Explicit Convection

Convective parameterizations are used in current operational hydrostatic models to account for the effects of convection since the model cannot resolve convective motions explicitly. However, high-resolution (one- to two-km) nonhydrostatic models can be run without CP schemes because the grid spacings are small enough to begin to resolve convective motions. For example, the resolution can be fine enough that entire grid boxes can be filled with updraft air and condensate while others are filled with downdrafts.

A rapid and tight coupling between the dynamics and the microphysics scheme results in

  • Explicitly simulated updrafts strong enough to lift hydrometeors up to the equilibrium level
  • Explicitly simulated downdrafts and their accompanying gust fronts

This allows a more realistic redistribution of heat and moisture than when a CP scheme is used. It also enables the winds and vertical motion to be modified directly by the convection. Explicit convection ultimately provides a direct prediction of convective precipitation. (Remember that CP schemes can only indirectly predict convective precipitation as a by-product of removing instability and thus do a poor job.)

The animation shows how the nonhydrostatic ARPS model uses explicit convection to realistically simulate the 3 May 1999 OKC tornadic supercells as compared to radar observations of the same storms.

Animation showing how the non-hydrostatic ARPS model uses explicit convection to realistically simulate the 3 May 1999 OKC tornadic supercells compared to radar observations

Note the outstanding forecast of the character of the convection but the inability to predict details such as the location of a particular cell at a particular time. Unfortunately, this ARPS simulation only succeeded when radar data from the initial storms was assimilated, pointing to difficulties in predicting convective initiation even using high-resolution models.