Dispersion Aerosol Indirect Effect from Turbulent Fluctuations: Insight from a Cloud Chamber Experiment

Cloud optical properties are determined not only by the number density n and mean radius <r> of cloud droplets, but also by the shape of the droplet size distribution, for example, its width σ_r. The change in cloud optical depth with changing n, due to the change in distribution shape is known as the dispersion effect. Droplet relative dispersion is defined as d = σ_r/<r>. Over a decade of field measurements have given conflicting results not only in the magnitude, but even the sign of the dispersion effect. We have investigated a commonly-used effective radius parameterization in a turbulent cloud created in a controlled laboratory environment. Stochastic condensation growth suggests d independent of n for a non-precipitating cloud, hence nearly zero albedo susceptibility due to the dispersion effect. However, for size-dependent removal, such as in a laboratory cloud or highly clean atmospheric conditions, stochastic condensation produces a weak dispersion effect. The albedo susceptibility due to turbulence broadening is observed to have the same sign as the Twomey effect and to augment it by order 10%. This laboratory quantification of the dispersion effect for non-precipitating clouds helps constrain at least one aspect of the full, multi-dimensional problem in the atmospheric context.