Cloud microphysical parameterizations are techniques for representing sub-grid-scale microphysical processes using grid-scale information in atmospheric cloud-scale-resolving numerical models. A multimoment cloud microphysical parameterization is one that uses two or more of the following moments including, but not limited to, number concentration (number of hydrometeors per unit volume), characteristic diameter of hydrometeor size distribution (inverse of slope of size distribution), mixing ratio (hydrometeor density divided by density of air), and reflectivity (related to radiation power scattered back to a radar) to predict the evolution of aerosols, clouds, and precipitation.

The prediction of changes in aerosols, clouds, and precipitation involves the development of parameterizations of processes such as nucleation (initiation of cloud particles with or without aerosols), vapor diffusion growth (loss or gain of hydrometeor mass owing to a phase change of water vapor), collection and breakup (one type of particle collecting another and destruction of hydrometeors by collisions), freezing (phase change of liquid water to ice), melting (phase change from ice water to liquid), and sedimentation (fallout of hydrometeors, usually at their terminal velocities). The prediction of aerosols and hydrometeors also requires the need for numerical approximations for positive-definite solutions to transport and turbulent mixing.