Clouds and precipitation have long been recognized as important sinks for atmospheric aerosol, with size and solubility-induced selective removal affecting both the aerosol size distribution and chemical nature of the aerosol population [cf., Junge, 1963; SMIC, 1971]. The role of clouds as instruments of differential transport of aerosols in the vertical has also been appreciated and studied throughout the history of research on aerosol-cloud interactions [cf., Rodhe, 1983; Kleinman and Daum, 1991; Cautenet and Lefeivre, 1994]. More recently, chemical processes within cloud drops have been shown, in a plethora of experimental and theoretical studies, to be capable of altering the size and composition of aerosol particles processed by them [e.g., Easter and Hobbs, 1974; Hegg et al., 1980; Flossmann et al., 1987; Hoppel et al., 1990, 1994; Leaitch, 1996; Bower et al., 1997; Bott, 1999]. Still more recently, support for nucleation of new particles in the near-cloud environment has been forthcoming [e.g., Hegg et al., 1990; Frick and Hoppel, 1993; Wiedensohler et al., 1997].

Taken as a whole, the above body of research suggests that there are essentially four processes by which clouds can impact aerosol populations:

• Vertical transport
• Scavenging processes (both with water and ice hydrometeors)
• Chemical processes in cloud drops
• Particle formation near clouds

Of course, these mechanisms commonly act in concert and frequently all can significantly impact aerosol populations within a single cloud event (as illustrated schematically in Figure 1). Indeed, it is useful to consider these processes as terms in a continuity equation for particles of a particular size range, i.e., to visualize the problem of cloud processing the way a prognostic model would treat it. Viewed from this perspective, all these processes are clearly operable simultaneously; it is only their relative magnitudes that change with conditions. Furthermore, given that a set of such continuity equations must be utilized (one for each size range), the size-dependent nature and impact of all of these processes becomes clear. For example, chemical processes which add mass to pre-existing particles act as a sink for the relatively small initial particles but are a source of larger particles. It also becomes clear that these processes can act not merely simultaneously but also synergistically, as when scavenging of pre-existing aerosols aids near-cloud particle nucleation. Nevertheless, in the space available, it is only feasible to address these processes separately and delineate the salient aspects of each.

Vertical transport

If vertical transport could be isolated from the other processes just introduced, it would be the most straightforward mechanism by which aerosol populations are altered by clouds. Convective clouds in particular offer a rapid pathway for the vertical transport of air from the boundary layer into the free troposphere. Because nearly all primary aerosol sources are surface-based, this rapid transport affords a means of getting aerosols into the free troposphere without the sometimes substantial changes incurred by aging processes. Furthermore, while such transported aerosols will typically be much reduced in concentration compared to boundary layer values, they can still be sufficiently high to significantly perturb concentrations aloft [e.g., Kleinman and Daum, 1991]. Nevertheless, as noted by Rodhe [1983], such cloud-convective transport must necessarily involve wet-scavenging processes, processes which will generally be the dominant modulators of the aerosol size distribution. Hence, in this brief review, we turn next to these processes.

Scavenging processes

Figure 1. A schematic diagram of the in- or near-cloud processes that can alter the aerosol size distribution.

The term 'scavenging processes' encompasses a wide variety of cloud-aerosol interactions, as illustrated in Figure 1. These include nucleation scavenging of cloud condensation nuclei (CCN), diffusion scavenging of interstitial aerosol by the activated cloud drops (or ice particles) and impaction of aerosols onto both ice and water hydrometeors sufficiently large to have appreciable fall velocities. Furthermore, as noted by Hudson and Frisbie [1991], the collision-coalescence process will act as an effective aerosol modification mechanism because the larger drops formed will yield, in principle, single aerosol particles upon evaporation either below cloud or in detraining air aloft. Similarly, collection processes, which involve capture of cloud drops by larger falling hydrometeors, would yield the same results if subsequent evaporation was significant. Hence, these processes all act as sort of aqueous-enhanced coagulation processes. In consequence, even processing by non-precipitating clouds can significantly alter the aerosol size distribution, though no aerosol mass may actually be removed from the atmosphere in the immediate process. For precipitating clouds, of course, actual removal in precipitation must be added to all the other processes that impact the aerosol size distribution.

The results of the action of these mechanisms will vary from case to case but several broad results can be generally invoked. First, there is a tendency for the size distribution to broaden as well as for number concentrations to decrease [e.g., Flossmann et al., 1985; Alheit et al., 1990; Pandis et al., 1990; Noone et al., 1992; Khain et al., 1999]. Second, at least for non-precipitating clouds, a marked minimum in the size distribution will form at the particle size corresponding to the effective maximum supersaturation (and the smallest activation radius) achieved in the processing cloud(s) [cf., Fitzgerald et al., 1998; Feingold et al., 1996]. This arises from the various cloud drop collection processes discussed above and, if rainout is minimal, can actually create a new mode in the size distribution (as can sulfate production, discussed below). There is considerable observational support for this nucleation minimum, at least in marine air [e.g., Hoppel et al., 1990]. However, the location (in size) and depth of this minimum will depend not only on the aerosol size distribution but on its chemical composition as a function of size. Hence, this topic leads us next into the general area of differential scavenging due to aerosol composition.

The activation of aerosol particles to form cloud drops, commonly referred to as nucleation scavenging, is by far the most important mechanism for aerosol mass scavenging and commonly results in a drastic depletion of the accumulation mode in cloud. The well-known Köhler equation, which governs this process, clearly shows the dependence of the process on aerosol composition and, in fact, numerous field studies have documented this dependence for atmospheric aerosols [e.g., Noone et al., 1992; Svenningsson et al., 1997; Martinsson et al., 1999]. These studies have demonstrated a favorable partitioning of soluble aerosols into cloud drops and, conversely, relatively inefficient scavenging of insoluble aerosols. However, a number of caveats must be imposed on this general picture. Atmospheric aerosols are commonly internal mixtures with both soluble and insoluble components. For example, elemental carbon, which is hydrophobic, is frequently coated with a soluble layer of sulfate or nitrate with a resultant higher nucleation scavenging efficiency than would be expected from its structure. Another factor, which is important to keep in mind, is the variability in the internal microphysical and dynamical structure of clouds. For example, Gieray et al. [1997], even after taking into account the composition and internal mixing of aerosols, found anomalies in scavenging which they attributed to spatial variability of water vapor supersaturation. Finally, much of both the experimental and field work just discussed dealt solely with "warm" or purely water clouds. Mixed-phase clouds present a still more complex picture.

Mixed-phase clouds are more prevalent than either pure water or ice clouds, much more so in the mid-latitude bands of the Northern Hemisphere where aerosol concentrations are highest. From the standpoint of aerosol removal from the atmosphere, such mixed-phase clouds are considerably more efficient than single-phase clouds due to the relative efficiency of the riming and accretion processes, particularly riming [e.g., Hegg et al., 1989; Respondek et al., 1995]. But beyond this is the issue of effective scavenging of otherwise poorly scavenged aerosols by ice particles. Insoluble aerosols such as soil dust, elemental carbon and relatively hygrophobic organics, largely impervious to nucleation scavenging in their own right, have long been known to be favorable nuclei for ice crystal growth from the vapor phase [cf., Pruppacher and Klett, 1997]. Furthermore, such crystals, when formed, have been demonstrated to be rather effective scavengers of interstitial aerosols, including insoluble ones [Bell and Saunders, 1995; Song and Lamb, 1994]. Hence, mixed-phase clouds will tend to have a less composition-dependent impact on aerosol scavenging than single-phase clouds.

Chemical processes in cloud drops

By far the most studied cloud chemical process capable of significantly modifying aerosol size is the aqueous conversion of SO₂ to sulfate. Evidence of the significance of this process has been accumulating steadily for decades and it may now be regarded as well established. However, both modeling and observational studies have suggested considerable variability as to precisely where in the size distribution the prospective new sulfate mass will reside [cf., Hegg et al., 1980, 1992; Yuen et al., 1994; Bower et al., 1997], with considerable dependence on the initial composition and size distribution of the cloud forming aerosol. Additionally, of course, will be factors external to the aerosol itself, such as maximum cloud updraft, that will determine the size demarcation between activated and inactivated particles and thus the initial aerosol sizes only above which appreciable aqueous oxidation can take place.

Nevertheless, once again some general (and interesting) conclusions can be drawn. Most studies have concluded that in-cloud sulfate production adds mass most effectively at sizes just above the minimum activation radius achieved in cloud. For example, chamber studies conducted by Hoppel et al. [1994] show this quite clearly, and it is suggested in several field studies as well [e.g., Hegg et al., 1980; Leaitch, 1997]. While there are both observational [cf., Bower et al., 1997] and theoretical [Flossmann et al., 1987] studies which suggest significant variations at further distances from the mean activation radius (larger), all studies are in agreement that the enhancement will occur for particles on the order of 0.1 µm or larger. This is of considerable significance in that all such particles will likely be effective CCN in most atmospheric clouds. Hence, while the number of CCN does not in principle change, their activation supersaturation will be lower, thus raising the possibility of activation in successor clouds with lower supersaturation [cf., Hegg, 1990; Kaufman and Tanre, 1994]. Indeed, numerous studies, both observational and theoretical, suggest substantial changes in the cloud condensation nuclei (CCN) activation spectrum due to cloud processing [e.g., Hegg et al., 1980; Fitzgerald et al., 1998; Broadbury et al., 1999; Zhang et al., 1999]. Feingold et al. [1996] have proposed that such sulfate production would increase the new size of CCN in a similar manner to the collision-coalescence process discussed above. It has been suggested that the location in the size distribution of the additional sulfate mass produced by aqueous sulfate production enhances the light-scattering efficiency of cloud processed aerosols and thus affects the direct aerosol forcing of climate [Heintzenberg and Lelieveld, 1992].

One aspect of most of the studies discussed above, which is of considerable moment, is their applicability to, or exclusive concern with, non-precipitating clouds. Certainly the removal of any substantial number of cloud drops, particularly the preferential removal of larger drops, will impact modification of size distribution and CCN activation spectrum. This can be seen in the study of Khain et al. [1999] where, by and large, the aerosol accumulation mode is much reduced rather than enhanced by cloud processing. Nevertheless, other studies have suggested significant cloud microphysical impacts due to sulfate production even taking precipitation into account [e.g., Flossmann et al., 1987; Kogan et al., 1994]. Furthermore, many clouds are largely nonprecipitating.

Particle formation

The observational record of enhanced particle concentrations in the vicinity of clouds is now quite lengthy [e.g., Dinger et al., 1970; Saxena et al., 1970; Saxena and Gravenstein, 1994; Hegg et al., 1990; Hoppel et al., 1994; Saxena, 1996]. Early work, which dealt largely with increases in CCN, predominantly hypothesized enhancements in pre-existing particle size due to in-cloud sulfate production to explain the phenomenon; although the possibility of shattering of supersaturated salt droplets formed by rapid evaporation of cloud drops (as per Dessens [1949]) was also considered [e.g., Radke and Hegg, 1972]. However, the viability of this latter process is not entirely clear [Mitra et al., 1992]. More recently, advances in particle sizing instrumentation have permitted more complete characterization of near-cloud particle formation and support the formation, at least on occasion, of quite small particles rather than the larger particles that constitute the effective CCN population. The mechanistic explanation for such small particle production has generally involved homogeneous nucleation from the gas phase. While there is some debate on the extent to, and manner in which the new particle formation is an example of "cloud processing," it seems quite clear that clouds are involved. For example, Shaw [1989] postulated that only in very clean air in which virtually all pre-existing aerosol had been removed by cloud scavenging could binary nucleation of H₂O-H₂SO₄ particles occur. Hegg et al. [1990], in a diagnostic modeling study of observed high ultra-fine particle concentrations above marine stratocumulus, found that high water vapor concentrations detrained from the cloud deck coupled with an unusually high actinic flux (conducive to OH formation and thus H₂SO₄ production) due to isotropic backscatter from the cloud droplets were the main contributors to particle formation, though somewhat reduced particle surface area played some role. Similar observations of new particle formation adjacent to stratiform clouds have been made by Frick and Hoppel [1993] and Kutz and Dubais [1997] with attribution to the same mechanisms. It is also possible that the nucleation is ternary, involving NH₃, rather than binary [cf., Coffman and Hegg, 1994; Weber et al., 1995].

A slightly different priority with respect to contributing factors is likely present in the observations of particle formation in evaporating Antarctic stratus by Saxena [1996] where low temperature and the low pre-existing aerosol surface area were no doubt of considerable importance. Similarly, the phenomenon observed by Perry and Hobbs [1994], of particle formation in air detraining aloft from convective clouds, was partially attributable to the low temperatures and low aerosol surface area aloft on the basis of diagnostic modeling carried out on the cases studied. The overall scenario suggested by Perry and Hobbs, in which convective clouds bring up SO₂ to higher levels aloft while simultaneously removing particles, followed by H₂SO₄ production and particle nucleation in the cold, moist detraining air, has been adopted to explain still larger-scale phenomena by Clarke et al. [1998]. These workers found evidence of particle formation aloft associated with detraining air from moderate to large convective clouds and proposed a scenario similar to the Perry-Hobbs scheme but applied to a much larger scale and involving growth after formation with gradual cyclic subsidence and re-introduction into the boundary layer as CCN. The extent to which this larger scale phenomenon can be considered as modulation of the aerosol size distribution due to cloud processing is not entirely clear but certainly clouds play a vital role in the process as presented by Clarke et al. [1998] and it is therefore included here.

Synthesis of mechanisms

Given the multiplicity of mechanisms that must be considered, what might one expect for an overall effect? Clearly this will depend on the particular scenario to be evaluated and, indeed, the phenomenology of cloud effects on the aerosol size distribution and composition is diverse. In many instances in which non-precipitating clouds of small extent and liquid water content form on aerosols, little effect is to be expected; in others, with high actinic flux at cloud top, sufficient SO₂ (and possibly NH₃) and detraining water vapor, nucleation is to be expected; sufficient SO₂ and time in-cloud should normally produce significant sulfate mass; rapidly precipitating clouds with little SO₂ present should strongly deplete the accumulation mode; and so forth.

Figure 2. A qualitative schematic illustrating the impact of the various processes discussed in the text on the aerosol size distribution. The magnitude of the individual effects will be scenario dependent and could be either substantially larger or smaller than those shown here.

Figure 2 serves as a schematic to illustrate the potential impact of all the processes discussed here, assuming favorable conditions. The actual spectral modification to be expected for any particular event will be, once again, dependent on the specific conditions for that particular scenario and how they amplify or retard the impact of each process. Perhaps the most important single point to keep in mind is that clouds can have a profound local impact on both the size distribution and composition of aerosols, and that on larger scales clouds are major factors in the determination of aerosol properties.

Acknowledgment

This paper was prepared with support from the Office of Naval Research, grant N00014-97-0132. The author thanks Professor Marcia Baker for useful comments.

References

1. Alheit, R.R., A.I. Flossmann, and H.R. Pruppacher, A theoretical study of the wet removal of atmospheric pollutants, IV, The uptake and redistribution of aerosol particles through nucleation and impaction scavenging by growing cloud drops and ice particles, J. Atmos. Sci., 47, 870-887, 1990.
2. Bell, D.A., and C.P.R. Saunders, An experimental study of aerosol scavenging by hexagonal plate ice crystals, Atmos. Res., 38,9-19, 1995.
3. Bott, A., A numerical model of the cloud-topped planetary boundary-layer: Chemistry in marine stratus and the effects on aerosol particles, Atmos. Environ., 33, 1921-1936, 1999.
4. Bower, K.N., et al., Observations and modeling of the processing of aerosol by a hill cap cloud, Atmos. Environ., 31, 2527-2543, 1997.
5. Bradbury, C., K.N. Bower, TW. Choularton, E. Swietlicki, W. Birmili, A. Wiedensohler, B. Yuskiewicz, A. Berner, U. Dusek, C. Dore, and G.G. Mcfadyen, Modeling of aerosol modification resulting from passage through a hill cap cloud, Atmos. Res., 50, 185-204, 1999.
6. Cautenet, S., and B. Lefeivre, Contrasting behavior of gas and aerosol scavenging in convective rain: A numerical and experimental study in the African equatorial forest, J. Geophys. Res., 99,13,013-13,024, 1994.
7. Clarke, A.D., J.L. Varner, F. Eisele, R.L. Wauldin, D. Tanner, and M. Litchy, Particle production in the remote marine atmosphere: Cloud outflow and subsidence during ACE1, J. Geophys. Res., 103, 16,397-16,409, 1998.
8. Coffman, D.J., and D.A. Hegg, A preliminary study of the effect of ammonia on particle nucleation in the marine boundary layer, J. Geophys. Res., 100, 7147-7160, 1995.
9. Dessens, H., The use of spider's threads in the study of condensation nuclei, Quart. J. Roy. Met. Soc.,75, 23-27, 1949.
10. Dinger, J.E., H.B. Howell, and T.A. Wojciechowski, On the sources and composition of cloud condensation nuclei in the subsidence air mass over the North Atlantic, J. Atmos. Sci., 27, 791-797, 1970.
11. Easter, R.C., and P.V. Hobbs, The formation of sulfate and the enhancement of cloud condensation nuclei in clouds, J. Atmos. Sci., 31, 1586-1591, 1974.
12. Feingold, G., S.M. Kreidenweis, B. Stevens, and W.R. Cotton, Numerical simulations of stratocumulus processing of cloud condensation nuclei through collision-coalescence, J. Geophys. Res., 101, 21,391-21,402, 1996.
13. Fitzgerald, J., J.J. Martin, W.A. Hoppel, G.M. Frick, and J. Gelbard, A one-dimensional sectional model to simulate multi-component aerosol dynamics in the marine boundary layer, 2, Modeling application, J. Geophys. Res., 103, 16,103-16,117, 1998.
14. Flossmann, A., W.D. Hall, and H.R. Pruppacher, A theoretical study of the wet removal of atmospheric pollutants, I, The redistribution of aerosol particles captured through nucleation and impaction scavenging by growing cloud drops, J. Atmos. Sci., 42, 583-606, 1985.
15. Flossmann, A. I., H.R. Pruppacher, and J.H. Topolian, A theoretical study of the wet removal of atmospheric pollutants, II, The uptake and redistribution of (NH₄)₂SO₄ particles and SO₂ gas simultaneously scavenged by growing cloud drops, J. Atmos. Sci., 44, 2912-2923, 1987.
16. Frick, G.M., and W.A. Hoppel, Airship measurements of aerosol size distribution, cloud droplet spectra, and trace gas concentrations in the marine boundary layer, Bull. Amer. Meteor. Soc., 74,2195-2202, 1993.
17. Gieray, R., P. Wieser, T. Englehardt, E. Swietlicki, H.C. Hansson, B. Mentes, D. Orsini, B. Martinsson, B. Svenningsson, K.J. Noone, and J. Heintzenberg, Phase partitioning of aerosol constituents in cloud based on single-particle and bulk analysis, Atmos. Environ., 31, 2491-2502, 1997.
18. Hegg, D.A., Heterogeneous production of cloud condensation nuclei I the marine atmosphere, Geophys. Res. Lett., 17, 2165-2168, 1990.
19. Hegg, D. A., P.V. Hobbs, and L.F. Radke, Observations of the modification of cloud condensation nuclei in wave clouds, J. Rech. Atmos., 14, 217-222, 1980.
20. Hegg, D.A., L.F. Radke, and P.V. Hobbs, Particle production associated with marine clouds, J. Geophys. Res., 95, 917-926, 1990.
21. Hegg, D.A., S.A. Rutledge, P.V. Hobbs, M.C. Barth, and O. Hertzman, The chemistry of a mesoscale rainband, Q. J. R. Meteorol. Soc., 115, 867-886, 1989.
22. Hegg, D.A., P.-F. Yuen, and T.V. Larson, Modeling the effects of heterogeneous cloud chemistry on the marine particle size distribution, J. Geophys. Res., 97, 12,927-12,933, 1992.
23. Hoppel, W.A., J.W. Fitzgerald, G.M. Frick, R.E. Larson, and E.J. Mack, Aerosol size distributions and optical properties found in the marine boundary layer over the Atlantic Ocean, J. Geophys. Res., 95, 3659-3686, 1990.
24. Hoppel, W.A., G.M. Frick, J.W. Fitzgerald, and B.J. Wattle, A cloud chamber study of the effect that nonprecipitating water clouds have on the aerosol size distribution, Aerosol Sci. and Technol., 20, 1-30, 1994.
25. Hudson, J.G., and P.R. Frisbie, Cloud condensation nuclei near marine stratus, J. Geophys. Res., 96, 20,795-20,808, 1991.
26. Junge, C.E., Air Chemistry and Radioactivity, Academic Press, New York, 1963.
27. Kaufman, Y. J., and D. Tanre, Effect of variations in supersaturation on the formation of cloud condensation nuclei, Nature, 369, 45-48, 1994.
28. Khain, A., A. Pokrovsky, and I. Sednev, Some effects of cloud-aerosol interaction on cloud microphysics structure and precipitation formation: numerical experiments with a spectral microphysics cloud ensemble model, Atmos. Res., 52, 195-220, 1999.
29. Kleinman, L.I., and P.H. Daum, Vertical distribution of aerosol particles, water vapor, and insoluble trace gases in convectively mixed air, J. Geophys. Res., 96, 991-1005, 1991.
30. Kogan, Y.L., P.K. Lilly, Z.N. Kogan, and V.V. Filyushkin, The effect of CCN regeneration on the evolution of stratocumulus cloud layers, Atmos. Res., 33, 137-150, 1994.
31. Kutz, S., and R. Dubois, Balloon-borne aerosol measurements in the planetary boundary layer: particle production associated with a continental stratiform cloud, Contrib. Atmos. Phys., 70, 109-116, 1997.
32. Leaitch, W. R., Observations pertaining to the effect of chemical transformation in cloud on the anthropogenic aerosol size distribution, Aerosol Sci. and Technol., 25, 157-173, 1996.
33. Lelieveld, J., and J. Heintzenberg, Sulfate cooling effect on climate through in-cloud oxidation of anthropogenic SO₂, Science, 258, 117-120, 1992.
34. Martinsson, B.G., et al., Droplet nucleation and growth in orographic clouds in relation to the aerosol population, Atmos. Res., 50, 289-315, 1999.
35. Mitra, S.K., J. Brinkmann, and H.R. Pruppacher, A wind tunnel study on the drop-to-particle conversion, J. Aerosol Sci., 23, 245-256, 1992.
36. Noone, K.J., J.A. Ogren, A. Holberg, J. Heintzenberg, J. Strom, H.C. Hansson, B. Svenningsson, A. Wiedensohler, S. Fuzzi, M.C. Facchini, B.G. Arends, and A. Berner, Changes in aerosol size- and phase distributions due to physical and chemical processes in fog, Tellus, 44B, 489-504, 1992.
37. Pandis, S.N., J.H. Seinfeld, and C Pilinis, Chemical composition differences in fog and cloud droplets of different sizes, Atmos. Environ., 24A, 1954-1969, 1990.
38. Perry, K.D., and P.V. Hobbs, Further evidence for particle nucleation in clear air adjacent to marine cumulus clouds, J. Geophys. Res., 99,22,803-22,818, 1994.
39. Pruppacher, H.R., and J.D. Klett, Microphysics of Clouds and Precipitation, 2nd Ed., Kluwer Acad. Publications, Dordrecht, 1997.
40. Radke, L.F., and D.A. Hegg, The shattering of saline droplets upon crystallization, J. Rech. Atmos., 6, 447-455, 1972.
41. Respondek, P.S., R.R. Alheit, and H.R. Pruppacher, A theoretical study of the wet removal of atmospheric pollutants, V, The uptake, redistribution, and deposition of (NH₄)₂SO₄ by a convective cloud containing ice, J. Atmos. Sci., 52, 2121-2132, 1995.
42. Rodhe, H., Precipitation scavenging and tropospheric mixing, In Precipitation Scavenging, Dry Deposition and Resuspension (H.R. Pruppacher, R.G. Semonin and W.G.N. Slinn, eds), Elsevier, New York, pp. 719-730, 1983.
43. Saxena, V.K., Bursts of cloud condensation nuclei (CCN) by dissipating clouds at Palmer Station, Antarctica, Geophys. Res. Lett., 23, 69-72, 1996.
44. Saxena, V.K., J.N. Burford, and J.M. Kassner Jr., Operation of a thermal diffusion chamber for measurements on cloud condensation nuclei, J. Atmos. Sci., 27, 73-80, 1970.
45. Saxena, V.K., and J.D. Grovenstein, The role of clouds in the enhancement of cloud condensation nuclei concentrations, Atmos. Res., 31, 71-89, 1994.
46. Shaw, G., Production of condensation nuclei in clean air by nucleation of H₂SO₄, Atmos. Environ., 23, 2841-2846, 1989.
47. SMIC, Inadvertent Climate Modification, MIT Press, Cambridge, 1971.
48. Song, N., and D. Lamb, Experimental investigation of ice in supercooled clouds, II, Scavenging of an insoluble aerosol, J. Atmos. Sci., 51, 104-116, 1994.
49. Svenningsson, B., H.C. Hansson, B. Martinsson, A. Wiedensohler, E. Swietlicki, S.I. Cederfelt, M. Wendisch, K.N. Bower, T.W. Choularton, and R.N. Colvile, Cloud droplet nucleation scavenging in relation to the size and hygroscopic behavior of aerosol particles, Atmos. Environ., 31, 2463-2475, 1997.
50. Weber, R.J., P.H. McMurry, F.L. Eisele, and D.J. Tanner, Measurement of expected nucleation precursor species and 3-500 nm diameter particles at Mauna Loa Observatory, Hawaii, J. Atmos. Sci., 52, 2242-2257, 1995.
51. Wiedensohler, A., et al., Night-time formation and occurrence of new particles associated with orographic clouds, Atmos. Environ., 31, 2545-2560, 1997.
52. Yuen, P.-F., D.A. Hegg, and T.V. Larson, The effects of in-cloud sulfate production on light-scattering properties of continental aerosol, J. Appl. Meteor., 33, 848-854, 1994.
53. Zhang, Y., S.M. Kreidenweis, and G. Feingold, Stratocumulus processing of gases and cloud condensation nuclei, 2, Chemistry sensitivity analysis, J. Geophys. Res., 104, 16,061-16,080, 1999.