Spatial Dependency of Urban Sprawl and the Underlying Road Network Structure

The spatial correlation between urban sprawl and the underlying road network has long been recognized in urban studies. Accessibility to road networks is often considered an approximation for the measurement of human mobility, which is a key factor in determining potential urban sprawl in the future. Despite the close relationship between urban development and road networks, the spatial dependency of these two spatial layers has never been systematically evaluated. This study conducted a comprehensive investigation on the spatial dependency between these two spatial layers using an urban expansion data set between 2000 and 2010 of East Asian regions and the road network data from OpenStreetMap. Four Chinese cities, namely Beijing, Shanghai, Chengdu, and Shenzhen, were selected to conduct the analysis. The spatial correlations between the urban sprawl and road networks were first quantitatively analyzed using Ripley’s cross-$K$ function. Highly significant spatial correlation has been observed in all four tested cities. A Bayesian network model was also developed to verify the predictability of urban sprawl using the spatial and structural features extracted from the existing road networks as well as the spatial pattern of the past built-up areas. The results show an affirmative answer to the predictability of urban sprawl by achieving an overall accuracy of 79% in classifying urban sprawl and undeveloped areas. Finally, the hidden dependencies among the urban sprawl and the extracted spatial features were interpreted and analyzed based on the Bayesian network structure learned from the data.