Abstract
A Monte-Carlo model by Lapidus and Shafrir for the temporal development of cloud droplet spectra is critically examined. Modifications are suggested to improve the statistical validity of the model.
Both the original model and the modified model only approximately simulate the coalescence process. It is demonstrated that when large collection kernels, such as the Golovin collection kernel, are used both models are inadequate. However, for smaller and more realistic collection kernels the models do not produce results that differ substantially from the standard solutions given by the coalescence equations.