Citation

Hsu CK. Travel Behav. Soc. 2024; 34: e100710.

Copyright

(Copyright © 2024, Elsevier Publishing)

DOI

10.1016/j.tbs.2023.100710

PMID

unavailable

Abstract

Various weather conditions, ranging from heavy rain, and gusts, to extreme heat, can influence road injury frequency through impacting driver behavior, with potentially increasing impacts in the context of a changing climate. Existing studies often used traditional season categories (e.g., four seasons) to stratify regression models, isolating certain weather effects on road injury frequency. However, this approach falls short in characterizing non-temperate climate regions, where traditional seasons may not apply. For instance, tropical summers often involve rainfall and carry different meteorological implications than the traditional definition of summer, which can have unique effects on road injury frequency. To this end, this study aims to investigate the varying levels and directions of weather effects on road injury frequency given different prevailing climatic conditions, with a particular geographic emphasis on non-temperate areas by using Kaohsiung, a tropical monsoon climate city, as the study area. This study first employed principal component analysis and cluster analysis to reveal the underlying climatic patterns unique to Kaohsiung in a data-driven manner. Subsequently, this study analyzed and compared the weather effects on road injury frequency across different climatic patterns. The findings show that adverse weather conditions, including heavy rainfall, sub-optimal temperatures, and strong wind, contribute to increased road injury frequency, especially in congested weekday traffic. Moreover, these weather effects on road injury frequency can vary in both magnitude and direction depending on the prevailing seasonal conditions. This echoes previous research findings regarding the effects of rainfall and temperatures while also shedding light on the understudied impact of wind. Based on these findings, this study confirms the effectiveness of integrating data-driven techniques and highlights practical policy implications concerning weather effects on road injury frequency. Further work should delve deeper into specific behavioral mechanisms through which weather factors influence road injury frequency given certain climatic conditions.

Language: en

Keywords

Cluster analysis; Principal component analysis; Road injury frequency; Seasonality; Tropical monsoon climate; Weather