Development of a new method to estimate the regional and local contributions to black carbon

Abstract

Abstract Purpose Much attention has been focused on the role of elemental carbon or its surrogate black carbon (BC) in PM-mediated toxicity. It is possible that its toxicity is related to the age or origin of the particles. This study presents methods to apportion ambient BC into its regional and local components and, once estimated, characterizes their seasonal and daily variability. Methods We estimated regional and local BC concentrations in Boston, MA using quantile regression of ambient BC on ambient sulfur. These estimates were compared to estimates obtained using alternative approaches, including quantile regression methods that used sulfate rather than sulfur as a predictor, the use of rural EC concentrations as proxies for regional BC, and source contributions estimated using EPA’s Positive Matrix Factorization (PMF). Results Regional BC was found to comprise on average 44% of ambient BC in Boston. Mean regional levels equaled 0.289 (±0.126) μg m−3, while average local BC concentrations equaled 0.372 (±0.231) μg m−3. Local BC concentrations varied by season, day of week and hour, with levels highest during weekdays, summer and between 6am and 2pm. Similarly, regional BC concentrations were highest during the summer months. Day of week differences in regional BC concentrations were small. Regional BC did not vary by hour of day. Our estimates agreed well with corresponding estimates of regional and local BC based on sulfate, rural EC concentrations, and PMF methods. Conclusions Quantile regression is an appropriate method to estimate local and regional BC concentrations. In Boston, we found that regional and local BC comprise approximately equal fractions of total ambient BC. Results suggest the utility of this method for epidemiologic studies and air quality management.