Characterizing growth types in a metropolitan region as additional factors for simulating a city urbanization

Bekasi city, in the Jakarta Metropolitan Region, is experiencing rapid urban growth and urban sprawl which have negative impacts on the natural resources, economy, public health, and community character. Good urban-growth model is important for the city planners to manage policies and practices. This study identified three urban growth types through a combination method of remote sensing and statistics. Two growth types were found, i.e. the infilling and edge-expansion, that were used as additional drivers of land-use/cover change model in this study. The model was trained using artificial neural network with two scenarios. First scenario used ordinary drivers, e.g. proximity to river, road, commercial areas, etc. whereas the second scenario add proximity to urban growth as well. Validation result found that the second scenario was more accurate than the first one. Another two scenarios were employed to simulate the future growth (2030 and 2050), i.e. business-as-usual and conservation scenarios. Conservation scenario that minimize the effects of diminishing vegetation and agriculture showed better result.