Improving Prediction of Mesoscale Convective Systems with Convective Parameterization

Despite the advantages of convection-permitting resolution for mesoscale systems (MCSs), such fine resolution is not always practical. This is especially the case when long time scales are being projected, as in seasonal or decadal predictions or climate change studies. Here we discuss a simple modification of convective parameterizations for the prediction of MCSs that is beneficial when grid spacing is too small for explicit resolution of deep convection but is sufficiently fine to resolve the overall structure of the MCS (i.e., around 10 to 30 km). The basic philosophy of the modification is to more tightly couple the convective updraft within the parameterization to the grid-scale microphysics. We test this approach in multi-year simulations using both the Kain-Fritsch and Grell convective parameterizations within the WRF-ARW model. The results show considerable improvement in the generation and realism of propagating MCSs in comparison to the unmodified convective parameterizations.