A study on air pollution exposure of “first and last mile” urban commuters under space-behavior dual verification based on big data, land-use regression model and space syntax

Abstract

Urban traffic-related air pollution (TRAP) exposure is a serious problem during daily commutes. Previous studies assess the TRAP exposure of commuters mainly based on fastest and lowest-dose routes. In addition, the first and last parts of the commutes (“first and last mile”) are often overlooked, wherein the commuters tend to choose active transport (walking, jogging and cycling) and are highly exposed to air pollution. This study uses a space syntax model to simulate the most likely actual commuting routes (behavior-based routes) of “first and last mile” in Changsha, China, and combine it with a land-use regression (LUR) model to assess the TRAP exposure. Furthermore, the air pollution exposure doses of commuters across the behavior-based routes are compared with those across the fastest and lowest-dose routes. The results indicate that only 5.9%–13% of lowest-dose routes do not overlap the fastest routes, and the differences in both average fine particulate matter (PM2.5) and nitrogen dioxide (NO2) exposure are below 1%. However, up to 53.1%–68% of the behavior-based routes do not overlap with the fastest and lowest-dose routes, and the average accumulated exposure of the behavior-based routes exceeds those of the other routes by 7.8%–12.9%. These findings suggest an underestimation of TRAP exposure and the potential benefits of choosing cleaner routes. The results also indicate that commuters living in sub-central areas are linked with relatively larger commuting time costs and higher accumulated exposure doses. This highlights the need to focus on the environmental issues suffered by the commuters in these areas. In addition, our study suggests that behavioral science methods, such as space syntax, can shed new light on air pollution exposure research and develop targeted interventions to address related environmental problems.