The effect of weather on the use of North American bicycle facilities: A multi-city analysis using automatic counts

This study investigates the impact of weather on the use of urban bicycle facilities in Montreal, Ottawa, Vancouver and Portland, as well as on the Green Route in Quebec. This research makes use of long-term hourly and daily counts collected automatically using inductive loop detectors. The count data locations are organized into two groups - utilitarian and recreational. Using regression models with autoregressive and moving average (ARMA) errors, the direct impact and lagged effects of weather variables on hourly and daily bicycle counts are investigated. Among the main findings, temperature and humidity are positively and negatively associated with cycling, respectively, with a non-linear association in most cases. Precipitation has a significant negative impact on cycling flows, and its effect was observed to increase with rain intensity. Lagged effects of rain were also observed, such as the effect of rain in the previous three hours, rain in the morning only, and rain in the afternoon only. Furthermore, urban bicycle flows are more sensitive to weather on weekends than on weekdays, and recreational facilities are more sensitive than utilitarian facilities.