1.4 Hygroscopicity of Bimodal Particle Distributions as Evidence of Aerosol-Cloud Processing in Two Aircraft Field Campaigns

Monday, 9 July 2018: 9:45 AM
Regency D (Hyatt Regency Vancouver)
Stephen R. Noble, DRI, Reno, NV; and J. G. Hudson

Cloud processing of atmospheric aerosols is one of many aerosol-cloud interactions leading to uncertainty in climate change. Three ways that clouds process aerosols are aqueous oxidation of trace gases dissolved within cloud droplets that add soluble material (chemical); collision and coalescence of droplets to combine soluble material (physical); and Brownian capture of interstitial particles to add soluble or less soluble material (physical). These processes have effects on particle sizes, concentrations, and solubility as well as effects on soluble trace gases. Because most droplets evaporate, the resulting particle is altered from the one initially activated. The cloud condensation nuclei (CCN) become larger and more easily activate at lower supersaturations in subsequent cloud cycles. This activation improvement impacts subsequent cloud droplet number and size, and drizzle amount (Hudson et al., 2015; Hudson et al., 2018), which have implications for climate. Aerosol-cloud processing separates the processed particles (accumulation mode) and unprocessed particles (Aitken mode), forming a gap and bimodal distributions. Hygroscopicity (κ) can be determined from these particle size and CCN distributions, measured by critical supersaturation. Overlaying particle size and CCN distributions and tuning the κ value to produce agreement gives κ for the distribution. Separate tuning for Aitken (unprocessed) and accumulation (processed) modes provides κ for each mode. κ differences between the two modes indicate cloud processing type. In two aircraft field projects; a polluted stratus cloud study (MArine Stratus/stratocumulus Experiment, MASE), and a clean summertime cumulus cloud study (Ice in Cloud Experiment-Tropical, ICE-T); differential mobility analyzers measured particle size distributions while the Desert Research Institute CCN spectrometer measured CCN distributions. κ was then determined for each mode. Collision and coalescence should combine CCN with similar κ. This process would be indicated by similar κ between modes. This occurred in 60% of bimodal distributions in ICE-T but in only 26% of bimodal distribution in MASE. Thus, collision and coalescence dominated in ICE-T. However, aqueous oxidation (chemical processing) dominated in MASE because the two modes had different κ for 74% of bimodal distributions (Fig. 1A). This is consistent with Hudson et al. (2015). Generally, adding hygroscopic material through cloud processing makes processed κ greater than unprocessed κ. However, when unprocessed κ is high, aqueous oxidation adds material that is less soluble (sulfates added to sea salts) and the resulting processed κ tends towards κ of the added material. This was evident in MASE (Fig. 1A). Chemistry measurements in MASE also point to aqueous oxidation of sulfur dioxide to form sulfates. Greater fractions of accumulation mode CCN to total CCN were associated with less SO2 and greater SO42- (Fig. 1B). This also corresponded to lower ozone and higher sodium chloride. The higher droplet pH from sea salt made ozone a more effective oxidizer which promoted oxidation to sulfate from sulfur dioxide. Distributions with a single mode in the Aitken range had more SO2 and ozone with less sulfate (Fig 1B). Cloud processing types in these projects appear to be affected by cloud type. Larger vertical motions (W) in cumulus clouds create larger droplets that coalesce, while shallow stratus clouds with limited W promote chemical processing, especially when polluted. Hygroscopicity of the particle distributions provided an opportunity to observe the extent of aerosol-cloud processing. These processes have implication on climate and air quality. Here, the evidence suggests both types of processing existed to some extent in both projects. However, more investigations among cloud types and in other regions are needed.

Hudson, J.G., S. Noble, and S. Tabor (2015), J. Geophys. Res. Atmos., 120, 3436–3452.

Hudson, J.G., S.R. Noble, and S. Tabor (2018), in review, J. Geophys. Res. Atmos.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner