SDI Explained

Chapter Two

Spatial Data Infrastructure (SDI) Explained

Introduction

Spatial Data Infrastructure (SDI), also termed Geospatial Data Infrastructure (GDI), is essentially the enabling environment, that supports easy access to, and utilization of, geographical data and information, thereby ensuring the inclusion of all members of society in decision-making based on spatial information. Even more succinctly, we may define SDI as the mechanisms for efficient production, management, dissemination and use of geospatial information. A broad overview of the components of SDI is provided in this chapter.

While most people who are reading this guide are likely to have an understanding of SDI, it may often be necessary to explain the concept to others, who may not have a technical background. These could include managers who need to be persuaded to understand the advantages of spatial considerations in development with a view the to changing the way that things are done, or those who will make decisions concerning the allocation of resources to SDI. This chapter then aims also to provide explanations that may be useful in communicating SDI and related concepts to people who have not had hands-on experience in managing and manipulating digital geographic information.

Explaining the concepts underlying the rationale for spatial data infrastructure

Before explaining what SDI is to someone who has not had exposure to the use of digital geographic information, it may first be necessary to explain what is meant by geospatial information or GIS, in such a way that the rationale for SDI is understood.

Box 1: Misconceptions regarding geographic information

Misconceptions abound concerning what actually constitutes geographic information. This lies at the heart of a statement made by a postal service official who said " We have a list of addresses to which mail goes in a particular area. We don't need a GIS." Underlying this assertion is the misconception that the address itself tells one where something is. This is actually not the case: the postal workers know where the properties associated with each of the addresses are located, and hence are able to deliver mail to the correct location. Of course this may have been learnt from actual visits to the properties, rather than from having the luxury of being able to look this up on a paper or digital map, or even being provided with the co-ordinates of each property using a GPS.

What is geospatial information?

Perhaps the simplest explanation is that geospatial data or information tells one something about a location on earth. For example, a settlement has a location and occupies a definable area, within which there may be water sources, farming areas, schools, market places etc. Information about each of these features, e.g. the settlement's total population, what crops are produced in a farming area, is also considered geospatial information, as it is information about the location. The spatial relationships between these features within the settlement area can be readily assimilated when depicted on a map.

An understanding of the spatial relationships between features is valuable in guiding planning and development. The importance of having geographic information is illustrated by the role it played in directing responses to the Mozambique floods of 2000: the extent of the flooding was assessed through the use of remotely sensing images. The physical extent of the flood, as depicted in these images, was overlaid on an existing geographic database, in order to evaluate the extent of the damage and to focus humanitarian assistance.

One can provide information about a location either by using a co-ordinate system to define positions on the earth (technical jargon for this is "spatial referencing by (geographic) co-ordinates"), or by linking the information to named locations (technical jargon for this is "spatial referencing by geographic identifiers"), such as the name of a settlement, the position of which may in turn be defined through reference to a co-ordinate system.

Note that in practice several terms are used synonymously to denote geospatial information: these include spatial information, geographic information, geographically-reference information, or geo-information.

Why the talk of "geospatial information" instead of talking about "maps"?

In the past geographical information was mostly presented in the form of paper maps, with which most people are familiar. Increasingly today, geographic information is being captured in digital form and used through a Geographical Information System (GIS). This change has changed the conception of what Geographic information (GI) is and has introduced new challenges in handling GI.

What is a GIS?

A Geographic(al) Information System, or GIS, may be described as a computer system capable of assembling, storing, manipulating, and displaying geographically referenced information, i.e. data associated with particular locations. Practitioners often refer to the "total GIS" as including operating personnel and the data that go into the system. The way in which the digital geographic data are structured makes it possible to use a GIS to perform complex analysis.

Has the advent of digital GIS technology made dealing with spatial information more complicated than it was in the old days dealing with paper maps?

A note on both the challenges and possibilities brought about by the use of digital technology to capture and manipulate geographic data may also contribute to developing an understanding of why spatial data infrastructure is needed. Here perhaps an analogy with word processing may be helpful.

Before the advent of the word processor, inserting a sentence on the 3^rd page of a 10-page document captured using a typewriter was a major undertaking, often necessitating the retyping of most of the document. Word processors took the pain out of the editing of documents. Likewise, the editing or correcting of geographic data is made dramatically easier by being able to edit a digital database, rather than having to recreate a map. But using the word processor brought other advantages too: for instance, one can easily perform a search for occurrences of a particular word or phrase in a long document. And digital documents can be reused - cut-and-paste functionality allows one to construct "new" documents rapidly, through integrating portions of existing documents. The ability to merge documents from different sources also facilitates collaboration in creating documents, where various people might be tasked with putting together different parts of a document. All this applies too to the capturing of digital geographic data.

But it is precisely in the possibilities offered by the ability to reuse existing documents (or geographic data) and work collaboratively on a greater scale, that new challenges arise. Anyone who has had to put together a document from several documents, authored by different people, must have encountered the need to adjust the fonts, paragraph numbering etc. More subtle than these cosmetic changes needed to make the document appear coherent, and also more difficult to undertake, is ensuring that the terminology and the way in which terms have been used by the various authors, is harmonised throughout the new document. The various documents to be integrated may be in different file formats.

These kinds of issues are also encountered when digital geographic data from different sources is brought together: This is the challenge that has been brought by the greater availability of digital technology to manage geographic information.

The way to avoid these kinds of difficulties, associated with bringing together a variety of documents or data sources, is by anticipating the aspects that will need to be harmonised afterwards. The ideal situation involves obtaining an agreement before work begins, regarding how the authors (or data capturers) will construct a component of the document (or dataset).

Geographic datasets are in general far more complex, time-consuming and costly to collate, than the capturing of words electronically, using a word processor. Therefore harmonizing geographic data from a variety of sources is also then far more complex, costly and time-consuming, than adjusting styles in a document.

Although apparently complex, it should be emphasised that this kind of harmonization is achievable. Examples of this include the Country-at-a-glance initiative, undertaken in Ghana (see Box 2) and the integration of topographic data, cadastral data and demographic data derived from the 1996 census conducted in South Africa, in order to demarcate electoral wards and plan where voting stations would be established in the general election of 1999.

Box 2: The Ghana - Country- at-A-Glance

The Ghana - Country at A Glance (G-CAG) was developed as a synoptic, inter-operable, and geographical database at the equivalent mapping scale of 1:000 000, to assist in national-level environmental management and planning. It evolved as a logical extension of the Environmental Information System Development (EISD) component of the Ghana Environmental Resource Management Project (GERMP).

One of the main aims of the initiative was to use it as an introduction to the detailed data sets that are available at the custodian organisations. A major task was to harmonise the various data sets data to a general standard, yielding a homogeneous output. Apart from having a higher geographical resolution, the original data sets also contained more complex information. It was also a way to formalise and standardise data format and distribution processes. The CAG may be considered as a reference for existing databases in the country, their contents and where to obtain them. This will prevent organisations from re-creating already existing data sets and promote inter-organisational co-operation.

The G-CAG database was designed to harmonise identified data layers and features required to conduct environmental analysis. Various institutions within the EISD framework, with the appropriate mandates, had produced each of the various types of information independently. The main part of the data was derived from detailed 1:250,000 databases generated by these institutions. Important data sets from other sources are also included. The database contains 51 geographically referenced and harmonised data sets covering 11 geographical themes.

From a practical project management standpoint, in terms of time and resources, the actual data manipulation was undertaken by one designated organisation with the capacity and skill sets required for such an undertaking. However, the various stakeholders first agreed upon the broad principles that would govern the process, and strategic as well as decisions on approach. All information was thoroughly checked by the custodians and approved before inclusion in the final database. Original input data was either derived directly from databases available at the custodian organisations or from international data sets. In the latter case the data was validated and approved by a national organisation having specialist competence in that particular field before inclusion.

The advantage of having a digital (geographic) dataset, as discussed above, can be summarized as followings:

Easy storage
Easy dissemination
The facilitation of data exchange/sharing
Faster and easier updating and correcting information
The ability to integrate data from multiple sources and
The customisation of products and services.

While advantageous to adopt, the usage of this new technology poses new challenges to the user community. For example, from the view point of the data producers they are now required to provide more detailed metadata (see below for an explanation of metadata). The end user is now required to have the technical knowledge necessary to assess from the metadata, how appropriate the data set is for his or her own use.

Explaining SDI

What is "SDI"?

There are numerous definitions available for SDI; please refer to Appendix 2.1 for a listing of some definitions which have been used. The fact that there are so many views or definitions is an indication that there is not a universal understanding of exactly what SDI entails, which in turn is rooted in the fact that different countries, or even different sectors within a country, may have differing needs. Consequently, the motivation for SDI development may vary from country to country (see examples of this in Chapter 6, Getting Started). Note that some of these definitions emphasise various components of SDI, while others place emphasis on the purpose of developing SDI. However, the gist of all these definitions of SDI comes down to the fact that SDI is the framework of elements/factors that are needed by a community, in order to make effective use of spatial or geographic data.

But these definitions, or even a list of the components of SDI, will not necessarily resonate with someone who has not had practical experience in assembling spatial information, in order to address a particular problem. It may be helpful to use other ways and examples to explain SDI.

One might employ a story or scenario, effectively providing an "operational definition" of what is meant by SDI.

Using a scenario related to a topical issue could be particularly persuasive, as could be using a past event, where difficulties might have been encountered in assembling information necessary to solve a particular problem.

As an example, linked to the illustration of how geographic information was used following the floods in Mozambique, one might stimulate thought using the "story" in Box 3.

The development and use of an analogy with some other kind of infrastructure may also be helpful.

This may assist both in explaining the notion of SDI, as well as the need for coordinated development of and investment in SDI at a national level.

The explanation rests on an understanding of what is meant by infrastructure; a definition of infrastructure is as follows: the basic systems and services, such as transport and power supplies, that a country or organization uses in order to work effectively (from the Cambridge Advanced Learner's Dictionary). Other infrastructures often referred to include the health infrastructure, educational infrastructure or telecommunication infrastructure. Spatial data infrastructure can be seen as an infrastructure in the same sense: just as the ability to access and use the road network is necessary for undertaking a variety of economic activities, so too is the ability to access and use geospatial information necessary to plan and work effectively.

In general, there is considerable investment by government in developing these infrastructures, and the country's government will put in place measures to ensure coordination in the continual development of the infrastructure, which is likely to involve many players, simply because of the scale of the development required.

In general too, the development of many national infrastructures required interventions to bind infrastructures, which evolved independently on a smaller scale, into a single connected, coherent infrastructure. An example of this is the simultaneous development of railway lines with different gauges by different companies connecting various centres: to exploit the railway infrastructure optimally required that the gauges be standardized. Further, the "owners" of the various portions of railway infrastructure had to come to agreements on the use of "their" infrastructure by other service providers, and even details like timetables for use of the lines had to be agreed upon.

Another angle on explaining SDI, is to cast it in the light of effective management of resources.

Considerable investment may go in to the gathering of information, which implies that information is a resource which needs to be looked after, in much the same way as other large capital investments need to be maintained, e.g. a bridge that has been constructed, in order to ensure continued use, to provide value commensurate with the initial expenditure. SDI thinking goes about ensuring that the cost-benefit analysis associated with creating an information resources in the first place was carefully thought through, as well as about having plans in place to ensure that the information resources continues to be useable and useful.

In economic terms, one might also describe the impact of SDI as to reduce the transactional costs associated with the use of geospatial information.

Unless geospatial information is readily available in a format suitable for immediate use, there may be significant costs associated with obtaining it (consider the time spent in locating data, and a possible cost associated with the delay in obtaining it) or manipulation to get it into a form in which it can be used. A coherent SDI reduces these transactional costs, thereby contributing to efficiency.

"Why talk about SDI when we simply need data?"

There are times when one might encounter a push to create a centralized one-size-fits-all spatial database or "databank", to "solve" all the information needs of a country. To counter this it may be helpful to point out that the existence of geographic data and information does not alone ensure that it is used in decision-making and rational choices regarding the allocation of resources. Several factors come in to play, if information is to be used and reused:

To be used, people need to know that the data exist, and where to obtain it.
Then, they need to be permitted to access and use the data.
Further, they need to know something of the history of the data capture, in order to interpret it correctly, trust it and be able to integrate it meaningfully with data from other sources.
One may even depend on certain other data sets, in order to make sense of data, e.g. the listing of the population of various municipalities will be of limited use, unless one also knows where the municipal boundaries are.

Components of Spatial Data Infrastructure

As mentioned above, several factors determine a country's (or region's) ability to make effective use of available spatial or geographic information, namely:

Clearly defined core (or base) spatial data sets,
The adherence of geographic datasets to known and accepted standards,
Accessible documentation about existing geo-information (metadata),
Policies and practices which promote the exchange and reuse of information, as well as
Adequate human and technical resources to collect, maintain, manipulate and distribute geo-information.

These elements of SDI are elaborated on in the sub-paragraphs that follow. These sections also incorporate analogies, which once again may prove useful in providing explanations to people without a strong geographic information background.

Geospatial data development - building data for multiple uses

What data do we mean here?

Data sets, which may be used for many different purposes and in many different applications, are often referred to as base data, core data, fundamental data or reference data. A discussion on the distinctions sometimes made between reference, core, foundation and framework datasets may be found in chapter 2 of the SDI Cookbook.

This commonly used data would not in general require specialist subject knowledge of the field. For example, a dataset describing roads could be relevant to both disaster response applications, as well as the planning of where a new school should be located, while the principle user might be agency directly involved in maintaining and developing the road infrastructure (and most likely, developing and maintaining the road data set).

How does one know what data sets are core data sets?

This links to chapter 3 of this guide, namely identifying data needs. In undertaking a data needs assessment, certain data sets will emerge as being widely needed, for a variety of purposes, by many agencies. These then are the core data sets. It makes sense to prioritise their development, because they will be used widely.

To give an analogy, the letters A to Z can be regarded as core datasets of English language, which can be combined and re-used many times to provide different words, following standard spelling rules.

For example, in Nigeria's (draft) National Geoinformation Policy (see Appendix 2.2), the following fundamental or core datasets are identified:

a. Geodetic control database

b. Topographic database/DEM (at the scale of 1:50000 pending availability of 1:25000 national coverage)

c. Digital imagery and image maps

d. Administrative boundaries' data

e. Cadastral databases

f. Transportation (roads, inland water ways, railways, etc.) data

g. Hydrographic (rivers, lakes, etc.) data

h. Land use/land cover data

i. Geological database

j. Demographic database

The policy also states that this list of fundamental datasets will periodically be revisited, in order to make adjustments if necessary, in accordance with evolving national needs.

Who should develop a particular data set?

Even if there is agreement beforehand between a number of agencies, that a particular data set is needed by all of them, and that they will in fact share this data, the responsibility for development - and maintenance - of the data set needs to reside with a particular agency or organisation, the data custodian. The ideal would be to assign this responsibility to an agency, which is absolutely dependent on this data for its operations, and which could generate this data, as part of its business process. This means that it is likely that the agency will prioritise the development and updating of this data.

For example, a study conducted in Uganda in 2001 lead to concrete recommendations regarding the custodians of certain datasets, even though these agencies might not undertake the actual data capturing themselves. For instance, it is recommended that the Ministry of Local Government be the custodian for datasets on administrative units, while the Forest Department take responsibility for data pertaining to protected areas.

How does one ensure that many different users can use the data developed?

There are a number of factors that contribute to the possibility of multiple uses of the data.

Consultation with potential data users, prior to data development, can ensure that data is developed which will meet their requirements.
Standardization of the data developed, is basic to its correct interpretation and the integration of data from various sources. An analogy here is the ability to put various parts together, which may have been machined in different places, to assemble a car, or even simply to replace a part with another, and still have a working vehicle (or better still, have a vehicle in better working order than before). The key here is that the parts have been manufactured to comply with certain standards.
Metadata (information about the data - see below) will of course be needed.
Perhaps almost too obvious to be taken into account at times, is the fact that there has to be a way of distributing, or providing access to, the data, to all parties who would like to do this e.g. can access be provided online, through a Web Mapping Service? Or can the data be transferred via the Internet (e.g. ftp)? Or, is it possible to transfer this via CD-ROM?

Note the point that the factors listed above have a remarkable correspondence with the components needed for SDI - not an accident at all.

Geospatial information standards and standardization

It is likely that any organisation will encounter a need to obtain - or wish to share - information beyond their current information community, at some point in the future. As mentioned above, the ability or lack of ability to do this easily, or at all, depends partly on the nature of the datasets. Through standardization, one facilitates the use of a wider range of data. In developing standards for geographic data, one should look beyond the immediate information community of which one forms part, to standards in place or in development in other sectors, neighbouring countries or even regions.

How are standards developed?

The development of formal standards through national standards bodies as well as through international standards organisations (e.g. ISO and IEC - see boxes 5a and 5b) is achieved through a consultative process, generally requiring the honing of consensus on the nature of the standard under development. At a sub-regional level, there is an initiative towards sub-regional standardization being taken through SADCSTAN (see http://www.sadcstan.co.za, checked 28 October 2003), set up in terms of a Memorandum of Understanding signed by Ministers of Trade and Industry for SADC countries. The national standards bodies are members of SADCSTAN. Most often, stakeholders and role-players would constitute a committee and/or working group, to develop a standard or set of standards. Also built into both national and international standards systems, is the fact that a standard is not static, but there is an obligation to review all existing standards on a regular basis.

Informal standards also tend to evolve through a consensus process involving the players who stand to benefit most from adherence to a particular standard.

How are standards implemented?

The implementation of new standards may take some time, as there is a cost associated with implementation, and actual changes may need to be made to data or information adhering to "old" standards.

To encourage the adoption and implementation of standards, the process needs to be made as easy as possible. For example, the supply of software, which "forces" adherence to a standard, can accelerate the uptake of a standard. The best example here relates to the widespread adoption of the FGDC's Content Standard for Digital Geospatial Metadata (see http://www.fgdc.gov/metadata/metadata.html, checked 28 October 2003): the main driver for this was the availability of a free, easy-to-use capturing tools. Another example is the widespread use of ZIP software for compressing files.

Box 3a: About ISO's development of standards

ISO denotes the International Organisation for Standardization, based in Geneva. International Standards are developed through a consultative process involving its members, which are the standards bodies of various countries. Other organisations (e.g. international scientific organisations, UN bodies) may join as liaisons. The development of standards in particular areas is the work of a Technical Committee (abbreviated to TC). In the case of geographic information (or geomatics), the TC is TC 211. All standards numbered, e.g. ISO 9000 series and ISO 14000 series are well known. Standards pertaining to geographic information will fall in the range 19100 to 19199., and are hence referred to as the ISO 19100 series (or family) of standards. The home page of TC 211 is http://www.isotc211.org , although the documents relating to standards under development are accessible only to members of the TC. African countries which are members of TC 211 at present (March 2003) are: Mauritius, Morocco, South Africa, Tanzania and Zimbabwe.

What kind of standards does one need to implement?

Increasingly, the way in which the data is stored, which may be software dependent, is no longer a major stumbling block to the sharing and integration of geographic data. More important is having an understanding of what the data represents, and how it does this. For example, unless there is a standard understanding of what is meant by "forest", there may be a misinterpretation of land cover data, as an area labelled as covered by "forest" may to someone else appear to be covered by "shrub land". Some crucial aspects to look at include the following:

Geographic referencing: in order to be able to bring together (technical jargon often used for this is "overlay") different datasets, which cover the same (or adjoining) areas, one needs to know how the position of features has been defined, that is, one needs to know the projection and datum, and details of the co-ordinate system, to ensure the correct spatial relationships between features in different datasets. The AFREF project aims to develop standardised spatial referencing systems for Africa (see Appendix 2.3)

Box 3b: About OGC's development of standards

OGC, or the Open GIS Consortium, is primarily a grouping of industry partners, developing specifications for geographic information. Several different membership options are available for organisations wishing to join and participate in OGC. At present, there are no distinctly African members of OGC. OGC's home page is at http://www.opengis.org/ , and provides information on their programme of work, products which claim conformance to OGC specifications and the specifications themselves, once they are finalised. There is a close relationship between OGC and ISO/ TC 211, resulting in an effective joint development of certain standards.

The data content: what features are included in the dataset, how are these defined, and what is the relationship between them? A data dictionary (or feature catalogue), which accompanies a dataset, may ensure that the data is correctly understood, but unless the features are standard, it will not necessarily enable meaningful results to be obtained in combination with another data set.
The resolution or scale of the geographic data: in general, only datasets of comparable scale or resolution may be combined for the purposes of analysis.
Metadata, or data about data: all the above might be carried in the documentation about a dataset, but for ease of understanding and comparability, this information is recorded, i.e. the metadata, should be recorded in a standard way.

Metadata - describing geospatial data

Why is metadata needed?

The recording of metadata, or data about data, serves a number of purposes. Information about a dataset may be necessary in order to

locate appropriate data,
evaluate whether the dataset meets one's requirements,
extract the relevant data and
actually make full use of the data in an application.

There is a helpful discussion on metadata in Chapter 3 of the SDI Cookbook. There are many useful references on the benefits of recording metadata, to both producers and users of the associated data.

What metadata is needed?

Different information about the dataset is needed to support each of the above. Again an analogy may be helpful: for example, a few simple characteristics of a book (the title, author, year and place of publishing etc) may be recorded in a library's catalogue, to facilitate locating and obtaining a particular book. The dust-jacket of a book itself often contains more information on the content of the book as well as information about the author, which is useful in order to evaluate whether the book is suited to the would-be reader's requirements.

How should the metadata be structured?

Mention has already been made of the fact that it is useful to record metadata in a standard way, to enable a potential user to make a more rapid evaluation of whether the dataset will meet his or her needs, that is, there is need for the content of a metadata record to be standardized. Internationally, people who work with geographic datasets have been at the forefront of developing standardized metadata content.

However, metadata needs also to be structured in a way that supports automated indexing, searching and retrieval of information, if it is to be made accessible through digital catalogues on the Internet. This is most often implemented through the provision of a standard metadata capturing tool.

Examples of metadata pertaining to a spatial datasets may be found in Appendix 2.4.

Cataloguing geospatial data

Why catalogue geospatial data?

The capturing of metadata relating to geospatial datasets is necessary, but not sufficient on its own, to ensure wider knowledge of a dataset, and hence wider usage. This metadata needs to be made available to potential users, together with search facilities, which enable a user to identify the datasets that most closely match their requirements.

What is meant by a "distributed catalogue"?

There are many producers of datasets, and once they have captured the metadata relating to datasets they create and/or maintain, the metadata collected needs to be accessible to a potential user. However, someone looking for data would want a "one-stop-shop", that is, they would rather not have to look in many different places for metadata. This is made possible with a "distributed catalogue", which makes it possible for a user to query collections of metadata, which reside on many different servers. This means that the publishers of metadata can maintain and post metadata to their own server, rather than having to transfer records to a server running the catalogue service. From the FGDC Clearinghouse in the USA, for example, it is possible to access metadata records on servers in Ethiopia, Kenya, Senegal and South Africa. A trivial analogy is provided by Internet search engines such as Yahoo or Google, which direct the searcher to Web-pages according to their search criteria, which are housed on servers all over the world.

Providing access to geospatial data via the web

What is the purpose of Web mapping?

It is extremely useful to be able to see geospatial data portrayed in the form of a map. If one has the appropriate GIS software, it is possible to obtain the dataset from the producer and map the information. With Web mapping, it is not necessary to have to obtain a data set and use own software to portray this as a map, beyond an internet browser. If through a catalogue one locates a dataset of interest, this may also be viewed over the Web. Note that often merely being able to view geospatial data in the form of a map may be all that is required in order to plan or make a decision. This greatly increases the number of potential users of geospatial data, as this group is no longer limited to those who have the relevant GIS software and expertise to be able to manipulate digital geospatial datasets.

How can maps be provided through the Web?

Many software products are available to publish geospatial data in the form of maps through the Web. A significant contribution of OGC has been to define specifications for web mapping interfaces. This has opened to way for the visual overlay of geographic information residing on different servers. Examples of this in action may be found at http://clearinghouse5.fgdc.gov/multiviewer/viewer.php?type=africa (checked 28 October 2003).

Data policies and legislation

Why are data policies and legislation important?

Many readers will have had first-hand experience of someone refusing to share data, where technology was certainly not a barrier. Various explanations for the refusal might be offered, but these would often come down to either an explicit restriction on providing the data to other parties, or to an absence of policy relating to provision of the data altogether.

What kinds of policies are relevant to SDI?

A wide range of policy may impact on the ability to use geospatial data. These include:

Policy or legislation relating to the right (or otherwise) to access information: sometimes countries have legislation that defines the rights people have to obtain information held by both public and private sector bodies (e.g. South Africa's Promotion of Access to Information Act, Act 2 of 2000). This is obviously a factor, which promotes the interchange and distribution of geospatial information.
Pricing policies: pricing policies may provide for a low or negligible cost associated with the acquiring of geospatial data which has been captured using public funding, or for full or partial cost recovery. Higher prices of data are likely to limit its distribution, but the absence of a homogenous policy in relation to cost recovery by public agencies can inhibit the flow of information even more.
Policy relating to the use of spatial data: the position regarding ownership of copyright on geospatial data, as well as liability in relation to decisions taken on the basis of geospatial information, may also affect the use and reuse of geospatial datasets.
Legislation and policy relating to other areas may have implications for SDI development: this is perhaps best demonstrated by examples. For instance, in many countries, legislation relating to the obligation to undertake Environmental Impact Assessments relating to developments in planning implies a need for the availability and use of geospatial information. In Uganda, the obligation to produce a State of the Environment Report at regular intervals implies a need for certain geospatial information to be made available. South Africa's Local Government Municipal Demarcation Act (see Appendix 2.5) requires bodies to provide to the Municipal Demarcation Board information needed for making rational decisions regarding the boundaries of local authorities.

Chapter 5 of this guide provides a more in-depth study of the policy element of SDI and the development of policy.

Partnership and leadership

Why is partnership an important component of developing SDI?

As a single agency is unlikely to have all the resources, or even skills and knowledge required to undertake the development of all aspects of SDI, the partnership of agencies and organisations is called for. Not only does the establishment of a partnership of organisations working together to create SDI mean that a greater amount and wider range of resources can be brought to bear on its development, but having organisations working together at the outset, is vital to ensuring that SDI develops in a way that will support all the partners in their use of data. It may be appropriate to involve both public and private partners, as well as academia and individual experts in a consortium approach to developing the SDI needed by a country. An example of a public/private sector partnership is the development of the 1:50000 Digital Topographic Maps in Kenya. In this case the National Mapping Agency (Survey of Kenya) and World Agro Forestry Centre (ICRAF) pooled resources in the collection of the data required to produce the 1:50 000 map.

The importance of partnerships in developing SDI is sufficiently important to merit an entire chapter in this guide - see chapter 4 on the institutional framework for developing SDI.

What kind of leadership is needed?

While it is argued above that the co-operation of many partners is needed to achieve SDI, it is also important to ensure that activities to develop aspects of SDI remain co-ordinated and focussed. An overarching vision or goal to which all partners subscribe is important, such as Senegal's Plan Géomatique National. The designation of a lead agency from among the partners, with dedicated resources to be able to provide co-ordinating mechanisms, is likely to expedite the development of SDI. In the case of Nigeria, the (Draft) National Geoinformation Policy designates the National Space Research and Development Agency (NARSDA) as the lead agency, which will co-ordinate the activities of a National Geospatial Data Infrastructure committee. An additional leadership role is to keep partners inspired, and to promote continuously the vision or goal of SDI development activities.