Abstract The studies of polycentric urban spatial structure in the traditional perspective generally focus on the static spatial distribution of population density. They used census data to extract, and used mathematical model to simulate these distributions, and then find the evidence of the polycentric characteristics. In contrast, the spatial association perspective tends to build urban spatial network based on "flow" and "spatial interaction", and then identify the polycentric characteristics from the nodes of these networks. Taking Shenyang as example, absorbing the experience of the famous research teams POLYNET, an empirical study on polycentric urban spatial structure in the perspective of spatial association was implemented in this paper. The analyzing process was based on three investigating aspects including population density which stood for the traditional perspective, jobs- resides spatial network and enterprise spatial organization network which stood for the spatial association perspective. Population centers, job centers, and industry centers were identified from these three investigating aspects respectively. Finally the conclusions were drawn by comprehensive analysis of these centers. From the comprehensive analysis of these centers, we found that:①urban core area presents criss-cross structure composed by a couple of job center and a couple of population center; ②Hunnan new district shows a certain degree of centrality; ③there is an obvious spatial mismatch among the population centers, job centers, and industry centers in the urban periphery; ④Shenyang owns an unbalanced polycentric urban spatial structure. This paper demonstrated that, with the help of the perspective of spatial association, we could find new results and interpret new connotation of polycentric urban spatial structures,which is difficult for using traditional perspective. Thus, it is an effective tool to study the polycentric issues in the urban scale.
2014, Vol. 29 
