (represents local government), Intra-governmental Group on Geographic Information (represents central government)
• Private sector organisations: Property Intelligence Plc, Landmark Information Group Ltd
• Key government departments and agencies: Central Information Technology Unit, Land Registry, Ordnance Survey, British Geological Survey, Office for National Statistics, Post Office
• Representatives from Ordnance Survey of Northern Ireland and from Registers of Scotland

The two Project Boards manage their programme of work appropriate to the further development and implementation of their specific project. Each has a group below it tasked with specific deliverables. The NGDF Central Management Team manages the financial affairs, promotion and co-ordination of the NGDF programme.

Remit and Strategy
The Mission of the NGDF is

to develop an over-arching framework to facilitate and encourage efficient linking, combining and widespread use of geospatial data which is fit for purpose.

And that of the Management Board to

enable the unlocking and improvement of geospatial information for the benefit of the citizen, business growth and good government

The Strategic Plan, revised last year, sets out the strategic objectives, targets and deliverables. The strategic objectives fall into 5 groups

• To improve knowledge about what GI currently exists and encourage easier access to it
• To enable easier integration of GI through the use of standards and guidelines
• To encourage more widespread use of geospatial information
• To provide users with assurance that the information is consistent and of defined quality
• To provide advice to government on geospatial information

The key deliverables in Phase 1 (between 2000 and 2001) are a Gateway for delivery of services, a metadata service, and a set of NGDF-recommended data standards and guidelines.

In contrast to some other SDI’s, NGDF is NOT a database, a set of basic data, or a government-driven initiative. It concentrates on creating an environment rather than a structure.

Projects
The majority of the work is conducted through two major projects. The first is the Metadata Project, and the second the UK Standard Geographic Base (UKSGB) Project.

The Metadata Project has concentrated on setting up the Data Locator service, and populating it with metadata. The UKSGB Project is a more ambitious project that will seek eventually to define a standard set of geographical units for the UK. The service delivered from this project is called the Data Integrator. Both services are delivered through askGIraffe and are discussed below.

askGIraffe – The Service
NGDF delivers its services through askGIraffe. askGIraffe aims to become the first choice web-site to provide the gateway to geospatial information provided by the public and private sector across the UK. This gateway allows access to the Data Locator service and the Data Integrator service.

askGIraffe Data Locator
The provision of metadata has been on the agenda for many years. Government had a metadata service called SINES (Spatial INformation Enquiry Service) running with telephone enquiries in 1994. This was later extended to email and on-line searches. However, this was limited to government datasets, and considerable effort was expended to bring the metadata into one format. As a result it was not easily maintained.

With the exponential growth in the use of the Internet and its adoption as a service within organisations (Intranets) a means is now available to use World Wide Web based search tools to locate sources of geospatial information. This increases the importance of providing standard methods for describing data that can be analysed by computer based systems. The askGIraffe Data Locator service exploits this technology. It provides a facility to search in a variety of different ways, through a map, through an index, through keywords or through a simple phrase. The information returned falls into broad categories; identification, subject matter, originator, access use and restrictions, time of capture, geographic extent and data supplier. From there you can either follow the hypertext links to the data provider’s web site or e-mail, or you can phone, fax or write to them to access the information.



The askGIraffe Data Locator uses the NGDF Gateway Metadata Infrastructure that provides a national focus but enables local maintenance of the information (See Figure 1).

Unlike SINES which was one central database, the NGDF Gateway supports routing facilities to service providers that operate their own node and through these to data providers. NGDF also provides an NGDF node for those data providers unable to support their own service or find an appropriate service provider. Service providers are free to support enquires from sources other than the NGDF gateway. NGDF also provides support and encouragement to adhere to the standards and a central resource where minor data providers can register their metadata. A helpdesk facility and impartial advisory service are also available.

askGIraffe Data Integrator
The majority of geospatial information is not directly referenced using geographic co-ordinates but is indirectly referenced using an address, a postcode or the name of a region. The UK is made up of many different indirectly referenced systems; some which are used for purely operational purposes such as Health Authority areas and National Park areas, others are more widely used, for example administrative, postal, electoral and census areas. The problem is made more complicated by the fact that each of the reference systems is made up of spatial units. In the case of administrative areas, they include units such as unitary authorities, districts and wards and in postal reference systems, there are postal areas, districts, sectors and units. Indirect spatial referencing is fine for tabulation (such as health authority league tables published in the newspapers) but users of such information often need to do two things with it

• Tabulate or aggregate the information on a spatial base that is different from the one on which it was collected, or cross-refer between two sets of data collected on different systems. For example comparing data by postcode is not always very meaningful, but presenting the same data by ward, constituency or school catchment area can transform them into powerful information that is relevant to millions of people. Seemingly uninteresting data can be turned into something that people can use for life-changing decisions.
• Attach the data to a base map. Tables are precise, but do not allow the reader to grasp spatial relationships very well. Shaded (choropleth) maps communicate new insights into physical, social and economic patterns. They also enliven the presentation of statistics and simplify their comparison. In the information age, geographical presentation of detailed and dynamic data should be increasingly feasible for everyone, and it is now becoming expected.

Although there are many base-geographies, there is an infinite number of statistics and other information that can be poured into them. These statistics originate from a vast range of sources and it is therefore crucial that statistics from an organisation can be readily and accurately linked to geographies provided from another organisation. Linking them is a trivial matter if the common key is identical in all respects.

There is however a major problem, which is rooted deeply in the way that the standard geographic bases for the UK are established: for good reasons they are not permanently fixed in terms of names, codes or positions of boundaries and their position changes over time (see Figure 2). As a consequence, linking indirect spatial references by the two methods mentioned above is fraught with pitfalls for the unwary which can introduce mismatches and imprecise matches. Whilst these problems are merely tiresome when dealing with only a handful of spatial units such as counties, they can become insurmountable when dealing with statistics for thousands of small areas such as wards or over a million of areas such as unit postcodes.

The UK Standard Geographic Base (UKSGB) aims to provide a core set of units that can be used to relate with each other and with specialised referencing systems. The UKSGB currently covers the administrative, postal, electoral and census geographies. Each of these geographies is a hierarchy built up from a jigsaw puzzle of small areas.

• Electoral geography comprises wards (known as divisions in Wales and the Isle of Wight) for the election of local councillors. Wards and divisions are grouped together to form parliamentary constituencies (national and European) for the election of MPs and MEPs.
• Administrative geographies for local and national government, health, education, etc. are built up from electoral wards/divisions.
• Postal geography is built up from unit postcodes, each of which is a cluster of delivery points.
• Census geography based on 1991 Enumeration Districts.

Each of these small areas has a well-defined boundary (except the Unit Postcode) and is assigned a unique code. Sometimes a name is assigned as well.

In order to help the development of the Data Integrator service, two underlying initiatives have been developed

• comprehensive best practice guidelines. In essence best practice is to learn from the experience of others. Adopting best practice helps organisations become more efficient and effective. Shrugging off the burdens of inefficiency, redundancy, error and inaccuracy allows resources to be redeployed thus enabling your organisation to achieve better performance. Whilst best practice is about what to do, an important element of best practice is improved understanding of why things are best done in a certain way. Consequently the best practice guidelines explains what the UKSGB is and why it has certain characteristics.
• A British Standard (BS7975) for the UK Geographic Base is in preparation. This British Standard will provide a basis for describing and referencing all geographic information for the United Kingdom in a consistent manner. It will define terms for identifying, labelling and positioning instances of geographic objects. It will also define the concepts of geographic referencing and introduce the concept of geographic bases, which are aggregations of geographic information relevant to a given theme or purpose. The use of geographic information will be enhanced if there is a common basis for describing it, referencing it, structuring it and combining it through the use of this standard. The standard will consequently promote the connectivity of geographic information from a wide variety of sources and formats and the creation of geographic bases for a number of purposes. NGDF is working with others to create this standard; The Department of Trade and Industry is funding the drafting of the standard; the British Standards Institute will be responsible for ratifying the standard; and those in the geographic information community will participate in the public consultation period during Autumn 2000.

The Data Integrator service brings the initiative into the public domain for the first time. It defines a set of “core spatial units” - commonly used geographic units in the UK - and information about them. It also provides information about the key geographic data products that are available to support the referencing and analysis of these units, with contact points for each.

How does NGDF compare with other NGDI’s?
In 1998 NGDF commissioned a study to look at other NGDI’s and seek lessons for NGDF in the future (Masser, 1998). This interesting study outlines a comparative methodology. From this come several key differentiators of NDGI’s. These are

• Geographical and Institutional Context
• Driving Forces
• Status
• Scope
• Access to public information
• Resources

The NGDF differs significantly from other NGDI’s in most of these areas and some of these differences are rooted in national differences in government structure, culture, the existing role of the private sector and the status quo:

Rackham & Rhind (1998) also considered the position of NGDF in relation to other NDGI’s and found that

• “the British organisation attempts to include the public and private sector within its decision-making Board (the Canadian and Dutch SDIs have separate structures and the US one has a primary focus at federal government level);
• the Dutch and US organisations have the greatest level of formal political support (and hence some centrally-contributed financing);
• the UK initiative is intermediate between the ‘can do’ approach in Canada and the committee-rich structure in the USA (see, for instance Federal Geographic Data Committee (FGDC) 1998 which states: ‘One of the FGDC’s main roles is to help implement standards that define complex geospatial data types that are broadly accepted and used by professionals in the federal, state local and private sectors...In order to carry out this role, much of the FGDC’s day to day work involves coordinating committees and working groups.’);
• the NGDF seems much more explicitly underpinned by the need to foster information services and facilitate business, rather than by a concern for data per se, than do some other initiatives.’’

The future
The National Geospatial Data Framework has been subject to many changes in direction and function during its five-year life. This shows an organisation that is prepared to change with emerging technology and new thinking. This is a continuing process.

During the latter part of 2000, the Board asked for a strategic review of the future of NGDF. The conclusion of this was that the NGDF should cease to exist as an entity during 2001. The NGDF activities would be delivered by alternative methods

• the objectives and operation of AGI and NGDF will converge, [Note: The AGI is a not for profit, independent organisation that is the national centre for GI. It represents the broad interests of the GI community in the UK.];
• AGI will adopt all the existing NGDF operations such as askGIraffe; and
• a strategic alliance will be formed to deliver NGDF’s ‘joined up geography’ objectives.

This new arrangement will present a clearer, more focused GI image to those inside and outside the GI industry and will allow ownership of the initiative to be taken by the wider GI community in the future.

References

• Chorley R, 1987, Handling geographic information, Committee of Enquiry into the Handling of Geographic Information, HMSO, ISBN 0117520152
  
  
• Elliott L, 2000, Unlocking Geospatial Information? – askGIraffe for the key’, AGI 2000 Conference Paper, Association of Geographic Information
  
  
• Hobman E, 1997, The National Geospatial Data Framework : From Concept to Reality, AGI 1997 Conference Paper, Association of Geographic Information
  
  
• Masser I, 1998, An International Overview of Geospatial Information Infrastructures: Lessons to be learnt for the NGDF, NGDF Paper
  
  
• Nansen B, Smith NS, & Davey A, 1995, What is the British National Geospatial Database?, AGI 1995 Conference Paper, Association of Geographic Information
  
  
• Rackham L & Rhind D, 1998, Establishing the UK National Geospatial Data Framework, Presented at SDI, Ottawa, Canada, June 1998
  
  
• Rhind D, 1997, Overview of the National Geospatial Data Framework, AGI 1997 Conference Paper, Association of Geographic Information