With the prevalence of smart city, big data and sharing economy, bike-sharing as a mode of transport to lessen carbon dioxide, promote public health, and solve the problem of the first/last mile connection to other transition modes has been popular in China since 2008. This paper conducts a spatiotemporal analysis of bike-sharing in areas under different land uses, especially the commercial area, in Tianhe district, Guangzhou. The data used in this project consist of over 267 thousand sharing bikes recorded on two weekdays plus one weekend in October and November, 2017. Line tracking and density analysis tools are used in the geographic information system to simulate cycling tracks and carry out a spatial visualization analysis of different time and distance distributions of cycling flows during the day. Results show that, in the time aspect, travel of bike-sharing on weekdays happens mainly in the morning rush hours, while travel at the weekend could be more dispersed. In the space distribution aspect, sharing bikes on weekdays are centrally distributed along the Guangyuan expressway while at weekends, bicycles tend to spread out in the periphery, especially the Tianhe road and Zhujiang New Town trading area. The sharing-bikes in the commercial area are characterized by low turnover rate as well as low usage. The paper suggests that the future allocation of sharing bikes should meet the spatiotemporal characteristics of cycling. The present findings could be useful for bike-sharing enterprises and the authority who want to promote the use of sharing bikes, lower marginal operating costs and further encourage the practice of human-oriented planning that addresses citizens' everyday experience and quality of life.

bike-sharing system  /  spatiotemporal characteristics  /  smart city