Weather and cycling: Mining big data to have an in-depth understanding of the association of weather variability with cycling on an off-road trail and an on-road bike lane

Abstract

Although cycling is an easy and popular form of physical activity and urban travel, barriers exist. In particular, cycling is more likely and more severely to be affected by inclement weather than the motorized modes as the cyclists are entirely exposed to outdoor environment. Understanding the weather-cycling relationship is of great importance to academics and practitioners for cycling activity analysis and promotion. This study contributes to an in-depth understanding of how the changes in weather conditions affect cycling on an off-road trail and an on-road (bridge) bike lane at both daily and hourly scales across four seasons. The paper compares the weather-cycling relationship based on day of week and time of day combinations. The autocorrelation effect of cycling itself and the lagging effect of weather elements are also examined. The findings indicate that cycling is significantly self-dependent especially at the finer temporal scales. Weather have a very different influence on bicycle usage of off-road trails versus on-road bike lanes. When it rains its negative impact not only continues but also significantly affects the cycling within previous one hour. At the daily level, weekend cycling on the trail is less likely to be affected by weather as compared to cycling on the bike lane, whilst inverse is true for weekday cycling. Cycling is most likely to be affected by weather conditions in spring and least likely to be affected in winter. Cycling pattern which is more unrelated to weather at the aggregated level tends to be more flexibly adjusted according to the real-time weather conditions at the disaggregated level. Cyclists on weekends especially during the weekend peak hours (11 AM–4 PM) tend to have more flexibility to adjust their cycling schedule before or after the adverse weather conditions than on weekdays. In addition, cyclists with utilitarian purposes are more likely to shift from cycling to other modes (e.g., transit) due to real-time bad weather conditions in weekdays than in weekends, especially during weekday peak hours (7–9 AM and 4–6 PM). The results provide weather officials, transport agencies and research institutions with valuable information for cycling activity analysis and promotion by considering the effects of weather events especially rainfall.