|
The eastern coastal region of China (ECR) has pioneered the nation in urbanization processes and economic development. The support vector machine (SVM) method was used to extract urban areas in ECR from DMSP-OLS night time lights and SPOT vegetation (VGT) images for the period of 1998-2010. Spatial-temporal urban areas changes were revealed using dynamic degree model, urbanization intensification index and the centriod shift model. The grey relational analysis was adopted to indentify the most influenced social-economic factors of urban expansion and the artificial neural network (ANN-BP) model was used to explore their relationship. Results showed that the average annual growth rate of urban sprawl in ECR was about 13.31 %, and the lowest value of 8.11% was found in 2007. Urban expansion of Beijing-Tianjin-Hebei metropolitan region (Jing-Jin-Ji) and the Pearl River Delta characterized from the center to surrounding areas, whereas the urban land development of the Yangtze River Delta showed a connecting pattern. During 2007-2010, the rate of urban expansion was 11237km2 per year, and the urban sprawl intensity reached the peak value of 2.565. The urban development center gradually shifted to the northeast in Beijing, while the urban land moved towards southwest in Yangtze River Delta and the Pearl River Delta. Urban expansion was closely related to the soci-economic factors including GDP, fiscal revenue, construction output, total fixed asset investment and annual per capita disposable income of urban residents derived from gray correlation model. The simulated urban area by BP neural network model had high correlation with actual values.
|
