Acta Geographica Sinica ›› 2021, Vol. 76 ›› Issue (2): 459-470.doi: 10.11821/dlxb202102015

• Industrial and Regional Development • Previous Articles     Next Articles

Spatial characteristics of land use based on POI and urban rail transit passenger flow

PENG Shiyao1(), CHEN Shaokuan2, XU Qi1(), NIU Jiaqi3   

  1. 1. Key Laboratory of Transport Industry of Big Data Application Technologies for Comprehensive Transport, Ministry of Transport, Beijing Jiaotong University, Beijing 100044, China
    2. Integrated Transportation Research Centre of China, Beijing Jiaotong University, Beijing 100044, China
    3. School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China
  • Received:2019-07-08 Revised:2020-10-28 Online:2021-02-25 Published:2021-04-25
  • Contact: XU Qi;
  • Supported by:
    Fundamental Research Funds for the Central Universities(2019JBM034);National Natural Science Foundation of China(71621001);National Natural Science Foundation of China(71890972/71890970)


The integrated development of urban rail transit and land use nearby is one of the most important issues for sustainable development of cities. To improve the sustainability of urban rail transit and the rationality of land resource allocation, it is of great importance to understand the dependence relationship between passenger flow of urban rail transit and functions of land use. Regression analysis is the main method to study this relationship. However, the descriptions of land use in existing research are mostly based on sketchy data such as land area, which is difficult to reveal the impact mechanism and spatial effects of land use of various attributes on passenger flow. To this end, this study utilizes Point of Interest (POI) data of Baidu Map to describe land use information, and proposes a fine-grained description method of land use function within the attraction scope of urban rail transit station. Based on the case of Beijing Subway, global regression models with constant parameters and local regression model with variable parameters are employed to study the dependence relationship and spatial effects of coarse and fine-grained land use with outbound passenger flow at morning peak. The case study of Beijing Subway shows that comprehensively considering the tradeoff between the explanatory power and complexity of models, and the effect of dealing with spatial dependence and heterogeneity, the geographically weighted regression (GWR) model with variable parameters has the best estimation compared with the global model with constant parameter. Its interpretation ability is 84%, and Moran's I index of residuals is 0.0001, which can describe the spatial heterogeneity of the dependence of station outbound passenger flow and POI. The results also display that the Beijing's urban rail transit station basically covers the social and economic center of the central city. These areas are usually developed in a high-intensity hybrid manner for land development. Moreover, the impact and spatial characteristics of land use with different attributes and functions on the morning peak outbound passenger flow are significantly different. For example, the morning peak outbound passenger flow is closely related with the land for commercial and business facilities, administration and public services, which are related to housing and employment, and the commuter between the two places. At the fine-grained level, the outbound passenger flow is more dependent on POI of office buildings and government agencies, which are significantly distributed in the central city functional areas and urban core areas with dense employment. The local model with variable parameters based on fine-grained POI can better identify the impact and spatial heterogeneity of various types of land use on station passenger flow. The case study indicates that the dependence of station passenger flow and land use is the superposition of impacts and spatial effects of various attribute functions of land use.

Key words: urban rail transit, land use, passenger flow, Point of Interest (POI), geographically weighted regression