Aqueous-phase reactive uptake of dicarbonyls as a source of organic aerosol over eastern North America

Abstract

We use a global 3-D atmospheric chemistry model (GEOS-Chem) to simulate surface and aircraft measurements of organic carbon (OC) aerosol over eastern North America during summer 2004 (ICARTT aircraft campaign), with the goal of evaluating the potential importance of a new secondary organic aerosol (SOA) formation pathway via irreversible uptake of dicarbonyl gases (glyoxal and methylglyoxal) by aqueous particles. Both dicarbonyls are predominantly produced in the atmosphere by isoprene, with minor contributions from other biogenic and anthropogenic precursors. Dicarbonyl SOA formation is represented by a reactive uptake coefficient γ = 2.9 × 10 −3 and takes place mainly in clouds. Surface measurements of OC aerosol at the IMPROVE network in the eastern U.S. average 2.2 ± 0.7 μg C m −3 for July–August 2004 with little regional structure. The corresponding model concentration is 2.8 ± 0.8 μg C m −3, also with little regional structure due to compensating spatial patterns of biogenic, anthropogenic, and fire contributions. Aircraft measurements of water-soluble organic carbon (WSOC) aerosol average 2.2 ± 1.2 μg C m −3 in the boundary layer (<2 km) and 0.9 ± 0.8 μg C m −3 in the free troposphere (2–6 km), consistent with the model (2.0 ± 1.2 μg C m −3 in the boundary layer and 1.1 ± 1.0 μg C m −3 in the free troposphere). Source attribution for the WSOC aerosol in the model boundary layer is 27% anthropogenic, 18% fire, 28% semi-volatile SOA, and 27% dicarbonyl SOA. In the free troposphere it is 13% anthropogenic, 37% fire, 23% semi-volatile SOA, and 27% dicarbonyl SOA. Inclusion of dicarbonyl SOA doubles the SOA contribution to WSOC aerosol at all altitudes. Observed and simulated correlations of WSOC aerosol with other chemical variables measured aboard the aircraft suggest a major SOA source in the free troposphere compatible with the dicarbonyl mechanism.