Detecting urban dynamics with taxi trip data for evaluation and optimizing of spatial planning: The example of Xiamen City, China

Abstract

Commonly, it is very hard to examine underlying urban dynamics due to rapid spatial expansion and land use variations. In this paper, the origin-destination (OD) data extracted from taxi trip data collected in Xiamen, China, covering 30 days was utilized to detect the underlying dynamics of Xiamen City. Specifically, we discretized the study area into 400m*400m grids so that the number of originating points and destination points of the taxi trips could be counted separately within each single grid. Then, heat maps of the taxi mobility were made to achieve a general understanding of urban dynamics. Secondly, we took advantage of the concept of complex networks to analyze the daily taxi trip data. Using a method of community detection, we divided the study area into six main sub-regions called functional self-sufficient zones (FSZs) in which spatial associations are tight and dense. The features of these FSZs helped us to gain a deeper understanding of urban dynamics. Finally, based on this understanding, we further evaluated and optimized the urban spatial planning of Xiamen. Balancing land use allocation was suggested to enhance the multi-centric structure and reduce congestion. This study provides a relevant contribution by exploring the potential of applying taxi trip data to identify urban dynamics revelations and urban planning optimization solutions.