Test of a convectively forced gravity wave drag parameterization in a general circulation model

The influence of gravity wave drag induced by cumulus convection (GWDC) on a simulated boreal summer climate was evaluated in a general circulation model. For this, the GWDC scheme developed by Chun and Baik was implemented into a version of the National Centers for Environmental Prediction (NCEP) global spectral model (GSM). Ensemble simulations with the two different convection schemes, the simplified Arakawa-Schubert (SAS) scheme and Community Climate Model (CCM) convection scheme, were conducted for the boreal summer of 1996. A cloud factor to modulate the stress intensity with respect to the cloud type was introduced in this study, in order to prevent unrealistic behaviors of the GWDC scheme in GSM. The effect of gravity wave drag on the zonal mean of wind and temperature fields was focused. On the whole, the effect of GWDC in this study is positive on the simulated seasonal climate. It is evident that biases in temperature in the polar region as well as in the zonal and meridional winds in the upper atmosphere are reduced. The percentage of reduction of the bias in zonal winds is about 10–20%. Such a response of the GWDC forcing widely appears not only in tropical regions but also in mid-latitude regions. These characteristics are prominent in the case of the SAS scheme, which is due to the various convective cloud types. The magnitude of GWDC forcing is generally small, but still positive, in the case of the CCM scheme, which is due to rather homogeneous cloud types. It is also found that the role of a particular GWDC forcing depends upon the inherent systematic biases of a particular model. It is concluded that incorporation of the GWDC parameterization in GCMs should be taken into account to improve the seasonal prediction.