Particle‐resolved simulation of aerosol size, composition, mixing state, and the associated optical and cloud condensation nuclei activation properties in an evolving urban plume

Abstract

The recently developed particle‐resolved aerosol box model PartMC‐MOSAIC (Particle Monte Carlo model–Model for Simulating Aerosol Interactions and Chemistry) was used to rigorously simulate the evolution of aerosol mixing state and the associated optical and cloud condensation nuclei (CCN) activation properties in an idealized urban plume. The model explicitly resolved the size and composition of individual particles from a number of sources and tracked their evolution due to condensation, evaporation, coagulation, emission, and dilution. The ensemble black carbon (BC)–specific absorption cross section increased by 40% over the course of 2 days due to BC aging by condensation and coagulation. Threefold and fourfold enhancements in CCN/CN ratios were predicted to occur within 6 h for 0.2% and 0.5% supersaturations (S), respectively. The particle‐resolved results were used to evaluate the errors in the optical and CCN activation properties that would be predicted by a conventional sectional framework that assumes monodisperse, internally mixed particles within each bin. This assumption artificially increased the ensemble BC‐specific absorption by 14–30% and decreased the single scattering albedo (SSA) by 0.03–0.07, while the bin resolution had a negligible effect. In contrast, the errors in CCN/CN ratios were sensitive to the bin resolution for a chosen supersaturation. For S = 0.2%, the CCN/CN ratio predicted using 100 internally mixed bins was up to 25% higher than the particle‐resolved results, while it was up to 125% higher using 10 internally mixed bins. Neglecting coagulation overpredicted aerosol water content and number concentrations (<0.2 μm), causing errors in SSA from −0.02 to 0.035 and overprediction of CCN concentrations by 25–80% at S = 0.5%.