There is evidence that aerosols do have an effect on climate via cloud modification. The observations of ship tracks (e.g., King et al. [1993]), modified droplet fields embedded within stratocumulus clouds underneath which ships have passed, are cited as evidence that aerosols can affect cloud properties. Han et al. [1994] report the results of a survey of effective cloud drop radii derived from satellite data. Their analyses indicate systematic differences in drop size between continental and marine water clouds and between marine clouds in the Northern and in the Southern Hemispheres, with smaller drop radii associated with regions affected by anthropogenic pollution.
Despite evidence for the effect of anthropogenic aerosols upon cloud radiative properties, it is difficult to estimate the magnitude of possible indirect aerosol effects, and quantification of uncertainties in such estimates is not possible ( Penner et al. [1993a]). Many of the critical relationships are not well understood, and a variety of climate ``feedback'' processes have been proposed (e.g., changes in cloud lifetimes) that would alter conclusions drawn from considering only initial effects. Indirect effects depend on the relationship between changes in aerosol (and aerosol precursor) emissions and changes in the atmospheric concentrations of that subset of the aerosol population active as CCN. The effects are also impacted by the relationship between CCN and the hydrologic cycle. A review of the research history and a summary of current understanding of CCN is presented in Hudson [1993], who concludes that significant expansion in the knowledge base of CCN is required before any reduction in uncertainty of climate change estimates can be achieved. Fundamental research in cloud physics (reviewed by Rasmussen and Hallett elsewhere in this volume) is also an integral part of developing estimates of the indirect forcing.
Among the research needs discussed by Hudson [1993] are methods for differentiation between natural and anthropogenic CCN and chemical information to assist in determining CCN sources. Of particular interest are the sources of CCN in clean ocean regions, since the optical properties of marine stratocumulus clouds are particularly susceptible to modification ( Twomey [1991]). Sulfate is a major component of the marine aerosol, and dimethylsulfide (DMS) has been proposed as the major source of non-sea-salt sulfate and thereby of cloud condensation nuclei. Hegg et al. [1991] present observational support for relationships between DMS and CCN concentrations. Hoppel et al. [1994] have suggested that the cycling of aerosol though nonprecipitating clouds produces the bimodal size distribution typically observed in clean marine regions and present laboratory confirmation of this processing. Within cloud, aqueous conversion of sulfur gases produces (nonvolatile) sulfate, which is deposited on the original CCN when the cloud evaporates. This larger CCN is then active at smaller supersaturations, thus enhancing the CCN spectrum. Scavenging of interstitial aerosol also adds mass to the CCN and depletes the small particle population; in this scenario, the aerosol distribution is maintained by production of new particles during S gas oxidation ( Kreidenweis et al. [1991]).
Although much attention has focused on the role of biogenic and anthropogenic sulfate as sources of CCN, other species should also be considered. Hudson et al. [1991] showed that significant number concentrations of CCN are produced from biomass burning. Novakov and Penner [1993] present observational evidence that organic species can contribute substantially to CCN number concentrations. In view of the prevalence of, and anthropogenic contribution to, organic compounds in particulate matter, this possibility deserves further investigation. One of the interesting observations to come out of the study of the Kuwait smoke was that a surprisingly large component of the aerosol were effective CCN ( Hudson and Clarke [1992]). This characteristic decreased the atmospheric lifetime of the particles so that severe climate effects of the smoke were confined to the vicinity of the source. However, it was suggested that anomalous precipitation events in Asia downwind of the plume were triggered by the fires ( Parungo et al. [1992]). These last observations also serve to point out that, in addition to sources, CCN abundance is affected by removal processes. Since the most important sink is precipitation, CCN can affect their own removal rates by influencing drizzle production. An improved understanding of the interaction of processes which control the CCN population is needed ( Baker [1993]).