GISdevelopment.net --> Policy --> Geographic Information Infrastructure
Introduction
At the most senior levels of government, high expectations have been expressed about the beneficial effects of the 'information society' or 'information economy', for example, for the delivery of healthcare, transportation management, life-long learning, sustainable development , etc. The role of the private sector in providing the communications infrastructure and the so-called 'value-added' information services is emphasised strongly. Recognising that government itself is a very large source as well as user of such data, efficient and easy access to these sources becomes a high priority. These expressions of policy signal that facilitating access to government owned data results in increasing positive externalities by reducing transaction costs to society as whole. These expectations have been articulated in, for example, EC (1994), Executive Order 12906 (1994), and in EC (1998).
The vision presented in these policies emphasizes the benefits of a private publishing sector using access to government held data sets for those "value added" products and services. However, the evidence is that this sector is slow coming off the ground. See for example Lawrence (1998) or Longhorn (1998). More immediate benefits seem to be possible if governments would focus their efforts on increasing efficiency and effectiveness of its own information and data use. This can be done by means of promoting sharing of data and information through appropriate domain and enterprise Geospatial Data Infrastructures (GDI) between ministries and departments/agencies of government.
The idea being that if government has its own geospatial data development, supply and access in order it would, by example, encourage the usefulness and efficiencies of accepting related national GDI standardisation by other sectors in the economy.
This paper explores the concept of Geospatial Data Infrastructure in this context, discusses the limiting factors in the development and implementation of (national) data standards, reflects on the process of developing the NGDI and draws some conclusions about what will likely drive the development of NGDI in most countries.
What is GDI?
The notion of the sharing of existing data through Information Infrastructures emerges as a significant matter of efficiency, and as a generator of positive externalities. See for example Branscomb (1982).
We will use the term Geospatial Data Infrastructure (GDI), as opposed to the terms Geographic Information Infrastructure or Spatial Information Infrastructure. Furthermore we like to follow Groot and McLaughlin (2000), introducing the related concept of a Geospatial Data Service Centre (GDSC), for enterprise-wide or domain-oriented activities.
The same authors define the purpose of GDI as: to facilitate access to and responsible use of geospatial data at affordable prices. National Geospatial Data Infrastructure, Regional Geospatial Data Infrastructure, and Global Geospatial Data Infrastructure are special cases to be defined in terms of what gives the GDI its regional, national or global character. GDI is seen in this respect as a generalised concept, which can be implemented at the enterprise level, the level of broad application domains such as coastal zone management, urban management or physical planning, or in the national or regional context.
Geospatial Data Infrastructure encompasses the networked geospatial databases and data handling facilities, the complex of institutional, organizational, technological, human, and economic resources which interact with one another and underpin the design, implementation and maintenance of mechanisms facilitating the sharing, access to, and responsible use of geospatial data at an affordable cost for a specific application domain or enterprise.
Fig. 1 shows the idea of GDI for the environment and physical planning department (as example) in a major municipality. This forms then the application domain Environment and Physical Planning.
On the right hand side are the individual applications within the domain with their GIS systems, which all need routine supply of directly applicable data. This stream of requirements is being met through a Geospatial Data Service Centre (GDSC). The GSDC harmonises / standardises all data for its application domain. It ensures they are described in a (national) meta data standard to facilitate the sharing of these resources in the domain and amongst other potential users.
Fig. 1: The foundation data for the GDI in relation to the Geospatial Data Service Centre.
The GDSC also enforces the information policies that control access, use and pricing, in keeping with legislation and overall government policy or enterprise regulations. It is fully accountable for the total integrity of the infrastructure. It is staffed by a cross section of technical, administrative, legal professionals as well as specialists in Geoinformatics who understand the language and content of the application domain. The latter are essential for the data model and standards development in a manner, which promotes the broadest possible use, and thus sharing, of data resources within the domain.
There is no a priori theoretical classification of Application Domains. Usually the manager of this Environment and Physical Planning department sees the need to apply Information and Communication Technology, such as GIS, to improve operational efficiency and effectiveness. GI Systems are acquired and a number of GIS applications are implemented, typically independently of each other. Next the manager begins to discover that these applications draw heavily on the budget and cost effectiveness remains low.
Further investigation by the manager indicates that in all likelihood low cost effectiveness is, in large part, caused by the excessive proportion of time being spent trying to find or develop data and information relative to its actual operational use. Studies in the oil and gas sector carried out in Western Canada in the late 1980s and early 1990s indicate that exploration geologists and - geophysicists spent on the whole more than 60 % of their time searching for data and only about 20 % doing something useful with that. The major oil and gas companies in Western Canada jointly created the Canadian Oil and Gas GIS (Canoggis) in the early 1990s to address this problem. Canoggis is an enterprise or domain Geospatial Data Infrastructure for the Western Canadian Oil and Gas industry along the lines described in Fig. 1.
The effects of this GDI did not take long to emerge. Within three years the cost of access to data had been reduced by a factor of 10! Furthermore the number of participants that started the Canoggis GDI expanded from less than 10 to approximately 50 within three years. This seems evidence that it is better to start up a clearly focussed application domain GDI with a limited number of participants and really make that operational, than to try to convince a much larger group of potential users of the acceptance of the necessary standards and protocols required in the GDI. Apparently other potential users are much easier to accept the standards and protocols of something that demonstrably works.
Interesting in this development is that the oil and gas industry is notoriously secretive about its data and information resources. Yet it moved decisively in the direction of creating jointly an information infrastructure to support the sharing of a growing part of their data assets. They had understood that sharing knowledge and information/ data resources is economically interesting. High quality exploration staff having efficient access through the sharing of the data resources, rather than continuing excessive secrecy better serve their competitive edge (see http://www.geoconnections.org/iacg/gis/ind/n235n.htm ).
Establishing such an infrastructure is complex, takes time, money and effort. Therefore, unless management feels the pressure to improve efficiency and effectiveness it will not be motivated to initiate the design and implementation of the Application Domain GDI. This is an unfortunate reality in establishing National Geospatial Data Infrastructure (NGDI).
In addition to the Canoggis case, there are a number of case histories that are worthwhile taking note of. For example North Carolina (USA) has a long history in building the geospatial data infrastructure concept. See Siderelis (2000). Its success can be attributed at least in part to the commitment of people at all levels in the state who exploited opportunities to create an information infrastructure; an emphasis on using the infrastructure for addressing critical issues facing the state; enthusiastic focus on multi-lateral collaboration; and the spin-off benefits of being located in a technologically progressive state where political support for technology initiatives is high.
Another interesting case describing the evolution of an international GIS initiative in the Baltic region towards a GDI is for example Langaas (1998).
Developing the National Geospatial Data Infrastructure (NGDI)
Governments need geospatial data, referenced and defined in the national context, to govern. Examples are in legislative and policy development, for the allocation and management of natural resources, for defence and public safety, in support of a variety of regulatory activities, and generally in promoting a better understanding of the physical, economic and human geography of the nation. To satisfy these requirements, the data is organised in a national co-ordinate system or "geodetic datum", while the data definitions are made consistent for the whole country. There is often a temporal dimension as well recording how some feature has changed over time.
What makes a GDI "National"? The National Geospatial Data Infrastructure seeks to support the sharing of data in the national context by means of a set of standards, such as: national spatial reference systems, a national topographic template, national elevation model, any other standardized spatial data set of national scope such as geographical names, administrative boundaries, certain thematic data sets (soils, hydrology, vegetation population, etc.), and meta data standards to describe in a consistent manner each of the GDI holdings. See Groot and McLaughlin(2000).
Many enterprise- and domain-oriented GDIs will have used foundation data, which adhere to these standards, and which are produced by National Mapping Agencies (NMAs). Linking these GDIs into a NGDI is not necessarily onerous as long as standardized data descriptions (metadata) are being applied. This requires the application of national meta standards - sometimes a time-consuming retrofitting exercise.
It is interesting to note that most managers of enterprise or application domain GDIs give little attention to the potential interest of other users for their application data and the commercial opportunities this may create for such data holders. This is understandable because distribution of domain application data is usually not part of their mandates. This is mainly the result of old distribution paradigms, which dictated difficult to justify investments in some analogue distribution facility. Jack Dangermond's initiative for the Geodata program (see http://www.geographynetwork.com) demonstrates that at relatively small investment it is now possible to distribute such data for a fee. It uses the internet facilities to let data holders make their data broadly (commercially) available. The metadata standard offered by ESRI is however at this point dealing with too highly generalised data to be useful in very specialised applications. However, for many less sophisticated applications it is very useful. Taken in this context it may be well considering investing in the application of, more detailed, national, metadata standards so that the data in an application domain can also be useful to more sophisticated users who may also be more prepared to pay sophisticated prices.
Market differentiation in the NGDI
Opening up accessibility to government and other geospatial data is accompanied by the emergence of an interesting market differentiation. Take for example access to the geospatial positioning infrastructure through the growing availability of GPS receivers of a wide range of prices and precision. In the past geospatial positioning was almost exclusively entrusted to well-trained or even certified surveyors and navigators. Now the public can buy at low cost a GPS receiver for applications not requiring high precision, or with more sophisticated instrumentation for car and other navigation. There still is a requirement for high precision GPS or even ground survey but that demands more expensive equipment and higher trained technicians and professionals. Hence an unexpectedly broad market has emerged for these low- cost instruments while a smaller market segment for high precision positioning continues. The GPS infrastructure and for higher accuracies the so called Active Control Systems which have replaced in many industrial countries the conventional triangulation network are the foundation for this expansion and popularisation of the positioning capabilities.
Similar developments can be expected with ubiquitous access to geospatial data from any source combined with low- cost GIS software available on- line through the web. There will be a kind of popularising or may be de-professionalising of use which is yet unpredictable. ESRI's Geodata project is beginning to demonstrate this phenomenon. Hence we should expect a growing "popular" demand while the market for professional engineering and public administration applications will continue to demand highly reliable, timely and up to date data and information services. The sequel of this paper will deal with the requirements and evolution of latter part of the NGDI. It will address the following issues:
Generally speaking the National Surveys are the relevant sources for the Framework Data. (See Fig. 1). A special subset of these data is the Foundation Data (FD) which is the fundamental geographical reference for all other thematic application data. NMAs are usually responsible to produce, maintain and distribute the FD.
Within the context of the efficiency and effectiveness of government itself and the previously mentioned political expectations, optimal performance of NMAs and responsiveness to their client community is obviously of paramount importance in reducing transaction costs in society. Since their inception (some, such as the Survey of India, as far as 400 years ago but most in the mid- 1800's) they have enjoyed a monopoly in the technological and industrial organisation of national surveying and mapping activity.
Beginning in the mid-1970s these monopolies were increasingly challenged, with the growing proliferation of Information Technology (IT). The surveying and mapping technology has become increasingly embedded in software and accessible to non-specialists in the NMAs client community. Furthermore the client community obtained growing access to substitute products for the standardised topographic bases or thematic framework data sets (for example from Remote Sensing ). Both challenge the monopoly of the NMAs. Furthermore, the client community is increasingly changing to include users interested in the digital data for application in Geographic Information Systems (GIS).
These effects of the ICT coincided with growing government deficits, which for most countries has led to a 20 year period of budget reductions, demands for revenue generation and, in the context of reform of government, privatisation. See for example Groot(1998). More importantly these radically changing circumstances led to critical review of the relevancy of existing and emergence of new mandates. See for example: Canadian Government (1986), Department of the Environment (1987), Mapping Sciences Committee. (1990), ANZLIC. (1996).
In spite of these studies NMAs all over the world have been slow in responding to these changes and many are experiencing major problems pushing through the restructuring necessary to move from automation to informatisation . One of the more perplexing problems NMAs have been faced with is the definition of the content and formats of the Foundation Data, especially the topographic component of this. At first, these organisations implemented the IT to make the existing production lines of their standard topographic map series more efficient. However, these are extremely complex products, time consuming and expensive to produce and keep up to date. Budgets and human resources could realistically speaking never be sufficient to assure both timeliness and currency. A completely new design concept is required to meet the new requirements. As early as 1986 the Ordnance Survey had been challenged by the National Joint Utilities, the Municipalities and the Land Registry to produce the national topo data base at large scales for the whole country by 2000. These major users had demonstrated that a highly simplified data model should replace the traditional map content model. The simplification would be useful in practice and would make completion of the database by the year 2000 also feasible. See Department of Environment (1987).
Meanwhile progress has been made in defining the least complex and topo data set that will serve as a geometric framework for a multitude of applications. In the UK the term topographic template has found acceptance. See for example Smith and Rhind (1999). The Netherlands is taking the definition of this dataset a step further by seeking to give it a legal authoritative status such as the data in the Kadaster or the Register of Persons or the Register of Companies. This would have far reaching consequences in all Government Ministries who will be compelled to use this data only for refernccing purposes. It also would have far reaching consequences for the National Mapping Agency, which would become a legal as well as natural monopoly, a status which has proven to be not conducive to client orientation and efficiency.
The latter is important because poor performance by NMAs leads to the client community, especially the GIS community, falling back on cheaper and often simpler but more up to date substitute products from whatever sources may be available. This in turn leads to the loss of the positive externalities, to costly duplication, reduced and incompatible accessibility of existing (government owned) data, increased transaction costs and possibly reduced timeliness in decision making processes. In many ways it impedes full exploitation of ICT for the benefit of society in geospatial data applications.
It is significant that in those countries where NMAs have successfully adapted to the dynamics of the ICT environment and the imperatives of government reform, the informatisation has been accompanied, and even led, by institutional and regulatory reform. The organisation was thereby placed in a regulatory environment in which both management and staff are motivated to be innovative and efficient and are rewarded for this. For a detailed analysis of the relationship of government goals with respect to Geospatial Information, the economic efficiency in pricing and distribution, and the regulatory position of the NMAs, see Groot (2001).
What can be done to advance the NGDI concept?
There are, worldwide a large number of NGDI initiatives underway. See for example Onsrud (1999). Yet the case history literature is still not very complete. Nevertheless one of the most comprehensively documented developments towards a GDI for land administration is probably in New Brunswick. See Finley (2000). While New Brunswick is a small, rural province in Canada and its experience will not be readily replicated elsewhere, it does provide some important lessons. For example, much of its success to date can be attributed to the following factors:
The essence of the GDI concept is that there is no master architect. See (McLaughlin and Groot(2000)). There cannot be, nor will there be, a single organization responsible for designing and implementing some kind of GDI blueprint, especially at the national (NGDI) level. Instead, we can imagine an almost organic web of partnerships and relationships evolving purposefully within a given jurisdiction. It will sometimes be pushed by technology, sometimes pulled by market requirements. But at some point there will be a sufficient inter-connectedness of databases, a level of access to the data and use of the data, as well as the maturing interest of stakeholders, to participate and invest in the partnerships required for a nascent NGDI to be recognized. Having said this, the evolution of any GDI concept will most likely emerge from a combination of 'top-down' and 'bottom-up' strategies, the specific mix of which will vary significantly from one jurisdiction to another.
The top-down approach will entail defining strategic goals, assessing priorities, developing implementation plans, and obtaining core funding. It will require some kind of institutional framework (lead agencies, steering committees, working groups for standards, etc., and monitoring arrangements). Examples of outputs which may be expected from the top-down approach are defining the fundamental geospatial data sets, building the clearinghouses, establishing metadata standards and access protocols, and resolving the information policy issues. Core funding for geospatial database development and networking, as well as for subsequent maintenance and upgrading, will also be key to defining the initial directions and priorities. This funding will invariably result from a combination of direct public investment and revenues derived from the sale of information products and services, as described in detail in Rhind (2000). While the debate continues over whether one should charge for public information, the reality is that some combination of direct investment and charging for products and services will be required.
The bottom-up approach will recognize the multiple local initiatives to build application-specific and enterprise-wide geospatial databases and will encourage their evolution towards a universal framework for accessing, combining and using the data through a process of proselytizing, providing incentives (especially shared financing) and imposing regulations (e.g., through standards regimes). Money talks and access to shared funding will be of special significance in advancing the NGDI agenda; but even more important will be the concrete evidence of the power and potential of an integrated infrastructure.
It is at least debatable if these conclusions will be as valid for a large and complex country such as India as they are for countries with a different culture of government, different relationships between levels of government, as well as between the governments at various levels and the private sector. But one of the foremost differences must lie in the information density of the Indian sub- continent which can only be dealt with at the local level.
Obviously we do not know the Indian context and conditions sufficiently well to suggest some observations about advancing the NGDI in India. Yet our sense is that considering the decentralizing and customizing character of the ICT, (one could almost say the anarchistic nature of these technologies), the role of the National Government should probably be relatively light. By this we mean that the rapid provision of the national foundation data sets and the definition of the hierarchy of a national metadata standard, and the development of Clearinghouse functions should have priority, developed on the basis of a business plan, as described earlier. Beyond that, the national role should probably be one of encouragement in decentralizing GIS applications in manageable application domains within the GDI concept and developing the top- down, bottom- up relationships that are necessary for that.
Bibliography
Advancing the concept National Geospatial Data Infrastructure: Reflections on the "bottom line"
Richard Groot and Yola Georgiadou
Richard Groot and Yola Georgiadou
Introduction
At the most senior levels of government, high expectations have been expressed about the beneficial effects of the 'information society' or 'information economy', for example, for the delivery of healthcare, transportation management, life-long learning, sustainable development , etc. The role of the private sector in providing the communications infrastructure and the so-called 'value-added' information services is emphasised strongly. Recognising that government itself is a very large source as well as user of such data, efficient and easy access to these sources becomes a high priority. These expressions of policy signal that facilitating access to government owned data results in increasing positive externalities by reducing transaction costs to society as whole. These expectations have been articulated in, for example, EC (1994), Executive Order 12906 (1994), and in EC (1998).
The vision presented in these policies emphasizes the benefits of a private publishing sector using access to government held data sets for those "value added" products and services. However, the evidence is that this sector is slow coming off the ground. See for example Lawrence (1998) or Longhorn (1998). More immediate benefits seem to be possible if governments would focus their efforts on increasing efficiency and effectiveness of its own information and data use. This can be done by means of promoting sharing of data and information through appropriate domain and enterprise Geospatial Data Infrastructures (GDI) between ministries and departments/agencies of government.
The idea being that if government has its own geospatial data development, supply and access in order it would, by example, encourage the usefulness and efficiencies of accepting related national GDI standardisation by other sectors in the economy.
This paper explores the concept of Geospatial Data Infrastructure in this context, discusses the limiting factors in the development and implementation of (national) data standards, reflects on the process of developing the NGDI and draws some conclusions about what will likely drive the development of NGDI in most countries.
What is GDI?
The notion of the sharing of existing data through Information Infrastructures emerges as a significant matter of efficiency, and as a generator of positive externalities. See for example Branscomb (1982).
We will use the term Geospatial Data Infrastructure (GDI), as opposed to the terms Geographic Information Infrastructure or Spatial Information Infrastructure. Furthermore we like to follow Groot and McLaughlin (2000), introducing the related concept of a Geospatial Data Service Centre (GDSC), for enterprise-wide or domain-oriented activities.
The same authors define the purpose of GDI as: to facilitate access to and responsible use of geospatial data at affordable prices. National Geospatial Data Infrastructure, Regional Geospatial Data Infrastructure, and Global Geospatial Data Infrastructure are special cases to be defined in terms of what gives the GDI its regional, national or global character. GDI is seen in this respect as a generalised concept, which can be implemented at the enterprise level, the level of broad application domains such as coastal zone management, urban management or physical planning, or in the national or regional context.
Geospatial Data Infrastructure encompasses the networked geospatial databases and data handling facilities, the complex of institutional, organizational, technological, human, and economic resources which interact with one another and underpin the design, implementation and maintenance of mechanisms facilitating the sharing, access to, and responsible use of geospatial data at an affordable cost for a specific application domain or enterprise.
Fig. 1 shows the idea of GDI for the environment and physical planning department (as example) in a major municipality. This forms then the application domain Environment and Physical Planning.
On the right hand side are the individual applications within the domain with their GIS systems, which all need routine supply of directly applicable data. This stream of requirements is being met through a Geospatial Data Service Centre (GDSC). The GSDC harmonises / standardises all data for its application domain. It ensures they are described in a (national) meta data standard to facilitate the sharing of these resources in the domain and amongst other potential users.
Fig. 1: The foundation data for the GDI in relation to the Geospatial Data Service Centre.
The GDSC also enforces the information policies that control access, use and pricing, in keeping with legislation and overall government policy or enterprise regulations. It is fully accountable for the total integrity of the infrastructure. It is staffed by a cross section of technical, administrative, legal professionals as well as specialists in Geoinformatics who understand the language and content of the application domain. The latter are essential for the data model and standards development in a manner, which promotes the broadest possible use, and thus sharing, of data resources within the domain.
There is no a priori theoretical classification of Application Domains. Usually the manager of this Environment and Physical Planning department sees the need to apply Information and Communication Technology, such as GIS, to improve operational efficiency and effectiveness. GI Systems are acquired and a number of GIS applications are implemented, typically independently of each other. Next the manager begins to discover that these applications draw heavily on the budget and cost effectiveness remains low.
Further investigation by the manager indicates that in all likelihood low cost effectiveness is, in large part, caused by the excessive proportion of time being spent trying to find or develop data and information relative to its actual operational use. Studies in the oil and gas sector carried out in Western Canada in the late 1980s and early 1990s indicate that exploration geologists and - geophysicists spent on the whole more than 60 % of their time searching for data and only about 20 % doing something useful with that. The major oil and gas companies in Western Canada jointly created the Canadian Oil and Gas GIS (Canoggis) in the early 1990s to address this problem. Canoggis is an enterprise or domain Geospatial Data Infrastructure for the Western Canadian Oil and Gas industry along the lines described in Fig. 1.
The effects of this GDI did not take long to emerge. Within three years the cost of access to data had been reduced by a factor of 10! Furthermore the number of participants that started the Canoggis GDI expanded from less than 10 to approximately 50 within three years. This seems evidence that it is better to start up a clearly focussed application domain GDI with a limited number of participants and really make that operational, than to try to convince a much larger group of potential users of the acceptance of the necessary standards and protocols required in the GDI. Apparently other potential users are much easier to accept the standards and protocols of something that demonstrably works.
Interesting in this development is that the oil and gas industry is notoriously secretive about its data and information resources. Yet it moved decisively in the direction of creating jointly an information infrastructure to support the sharing of a growing part of their data assets. They had understood that sharing knowledge and information/ data resources is economically interesting. High quality exploration staff having efficient access through the sharing of the data resources, rather than continuing excessive secrecy better serve their competitive edge (see http://www.geoconnections.org/iacg/gis/ind/n235n.htm ).
Establishing such an infrastructure is complex, takes time, money and effort. Therefore, unless management feels the pressure to improve efficiency and effectiveness it will not be motivated to initiate the design and implementation of the Application Domain GDI. This is an unfortunate reality in establishing National Geospatial Data Infrastructure (NGDI).
In addition to the Canoggis case, there are a number of case histories that are worthwhile taking note of. For example North Carolina (USA) has a long history in building the geospatial data infrastructure concept. See Siderelis (2000). Its success can be attributed at least in part to the commitment of people at all levels in the state who exploited opportunities to create an information infrastructure; an emphasis on using the infrastructure for addressing critical issues facing the state; enthusiastic focus on multi-lateral collaboration; and the spin-off benefits of being located in a technologically progressive state where political support for technology initiatives is high.
Another interesting case describing the evolution of an international GIS initiative in the Baltic region towards a GDI is for example Langaas (1998).
Developing the National Geospatial Data Infrastructure (NGDI)
Governments need geospatial data, referenced and defined in the national context, to govern. Examples are in legislative and policy development, for the allocation and management of natural resources, for defence and public safety, in support of a variety of regulatory activities, and generally in promoting a better understanding of the physical, economic and human geography of the nation. To satisfy these requirements, the data is organised in a national co-ordinate system or "geodetic datum", while the data definitions are made consistent for the whole country. There is often a temporal dimension as well recording how some feature has changed over time.
What makes a GDI "National"? The National Geospatial Data Infrastructure seeks to support the sharing of data in the national context by means of a set of standards, such as: national spatial reference systems, a national topographic template, national elevation model, any other standardized spatial data set of national scope such as geographical names, administrative boundaries, certain thematic data sets (soils, hydrology, vegetation population, etc.), and meta data standards to describe in a consistent manner each of the GDI holdings. See Groot and McLaughlin(2000).
Many enterprise- and domain-oriented GDIs will have used foundation data, which adhere to these standards, and which are produced by National Mapping Agencies (NMAs). Linking these GDIs into a NGDI is not necessarily onerous as long as standardized data descriptions (metadata) are being applied. This requires the application of national meta standards - sometimes a time-consuming retrofitting exercise.
It is interesting to note that most managers of enterprise or application domain GDIs give little attention to the potential interest of other users for their application data and the commercial opportunities this may create for such data holders. This is understandable because distribution of domain application data is usually not part of their mandates. This is mainly the result of old distribution paradigms, which dictated difficult to justify investments in some analogue distribution facility. Jack Dangermond's initiative for the Geodata program (see http://www.geographynetwork.com) demonstrates that at relatively small investment it is now possible to distribute such data for a fee. It uses the internet facilities to let data holders make their data broadly (commercially) available. The metadata standard offered by ESRI is however at this point dealing with too highly generalised data to be useful in very specialised applications. However, for many less sophisticated applications it is very useful. Taken in this context it may be well considering investing in the application of, more detailed, national, metadata standards so that the data in an application domain can also be useful to more sophisticated users who may also be more prepared to pay sophisticated prices.
Market differentiation in the NGDI
Opening up accessibility to government and other geospatial data is accompanied by the emergence of an interesting market differentiation. Take for example access to the geospatial positioning infrastructure through the growing availability of GPS receivers of a wide range of prices and precision. In the past geospatial positioning was almost exclusively entrusted to well-trained or even certified surveyors and navigators. Now the public can buy at low cost a GPS receiver for applications not requiring high precision, or with more sophisticated instrumentation for car and other navigation. There still is a requirement for high precision GPS or even ground survey but that demands more expensive equipment and higher trained technicians and professionals. Hence an unexpectedly broad market has emerged for these low- cost instruments while a smaller market segment for high precision positioning continues. The GPS infrastructure and for higher accuracies the so called Active Control Systems which have replaced in many industrial countries the conventional triangulation network are the foundation for this expansion and popularisation of the positioning capabilities.
Similar developments can be expected with ubiquitous access to geospatial data from any source combined with low- cost GIS software available on- line through the web. There will be a kind of popularising or may be de-professionalising of use which is yet unpredictable. ESRI's Geodata project is beginning to demonstrate this phenomenon. Hence we should expect a growing "popular" demand while the market for professional engineering and public administration applications will continue to demand highly reliable, timely and up to date data and information services. The sequel of this paper will deal with the requirements and evolution of latter part of the NGDI. It will address the following issues:
- What is the role of National Mapping Agencies in building the NGDI?
- What can be done to advance the NGDI concept?
- What is the "bottom line" to achieve NGDI in due time?
Generally speaking the National Surveys are the relevant sources for the Framework Data. (See Fig. 1). A special subset of these data is the Foundation Data (FD) which is the fundamental geographical reference for all other thematic application data. NMAs are usually responsible to produce, maintain and distribute the FD.
Within the context of the efficiency and effectiveness of government itself and the previously mentioned political expectations, optimal performance of NMAs and responsiveness to their client community is obviously of paramount importance in reducing transaction costs in society. Since their inception (some, such as the Survey of India, as far as 400 years ago but most in the mid- 1800's) they have enjoyed a monopoly in the technological and industrial organisation of national surveying and mapping activity.
Beginning in the mid-1970s these monopolies were increasingly challenged, with the growing proliferation of Information Technology (IT). The surveying and mapping technology has become increasingly embedded in software and accessible to non-specialists in the NMAs client community. Furthermore the client community obtained growing access to substitute products for the standardised topographic bases or thematic framework data sets (for example from Remote Sensing ). Both challenge the monopoly of the NMAs. Furthermore, the client community is increasingly changing to include users interested in the digital data for application in Geographic Information Systems (GIS).
These effects of the ICT coincided with growing government deficits, which for most countries has led to a 20 year period of budget reductions, demands for revenue generation and, in the context of reform of government, privatisation. See for example Groot(1998). More importantly these radically changing circumstances led to critical review of the relevancy of existing and emergence of new mandates. See for example: Canadian Government (1986), Department of the Environment (1987), Mapping Sciences Committee. (1990), ANZLIC. (1996).
In spite of these studies NMAs all over the world have been slow in responding to these changes and many are experiencing major problems pushing through the restructuring necessary to move from automation to informatisation . One of the more perplexing problems NMAs have been faced with is the definition of the content and formats of the Foundation Data, especially the topographic component of this. At first, these organisations implemented the IT to make the existing production lines of their standard topographic map series more efficient. However, these are extremely complex products, time consuming and expensive to produce and keep up to date. Budgets and human resources could realistically speaking never be sufficient to assure both timeliness and currency. A completely new design concept is required to meet the new requirements. As early as 1986 the Ordnance Survey had been challenged by the National Joint Utilities, the Municipalities and the Land Registry to produce the national topo data base at large scales for the whole country by 2000. These major users had demonstrated that a highly simplified data model should replace the traditional map content model. The simplification would be useful in practice and would make completion of the database by the year 2000 also feasible. See Department of Environment (1987).
Meanwhile progress has been made in defining the least complex and topo data set that will serve as a geometric framework for a multitude of applications. In the UK the term topographic template has found acceptance. See for example Smith and Rhind (1999). The Netherlands is taking the definition of this dataset a step further by seeking to give it a legal authoritative status such as the data in the Kadaster or the Register of Persons or the Register of Companies. This would have far reaching consequences in all Government Ministries who will be compelled to use this data only for refernccing purposes. It also would have far reaching consequences for the National Mapping Agency, which would become a legal as well as natural monopoly, a status which has proven to be not conducive to client orientation and efficiency.
The latter is important because poor performance by NMAs leads to the client community, especially the GIS community, falling back on cheaper and often simpler but more up to date substitute products from whatever sources may be available. This in turn leads to the loss of the positive externalities, to costly duplication, reduced and incompatible accessibility of existing (government owned) data, increased transaction costs and possibly reduced timeliness in decision making processes. In many ways it impedes full exploitation of ICT for the benefit of society in geospatial data applications.
It is significant that in those countries where NMAs have successfully adapted to the dynamics of the ICT environment and the imperatives of government reform, the informatisation has been accompanied, and even led, by institutional and regulatory reform. The organisation was thereby placed in a regulatory environment in which both management and staff are motivated to be innovative and efficient and are rewarded for this. For a detailed analysis of the relationship of government goals with respect to Geospatial Information, the economic efficiency in pricing and distribution, and the regulatory position of the NMAs, see Groot (2001).
What can be done to advance the NGDI concept?
There are, worldwide a large number of NGDI initiatives underway. See for example Onsrud (1999). Yet the case history literature is still not very complete. Nevertheless one of the most comprehensively documented developments towards a GDI for land administration is probably in New Brunswick. See Finley (2000). While New Brunswick is a small, rural province in Canada and its experience will not be readily replicated elsewhere, it does provide some important lessons. For example, much of its success to date can be attributed to the following factors:
- The province had a vigorous agenda for more than two decades related to building and linking its geospatial databases and providing effective public access to them. This agenda has been strongly supported politically, especially by a recent Premier who gave the highest priority to enhancing the province's information infrastructure. This points to the importance of a long-term strategic vision and high-level political support.
- For more than a decade New Brunswick has had a lead agency responsible for (1) designing and implementing the GDI concept; (2) coordinating the development of standards and protocols; (3) building and sustaining core data sets; and (4) providing online public access. This suggests the importance of a lead agency.
- The province has concentrated on building key data sets of particular importance to the economic and social development of the province. This has included parcel-based data sets in support of reforming the province's land administration systems and selected geospatial data sets required for effective resource management (especially in support of integrated forest management practices). This suggests the necessity for a focus on key priorities.
- The development and maintenance of the geospatial data infrastructure in New Brunswick is driven by a multi-year business plan. The lead agency, Service New Brunswick, is required to be self-sufficient. This has contributed to the emphasis on well-documented business cases for data and networking priorities, and on the funding strategies for ongoing upgrading and maintenance of the infrastructure. This leads to the importance of a business focus.
The essence of the GDI concept is that there is no master architect. See (McLaughlin and Groot(2000)). There cannot be, nor will there be, a single organization responsible for designing and implementing some kind of GDI blueprint, especially at the national (NGDI) level. Instead, we can imagine an almost organic web of partnerships and relationships evolving purposefully within a given jurisdiction. It will sometimes be pushed by technology, sometimes pulled by market requirements. But at some point there will be a sufficient inter-connectedness of databases, a level of access to the data and use of the data, as well as the maturing interest of stakeholders, to participate and invest in the partnerships required for a nascent NGDI to be recognized. Having said this, the evolution of any GDI concept will most likely emerge from a combination of 'top-down' and 'bottom-up' strategies, the specific mix of which will vary significantly from one jurisdiction to another.
The top-down approach will entail defining strategic goals, assessing priorities, developing implementation plans, and obtaining core funding. It will require some kind of institutional framework (lead agencies, steering committees, working groups for standards, etc., and monitoring arrangements). Examples of outputs which may be expected from the top-down approach are defining the fundamental geospatial data sets, building the clearinghouses, establishing metadata standards and access protocols, and resolving the information policy issues. Core funding for geospatial database development and networking, as well as for subsequent maintenance and upgrading, will also be key to defining the initial directions and priorities. This funding will invariably result from a combination of direct public investment and revenues derived from the sale of information products and services, as described in detail in Rhind (2000). While the debate continues over whether one should charge for public information, the reality is that some combination of direct investment and charging for products and services will be required.
The bottom-up approach will recognize the multiple local initiatives to build application-specific and enterprise-wide geospatial databases and will encourage their evolution towards a universal framework for accessing, combining and using the data through a process of proselytizing, providing incentives (especially shared financing) and imposing regulations (e.g., through standards regimes). Money talks and access to shared funding will be of special significance in advancing the NGDI agenda; but even more important will be the concrete evidence of the power and potential of an integrated infrastructure.
It is at least debatable if these conclusions will be as valid for a large and complex country such as India as they are for countries with a different culture of government, different relationships between levels of government, as well as between the governments at various levels and the private sector. But one of the foremost differences must lie in the information density of the Indian sub- continent which can only be dealt with at the local level.
Obviously we do not know the Indian context and conditions sufficiently well to suggest some observations about advancing the NGDI in India. Yet our sense is that considering the decentralizing and customizing character of the ICT, (one could almost say the anarchistic nature of these technologies), the role of the National Government should probably be relatively light. By this we mean that the rapid provision of the national foundation data sets and the definition of the hierarchy of a national metadata standard, and the development of Clearinghouse functions should have priority, developed on the basis of a business plan, as described earlier. Beyond that, the national role should probably be one of encouragement in decentralizing GIS applications in manageable application domains within the GDI concept and developing the top- down, bottom- up relationships that are necessary for that.
Bibliography
- ANZLIC. (1996). Australia New Zealand Land Information Council, Spatial Data Infrastructure for Australia and New Zealand, a discussion paper.
- Branscomb, A. (1982). 'Beyond Deregulation: Designing the Information Infrastructure'. The Information Society Journal, 1/ 3: 167-190.
- Canadian Government (1986). Management of Government: Major Surveys, A Study Team Report to the Task Force on Program Review, July 31, 1985. Canadian Government Publishing Centre, Ottawa K1A OS9, Canada.
- Department of the Environment (1987). Handling Geographic Information, Report of the Committee of Enquiry chaired by Lord Chorley. London, Her Majesty's Stationery Office, pp.67.
- EC (1994). Europe and the global information society: Recommendations to the European Council, May 1994, Office for official publications of the European Union, 1994.
- EC (1998). Public sector information: A key resource for Europe. Green Paper on public sector information in the information society. Office for official publications of the European Communities, 1999. http://www.echo.lu/legal/en/access.html.
- Executive Order 12906 (1994). Coordinating geographic data acquisition and access to the National Spatial Data Infrastructure. President Clinton, Federal Register 59, 17671- 17674.
- Groot R. (1998), Spatial Data Infrastructure (SDI) for sustainable land management; ITC Journal 19997- 3/ 4 pp. 287- 294.
- Groot R. and McLaughlin J. D. (Eds) (2000). Geospatial Data Infrastructure: Concepts, Cases and Good Practice; Oxford University Press, Oxford, UK.
- GrootR. (2001) Reform of government and the future performance of National Surveys; Special issue on Cadastre of the Journal for Computers, environment and urban systems, Pergamon. (in press)
- Langaas S. (1998): Transboundary European GIS databases: review of the Baltic experiences in Burrough P. and Masser I. European Geographic Information Infrastructures: opportunities and pitfalls. Taylor and Francis, London.UK. pp.31- 45.
- Lawrence V. (1998), European geoinformation data publishing: understanding the commercial industry; in Burrough P.and Masser I., European Geographic Information Infrastructures: opportunities and pitfalls.Taylor and Francis, London.UK pp.87-101.
- Longhorn R. (1998) Data integration for commercial information products: experiences from the EC's IMPACT-2 programme; in Burrough P.and Masser I. European Geographic Information Infrastructures: opportunities and pitfalls; Taylor and Francis, London.UK pp.101- 113.
- Mapping Sciences Committee. (1993). Towards a Coordinated Spatial Data Infrastructure for the Nation. Report of the Mapping Sciences Committee, Board on Earth Sciences and Resources, Commission on Geosciences, Environment and Resources, National Research Council, National Academy Press, Washington, DC, USA.
- Mapping Sciences Committee. (1990). Spatial Data Needs: The Future of the National Mapping Programme. National Research Council, National Academic Press, Washington D.C.
- McLaughlin J. and Groot R. (2000) Concluding remarks, in Groot R. and McLaughlin J. D. (Eds) (2000). Geospatial Data Infrastructure: Concepts ,Cases and Good Practice; Oxford University Press, Oxford, UK.pp.273- 275.
- Onsrud H. (1999) URL http://www.spatial.edu/harlan/gsdi/GSDI.html
- Rhind, D., (2000). Funding a National Geospatial Data Infrastructure, in Groot R.and McLaughlin J. D. (Eds.), Geospatial Data Infrastructure: Concepts, cases and good practise. Oxford University Press pp. 48- 51
- Smith and Rhind (1999). "Characteristics and sources of framework data" chapter 47, p. 655, in Longley P., Maguire D., Goodchild M., and Rhind D. W.(eds), Geographcal Information Systems: Principles, techniques, management, and applications.