Bimodal CCN from Cloud Processing with Effects on Cloud Microphysics

Hudson et al. (2015) (H15) showed that bimodal cloud condensation nuclei (CCN) spectra caused by cloud processing have opposite effects on stratus (Marine Stratus/Stratocumulus Experiment; MASE) and cumulus clouds (Ice in Clouds Experiment-Tropical; ICE-T).  Noble and Hudson (2016; this conference) showed that theoretical predictions of droplet concentrations (N_c) and mean diameters (MD) agree with those observations; bimodal CCN make greater N_c with smaller MD in stratus than do unimodal CCN spectra with the same concentrations.  However, in cumulus clouds, the greater vertical winds (W) cause bimodal CCN to produce lower N_c with larger MD compared to droplet spectra grown on unimodal CCN.  Here Figure 1 shows that this same cloud processing influence on CCN extends all the way to drizzle drop concentrations (N_d).  As in H15 these comparisons are between measured CCN spectra closest to the measured clouds. In typical cloud microphysics fashion competition among droplets reverses this CCN influence for larger cloud droplets and drizzle.  Thus, higher concentrations of small droplets lead to lower concentrations of large droplets and drizzle compared to cases of lower concentrations of small droplets that thus allow greater concentrations of larger droplets and drizzle.  Thus, Figure 1 shows opposite N_d responses to CCN than was the case for smaller, but more numerous, small cloud droplets of H15.  Therefore, lines representing droplet concentrations associated with the various CCN spectra cross each other at intermediate sizes to produce the Figure 1 observations.  More bimodal CCN were associated with stratus clouds that contained lower N_d (Figure 1A) whereas in the cumulus clouds more bimodal CCN spectra were associated with clouds that contained higher N_d (Figure 1B).  Figure 1A shows a nearly the same hierarchies for CCN modality and N_d.  The most unimodal CCN (gray) were associated with clouds that contained the highest N_d while the most bimodal CCN (black) were associated with clouds that contained the lowest N_d. The blue line (5) is only one of the 8 lines that is out of order in Figure 1A.  Figure 1B shows that the four most bimodal CCN spectra (black, red, green and yellow) are associated with clouds that have two orders of magnitude greater N_d than the clouds associated with the most unimodal CCN (blue, pink, cyan and gray).  Droplets larger than 210 µm were not even detected in clouds associated with CCN ranked 6 (pink) and 7 (cyan) in terms of bimodality.  Furthermore, the most bimodal CCN (black) were associated with clouds that had the highest N_d for sizes larger than 300 µm while the most unimodal CCN spectra were associated with clouds that contained the lowest N_d for all measured sizes.  Thus, cloud processing of CCN seems to enhance first (cloud albedo; H15) and second (cloudiness; Figure 1) indirect aerosol effect (IAE) in stratus and reduce both IAE in cumuli. The main reason for the noted differences between these cloud systems is that chemistry in the dominant processing mechanism in stratus whereas drop coalescence is the predominant process in cumuli because of their higher W, supersaturation (S), MD, LWC and thickness compared to stratus.    

On the other hand, since cloud processing makes the best CCN (largest particles; lowest critical supersaturations [S_c]) they can suppress cloud S, especially in stratus, which already have lower S due to lower W. This renders many higher S_c CCN irrelevant to cloud microphysics in the very clouds that are the most important for IAE, marine stratus.  Since anthropogenic CCN that cause IAE generally have high S_c cloud processing of CCN would thus seem to moderate IAE, which would be IAE buffering.  Added to this is the fact that many high S_c particles are removed by Brownian capture by cloud droplets.  This is especially so for stratus, where the low S allows more unactivated interstitial material that is eligible for Brownian capture.  Thus, cloud processing may actually reduce IAE. 

Fig. 2 substantiates that cloud processing is the primary cause of aerosol bimodality from surface particle size spectra and cloud remote sensing at the Oklahoma DOE ARM site .  N_u is the concentration of particles within the unprocessed mode, those that are smaller than the Hoppel minima.  Hoppel minima denote the separation between the two observed size or critical S (S_c) modes, which thus define aerosol bimodality.  N_p is particle concentration of the larger size (or lower S_c) mode, those larger than Hoppel minima that result from cloud processing. Thus, smaller (N_u-N_p)/(N_u+N_p) indicates bimodality and higher (N_u-N_p)/(N_u+N_p) indicates unimodality. Figure 2 shows a bimodal aerosol response to the lowest cloud base altitudes (CBA; black) and to the greatest cloud fractions (CF; blue).  The red data, which is the same for the highest CBA and lowest CF shows that the aerosol responds toward unimodality for either of these clear air situations.  Figure 2 displays mean aerosol modality for various hours after (positive time lags) or before (negative time lags) the cloud characteristics were observed at zero hour.  Two different instruments (ceilometer and LIDAR) show these same aerosol responses to CBA and four different instruments (including ceilometer and LIDAR) show these same aerosol responses to CF.  There are many other figures that further substantiate the Figure 2 indications.  This study demonstrates that more economical surface measurements, compared to aircraft measurements, are useful for advancing understanding of cloud processing and its effects on IAE.

The fact that clouds make the best CCN must be accounted for in order to understand and quantify IAE.  Significantly more research will be needed to even determine the direction of the effect of cloud processing on IAE let alone the magnitude of this effect on IAE.

Hudson et al. (2015), JGRA, 120, 3436-3452.