Authors: Yuqing Long*, Peking University, Yanguang Chen, Peking University
Topics: Applied Geography, Spatial Analysis & Modeling, China
Keywords: China; Urban road network; Multifractals; Agglomeration and diffusion
Session Type: Paper
Presentation File: No File Uploaded
Urban road network is typically a complicated open giant system with scale-free characteristic. Traditional single fractal model may not capture its spatial complexity effectively. In this paper, we perform a multifractal analysis for urban road network by box-counting method. Two sets of fractal parameters, including global and local parameters are employed to spatially analyze urban road network of 12 representative cities in China. Taking the urbanized area defined by City Cluster Algorithm as city boundary, we find that the spatial structure of road network in all of these 12 cities take on a clear fractal structure with multifractal attributes. First, most of the capacity dimensions D0 are near 1.9, and the information dimensions D1 and correlation dimensions D2 are near 1.8, suggesting a high coverage of roads in cities. Second, the dimension spectrum curves show several abnormal aspects. On the one hand, the upper limit of the global dimension severely breaks through the Euclidean dimension 2, which indicates the disordered expansion of roads in suburbs. On the other hand, the singularity spectrums of local dimension incline to the left, showing an agglomerative pattern of road distribution. We reveal some structural problems in 12 different cities by transversal comparison. The conclusions may be instructive for spatial optimization of urban road network in the future.