| description abstract | Any population of cloud droplets forming on polydisperse condensation nuclei is thermodynamically unstable. There is no value of the supersaturation for which the growth rate of all the droplets is zero, so that if some droplets are in equilibrium, then some must have positive and some negative growth rates. Droplets with positive growth rates will continue to grow at the expense of those with negative growth rates. This effect has been termed the ripening process, and has been postulated to be a potential mechanism to explain broad droplet size distributions in stratiform clouds. In this paper multiple parcel trajectories are used, derived using a simple representation of the turbulent dynamics, to examine the time evolution of the droplet size distribution in a nonentraining stratiform cloud. It is shown that the magnitude of the effect is critically dependent upon the mean parcel in-cloud residence time. The simulations suggest that, for a stratiform clouds of h = 400 m thickness, and a vertical wind standard deviation of σw = 0.6 m s?1 (typical for stratocumulus clouds in a fairly vigorous, well-mixed boundary layer), the ripening effect is negligible, in that the droplet size distribution changes little with time. However, clouds with low σw = 0.2 m s?1 (typical of weaker stratus clouds) show a marked spectral ripening effect over a period of several hours. Ripening is observed in the numerical model in both clean and polluted aerosol distributions. Autoconversion rates calculated from the droplet size distributions increase markedly with time as ripening takes place. It is suggested that to accurately model droplet size distributions in stratus cloud, it may be necessary to take into account the distribution of in-cloud parcel residence time. | |
