Estimating passenger transfer time and waiting time in metro-bus bimodal transit systems using multi-source data

Accurate estimation of passenger transfer and waiting times is essential for improving multi-modal transit integration and passenger satisfaction. This study develops a data-driven framework using smart card, vehicle location, and geographic information data to reconstruct stop-level passenger trip chains in metro-bus systems. The framework addresses missing bus boarding-stop information, bidirectional metro-bus transfer identification, and decomposition of transfer time into walking and waiting components. Hour-specific dynamic transfer time thresholds are used to identify transfers, and individual transfer waiting times are estimated after transfer time decomposition. The methodology is tested on a bus-metro bimodal network in Shenzhen, China. Results reveal spatially imbalanced transfer passenger flows and heterogeneous transfer times. Metro-to-bus waiting time is not linearly correlated with walking distance, suggesting that walking distance alone cannot explain variations in transfer waiting time. The findings demonstrate the value of integrated multi-source data for analyzing and improving multi-modal transit transfers.