Using decision tree analysis to identify the determinants of residents’ CO2 emissions from different types of trips: A case study of Guangzhou, China

Abstract

Understanding the influencing factors of residents’ CO2 emissions from travel is of importance for developing low-carbon transportation and land-use policies. Based on the 2015 travel survey data of Guangzhou, China, and decision tree analysis, this study identifies the determinants of CO2 emissions from different types of trips by quantifying the relative importance of the influencing factors. The results show that for different types of trips, the factors influencing CO2 emissions vary. Socio-demographics have more significant impacts on CO2 emissions from commuting and social trips than neighborhood built environments, while for CO2 emissions from recreational and daily shopping trips, built environments have more significant impacts than socio-demographics. Car ownership is the most crucial determinant of CO2 emissions for almost all types of trips except shopping trips that are most affected by built environments around neighborhoods, like distance to city public centers and bus stop density. The findings in our study confirm that there are complex interactions between the neighborhood built environment and the socio-demographics of residents, and some of these factors have non-linear effects. Positive and proper planning intervention on residential built environments is one of the critical means to achieve carbon emissions reduction goals.