Cloud Droplet Formation

Linking aerosol particles to cloud droplets is a weak point in estimates of the indirect aerosol effects. Accurate treatment of cloud droplet formation requires knowledge of the particle number concentration and size-distributed chemical composition of the aerosol and the vertical velocity on the cloud scale. Parameterizations based on the Köhler theory (Köhler 1923) have been developed to describe cloud droplet formation for a multimodal aerosol. This approach has been extended to include kinetic effects that consider mass accommodation of water at the gas-liquid interface and account for the fact that the largest particles may not have time to grow to their equilibrium size and activate. Competition between natural and anthropogenic aerosol particles, such as between sulfate and sea salt, is also considered (Forster et al. 2007).

Organic carbon is an important constituent of CCN, especially if it is surface active. Facchini et al. (1999) indicate that by lowering the surface tension of surface-active organic particles (e.g., obtained from fog water samples) the cloud droplet number concentration and cloud albedo can be enhanced, leading to a negative forcing as large as ~ -1 W m-2. In contrast, amphiphilic film-forming compounds may retard cloud droplet formation (Feingold and Chuang 2002). Delayed activation enables the growth of larger drops, which have formed earlier, and results in increased dispersion and enhanced drizzle formation. Chemical effects on cloud droplet formation, and thus on the indirect effect, may be as large as the effects of unresolved cloud dynamics (Lohmann and Feichter 2005). Whereas the effect of surface-active organics has recently been included in parameterizations of cloud droplet formation (Abdul-Razzak and Ghan 2004), other effects of organics, such as their film-forming ability have not yet been treated.

Application of parameterizations of cloud drop activation requires estimating cloud-scale vertical velocities in models which do not resolve these cloud scales. Recognizing that this information may not be available, some modelers assume an empirical relationship between modeled sulfate mass concentrations and droplet concentrations (e.g., Boucher and Lohmann 1995), which is equivalent to assuming there is only one single value of cloud updraft velocity for all clouds in the model. Others estimate vertical velocity based on turbulent kinetic energy calculated in boundary layer models (e.g., Lohmann et al. 1999). The latter represents a step in the right direction, but it does not account for the fact that cloudy updrafts are at the tail of the probability density function (PDF) of vertical velocity. Ghan et al. (1997), among others, assumed a normal distribution of vertical velocity with a mean value given by the

GCM grid point mean. They determined the velocity-weighted mean droplet concentration, taking into account the tails of their assumed PDF of vertical velocity. However, observed PDFs of vertical velocity in clouds in the boundary layer are multimodal and are better represented by double-Gaussian PDFs (Larson et al. 2001) with a mean that is a function of the root mean square vertical velocity rather than by a GCM grid point mean (Peng et al. 2005).

0 0

Post a comment