Closure investigation on cloud condensation nuclei ability of processed anthropogenic aerosols

Whether the cloud condensation nuclei (CCN) ability of aerosol could be predicted by compositions has been long debated. Measurements of sub-micron aerosol compositions and size-resolved CCN activation fraction were conducted at a mountain site (1344 m) near Beijing region during wintertime. The site was influenced in the noon-afternoon by ground anthropogenic sources through convective mixing (CM) and for certain period received aged pollutants by regional advection (RA). By comparing the measured CCN-derived hygroscopicity parameter (κCCNc) with that predicted using chemical composition in bulk (κchem), we found for CM period, κchem over-predicted κCCNc by 71±11% (25±13%) at SS=0.1% (0.3%); whereas for RA-only period, κchem under-predicted by 13±6% (18±11%) at SS=0.1% (0.3%). The former, representing fresher aerosols in smaller size, could be explained by not considering the size-resolved composition, as the bulk measurement mainly reflects the features of larger particle. The latter was proved to result from the depression of droplet surface tension by potential surface-active organics and the possible liquid-liquid phase separation occurring at moderate RH, and a use of depressed surface tension (than pure water) of 0.063±0.002 Jm-2 would reach an agreement. We propose that a hybrid approach combining size-resolved composition and reduced surface tension, for fresher and aged sources respectively, should improve the estimation of aerosol CCN ability.