Abstract: This paper gives both details and overview of research development and analytical methods in China’s urban systems since new China was founded. It is noteworthy that some new concepts and methods are introduced and well developed in this paper, which goes beyond the traditional hierarchical-and-order perspective. For example, we introduce the methodology and concepts from complex networks in the statistics physics field into urban studies. In addition to that, we create the urban network that is linked by basic activities among the Chinese cities at the prefecture level, by employing the technical tools such as GIS, Matlab and database analysis. Then this paper describes the changes and predicts the trends based on the empirical and comparative data of 2003 and 2007 in China. The conclusions can be drawn as follows. (1) According to the network structure we can divide urban China into three parts—the Northern China System, the Yangtze River system, and the Southern China System, which display the patterns of "three centers, several cores" in urban China. The directions of linkages among cities focus mainly on the large cities in the Bohai Rim, Yangtze River Delta, and Pearl River Delta. It is indicated that the hierarchical relation in the Bohai Rim can be easily distinguished and that the network linkages in Yangtze River Delta and Pearl River delta are gradually developed. In comparison with top 1 and top 5, it is shown that the directions of the linkage strength for each city in the network of top 10 tend to be more dispersed. In other words, the result is that the linkages in the network of top 1 and top 5 are inclined to be polarized and are likely to be widened in network of top 10. (2) Compared with the random network and the regular network (in this paper we choose the distance network in close relationship), the real networks such as top 1, top 5, top 10 have two typical characteristics—not only have high cluster coefficient in the distance network, but also have short path in random network. (3) It is deduced that the linkages in top 1, top 5, and top 10 tend to be decentralized as time goes by analyzing the cluster coefficient in 2003 and 2007, which means that zonal economic relationship is more open to each other. (4) Average path length decreases slightly in 2007 than in 2003, which reflects that there are close network linkages in urban China system.

Key words: urban system, complex network, degree distribution, cluster coefficient, average path length