Evaluation and Improvement of the Interchange from Bus to Metro Using Smart Card Data and GIS

Abstract

The ridership of a metro system depends largely on the feed from other modes, especially buses. Therefore the quality of the interchange between metro and buses is a significant factor affecting the attractiveness of a metro system. However, it is costly and time-consuming to evaluate the interchange quality between a metro station and its feeding bus stops citywide via interviews or field studies. This study provides a method to evaluate the interchange quality from bus to metro in both time and space dimensions using smart card data and geographic information system (GIS) tools. Alighting time tags from buses and boarding time tags at metro stations using smart card data are used to infer interchange time, and the straight-line distances and network distances between a metro station and its surrounding bus stops are extracted using GIS data. By combining interchange times and distances, metro stations citywide can be clustered into groups in terms of interchange characteristics, and stations with poor connections with buses are thus distinguished. Underlying causes of interchange problems, such as bus stops located too far away or low road connectivity in the vicinity of metro stations, can be initially determined. This method is applied in Beijing, and all the stations on the metro network are evaluated from the perspective of interchange quality with buses. From the perspective of interchange time, 82% of the total stations have connections with buses within 1 min walking, with an average of 48 s at the city level. The overall quality of bus–metro integration is good, but for approximately 5% of stations, passengers have to walk more than 15 min from bus to rail. Spatial analysis reveals that the average interchange distance at the city level is 343 m. The stations with long interchange times are successfully identified, and the reasons for the low connection quality are interpreted in combination with spatial analysis of the rail station and bus stops. Finally, the results are partially validated in a field study and measures to enhance bus–metro connections are provided.