Identifying the Group Differences in the Impact of Haze on Residents' Low-Carbon Travel

This paper matches the large-scale survey data and the corresponding historical weather data to explore how air pollution impacts on low-carbon travel choices. The K-means algorithm is employed to cluster the personal characteristics of residents into five groups according to their travel behavior. The authors take ordered Logit models to identify the group differences in the impact of haze on the five types of low-carbon travel choices, combining with the theory of responsibility attribution and protection motivation theory. The results show that haze has a significant impact on the two groups, namely young office workers and students. The other three groups will not consider the influence of haze when choosing travel vehicles, travel distance, and travel time. The quantity of personally owned automobiles also has a significant impact on the group differences in low carbon travel choices. It is indicated that low carbon travel policies should be considered in the group differences in the future, and efforts should be made from supply and demand sides to guide residents to choose low-carbon travel.