Investigate the influence of urban heat island effects on active travel behavior by leveraging social media data

This study investigates the impacts of urban heat island (UHI) effects on active transport by leveraging social media data. A multiscale geographically weighted regression (MGWR) model is utilized to investigate the spatial heterogeneity of UHI, built environment, and sociodemographic factors on travel frequency for both peri-summer and all-year trips. The investigation is showcased in Greater Melbourne, Australia, where Twitter posts related to active transport were collected and analyzed to identify active transporters’ travel frequency in different suburbs. The results demonstrate the effectiveness of the MGWR model, revealing that UHI has a significant negative impact on all suburbs, with greater intensity during peri-summer trips. Additionally, the density of tram stops has a positive influence, particularly during peri-summer trips. Furthermore, the parkland areas, population density, and young adults have positive effects, while the unemployment rate and dwellings with only one motor vehicle have negative impacts. This study contributes to the field of travel behavior analysis by utilizing social media data as a new and cost-effective source of travel behavior information. Moreover, it offers novel insights by examining the spatial variations of UHI effects on active transport, addressing a research gap in this area.