Measuring by Movements: Hierarchical Clustering of Cities in China Based on Aggregated Massive Positioning Data

Abstract

The world is becoming linked more and more. A shift of researching focus can be observed recently, from “city as a system” to “systems of cities,” given the context of fast-changing communicating technologies such as high-speed railways (physical) as well as social media over the internet (nonphysical). Flows play essential roles for a city network, indicating the trends of position and functions within the network. In this study, we adopt new type of aggregated positioning data of massive internet users in China to explore the spatial patterns of cities during the Spring Festival in 2015. By introducing new clustering algorithm highlighting spatial constraints, models output hierarchic results with vary regional zones containing different number of cities. The higher layer of results with less members is not similar to the conventional delineation according to the conditions of physical and economic geography of China. Nevertheless, the very differences suggest hidden forces driving cities connected intensely across the administrative boundaries such as sharing mutual regional cultures or employment markets. These facts grounded for a general picture for the study on polycentric urban regions over the whole national territory.