Parameterization of Convective Precipitation in Mesoscale Numerical Models: A Critical Review

Current approaches for incorporating cumulus convection into mesoscale numerical models are divided into three groups. The traditional approach utilizes cumulus parameterization at convectively unstable points and explicit (nonparameterized) condensation at convectively stable points, The fully explicit approach uses explicit methods regardless of stability. The hybrid approach parameterizes convective scale updrafts and downdrafts, but "detrains" a fraction of parameterized cloud and precipitation particles to the grid scale. This allows the path and phase changes of such particles to be explicitly predicted over subsequent time steps. The traditional approach provides the only alternative for numerical models with grid spacing too large to resolve mesoscale structure. As grid spacing falls below 50 km, the traditional approach becomes increasingly likely to violate fundamental scale-separation requirements of parameterization, particularly if mesoscale organization of convection is parameterized as well. The fully explicit approach has no such limits, but it has repeatedly failed in mesoscale models in the presence of large convective instability. Although it is preferable under certain specialized circumstances, the fully explicit approach cannot provide a general solution for models with grid spacing above 5–10 km. The hybrid approach most cleanly separates convective-scale motions from the slow growth, fallout, and phase changes of detrained hydrometeors that produces mesoscale organization of convection. It is argued that this characteristic removes the need to parameterize the mesoscale and thus reduces the scale-separation problems that may arise when the traditional approach is used. The hybrid approach provides in principle the preferred solution for mesoscale models, though such promise has yet to be fully realized. In the absence of large rotation, the fundamental assumptions of cumulus parameterization begin to break down once grid spacing falls below 20–25 km. For models with such resolution, the time scale of the convection being parameterized approaches the characteristic time scale of the grid, and parameterized and unparameterized convective clouds often exist simultaneously in a grid column. Under such ambiguous circumstances, successful simulations have been produced only because parameterized convection rapidly gives way in the, model to its grid-scale counterpart. It is essential to understand the interactions between implicit and explicit clouds that produce this transition, and whether they represent physical processes in nature, before cumulus parameterization can be widely used in such high-resolution models. In a broader sense, more detailed analysis of why convective parameterizations succeed and fall is needed.