Modeling the Influence of Weather on Transit Ridership: A Case Study from Brisbane,Australia

Abstract

This paper examines the influence that weather conditions exert on transit ridership in sub-tropical Brisbane, Australia. Drawing on transit smart card and weather station records covering a period of six months, the weather-transit ridership relationship across three different travel modes is modeled. Our results reveal that weather does impose an effect on transit ridership and that its influence varies by both transit mode and across weekdays and weekends. On weekdays, the heat index, an indicator of human-perceived temperature, was found to exert a negative effect on bus and train ridership, while ferry ridership was shown to be most sensitive to rainfall. On weekends, the ridership across all three transit modes decreased on rainy days. Results from quantile regression suggest that on weekdays during periods of higher ridership the influence of weather on transit ridership becomes stronger, while on weekends the influence of weather is lessened. We argue that our findings are important for transit agencies in their capacity to unveil how an important predictor of ridership variation, weather, exerts an influence across a transit system. These preliminary findings mark the first step in assembling the necessary evidence to build a more weather-resilient transit system.