| Abstract | Urban air pollution in general and traffic-related air pollution in specific has been increasingly concerned issues in the world due to its significant impacts on both environment and human health. The problem has long been studied in developed countries, but not yet in the developing countries. This study aimed to develop a framework for assessment of health impacts from traffic-related air pollution emissions at both local (roadside) and urban scales. This study used integrated monitoring and modeling approach to test a hypothesis that it is possible to develop a comprehensive framework to assess health effects of traffic-related air pollution in urban areas where multiple sources and influencing factors exist.
Air quality monitoring was designed focusing on traffic-related air pollutants including PM10, PM2.5, PM1, black carbon (BC), and BTEX (benzene, toluene, ethylbenzene, and xylenes) group. Simultaneously, the meteorological conditions and traffic flows were recorded at 2 selected roads in Hanoi. The monitoring was done in a winter period (31st December 2014 to 9th February 2015) and a transitional period (23rd September to 11th November 2015) at roadside of 2 roads (a busy road TC and small residential road NN) and one ambient site (AA) located in between the roads in a heavy traffic urban area of Hanoi. High levels of air pollutants were measured at the sites for all monitored pollutants with winter levels were generally above the transitional period levels. BTEX levels dropped significantly from the highest at the TC roadside (131±71 μg/m³ winter, 94±19 μg/m³ transitional), followed by the NN roadside (101±29 μg/m³ winter, 75±22 μg/m³ transitional) and the lowest at AA (30±15 μg/m³ winter, 14±5 μg/m³ transitional) while PM2.5 and BC were rather uniformly distributed among the 3 sites, with levels at AA approaching those measured at roadsides. The multivariate analysis (PCA) for the datasets of bi-hourly air pollutants, meteorology and traffic flows of 4 major vehicle categories (MC&DTC, PC&Taxi, Van&Pick, and Diesel), the BTEX species ratio, and the HYSPLIT airmass trajectories all showed the overwhelming influence of traffic emissions to BTEX and PM measured at the roadsides while air quality at the ambient site presented complex interactions of traffic emission nearby, meteorological conditions, and regional/long-range transport. Indoor and outdoor air quality monitoring at two representative residences in the study area showed the average of I/O ratio of >1.0 for BTEX species indicating the presence of indoor sources. For PM2.5, the I/O ratios were all below 1.0 for both roadside and non-roadside residences implying the dominance of pollutants intruded from outdoor sources.
The photochemical model system of WRF/CAMx was used to simulate PM air quality in the Hanoi Metropolitan Region domain (HRM) for the base year 2010. The model performance for PM simulation was satisfactorily evaluated with limited ground-based observations in term of MFB and MFE statistical criteria. The model captured reasonably the temporal variation of the observed PM levels but underestimated the point-based measurements which may be due to the grid averaging effect. A “what-if” scenario of depletion of 50% motorcycle population in the Hanoi Capital was analyzed following a master plan of development of public transport in the city. The results showed more reduction of annual PM2.5 concentration at urban districts, by 1.1±0.5 μg/m³ with the range of 0.9-1.7 μg/m³ while sub-urban areas experienced only a moderate reduction (<0.5 μg/m³) ranging between 0.2-0.5 μg/m³.
The risk of exposure to traffic-related pollutants was assessed using the monitoring data of BTEX and PM2.5. The results showed a severe risk for those living at roadside residences as compared to those living at non-roadside residences. The cancer risk from exposure to benzene at roadside residents was between 2.6 - 14x10-5 vs. 0.7 - 2.9x10-5 estimated for non-
iv
roadside residents. The health benefit from reduction in traffic emissions due to elimination of 50% MC population in Hanoi Capital was analyzed using the CAMx simulated PM2.5 and BenMAP model. BenMAP outputs showed that the benefits were not only confined to the Hanoi Capital but also spreading to neighboring provinces. The number of avoided mortality in the Hanoi Capital would be 3.9/100,000 people and that for the whole HRM domain would be 2.73/100,000 people. Using the value statistical life (VSL) transferred from the base VSL of OECD countries, the total monetary benefit for the scenario was ranged between $101-275 million USD in 2010 based on all-cause mortality, which account for 0.08-0.2% of total GDP in Vietnam in 2010.
A comprehensive framework for health assessment was developed which showed the need to use both monitoring and modeling tools to link between the traffic emissions and health effects. Future studies should be conducted to overcome several limitations of this study. Specifically, the air quality monitoring should be done over a long period and for additional traffic-related air pollutants such as PAHs. Exposure assessment should be conducted using personal monitoring and also with specific biomarker monitoring to better link to the traffic emissions. Street models could also be applied to study the detail exposure on different side of a studied road and relate the pollution levels to the health effects of people living at each side. A comprehensive questionnaire/survey and toxicological studies should be done to better assess the health effects of traffic emissions. |