Hoppel minima in detailed CCN spectra

Clear air aerosol spectra often display modes that have been attributed to cloud processing. Especially for bimodal distributions the sizes at the minimal concentrations between modes were used to infer effective supersaturations (S_eff) of nearby clouds (Hoppel et al. 1986). With particle composition assumed this Hoppel minimum diameter was converted to critical S, S_c by using particle solubility (κ) of that compound. This was based on the principle that lower S_c particles preferentially produce cloud droplets. Cloud processing then proceeded only on activated droplets grown on the lower S_c CCN subset. These physical (coalescence among droplets and Brownian capture of interstitial particles) and chemical (gas-to-particle conversion of sulfate or nitrate) processes then increased the soluble content of cloud droplets so that when they evaporated, as they usually do, the dry particles that emerged had even lower S_c than the CCN that had first produced the cloud droplets. These activated CCN thus increased in size (lower S_c) whereas the unactivated CCN did not change size or S_c. This resulted in a size distribution gap at the maximal S_c from which the activated CCN grew to droplets. The size at this Hoppel minimum then represented S_eff of the clouds that had processed the aerosol. Since this gap generally occurred at 70-80 nm diameter and since composition was generally assumed to be ammonium sulfate this inferred S_eff < 0.3% for the maritime stratus clouds under which the observations were made.  

Detailed spectra from the DRI CCNs now reveal bimodality in RICO (wintertime Caribbean cumuli), MASE (polluted California stratus), PASE (central Pacific), POST (California stratus) and ICE-T (summertime Caribbean cumuli). Since CCN measurements are in terms of S, particle composition is not required to estimate S_eff . In MASE and ICE-T there were also simultaneous DMA measurements to compare with CCN spectra by transposing size to S_c by using κ.  κ's that make DMA spectra agree with simultaneous CCN spectra (Fig.) thus provide CCN κ . DMA-CCN agreement was consistent for 227 MASE measurements and 50 ICE-T measurements. But for some bimodal spectra slightly different κ were indicated for the two modes. In Fig. 1a κ 0.35 works best for the higher S_c mode but κ 0.30 is best for the lower S_c cloud-processed mode. A somewhat consistent opposite pattern of κ differences between modes indicated chemical processing in MASE. Since most κ were found to be lower than ammonium sulfate (0.61) chemical processing should then make κ closer to 0.61 for the cloud-processed lower S_c mode. MASE above cloud measurements showed higher κ and less bimodality. These are consistent with the higher N_CCN above and that κ is lower in pollution and less cloud interaction than below cloud. That bimodality was not universal under the ubiquitous MASE stratus suggests less than a well-mixed boundary layer.

Alterations of CCN spectra by cloud processing have important indirect aerosol effect (IAE) implications. Cloud processing enhances the lowest S_c CCN, which are more likely to make cloud droplets. The lower concentrations of these low S_c CCN might then deprive higher S_c CCN, with higher concentrations, from forming cloud droplets. This could be another counter-IAE. At any rate cloud-processing changes alter cloud S_eff, droplet concentrations (N_c), mean diameter, spectral width and cloud albedo.

	flt	#	sec	κ	mode	Seff Hop	Seff spec	Nc	N1%
MASEbelow	9	135	7509	0.40±0.20	4.80±2.06	0.15±0.07	0.17±0.17	255± 88	703±263
MASEabove	9	92	1657	0.24±0.16	6.00±1.86	0.20±0.16	0.17±0.17	255± 88	1215±521
ICE-T	8	50	6162	0.34±0.22	3.26±2.23	0.52±0.26	1.31±0.54	164±72	189±100

Table 1.  Summary of detailed CCN spectral means and standard deviations. Mode is spectral rating (1 to 8); 1 two well separated modes; 8 strictly monomodal.

Hoppel et al., 1986: Geophys. Res. Lett., 13 (1), 125-128.