Ksemsan Suwarnarat, Director, Policy and Planning, Bangkok Metropolitan Administration, Thailand

Sadasivam Karuppannan, Planner, City of New Dehli, India

Haider W. Yaqub, Planner, City of Islamabad, Pakistan

Francisco Escobar, Ian P. Williamson, Ian D. Bishop, and Paul M.Yates, Department of Geomatics, The University of Melbourne, Australia

Urban population world wide is growing at a much faster rate than the population as a whole, and by larger increments than ever before. Improving social and economic conditions for all people and promoting sustainable development is increasingly an urban challenge. As cities grow, these improvements become increasingly complex . The world’s urban population in 1996 was estimated at 2.6 billion. Within 10 years more than half of the world population will be in cities, 3.3 billion of the 6.59 billion total. Urban areas are globally expected to double to more than 4 billion people by 2025, some 80 percent of them in developing countries.

These population changes have a profound effect on the demand for services, not only in large urban agglomerations - mega cities - but also in intermediate and smaller cities. The composition of the world’s largest cities has changed drastically over the past 50 years. In 1950, eleven out of 15 of the largest cities were in the developed world. In 1990, seven of the 15 large world cities were in developing countries. By 1994, 14 of the top 15 cities were over 10 million population. By 2005, 13 of the top 15 will be in developing regions and the 15^th will have nearly 15 million people. As illustrated in figure 1 most of the new mega cities will be in developing countries. By 2025 half of the world population will in cities.

Many of the cities within developing countries will house over ten million people making them congested, polluted and usually supporting a poor quality of life. To sustain these cities there will be a high demand for resources and services from within the urban and surrounding areas and an urgent demand to improve the management and delivery of the services. To meet this demand, systems and procedures will need to be developed that assist in providing the required resources and management in a more efficient and effective manner. A common component of these systems and procedures is land and all the issues related to the management of land. The wide range of spatial information technologies and tools available today provide considerable potential in supporting these requirements.

In recent times spatial information infrastructures have been built with the specific aim of supporting the management of cities in the developed world. Where geographical aspects are important for management, decision making and planning, these infrastructures have helped to improve the efficiency of these activities. Typically, more than ninety percent of information required for a city’s administration has a spatial component, such as parcels of land, road networks, utility infrastructure, emergency services, garbage collection and recreational areas. Increasingly spatial information technology and related information technology systems are basic infrastructure for managing cities in the developed world.

In contrast, for most cities of the developing world such spatial information related activities are undertaken manually if at all. These cities are mostly larger in population and area than in the developed world, are rapidly expanding and hence have an urgent need to make more efficient use of spatial information.

This paper examines issues relating to the design and implementation of a spatial information infrastructure for the city of Bangkok. Based on the achievements and prospects of the Bangkok experience, issues in the development of spatial information infrastructure for cities in developing countries are discussed. The paper has benefited from the direct experience of people involved in the Bangkok project.

Numerous differences exist between the cities of developing countries and those in the developed world. In particular, differences exist in the spatial information system requirements. Some important distinguishing characteristics of the cities in developing countries are:

1. The rapid growth in population is not matched by growth in delivery of land for housing, services, utilities and infrastructure important to sustain a reasonable quality of life. This is evident from the sprawl of informal settlements, increase in congestion, air and water pollution, poor infrastructure and dilapidated housing (Williamson, 1991). Typical mega cities in developing countries have populations of the order of ten million and grow at a rate of about 4 - 6% per annum - about half million people - requiring approximately 100,000 additional dwellings and additional basic infrastructure required to sustain a reasonable quality of life.
2. The growth of the city is dictated by the market forces rather than a strategic plan. Urban development is often uncoordinated and land speculation pursues quick financial gains. The suburban and rural areas are invaded by market-induced developments.
3. Laws and guidelines of land registration, planning and land management is diverse in developing countries. The reasons are many and varied between different countries with different traditions and political set up. Thus, spatial infrastructure cannot be easily standardized for developing countries.
4. Developed countries in general moved from prescriptive form of land use planning (e.g. Master Plan) towards market oriented ‘spot-zoning’ approach confirming to environmental guidelines. Prescriptive urban land use planning is adopted in most of the developing countries resulting into long term land use plans/master plans which are less market sensitive.
5. Almost in all cities in developing countries significant proportion of population live in slums, squatters and unauthorized settlements. Informal settlements is an unique phenomenon in cities of developing countries. Land tenure and ownership is obscured and unregulated.
6. Unplanned developments make it difficult to provide utilities, while providing these utilities at a later stage is costly. As a result of poor mapping and land administration system, the city administration usually does not know the location of all existing services. It follows that efficient maintenance is impossible. Without proper land information, the acquisition of land for public facilities or the undertaking of any city planning exercise is very cumbersome.
7. To cater for a growing demand for housing, informal settlements spring up in open spaces where there are less development controls. As a result, these settlements are of poor living environments.
8. There is little or no spatial information infrastructure. The biggest single barrier stopping the construction of a spatial infrastructure is a lack of data. Other limitations include inefficient processes for purchasing spatial information or GIS technology; lack of skilled personnel to establish and manage the infrastructure; and lack of funding or political will to support the construction of the infrastructure.

In addition to these specific characteristics, there is a growing expectation within the developed world, that cities in the developing world will meet externally prescribed standards in terms of environmental protection and working conditions. These changing external constraints imply improved information flow to administer effective plans for resource allocation and distribution (Lai, 1996). In order to do any progressive planning of these cities it is necessary to develop a spatial information system which can be an efficient tool for management of their resources; a tool which can assist in the planning and administration of cities.

The availability of spatial information for cities in developing countries is poor. In many cases the spatial data is in the form of unscaled sketches. Where maps exist they are often out dated or classified as restricted information and access by public departments is very difficult if not impossible. Where current and unclassified maps exist, they are usually of different scales aggravating the problem of sharing information efficiently among various sectors of the city. Digital representation of spatial data is even rarer due to the lack of appropriate equipment and trained staff; few cities are ready to begin generating digital spatial databases as the task is seen to be too large, too costly and too complex both technically and administratively. The design and implementation of a workable spatial data infrastructure is often a dream for the future.

Reproduction of maps, where they exist, is often done with ammonia printers with the original maps prepared and up-dated on transparent sheets. All cadastral data are stored in registers and updated manually where available. Where a digital database is being prepared for a city, map production often continues to be done manually. Information about state owned land or government property is poor. In many cities there are large military landholdings where information is restricted. Often maps showing the land ownership are in the form of sketches which are not to scale. Therefore, the exact demarcation of much of the land is impossible. Most information about land ownership and utilities are in the experienced hands of key personnel in separate departments. Due to the absence of any other information system they are treated as the primary information source. As a result this information can be easily lost.

Information about the location of underground utilities and facilities is often worse than the maps showing above ground features. Utilities and facilities are marked on plans but are rarely up-dated. In most cases the actual location of these service lines is different from the location appearing on those plans. Agencies responsible for the provision of different infrastructure maintain their own maps usually showing only the location of infrastructure under their control. These maps are often only in the form of sketches that are interpretable only by the staff of these organisations. For underground utility lines the depth of the lines is seldom or incorrectly shown on maps. It is often the case for these utilities that some lines run on top of another utility line at a different depth.

For obvious reasons, these management practices were possibly adequate when the size of a city is small, when land prices are low and the pace and life-style is slow. In these conditions, co-ordination between staff is possible and there is time to share information. For the large rapidly growing cities of developing countries today, which are becoming the engines of economic development, these management practices are woefully inadequate.

In order to assess the GIS implementation in a developing country, a case study of the Bangkok Land Information System (BLIS) is presented. The initial development strategy, the extent to which it partially succeeded, the reasons why it has not progressed as fast as hoped, and the lessons this provides for other developing countries, are addressed in the case study. As a result of the experiences gained in the BLIS, suggestions leading towards a successful approach to implementation of a spatial information infrastructure in developing countries is made. An important part of the suggested approach is to provide short term, useful and low cost outputs.

The City of Bangkok provides an interesting case study to examine the problems and strategies in developing an integrated spatial information system. Having realised the problem being faced by the city, the Bangkok Metropolitan Authority (BMA) decided to initiate a pilot GIS project (Suwarnarat, 1991). The Bangkok Land Information System (BLIS) was undertaken from 1989 to 1991 (Williamson and Mathieson, 1994). The primary objectives of the BLIS project were:

1. To educate, train and give experience to Thai Government officials in GIS;
2. To evaluate the need for an appropriate common base map for the city of Bangkok to be used by all partners of the BLIS program; and
3. To determine an achievable long term strategy for development of BLIS

These objectives were conceived on the basis that the GIS technology should integrate various agencies responsible for the management of the city. Also, the technology should be appropriate to the needs and resources available in the BLIS organisations, be relatively simple yet should be able to perform the task, be cost effective, and be application and user oriented.

Initially a team comprising the Metropolitan Electricity Authority (MEA), the Metropolitan Waterworks Authority (MWA), the Telephone Organisation of Thailand (TOT), the Department of Lands (DOL) and the Bangkok Metropolitan Administration (BMA) was put together to undertake the BLIS project with the assistance of Australian technical aid (Williamson and Mathieson, 1992 and 1994).

Early in the BLIS project, it was sought to establish the requirements for a common digital map base for Bangkok. The essential data items considered necessary by each BLIS partner organisation for its own purposes were identified (Suwarnarat, 1992). Although a common interest existed, there were also different points of view. The planners were satisfied with medium scale maps of scales 1:10000 and 1:4000, whereas most of the participants from the utility agencies required a scale of 1:1000. However the TOT insisted on a detailed base map of scale 1:250. Apparently, the TOT engineers mistook the GIS as a land parcel map which they would like to use for CAD (computer aided design) to produce engineering drawings.

Such misconceptions can arise because GIS, GPS (Global Positioning System), LIS (Land Information Systems), RS (Remote Sensing) and AM (Automated Mapping) are inter-related and often mistaken for each other, especially in developing countries where the technology is not as widely known or understood. This lack of understanding gave rise to protectionism and rivalry among the BLIS government agencies which were competing for budgets to develop their own computerised systems. In this environment a powerful authority with a misconception may get control over the budget and obstruct useful developments in other organisations. There were also administrative problems in the BLIS project team due to misunderstandings of the strengths and weaknesses of GIS vis-a-vis CAD. The GIS operators were often asked to modify maps or produce visual effects which were beyond the capabilities of the system. Differences in speed, accuracy and graphic qualities combined with unrealistic expectations which led to confusion among the potential end users. This situation was compounded by many of the senior personnel in the BLIS agencies having visited developed countries or international GIS conferences and seeing impressive GIS demonstrations, themselves often only pilot or demonstration projects. These expectations placed considerable pressure on the BLIS team.

After considerable discussion, various "compatibility" levels of different map scales required for urban planning and management were determined as illustrated in table 1. Informally, a generalised description was adopted.

In the context of this article, a land information system is defined as a GIS which utilises land parcels as the link to the non-graphic database attributes. The BLIS was conceived as a GIS designed primarily to serve the urban utility agencies (MEA, MWA and TOT) who share the same land information, such as the type of buildings and the name of the owners of land parcels, with the BMA. The BMA collect land tax, building tax and bill-board tax as well as a wide range of other services. While the utility agencies may not need land parcels to calculate the demand for their services or to plan for rate collection, the land parcel framework is very useful for the installation, operation and maintenance of their infrastructure. DOL was invited as a partner since it provided the base map.

The database was linked to numerous attributes with the most common information shared by the BLIS partners being the owners name, the area, location and type of building. Other specialised items such as cable sizes, water pipe dimensions and location of meters were input and managed by participating BLIS organisations, and were not the direct concern to the BLIS project team as a whole.

It was expected that significant cost savings and an improvement in efficiency would result from sharing substantial core data sets and particularly the base map, since it was acknowledged that data acquisition would be the most important cost factor in developing GIS in the BLIS partner organisations. The reality was that the BLIS team found that the conversion of the analogue map base into digital vector maps was an expensive, difficult and time consuming operation. Attribute data were often difficult to obtain.

At the end of the project, instead of establishing a common database, the MEA (which is part of central government and has access to significant funds) took the leading role in creating a digital base map to suit their specific needs. The MEA believed that an investment in GIS would be returned almost immediately through the collection of additional tariffs. This is similar to many developed countries where a utility has taken the initial lead, such as in Melbourne, Australia where the water authority took the lead. The MEA created it’s base map at a scale of 1:1,000 by digitizing the DOL maps which were from the same source as those used for BLIS.

The BLIS pilot project members are now separated with each pursuing their own goals in GIS development. However they still maintain a core group of participants in various government committees concerned with GIS/LIS and AM/FM technology. A national committee for the promotion of GIS was established in Thailand in 1996. The committee resembles the strategy proposed by Williamson and Mathieson (1994) but the focus of the committee has moved to the national level. A senior BMA official chairs the subcommittee responsible for GIS/LIS data in the Bangkok region. Encouragingly, every member of the original BLIS project is represented in this subcommittee.

Surprisingly, eight years after the initiation of the BLIS project there is still little to show in the BMA, although a great deal of experience has been gained by those involved with the project. It has been suggested by some of the former BLIS partners that the base map now produced by the MEA is not sufficiently accurate, but there is no serious challenger who is prepared to develop a more accurate map. Importantly, every member of the BLIS project has now agreed to use the MEA map. These recent developments have confirmed the benefits of the original BLIS pilot project.

After more than a decade of experimenting with GIS, the BMA now realises that GIS is an effective tool for planning and managing urban infrastructure and utility systems. However, the vision to develop and maintain a large common digital base map within the BMA has not been realised after nearly a decade of struggle. It is recognised as costly and administratively complicated. This has been an important lesson.

An alternative approach which some BMA officials believe has the best chance of success is the development of application oriented GIS to deal with specific issues and areas of concern. Using this approach the benefits of the use of GIS can be seen without waiting for a common base map to be established. A suitable purpose built data base can be completed more quickly than a comprehensive or multi-purpose data base. The GIS application for these individual projects may be relatively simple but may provide major benefits. For example the Ministry of Industry in Thailand found a digital terrain model showing locations of BOD loads useful to help locate waste water treatment plants (Suwarnarat et al, 1994).

The success of these small strategic applications can help establish the value of GIS in developing countries. For an organisation such as the BMA, the BLIS experience suggests such a strategic application oriented strategy is more relevant than the system-wide oriented strategy pursued in the BLIS project. This is certainly not the ideal but it is the reality. It is recognised in the context of developing countries that it is very important that the benefit and cost curves of the GIS development graph meet quickly and this can usually only be achieved through short term benefits.

To highlight the application oriented approach some examples of the use of GIS in strategic management in Bangkok are presented in the following section.

The metropolitan area of Bangkok consists of 40 districts under the overall control of a Governor. In order to demonstrate the potential of spatial information handling technology in a developing country, a simple project in producing maps from existing BMA data was undertaken. Within the City of Bangkok various socio-economic and demographic data for each district are collated annually and included, as a normal practice, in a City Yearbook. This data had been stored in a spread sheet and was made available along with a file (in DXF format) of the district boundaries. The boundary file had been digitized from a 1:20,000 source map.

The attribute data used includes many of the normal statistics collected in most large cities on a district-by-district basis such as population, employment, public facilities (schools, hospitals etc), land use, housing pattern, factory numbers, voting patterns, land tax returns, garbage collection rates, etc.

This data has been available annually in Bangkok for the last fifteen years. Until now this data has not been linked to any spatial reference or mapping system and consequently could not be fully utilised for decision making. It is the view of some of the senior BMA officials that the mapping of this data after simple analyses will not only be very useful to assist the management of Bangkok but that also be very influential in promoting the wider use of GIS in the BMA.

Using this data, the following set of maps were produced to assist decision making on various aspects of the management of the City by the Governor of Bangkok. Since the BMA operates according to official management plans, such as the "annual plan" and the "five year plan" these spatial analyses will be very useful for strategic planning. Bangkok is managed according to a five year plan which is prepared by officials under the supervision of the Governor and is approved by the City Council for implementation. An annual plan is controlled by the annual budget and is prepared in the same way as five year plan. The output from the following strategic analyses will make a valuable contribution to managing the City as well as promoting the benefits of GIS.

District officials are expected to be versatile, to know their district and their people, and to implement agreed plans efficiently and effectively. Simply a district official needs mapping and associated data at a scale of 1:1,000. Department officials, on the other hand, are professional people under a "permanent secretary" or city clerk. These officials are expected to be knowledgeable and have experience and expertise in the applications area for which they are in charge. They are more concerned with the long term strategic planning for which maps of 1:10,000 scale or smaller is sufficient.

The maps generated are thus suitable for Departmental officials. Maps have been prepared relating to garbage disposal, primary education and population density, but many other similar maps based on Yearbook data could have been just as easily prepared in conjunction with digital boundaries of corresponding geographies.

Examples of maps produced to assist in decision making are shown below :

1. Garbage disposal is one of the most costly services provided by the BMA. One person generates 500-2,000 Kg per year of garbage depending on their life style. The disposal of garbage by incineration costs $2 / kg but $0.2 / kg in case of land-fill or composting. The Governor or the appropriate senior official must know in which districts the generation of garbage and revenue are in balance. With the GIS the Governor can oversee the management of garbage disposal.
2. According to Thai law, primary education is compulsory with the BMA being responsible for providing this service. Individual district chiefs of the BMA manage this service. They need to know the supply and demand situation for their district. The Governor or the appropriate senior official need to know the relative proportion of schools and school age children among all districts.
3. The districts are of different sizes and have different populations. A GIS map with population density as an attribute is basic information indispensable for policy making by the Governor, as far as the districts are concerned since their annual budgets depend on it.

The BLIS group in the BMA is also looking at other possible techniques used to capture data for input to GIS through the use of raster images and other techniques to estimate attribute data such as population (Webster, 1996). Automatic conversion of raster data to vector data has not proved to be reliable. A trial scanning of a 1:4,000 map of Bangkok showed that about 30% of the buildings did not appear as closed polygons. Much touch-up work is required. However, scanning is a quick solution for the over-all view of the terrain and a broad coverage of an occurrence, such as a flood. New technologies in which resolution and ease of interpretation of remotely sensed and scanned images are emerging will broaden the applicability of these technologies.

Aerial photos are another type of raster data. Such data is used effectively as a back-ground to enhance vector maps. As mentioned, the cadastral base for the Department of Lands produced a vector land parcel map. Overlaying it on an orthoimage (rectified photomap) enables the visualisation of the relationship between the parcels and the surrounding geography. In 1996, the BMA obtained a new set of aerial photographs to support planning and related activities however they are yet to be used for security reasons. They also plan to use less expensive small format (4x4 square inchs) aerial photography to produce and update maps. Again the BMA has had this technology for many years but to date it has not been able to use it owing to internal reasons including delays in gaining approvals to take photographs for security reasons.

The Bangkok experience in the development of spatial information infrastructures provides a generic framework for cities in developing countries facing a similar challenge. Although urban administration and institutional arrangements differ among countries, the issues discussed here provide an insight into the technological, managerial and institutional issues involved in the adoption of spatial information systems for cities in developing countries.

Cheaper and faster computer technology is coming within the reach of most organisations in developing countries. It is therefore appropriate to consider the ways in which this technology has been implemented in developed countries - even though the implementation strategies may not be appropriate for cities in developing countries. In particular, future trends must be considered in order to understand the potential information infrastructure to which the developing world may aspire - while recognising that different problems and conditions exist in developing countries.

There is a range of GIS implementation strategies that have been adopted in developed countries. Three main steps of planning, analysis and implementation are of great importance in GIS implementation (Korte, 1994). The diffusion of GIS in an organisation has been extensively discussed (Campbell and Masser, 1995) and the main phases established by the NCGIA (1990). The adoption of new spatial information technologies in developing countries should also be considered in the context of these main steps and approaches. However these approaches should be modified by taking into account the following factors which can impede the adoption of spatial information technologies in developing countries.

As in the developed world, the successful implementation and use of spatial information infrastructures in developing countries is dependent on political and institutional support. Management support is necessary in building the infrastructure, understanding the limitations and providing the necessary flow of information to the system. This was particularly important in maintaining the momentum of the Bangkok project over many years. The project only received the necessary political support after a Professor of Surveying was seconded from a university to the BMA as a Deputy Governor. On the other hand, varying awareness of GIS in an organisation may bias the results. Staff will define products based on their personal awareness of GIS, on a vendors demonstration, on a demonstration they saw at a conference or during a visit to a developed country, not on an abstract need for information. For example Bangkok Metropolitan Authority requested 3-D perspective views of data, regardless of their actual value in decision-making process.

Multiplicity of agencies involved in urban management and their individual data requirements vary widely in terms of map scale, graphic elements, data accuracy and data currency. This makes the adoption of a common standard very difficult in any country, but especially in developing countries. A common data standard accepted by all potential users will reduce duplicity of data and data redundancies. As seen in Bangkok, utility agencies required larger scale maps than planning agencies. A map scale of 1:4000 or smaller scale is sufficient for strategic urban planning and urban monitoring but a scale of 1:1,000 or larger was required by the utilities. An agreement about the map scale at the very beginning of spatial infrastructure development is important as it will determine the application areas. In practise this is very difficult to gain agreement due to different institutional needs and jealousies.

A common practice in developed countries has been the establishment of data standards, fixed scales and data accuracy definition for national digital map bases. This task has normally been done by the National Mapping Agency (NMP). Lack of such organisations in most developing countries cause many problems in developing appropriate urban spatial information infrastructures. The BLIS is a good example of a strategy to develop appropriate standards in a developing country.

The successful development of spatial information infrastructures in developed countries has partly been due to much of the data being available in digital form or there being a political will and financial backing to build the required data sets. However, in developing countries the situation is very different. Data is generally not available in digital form and base maps are often out dated. Even where spatial data is in digital form few results have been derived. In order to have an efficient GIS operational in developing countries it is important to start with basic data collection. This requires time which few managers either have available or are willing to commit as the benefits derived are not immediate.

There often appears to be a political urgency which negates long term planning in developing countries. For example it is often stated that the GIS must be operational in say two years, but since this is virtually impossible, an appropriate strategy is not implemented thereby resulting in little progress ten years later. Making use of high resolution remote sensing provides great potential to alleviate some of these problems for medium and small scale GIS applications. On the other hand, developed countries build their digital systems from well established manual map systems which have often evolved over decades and sometimes centuries. Developing countries are becoming a good area to evaluate the potential of new high spatial resolution imagery in medium to small scale map making. By using this source of information, developing countries have the potential of saving time and money as they avoid the intermediate step of establishing manual map systems.

There is an acute shortage of trained professional GIS personnel capable of introducing spatial information technologies in developing countries. In the case of the Bangkok Metropolitan Administration, out of 20,000 employees there is not a single GIS professional. People unfamiliar with a new technology can also have serious concerns about being replaced by a computer and often view the technology as a competitor rather than an assistant. This was the case in developed countries in the early stages of GIS implementation processes in the public sector. Education, training programs and policies to extend the level of awareness of the technology appear to be absolutely essential and one of the best uses of international aid programs to promote GIS in developing countries. The installation of GIS hardware and software without extensive medium to long term education programs, as distinct from short term GIS training, often results in failure in developing countries.

Because of current practices in local governments in developing countries, it is often difficult to get budgets approved for the purchase of new technology, particularly, when senior administration has little experience with the technology. GIS requires investment in hardware, software, data collection, data maintenance and education and training, and it is very difficult to get support for such projects from senior management without the awareness of the potential of the new technology. There are, however, low cost desk top GIS software packages, with some of them being available through INTERNET at no cost. Such systems do not require specialised hardware, have simple, easily learnt interfaces, provide the essential spatial data handling procedures required of GIS software and allow rapid visualisation of information in a spatial context, and particularly for parcel based data which is the key for most urban applications. In particular they are good for raising awareness and gaining experience. For many urban applications, these simple desktop low cost GIS provide an important way forward for many urban applications in developing countries.

The bureaucratic procedures for approval and procurement of technology in developing countries are often very cumbersome. It requires a great deal of patience and energy in many developing countries to find a way through the political system to obtain support and funding, prepare tenders, evaluate tenders, award contracts and take delivery of the system. In most cases, by the time the contract is awarded, either the technology intended for use is out dated or the persons promoting such projects are transferred (and often both!). As a consequence, the project is often delayed. All through the developing world there are GIS and related technologies either under-utilised or lying idle for these reasons.

GIS is a new field with a limited clientele in many developing countries. Often there is little vendor support for GIS software. This is particularly the case in developing countries. Often, monopolistic conditions exist for a vendor who provides software and after sales support. This monopolistic phenomena often results in higher costs, poor services and limited expertise. It follows that many government departments do not want to invest in sophisticated GIS products. It is often cheaper and easier to implement simple off-the-shelf packaged software in developing countries.

Unfortunately there are few successfully implemented GIS projects in developing countries that provide guidance to others. Simply the establishment of spatial data infrastructures and GIS is very difficult in developing countries without a great deal of commitment from all involved. However, there are many cases where spatial data handling technology has been introduced in developing countries by developed countries in the form of international aid. It is important for the successful implementation of GIS in a country (developed or not) that the technology used should be well known and the people using it be well educated and trained in its use. Successful systems will need to manage with little initial data and lots of initial skepticism from the organisation. Systems should have a smooth learning curve, be relatively simple to learn and operate (measured in days not months) an ability to perform important basic functions with ease, and easy database building capabilities. Such systems should be capable of integrating and accessing existing data sources.

It is increasingly recognised that linkages between the various spatial information development areas leads to greater applications and benefits. Increasingly access to all these capabilities will be via a single interface, across an Intranet or the Internet, giving access to different features provided by a variety software products. In future, data from all platforms will be conveniently exchanged and interpolated for building various scenarios. Integrated systems will provide seamless decision support based on shared data using various forms of predictive modeling. Results will be presented through a multimedia interface and include direct comparison of results in spread sheet format with photo-realistic visualisation of relevant outcomes. Real time exploration of options in immersive multi-dimensional decision space may represent another possible mode of decision support.

These issues are the focus of a current research in developed countries. Developing countries have the opportunity to learn from developed countries mistakes to start their own process avoiding those mistakes. The rapid advances in network data access (Buttenfield, 1996), modeling and visualization (Pesce, 1995) suggest that the necessary integration tools are already here.

In coming years, a vast majority of the work force in developed countries will become proficient in the use of these new technologies which will have a major application in urban administration and management. Powerful computers with imbedded multi-media and GIS capabilities will be available as a common household tool. However this scenario will not necessarily occur in the developing world. The International Conference of the Internet Society, INET'98 (Geneva) has highlighted this issue as one of the main concerns in developing countries regarding technology diffusion (Hargittai, 1998). This has important implications for their adoption strategies.

The above technologies obviously have great potential in supporting improved urban management and administration. However the experience of developing countries, even with the most basic spatial information technologies, has not been good, for the reasons explained above. As such care should be taken by developing countries in adopting the latest technologies which are only now being adopted in the developed world.

Cities in the developing world are expanding at a much greater rate than in developed countries with resulting major environmental and management problems. While these cities are usually the engines of economic development in the respective countries, their infrastructure and quality of life is deteriorating. Addressing these issues in cities of the developing world is one of the great challenges facing all societies in the next millennium.

GIS and the underlying spatial data infrastructures and technologies being adopted in the cities of the developed world have the potential of significantly contributing to resolving many of these urgent issues and problems in the developing world. However care is required. The economic, social, institutional, legal and technical environment in the developing world is very different from that in the developed countries. As such the use of GIS and related technologies has had a chequered history in developing countries with many systems being under utilised or lying idle.

By reviewing the environment in developing countries, this paper has attempted to understand and explain the problems and issues in applying spatial information technologies, often taken for granted in the cities of the developed world. The paper reviews the Bangkok Land Information System project as an example to highlight the issues and to suggest some generic strategies.

The main conclusion from this research is that the development of a digital large scale parcel based map as basic spatial infrastructure for a range of GIS business applications is very difficult to achieve for many countries in the short to medium term. The main limitations are lack of resources and trained personnel, inefficient bureaucratic processes, lack of data, and lack of hardware and software vendor support.

In addition, it is suggested that small scale projects or business based GIS are the best option to introduce the GIS concept to cities to gain acceptance of the technology, as is shown by the Bangkok examples in the paper. In this regard, high resolution satellite imagery offers an opportunity to obtain the basic digital map base required to support small to medium scale urban GIS in the future.

It is acknowledged that the development of technology is moving at a rapid pace, and that while the above suggestions are considered appropriate in the foreseeable future, based on current technology, it is quite likely that new or improved spatial technologies will arise. This could have direct application in the cities of the developing world, even though personnel, institutional and financial issues will continue to constrain development.

This paper is the result of a continuing association between the University of Melbourne and the Bangkok Metropolitan Administration. That association has received support under the Targeted Institutional Links Program of the Department of Education, Employment, Training and Youth Affairs (DEETYA) of the Federal Government of Australia. In addition Khun Samphet of Chulalongkorn University is acknowledged for providing valuable assistance in the conversion of data and the preparation of maps.

Buttenfield, B. P., and Kumler, M. P., 1996, Tools for browsing environmental data: the Alexandria digital library interface: Third International Conference on GIS and Environmental Modelling, Santa Fe, 1996.

Campbell, H., & Masser, I., 1995, GIS and Organizations, London, UK; Bristol, PA: Taylor & Francis.

Hargittai, Eszter, 1998, "Globalization and Regional Implications: The Internet and Global Socio-Economic Development", INET'98, Geneva, July 21-24.

Lai, Poh-Chum & M. Wong., 1996, "Problems and prospects of GIS development in Asia" Ed. Tung Fung, etal, GIS in Asia, Hong Kong, 195-199.

NCGIA, 1990, Core Curriculum. 3 Vols: I. Intorduction to GIS. II. Technical Issues in GIS. III. Application Issues in GIS. Santa Barbara, CA. National Center for Geographic Information and Analysis. University of California.

Pesce, M., 1995, VRML Browsing and Building Cyberspace, New Riders Publishing.

Suwarnarat & Others, 1994, " The application of GIS in the control of water pollution" CATT 94 SEMINAR, Sirikit Convention Hall, Bangkok, June 27-29.

Suwarnarat, K., 1992, "An Improved Information System For The Urban Development Of Bangkok", ENTECH 1992 SEMINAR, Sirikit Convention Hall, Bangkok, October 18.

Suwarnarat, K.,1991, "BLIS, A New Tool for The Urban Development" Paper presented at The first International conference and Exhibition in Thailand on CAD-CAM-CAE-CIM-AEC-GIS Technologies and Application. Bangkok, July.

Webster,J.C., 1996, "The Potential of Urban Texture Measures in Monitoring Urbanisation from Space", GIS in Asia, Asia GIS/LIS,AM/FM and Spatial Analysis Conference, Hong Kong, published by GIS Asia Pacific Hong Kong Geographic Association., 309-322.

Williamson, I.P., 1991, "Land information management at the World Bank" The Australian Surveyor, 31(1), 41-51.

Williamson, I.P. & Mathieson, G.,1992, "The Bangkok Land Information System Project-Designing and Integrated Land Information System for a large City in The Developing World" CISM Journal, ACSGC, Vol.46.No.2,1992, 153-164.

Williamson, I.P. & Mathieson, G., 1994, "The Bangkok Land Information System-Past and Future" Developing World" The Australian Surveyor, December 1993,Vol. 38 No.4, 298-307.