A GIS-based Planning Support System for Environment-considered Urban Development in Developing Countries

本研究试图开发一个基于地理信息系统的规划支持系统，通过使用专家系统和元胞自动机建模并将它们与地理信息系统相结合，帮助亚洲发展中国家考虑环境的城市发展。结果如下：1。开发了支持环境区划的专家系统。知识库是从日本现有的四个环境管理计划中检测出来的。提出的评价方法允许在获得的数据范围内进行推断，并使用卫星数据代替在发展中国家难以获得的统计数据。在日本和韩国城市确认结果的合理性后，将适用于发展中国家的两个学科领域。应用结果表明了该系统的实用性和未来的发展方向。建立了基于元胞自动机（CA）的城市增长模型。通过在日本丰桥-丰川地区的应用，展示了这里开发的模型的再生产能力。通过对两个课题地区的实验应用表明，所提出的模型对预测未来城市扩张有一定的应用价值，但也有进一步改进的空间。EEZM和CA模型与GIS相结合，使决策者能够反映对各分区震级的价值判断，并进行各种基于场景的模拟。此外，通过模拟结果与区划图的叠加，提出了一种可提供不同类型信息的开发适宜性评价工具。该工具可以检测未来的毁林地区、未来有资格开发的地区和未来的危险区。通过在印度尼西亚望加锡市的应用，表明该系统是有效的，可以为考虑环境的城市发展提供一些建议。

This research is attempted to develop a GIS-based planning support system to aid environment-considered urban development in Asian developing countries, by using an expert system and cellular automata modelling and integrating them with a GIS.The results are as follows.1. An expert system for supporting environmental zoning (EEZM) is developed. The knowledge base is detected from four existing Japanese environmental management plans. The evaluation methods are proposed that allow the inference within the limits of acquired data and use satellite data as a substitute for statistical data that are hard to be obtained in developing countries. After the confirmation of the propriety of the results in Japanese and Korean cities, the system is applied to the two subjects areas in developing countries. The results of the application indicate the usefulness and future direction of the system.2. An urban growth model using cellular automata (CA model) is developed. The reproductivity of the model developed here is shown through an application to Toyohashi-Toyokawa region in Japan. It follows from an experimental application to the two subjects areas that the proposed model is of use to predict future urban expansion, but also has some room for further improvement.3. The integration of EEZM and CA model with a GIS makes it possible for decision-makers to reflect the value judgment for the magnitude for each zoning district and to perform various scenario-based simulations. Moreover, by overlaying simulation results and zoning map, a tool is proposed to provide different kinds of information to evaluate development suitability. This tool can detect future deforestation areas, areas eligible for future development, and future hazard areas. Through an application to Makassar in Indonesia, it is shown that the proposed system is useful and can give some suggestions for environment-considered urban development.