Limitations of using an equilibrium approximation in an aerosol activation parameterization

Because of its complex nature, attempts have been made to parameterize the process of cloud droplet formation and growth. The parameterization developed by Abdul-Razzak et al. [1998] and Abdul-Razzak and Ghan [2000], henceforth referred to as the ARG parameterization, is based on the adiabatic ascent of an air parcel, assuming that each particle is in equilibrium with its environment. Some of the limitations of the equilibrium assumption are evaluated here through direct comparison of a kinetic parcel model with the ARG parameterization. Conditions are described in which the supersaturation and the cloud droplet number concentration are underpredicted by the parameterization. The underprediction happens because of an overestimation of the condensation rate of water onto the larger particles. The effect is significant for conditions of lower updraft velocities (V < 50 cm s(-1)) and higher aerosol number concentrations (N-a > 500 cm(-3)). The parameterization behaves well for higher updraft velocities and lower aerosol number concentrations. The impact of this effect on the activation of sulfate aerosols in the presence of sea salt aerosol is also investigated. Using the ARG parameterization to diagnose N-d in the presence of sea salt aerosol leads to a greater reduction in the number of activated sulfate particles in lower updraft conditions than does using the kinetic parcel model. The relative contribution of sulfate particles to total Nd is also significantly lower for the ARG parameterization than it is for the parcel model in low to moderate updrafts. The error in the first indirect radiative forcing introduced by the equilibrium assumption is estimated to be 6 to 15 W m(-2) for typical marine clouds, depending on the updraft velocity.