1 June 2021 in International Journal of Heat and Mass Transfer
22 October 2026 in Computers, Environment and Urban Systems
15 June 2026 in SSRN Electronic Journal
Understanding how environmental conditions and urban design characteristics influence bike ridership is essential for promoting sustainable urban mobility; however, most studies focus on city-wide trends and overlook local variations in microclimate and context. This paper presents a methodology that integrates automated bike count data with high-resolution micrometeorological modeling and urban spatial indicators to quantify how thermal comfort, weather conditions, and streetscape characteristics affect bike ridership at hourly and seasonal timescales. The methodology addresses two critical data limitations in bike count analysis: (i) short temporal gaps due to sensor malfunctions, and (ii) spatial coverage constraints. We develop a temporal count prediction model to impute missing data and a new station prediction model to estimate counts at unmonitored locations. The proposed methodology demonstrates the benefits of using bike counter data and enables cities with different urban contexts to quantify local environmental barriers to biking, predict usage at potential counter locations. Applied to the Brussels Capital Region (BCR) using data from 18 bike counters (2022-2024), the study focuses on weekend leisure biking to isolate environmental comfort effects from commuting patterns. Statistical analysis reveals that rain and wind conditions strongly suppress biking (partial effect size η²: 23-42%), cold stress has larger suppressing effects than heat stress (η²: 16-17% vs. 2.4-2.5%), while greenness alone shows minimal independent boosting effects (η²: 0.2-1.16%), shade shows medium effects during spring and summer periods (η²: 11.4-13.4%), while during heat stress periods greenness show small but enhancing effects (η²: 0.1-1.3%).
25 October 2026 in Sustainable Cities and Society