Unraveling the relative contribution of TOD structural factors to metro ridership: A novel localized modeling approach with implications on spatial planning

Document Type

Journal Article

Publication Date


Subject Area

land use - impacts, land use - planning, land use - transit oriented development, mode - subway/metro, place - asia, place - urban, ridership - modelling, technology - passenger information, technology - ticketing systems


Transit-oriented development, TOD around metro station areas, TOD clusters, Metro ridership, Geographically temporally weighted regression


TOD (transit-oriented development) has gradually earned the reputation as a promising spatial planning strategy to encourage public transit usage. Many cities across the world, especially those in the global south, have established TOD projects around metro stations. From this standpoint, the paramount question of how TOD is conductive to metro ridership is at the heart of scholarly discourse. Theoretically, the territorial organization of a TOD comprises three structural factors –node (metro station), place (surrounding land use) and their feedback. An accurate judgement of the TOD-metro ridership relationship would not be achieved if we neglected any of the three structural factors forming the TOD architecture. However, the three TOD structural factors have not frequently been considered simultaneously in prior literature. A gap remains regarding the relative contribution of TOD structural factors to metro ridership across time and space. This paper aims to address this unresolved issue using a case study of the Hangzhou metropolitan area in China. As the dependent variable, metro ridership is measured using one week of smart card records, with the three TOD structural factors as explanatory variables and sociodemographic factors as control variables, described by a set of indicators. A novel localized modeling approach using variance decomposition of geographically temporally weighted regression is demonstrated to quantify the spatially and temporally varying relative contribution of TOD structural factors. The results show that both similarities and discrepancies were identified compared to earlier studies. Most importantly, new knowledge was gained, particularly that ‘metro station’ factors contribute the most to metro ridership, followed by the feedback factor, ‘surrounding land use’ factors and sociodemographic factors both on workdays and nonworkdays. Furthermore, our analysis highlights that the relative contribution of TOD structural factors presents noticeable spatial heterogeneities across metro station areas both on workdays and nonworkdays but exhibits only obvious temporal heterogeneity on nonworkdays. Finally, TOD clusters are generated based on the relative contribution of TOD structural factors with implications for spatial planning. This study is believed to open the door for framing locally representative strategies of TOD to stimulate the use of public transit.


Permission to publish the abstract has been given by Elsevier, copyright remains with them.


Journal of Transport Geography home Page: