The Capacity and Preliminary Improvement of Cumulus Parameterization on Typhoon Numerical Prediction

Cumulus parameterization defines the sub-grid scale convective activities in a numerical model and is crucial for typhoon numerical prediction. In this study, according to three typical typhoons made landfall at East China in 2006, 15 numerical experiments were implemented with three of the most popular cumulus parameterization schemes, namely, Kain-Fritsch, Betts-Miller and Grell for evaluation. It is concluded that Kain-Fritsch scheme is more stable and appreciable in heavy rainfall and track prediction. While admitting its advantages, it is also identified that Kain-Fritsch scheme is necessary to be improved in the case of weak environmental forcing. Particularly, the parameterization of convective parcel's temperature perturbation with environmental vertical velocity in convective trigger function is not so robust, which should be applicable only when the prerequisite of moisture transportation is satisfied. To avoid this deficiency, spatial temperature anomaly as well as the effect of moisture advection is taken into account in determining convective parcel's temperature perturbation, where the contribution from boundary layer heat and moisture flux is treated explicitly. In comparison with the original scheme, this technique can eliminate the convective instability more effectively and in favor of the rainfall simulation.

*This study is jointly supported by the Special Research Program 2005DIB3J104 sponsored by Ministry of Science and Technology of P. R. China and Shanghai Typhoon Research Fund