Examining the impact of urban-rural spatial structure on mobility networks: A case study of taxis

Urban livability and sustainability require a deeper understanding of the impact of urban spatial structure on human mobility. Here two mobility networks generated from two taxi data from two cities (Chengdu and New York City) are used to explore human intra-urban mobility. Through the network analysis, we found that many measures of two mobility networks present a two-segment feature. Combined with the map and population distribution, we demonstrate that the two segments correspond to the central and non-central regions of the city, respectively, which means hierarchical mobility related to urban-rural structure exists in both cities, although they have different levels of urbanization. We comprehensively analyze the impact of urban-rural spatial structure on the nature of mobility networks. The central and non-central regions differ significantly in the heterogeneity of human mobility and present different ranges and directions of trips, like the large degree nodes, showing a flared entry and exit pattern. We improved a gravity model by introducing flows to and from the node instead of population to predict the traffic between different nodes, and the prediction results also vary significantly between the two regions. These findings are valuable for understanding the relationship between urban-rural spatial structure and human mobility.