Unraveling the accessibility-usage mismatch: Identifying driving factors and weather-sensitive metro stations using GPS data for improved metro competitiveness

Abstract

With the rapid expansion of urban rail transit networks, an accessibility-usage mismatch has been revealed, where taxi trips have pick-up and drop-off points within metro catchment areas, yet commuters choose taxis (i.e., metro-replaceable trips). Existing literature fails to reveal the underlying reasons, fully account for influencing factors, or identify weather-sensitive metro stations. To address these, this study develops a comprehensive framework, including metro-replaceable trip identification, feature extraction, SHapley Additive exPlanations (SHAP) approach for identifying driving factors, and clustering algorithms for recognizing commuting patterns and weather-sensitive metro stations. The main findings are: (1) Weather conditions are the driving factors. (2) Rain prompts a shift from the metro to taxis, while in cold or windy weather, commuters prefer the metro. (3) Thresholds to alleviate the mismatch include travel distances over 7500 (morning) and 6000 m (evening); Metro-Taxi time ratio below 2.6 (morning); Taxi-Metro price ratio over 6.5 (evening); and departure times outside 7:40–8:40 and 17:45–18:55. (4) For commercial metro stations, density is more crucial, while diversity matters more for residential ones. (5) Weather-sensitive stations are located in densely populated residential or commercial areas, with entrances along Wuhan's primary roads. These findings offer valuable insights to promote green commuting and improve metro competitiveness.