Evaluation of 2008 season hurricane forecasts with a new shallow cumulus convection package in the NCEP GFS

The current operational shallow cumulus convection scheme in the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) uses a simple turbulent eddy diffusion approach with a specified eddy diffusivity profile for the transports of sensible heat and moisture within the cloud layer. While this scheme has been successful for destroying unrealistic moisture accumulation in the layer below the inversion by the additional diffusion of heat and moisture, it has suffered from the systematic underestimation of the low clouds especially over the eastern Pacific and Atlantic Oceans. To improve the current shallow convection scheme, we developed a bulk mass-flux parameterization for shallow cumulus convection. While the scheme is based on the Simplified Arakawa-Shubert (SAS) convection scheme, which has been developed for deep cumulus convection and is being used in the operational NCEP GFS, some aspects in the SAS scheme such as entrainment and detrainment specifications are modified for shallow convection situations.

The new shallow convection package also includes significant modifications in the present planetary boundary layer (PBL) and SAS deep convection schemes. In the modified PBL scheme, stratocumulus-top driven turbulence mixing is included, which is enhanced when the condition for cloud top entrainment instability is met. A local diffusion is used for the nighttime stable conditions. In the modified SAS scheme, the cloud top is not determined randomly any more but specified based on parcel buoyancy. Cloud water is detrained from every cloud layer, not massively dumped at cloud top. Finite entrainment and detrainment rates for heat, moisture, and momentum are specified. The effect of convection-induced pressure gradient force is included in momentum transport. An environmental moisture dependence of entrainment rate is also included. We will present comparisons of the performance of this package against the operational GFS for the 2008 hurricane season using a complete data assimilation and forecast system.