PNNL

Abstract

Recent laboratory and field measurements by a number of groups show that secondary organic aerosol (SOA) evaporates orders of magnitude slower than traditional models assume. In addition, chemical transport models using volatility basis set (VBS) SOA schemes neglect gas-phase fragmentation reactions, which are known to be extremely important. In this work, we present modeling studies to investigate the implications of non-evaporating SOA and gas-phase fragmentation reactions. Using the 3-D chemical transport model, WRF-Chem, we show that previous parameterizations, which neglect fragmentation during multi-generational gas-phase chemistry of semi-volatile/inter-mediate volatility organics ("aging SIVOC"), significantly over-predict SOA as compared to aircraft measurements downwind of Mexico City. In sharp contrast, the revised models, which include gas-phase fragmentation, show much better agreement with measurements downwind of Mexico City. We also demonstrate complex differences in spatial SOA distributions when we transform SOA to non-volatile secondary organic aerosol (NVSOA) to account for experimental observations. Using a simple box model, we show that for same amount of SOA precursors, earlier models that do not employ multi-generation gas-phase chemistry of precursors ("non-aging SIVOC"), produce orders of magnitude lower SOA than "aging SIVOC" parameterizations both with and without fragmentation. In addition, traditional absorptive partitioning models predict almost complete SOA evaporation at farther downwind locations for both "non-aging SIVOC" and "aging SIVOC" with fragmentation. In contrast, in our revised approach, SOA transformed to NVSOA implies significantly higher background concentrations as it remains in particle phase even under highly dilute conditions. This work has significant implications on understanding the role of multi-generational chemistry and NVSOA formation on SOA evolution in the atmosphere.