To travel or not to travel. ‘Weather’ is the question. Modelling the effect of local weather conditions on bus ridership

Abstract

While the influence of weather on public transport performance and ridership has been the topic for some research, the real-time response of transit usage to variations in weather conditions is yet to be fully understood. This paper redresses this gap by modelling the effect that local weather conditions exert on hourly bus ridership in sub-tropical Brisbane, Australia. Drawing on a transit smart card data set and detailed weather measurements, a suite of time-series regression models are computed to capture the concurrent and lagged effects that weather conditions exert on bus ridership. Our findings highlight that changes in particularly temperature and rainfall were found to induce significant hour-to-hour changes in bus ridership, with such effects varying markedly across both a 24 h period and the transit network. These results are important for public transport service operations in their capacity to inform timely responses to real-time changes in passengers’ travel demand induced by the onset of particular weather conditions.

Francisco Rowe

Professor of Population Data Science

My research interests include human mobility and migration; economic geography and spatial inequality; geographic data science.