Abstract
The condensation‐accretion (warm rain) and Wegener‐Bergeron (ice crystal) theories of rain formation are briefly outlined. In a number of situations the former process must be applicable. However, conventional models of the warm rain process lead to a 15–25 μm radius barrier to drop growth in observed cloud lifetimes. Modifications to the process to circumvent this difficulty have been only partially successful. These modifications will be critically reviewed. Giant hygroscopic and pre‐cipitable nuclei only explain a limited number of situations. Although uncertainties remain, especially in the treatment at small interaction distances, it is unlikely that changes in collision efficiencies alone can explain the observations, unless the modifications are due to turbulence effects. The stochastic nature of the accretion process also appears unsuccessful in explaining the observations, although there is some uncertainty concerning the applicability of the commonly used stochastic collection equation. It is important to model condensation and accretion simultaneously, but this consideration alone does not completely account for the short observed precipitation formation times. A recent model incorporating mixing processes has been successful in explaining the main characteristics of maritime cumuli clouds, but formation times for continental cumuli remain excessive. Although the considerations mentioned above reduce the discrepancy between theory and observation, it is not yet completely resolved. The importance of turbulence, the acceptability of the stochastic collection equation, and the suitability of present cumulus cloud models would appear to be the three most important unresolved questions.

This publication has 64 references indexed in Scilit: