convectionpermitting

Convection-permitting refers to numerical weather prediction models whose grid spacing is fine enough to explicitly simulate convective processes, rather than representing them with parameterization schemes. When the grid is too coarse, convection is approximated statistically; at convection-permitting scales some thunderstorm dynamics can develop and organize directly within the model.

Typical convection-permitting grid spacings are on the order of 3 to 4 kilometers or smaller, with many high-resolution studies using 1–2 kilometer grids. In some research contexts, resolutions below 1 kilometer are employed to resolve finer storm-scale features. The term emphasizes the scale at which convection can be represented without reliance on traditional convective parameterizations.

Benefits of convection-permitting models include more realistic initiation and organization of convection, improved timing and location of precipitation, and better depiction of mesoscale weather phenomena such as squall lines and organized storm systems. CPMs can enhance forecasts of local extreme events and better capture diurnal cycles in many regions.

Limitations and challenges include ongoing needs to parameterize subgrid-scale processes such as microphysics, boundary layer dynamics, and surface interactions; substantial computational demands; sensitivity to initial conditions and chosen physics schemes; and the necessity of ensemble approaches to quantify forecast uncertainty. Operational CPMs also rely on high-quality data assimilation and rapid forecast updates to realize their potential.

In practice, convection-permitting modeling is widely used in weather forecasting and atmospheric research, often described as convection-permitting or convection-allowing models. It represents a shift toward higher-resolution, process-based simulation of convective storms.