Regional modeling of secondary organic aerosol formation over eastern China: The impact of uptake coefficients of dicarbonyls and semivolatile process of primary organic aerosol.

Abstract

Capturing the secondary organic aerosol (SOA) concentration using the chemical transport model is difficult due to a large knowledge gap of its formation mechanism. Previous studies demonstrated the uptake of dicarbonyls and semivolatile process of primary organic aerosol (POA) emissions are the significant sources of SOA. However, the uptake coefficients of dicarbonyls have large uncertainties and the SOA from the semivolatile process of POA emission remains unclear. We applied the revised reactive uptake parameterization, with salting effects for dicarbonyls, and updated approaches for POA to the Community Multiscale Air Quality Modeling System (CMAQ) simulations for October 2014 to study their impacts on modeling the SOA formation over eastern China. We introduce a method of quantifying crystalized or deliquescent aerosols to further improve the parameterization. The revised glyoxal uptake coefficients results in higher glyoxal SOA in the Beijing-Tianjin-Hebei region, where is typically under low relative humidity (RH) and high aerosol pH conditions. It gives lower glyoxal SOA in the Pearl River Delta region, where is typically under high RH and low pH conditions. The updated parameterization gives negligible methylglyoxal SOA due to the low uptake coefficients. The implementation of semivolatile process of POA and the approach for potential SOA from combustion sources will largely decrease the predicted POA and increase the modeled SOA concentrations over eastern China. The increased SOA from POA emissions could improve the model performance for organic carbon and SOA. It slightly improves the performance in PM2.5 modeling by compensating the reduction of modeled POA. This study indicates the mixed impact of a parameterization considering salting effects on modeling the dicarbonyls SOA in key regions of eastern China. It also demonstrates the improved performance by implementing the POA approaches in aerosol modeling using CMAQ. Meanwhile, the uncertainty in the revised reactive uptake parameterization and POA approaches is discussed.