Prediction of Land use change in urbanization control districts using neural network - A Case Study of Regional Hub City in Japan

Land use is changeable in the urban area, depending upon the economical mechanism of market. The controlled urbanization area is made a region where the urbanization should be controlled by the city planning and zoning act. However, in the zone, there are also many areas where form regulation of the building is looser than the urbanization zone which should form a city area. Therefore disorderly development acts, such as location of the large-scale commercial institution and leisure facilities unsuitable for circumference environment, are accepted in the controlled urbanization area. On the other hand, energies decrease in existing village by population decrease and declining birthrate and a growing proportion of elderly people become a problem. In order to cope with this problem, it is important to understand the past conditions of land use for the urban planning. This paper describes the spatial structure of urbanization control districts based on the present conditions and the change structure of land use by using mesh data surveyed and the copy of the development permission register in a local hub-city in Japan. Land use forecasting systems are designed using neural network. Although land use is classified separately in every surveyed year, the common classification of land use is proposed, considering the similarity of spatial distributions and the physical meanings of land use. Then, the distribution by mesh at each division of land use is studied. Spatial distribution of land use and its transition are also discussed. Next, land use forecasting models are made out using neural network. The feature and structure of change in the land use of an area depends on whether development projects are carried out or not. Therefore, all of the meshes are divided into two groups, and forecasting models are designed. Though our proposed approach is a macroscopic forecasting method of land use, it is useful in the investigation of urban policies for development projects and in the evaluation of their effects.