In general, methods to install suitable magnitude of potential temperature perturbations around the level of free convection (LFC) are often used to activate a cumulus cloud under the atmospheric environment in a cloud resolving model. However, the generated convections may include unrealistic structure and unnatural gravity waves are often generated, because the presented perturbation does not entirely match with actual initial structure of convections. There are a lot of studies using such initial perturbations, but the best settings of perturbations are not yet clear.
Considering the mechanisms that convection cells within a cumulonimbus cloud are occurred by lifting of low-level unsaturated air mass, a new method for generating numerical initial disturbance of cumulus convections is introduced. In the method, blending ratio of temperature and water substances between the environmental air and a lifting air profiles at each location is designed by using the Gaussian function.
We examined the reproducibility of single cell convective clouds with the two kinds of initial perturbations generating methods; the new method (lifting method) and conventional method with an ellipsoidal shape of heating around LFC (heating method). The lifting method was able to reproduce the convective cloud with accumulated heating one-fifth of that in the heating method. In the lifting method, unrealistic structures of initial stage of convection are not found. On the other hand, the heating method includes unrealistic structures such as low-pressure in the low-level in the earlier stage, and updraft associated with gravity waves in the upper troposphere in the decaying stage of convections. Consequently, the lifting method is able to activate more natural convective clouds with smaller heating in association with the given environment.