Diagnostic analysis of the effects of weather condition on pedestrian crash severity.

Pedestrians are vulnerable to severe injury and mortality in road crashes. Numerous studies have attempted to identify factors contributing to crashes and pedestrian injury risks. As an active transport mode, the act of walking is sensitive to changes in weather conditions. However, comprehensive real-time weather data are often unavailable for road safety analysis. In this study, we used a geographical information system approach to integrate high-resolution weather data, as well as their corresponding temporal and spatial distributions, with crash data. Then, we established a mixed logit model to determine the association between pedestrian crash severity and possible risk factors. The results indicate that high temperature and the presence of rain were associated with a higher likelihood of Killed and Severe Injury (KSI) crashes. Also, we found the interaction effects of weather condition (hot weather and presence of rain) on the association between pedestrian crash severity and pedestrian and driver behaviors to be significant. For instance, the effects of jaywalking and risky driving behavior on crash severity were more prevalent under rainy conditions. In addition, the effects of driver inattention and reckless crossing were more significant in hot weather conditions. This has critical policy implications for the development and implementation of proactive traffic management systems. For instance, real-time weather and traffic data should be incorporated into dynamic message signs and in-vehicle warning systems. Doing so will enhance the levels of safety awareness of drivers and pedestrians, especially in adverse weather conditions. As a result, pedestrian safety can be improved over the long term.