AAAR 34th Annual Conference
October 12 - October 16, 2015
Hyatt Regency
Minneapolis, Minnesota, USA
Abstract View
CTM-Based Regression for Social Cost Accounting of Individual Emission Sources for PM2.5 Pollution
JINHYOK HEO, Peter Adams, H. Oliver Gao, Cornell University
Abstract Number: 348 Working Group: Source Apportionment
Abstract Due to long-distance transport and secondary formation, fine particulate matter (PM2.5) and its precursor emissions originated from numerous external or cross-regional sources generally account for a significant fraction of the social costs (or public health burden) of PM2.5 pollution within a region or an urban area. For quantifying the contribution of individual sources to the social costs, state-of-the-art chemical transport models (CTMs), though most desirable for accurate quantification of the relationship between an emission source and its social costs, are less than ideal and practical due to their technical complexity and extremely high computational costs, hence too cumbersome for practitioners and policy makers. Recently, CTM-based regression models (with robust out-of-sample validation after calibrated to a CTM-generated database) have been developed, which require minimal computation without sacrificing the technical rigor and accuracy of CTMs. Using such models, this study aims to provide space-and-time resolved social cost accounting of emission sources for PM2.5 pollution at a receptor location. Specifically, we focus on quantifying the fractional contributions of four species (elemental carbon (EC) and inorganic particulate matter precursors (SO2, NOx, and NH3)) emitted anywhere in the United States. As an application, we estimate the fractional public health burden in New York City attributable to inorganic PM2.5 pollution from local and regional emission sources. Preliminary results indicate that local emissions (i.e. emissions within New York City) are responsible for 9% of the total burden, showing its dominant fraction is originated from external sources. Emission sources that cumulatively account for 50% of the burden are as far as 880 km away. Accounting for 90% needs to include sources up to 1,700 km away. Such detailed social cost accounting with improved space-and-time resolution of individual emission sources are critical for successful identification and design of cost-effective and equitable air quality control strategies at various levels.