Evaluation of a highly condensed SAPRC chemical mechanism and two emission inventories for ozone source apportionment and emission control strategy assessments in China.
Academic Article
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
The response of summertime O3 to changes in the nitrogen oxides (NOx) and volatile organic compounds (VOC) emissions, and contributions of different NOx and VOC sources to O3 in China are studied using a highly condensed photochemical mechanism in the Statewide Air Pollution Research Center (SAPRC) family (CS07A) and two popular anthropogenic emission inventories, the Multi-resolution Emission Inventory for China (MEIC) and Regional Emission inventory in ASia (REAS). Although CS07A predicts slightly lower O3 concentrations than the standard fix-parameter version of the SAPRC-11 mechanism, the two mechanisms predict almost identical relative responses to daily maximum 8-hour O3 (O3-8h) due to NOx and VOC emission reductions. A source-oriented version of the CS07A is applied to determine source contributions of NOx and VOCs to O3 using MEIC and REAS. The two inventories lead to similar model performance of O3, with MEIC predicting higher O3 in Beijing and Shanghai, especially on high O3 days. Source apportionment results show that industry and transportation are the top two contributors to non-background O3 for both inventories, followed by power and biogenic emissions. In general, the two inventories lead to similar source contribution estimations to O3 attributable to NOx. However, their estimations of relative contributions to VOC-related O3 differ for the industrial and transportation sectors. Differences in the source apportionment results are more significant in some urban areas, although both emissions capture the spatial variations in the source contributions. Our results suggest that future emission control policies should be assessed using multiple emission inventories, and the condensed CS07A is suitable for policy applications when a large number of simulations are needed.
