Use of GIS Based DSS for Sustainable
Development: Experience and Potential
Use of GIS Based DSS for Sustainable
Development: Experience and PotentialThis report is based on a workshop held at the UNU/IIST in Macau, sponsored by IDRC. TuBao Ho, Vietnam, J. G. Krishnayya, India, and Subhash Bhatnagar, India who are members of WG 9.4 also took part in the Workshop. Several participants and authors of papers presented at the workshop contributed to this report through a discussion on a list server.
Development should be sustainable, ie. development that meets the needs of the present without compromising the ability of future generations to meet their own needs (Brundtland). Indigenous peoples have lived in harmony and stability with their environment for centuries, and this knowledge must be incorporated into the decision making process. Methods of development must be appropriate to the people and the environment in which they are applied. Decisions must be made not by aid agencies but by the people in the countries themselves. It is the involvement of the people themselves at all levels that is absolutely critical in making development sustainable.
Development involves making decisions, making choices amongst alternative possible development paths, selecting one line of action which returns benefit to the developing country. It is important that these decisions are made well, using the best information, methods and tools available. For sustainable development we need to make decisions that do not have long term negative effects, and assess the long term effects and impacts as well as the short term benefits. Because predictions can only be approximate, we cannot make decisions once and for all time, they need to be revisited and revised as the real consequences of decisions are revealed in practice. We describe geographic information systems (GIS) as an important recent approach to support decisions in development.
Geographic information systems (GIS) provide the ability to store and display maps and associated information. There are two methods for doing this - as "networks" in which the geographical area is divided into a number of polygons whose corners and interconnections are stored, and as "rasters" in which the area is divided into many small uniformly sized rectangles. Over these can be laid further networks of roads, rivers, boundaries and so on, as well as the locations of towns and similar features. Areas can be divided along theme lines, dividing the area according to certain criteria and displaying the theme map to highlight these areas. These two representations are equivalent and can be converted one to the other, but the representations favour different calculations. Examples of the network approach are Themaps (J. G. Krishnayya, Systems Research Institute, Pune, India) and Redatam, and of the raster approach is Idrisi (J. Ron Eastman, The Clark Labs, Clark University, USA). GIS systems have arisen from a number of application areas: cartography, surveying, civil engineering. However the key thing that has led to their rise to prominence have been the development of remote sensing and the increase in performance and decrease in costs of the technology necessary to make GIS systems usable.
Geographic information systems are capable of integrating geographical data with other data from various sources to provide the information necessary for effective decision making in planning sustainable development. Typically a GIS systems serves both as a tool box and a database. As a tool box, GIS allows planners to perform spatial analysis using its geoprocessing or cartographic modelling functions such as data retrieval, map overlay and connectivity. Of all the geoprocessing functions, map overlay is probably the most useful tool for planning and decision making -there is a long tradition of using map overlays in land suitability analysis. Decision makers can also extract data from the database of GIS and input it to other modelling and analysis programs together with data from other database or specially conducted surveys. It has been used in information retrieval, development control, mapping, site selection, land use planning, land suitability analysis, and programming and monitoring.
Brent Hall, University of Waterloo, Canada and his team reported on use of GIS in tourism planning and in the consideration of the environmental impact of tourism on fragile reefs in the Cayman Islands. In China the use of GIS for disaster monitoring using remote sensing data has been very effective. It is clear that where the financial resources are available, the use of GIS based DSS can be effective but this level of financial provision and administrative experience is not easily found in DCs.
Bhatnagar reported on the use of a DSS for planning the locations of new infrastructure in a district in India. He highlighted the need for cleaning up data and changing attitudes of decision makers to accept rational ways of decision making. Often data just cannot be collected, but decisions still have to be made, as has been seen in the planning of water resources in Zaire (M. K. Luhandjula, University of Kinshasa, Zaire)
The capturing of indigenous knowledge is seen as important, but in doing this the intellectual property right should be protected, as is being done in the Honey Bee network in India (Anil Gupta & Srinivas Chokkakula, IIM, Ahmedabad). Internationally recognised IPR legal instruments are being used.
In China rampant land development as a result of economic reform in 1978 has led to the encroachment of agricultural land in the Pearl River Delta. Official figures show that arable land in the Pearl River Delta has decreased from 1,044.7 thousand hectares in 1980 to 898.2 thousand hectares in 1991, a decrease of 14.0% in ten years.
It is clear that the interface of decision support systems is all important, so that their intended users find them easy to use and to understand. This has led the developer of a system to help regional decision making in the Philippines to select software for its superior interface, even when other aspects of the needed functionality would suggest different choices.The uptake of the decision support technology is conditioned by prior experience in using information technologies. Specialist GIS advisors in Egypt found that they were asked not to give advice on GIS but on the use of computers for more routine office functions. In the Philippines support is being provided not only for the annual round of financial planning but also for the routine all-year-round operational activities.
From all these case studies we have learnt a number of lessons which are discussed below.
The lack of available data is one of the major hindrances in the use of GIS for SD. Data is vital. In developed countries, most data needed is readily available thus making the establishment of a DSS relatively easy, but data is not so readily available in the developing countries. The most readily available data are those from remote sensing, but these are mainly limited to land cover information from which a very limited amount of information can be extracted. Base maps are often lacking or outdated, compiled by different agencies with different accuracy and map scales and geocoding systems making them difficult to be integrated into the system. Nevertheless much useful planning can be undertaken, as for example in China where satellite remote sensing data has been used in planning for disasters and changes in land use at the national level (He, Zhu Honglei, State Key Lab of Resource and Environment Information System, Chinese Academy of Sciences, Beijing, China) and even at the local level (Chen & Liao, Guangzhou Institute of Geography, China). Socio-economic data, are generally lacking and are often limited mainly to census data, though this can be very useful (Bowerman, University of Waterloo, Canada & Poulard). Socio-economic data requires field surveys which are expensive and time consuming. However, the main obstacle still lies in government recognition of the need for statistical information for planning and the willingness to mobilise resources in collecting it.
It is not only the availability of data which is a problem but the quality too. In India (Bhatnagar, I.I.M., Ahmedabad, India) it was found that locally collected data could be up to 30% in error: the solution seemed to lie in making the collectors of the data also the beneficiaries, so that they had a stake in the quality of the data collected.
The currency of data is very important in decision making and there need to be institutional arrangements to determine, coordinate, and monitor the frequency of data updating, and verifying the quality of the data collected.
The centrality of data to the adoption of DSS and the high costs and lead time to acquire data make if highly desirable to ensure that data is seen as a national asset serving multiple purposes. A first step towards acquiring this asset is coordination. Early and relatively cheap measures would include a national register of available data, to forestall repeated and duplicated acquisition. It might also be possible to encourage projects to extend their activities to acquire, at low incremental cost, additional data which is highly likely to be used by other projects. A second step is encouragement of adoption of standards for the content and representation of data, to provide a formal guarantee that the data will be applicable to other projects.
The current practice of decision making in developing countries has not advanced much in comparison to the tools available to help. The skills of planners and the planning system may not be ready to utilise the data and functions available, and may not yet be aware of the benefits and potential applications of technology. Little effort has been spent on transforming data into information for making decisions. The result of this is that decision making could be made in the interests of a few dominant stake holders.
There is a general shortage of trained manpower even in the developed countries. This shortage is more severe in the developing countries both in absolute numbers and relative terms. The problem of training is more severe in the developing countries because of the lack of expertise and shortage of funds in universities who do not lead in the teaching and research of DSS. Very often, it is the government agencies which buy and use the latest systems through funding from international agencies
Training programmes are needed for five major groups of users - policy makers, decision takers, programmers, technicians, and educators.
In the Philippines we have found (Gamboa, EcoKnowMICS Project, Manila, Philippines) that developing human resources is a slow process, requiring many years during which the relevant data would also be gathered and decision support systems developed to fit local needs. It will not necessarily be possible to transport training programmes from developed countries, since the decision making processes may be different. It arises partly from the need to handle uncertainty and risk, but also arises from cultural differences - how decisions are made and agreement reached may be very different.
The strong influence of leadership and organisational setting on the effective use and introduction of computers is very well documented. A few key individuals interested in computers become instrumental in the initial acquisition of equipment and guide its applications. The function of the leadership is to set clear goals and objectives, to win acceptance among information system users for such goals and objectives, and to provide commitment to achieve project goals and tasks. Another critical function of leadership is coordination of different departments sharing the information system.
Lack of prior computer use can also be critical, both in ensuring awareness of their potential, and for the infrastructure to support their use. DSS projects are very often initiated by international assistance agencies and there is a general failure to take account of the organisational setting and personal motivations of those involved. There is evidence of large investments having been made to acquire technology, but there is less evidence that the systems are functioning satisfactorily and contributing to national development efforts. Moreover, problems often arise in maintenance costs when the international assistance left the project.
Software for large scale systems is mainly purchased from the developed countries. It is expensive and consumes much foreign currency which is often in short supply. There is a general lack of locally developed software. Attempts have been made to use low cost commercial software to perform DSS tasks, the most popular are the combinations of commercial CAD packages such as AutoCAD with commercial database packages such as dBASE III. These systems, although limited, can make decision support available to departments and agencies with little funding. However, these low cost software systems still need to be purchased from the developed countries.
There have been quite a number of software developments in the developing countries. However, their developments are fragmented and most involve one to two researchers. They do not have the manpower and institutional set up to develop and maintain software like the commercial packages from the developed countries. (Krishnayya, Systems Research Institute, Pune, India). There may be a need for different researchers in a country or the region to pool their manpower and resource together to develop a package that can have good documentation, manuals, and support, similar to the commercial packages in the developed countries. Networks need to be established within the developing world and with the developed world. Already there are initiatives to do this in same regions (Noussi, African's Computer Science Institute, Gabon), and this workshop has lead to further trans-national networking.
Usability, and particularly the natural language of the interface, is a barrier to the adoption of technology. Most of the imported programs and manuals are written in English, but most of the users, and particularly the decision-makers, have limited understanding of English. User-friendly application programs which hide the technology form the users, with instructions or pull-down menu written in local languages need to be developed in order to enable local planners and decision-makers to use decision support systems.(Pat Hall)
Most of the DSS hardware and software used currently is imported from developed countries. It often takes a long time to repair a piece of hardware, particularly when the necessary components are not readily available locally. Equally, it is difficult to consult software companies when problems arise. Most of the service and expertise are also mainly concentrated in the large cities, especially primary cities, making hardware and software maintenance more problematic for sites located elsewhere. Systems must be available on low-end platforms like PCs, and must be fully serviceable in country, as would arise with locally produced software. Large countries with a substantial requirement for DSS and GIS systems should be encouraged to develop suitable software locally. Funding to acquire the system is mainly available through central government funding or international assistance, but little is available to maintain the system. Very often, the system cannot be in full operation because one or two terminals and peripherals are out of order and the agency responsible does not have funds to repair them. More serious is that there may not be funding and institutional arrangement to update the data after they are created. As decision making needs up-to-date information, the system will be useless if its data are not updated. The development of DSS should be considered as a continuous process and not just a one-off project. The sustainability of the DSSs themselves is important.
The workshop has established a shared understanding of the current state of development and application of Decision Support Systems for Sustainable Development. The papers written for the workshop will be available through the IDRC library, and a book will be produced documenting a comprehensive range of case studies of decision making for sustainable development and the methods and tools that were used there. This has left us with a very strong foundation from which to move forward. We recommend that financial and organisational support is found for the following actions:
Converted to HTML by Dr. David R. Newman, Queen's University Belfast.