Cloud Processing of CCN

Poster PDF

Description

High-resolution measurements by the Desert Research Institute (DRI) cloud condensation nuclei (CCN) spectrometers in two aircraft field projects; MASE and ICE-T often showed bimodality that had previously been observed in submicrometer aerosol size distributions of differential mobility analyzers (DMA). Sizes at the minimum concentration between the two modes inferred cloud supersaturation (S) (Hoppel et al. 1985, JGR) but this required particle hygroscopicity to convert size to S actually critical S, Sc of the particles. Since CCN composition is difficult to determine it was usually assumed to be ammonium sulfate, which is not always so. However, since CCN measurements are already in terms of Sc they do not require particle hygroscopicity to infer cloud S. Aerosol bimodality is caused by processes within cloud droplets; gas-to-particle conversions (mainly sulfate), Brownian capture of interstitial haze particles by cloud droplets, and coalescence among cloud droplets. Aerosol bimodality results because these processes are restricted to the activated cloud droplets, which have grown on the lower Sc CCN. Since all of the cloud processes increase the mass of soluble material within cloud droplets they move the Sc distribution toward lower values. When the droplets evaporate a size/Sc gap ensues because the size and Sc of the CCN group that had produced the unactivated haze droplets is unchanged. The DRI CCN spectrometers observed a great deal of spectral shape variability from very bimodal to very monomodal often within close proximity. Cloud S Hoppel minima could be ascertained for 63% of 325 measurements. These estimates were consistently lower than effective S (Seff) determined by comparing ambient CCN spectra with nearby cloud droplet concentrations (Nc). MASE polluted stratus averages were 0.15 and 0.23% and for the ICE-T cumuli 0.44 and 1.03%. This S disagreement between the two methods is first due to the fact that Hoppel minima include the effects of cloud processing, which is the most effective for the larger droplets grown on the lower Sc CCN mainly because of their greater surface areas. Significantly lower concentrations within the more bimodal spectra compared with the monomodal spectra indicated active physical processes because chemistry does not reduce CCN concentrations. Opposite relationships between CCN modality and cloud droplet concentrations (Nc); greater Nc for more bimodal spectra in MASE and lower Nc for more bimodal spectra in ICE-T indicated more chemistry and Brownian capture in MASE and more coalescence in ICE-T.