Analysis of metro station ridership considering spatial heterogeneity

Zuoxian Gan, Tao Feng, Min Yang (Corresponding author), Harry Timmermans, Jinyu Luo

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

This study aims to explore the role of spatial heterogeneity in ridership analysis and examine the relationship between built environment, station attributes and urban rapid transit ridership at the station level. Although spatial heterogeneity has been widely acknowledged in spatial data analysis, it has been rarely considered in travel behavior studies. Four models (three global models-ordinary least squares (OLS), spatial lag model (SLM), spatial error model (SEM) and one local model-geographically weighted regression (GWR) model) are estimated separately to explore the relationship between various independent variables and station ridership, and identify the influence of spatial heterogeneity. Using the data of built environment and station characteristics, the results of diagnostic identify evidence the existence of spatial heterogeneity in station ridership for the metro network in Nanjing, China. Results of comparing the various goodness-of-fit indicators show that, the GWR model yields the best fit of the data, performance followed by the SEM, SLM and OLS model. The results also demonstrate that population, number of lines, number of feeder buses, number of exits, road density and proportion residential area have a significant impact on station ridership. Moreover, the study pays special attention to the spatial variation in the coefficients of the independent variables and their statistical significance. It underlines the importance of taking spatial heterogeneity into account in the station ridership analysis and the decision-making in urban planning.

Original languageEnglish
Pages (from-to)1065-1077
Number of pages13
JournalChinese Geographical Science
Volume29
Issue number6
Early online date3 Jul 2019
DOIs
Publication statusPublished - 1 Dec 2019

Fingerprint

station
analysis
travel behavior
regression
residential area
urban planning
statistical significance
spatial data
spatial variation
diagnostic
decision making
data analysis
road
China
performance
evidence
built environment
global model
indicator
attribute

Keywords

  • built environment
  • rapid transit ridership
  • spatial heterogeneity
  • spatial models
  • station level

Cite this

Gan, Zuoxian ; Feng, Tao ; Yang, Min ; Timmermans, Harry ; Luo, Jinyu. / Analysis of metro station ridership considering spatial heterogeneity. In: Chinese Geographical Science. 2019 ; Vol. 29, No. 6. pp. 1065-1077.
@article{6c41d849d07b4767b8b8f01ef5bf61de,
title = "Analysis of metro station ridership considering spatial heterogeneity",
abstract = "This study aims to explore the role of spatial heterogeneity in ridership analysis and examine the relationship between built environment, station attributes and urban rapid transit ridership at the station level. Although spatial heterogeneity has been widely acknowledged in spatial data analysis, it has been rarely considered in travel behavior studies. Four models (three global models-ordinary least squares (OLS), spatial lag model (SLM), spatial error model (SEM) and one local model-geographically weighted regression (GWR) model) are estimated separately to explore the relationship between various independent variables and station ridership, and identify the influence of spatial heterogeneity. Using the data of built environment and station characteristics, the results of diagnostic identify evidence the existence of spatial heterogeneity in station ridership for the metro network in Nanjing, China. Results of comparing the various goodness-of-fit indicators show that, the GWR model yields the best fit of the data, performance followed by the SEM, SLM and OLS model. The results also demonstrate that population, number of lines, number of feeder buses, number of exits, road density and proportion residential area have a significant impact on station ridership. Moreover, the study pays special attention to the spatial variation in the coefficients of the independent variables and their statistical significance. It underlines the importance of taking spatial heterogeneity into account in the station ridership analysis and the decision-making in urban planning.",
keywords = "built environment, rapid transit ridership, spatial heterogeneity, spatial models, station level",
author = "Zuoxian Gan and Tao Feng and Min Yang and Harry Timmermans and Jinyu Luo",
year = "2019",
month = "12",
day = "1",
doi = "10.1007/s11769-019-1065-8",
language = "English",
volume = "29",
pages = "1065--1077",
journal = "Chinese Geographical Science",
issn = "1002-0063",
publisher = "Springer",
number = "6",

}

Analysis of metro station ridership considering spatial heterogeneity. / Gan, Zuoxian; Feng, Tao; Yang, Min (Corresponding author); Timmermans, Harry; Luo, Jinyu.

In: Chinese Geographical Science, Vol. 29, No. 6, 01.12.2019, p. 1065-1077.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Analysis of metro station ridership considering spatial heterogeneity

AU - Gan, Zuoxian

AU - Feng, Tao

AU - Yang, Min

AU - Timmermans, Harry

AU - Luo, Jinyu

PY - 2019/12/1

Y1 - 2019/12/1

N2 - This study aims to explore the role of spatial heterogeneity in ridership analysis and examine the relationship between built environment, station attributes and urban rapid transit ridership at the station level. Although spatial heterogeneity has been widely acknowledged in spatial data analysis, it has been rarely considered in travel behavior studies. Four models (three global models-ordinary least squares (OLS), spatial lag model (SLM), spatial error model (SEM) and one local model-geographically weighted regression (GWR) model) are estimated separately to explore the relationship between various independent variables and station ridership, and identify the influence of spatial heterogeneity. Using the data of built environment and station characteristics, the results of diagnostic identify evidence the existence of spatial heterogeneity in station ridership for the metro network in Nanjing, China. Results of comparing the various goodness-of-fit indicators show that, the GWR model yields the best fit of the data, performance followed by the SEM, SLM and OLS model. The results also demonstrate that population, number of lines, number of feeder buses, number of exits, road density and proportion residential area have a significant impact on station ridership. Moreover, the study pays special attention to the spatial variation in the coefficients of the independent variables and their statistical significance. It underlines the importance of taking spatial heterogeneity into account in the station ridership analysis and the decision-making in urban planning.

AB - This study aims to explore the role of spatial heterogeneity in ridership analysis and examine the relationship between built environment, station attributes and urban rapid transit ridership at the station level. Although spatial heterogeneity has been widely acknowledged in spatial data analysis, it has been rarely considered in travel behavior studies. Four models (three global models-ordinary least squares (OLS), spatial lag model (SLM), spatial error model (SEM) and one local model-geographically weighted regression (GWR) model) are estimated separately to explore the relationship between various independent variables and station ridership, and identify the influence of spatial heterogeneity. Using the data of built environment and station characteristics, the results of diagnostic identify evidence the existence of spatial heterogeneity in station ridership for the metro network in Nanjing, China. Results of comparing the various goodness-of-fit indicators show that, the GWR model yields the best fit of the data, performance followed by the SEM, SLM and OLS model. The results also demonstrate that population, number of lines, number of feeder buses, number of exits, road density and proportion residential area have a significant impact on station ridership. Moreover, the study pays special attention to the spatial variation in the coefficients of the independent variables and their statistical significance. It underlines the importance of taking spatial heterogeneity into account in the station ridership analysis and the decision-making in urban planning.

KW - built environment

KW - rapid transit ridership

KW - spatial heterogeneity

KW - spatial models

KW - station level

UR - http://www.scopus.com/inward/record.url?scp=85068869922&partnerID=8YFLogxK

U2 - 10.1007/s11769-019-1065-8

DO - 10.1007/s11769-019-1065-8

M3 - Article

AN - SCOPUS:85068869922

VL - 29

SP - 1065

EP - 1077

JO - Chinese Geographical Science

JF - Chinese Geographical Science

SN - 1002-0063

IS - 6

ER -