Shallow convection has significant impacts on large scale circulations in tropics and the mid-latitudes by affecting the vertical transport of thermodynamic properties. By enhancing turbulence between lifting condensation level (LCL) and inversion level, this process does not allow the excess moisture trapped near the surface in synoptically inactive regions. Thus, it plays an important role in cloud development and the earth's radiation budget.

Recently, a mass flux approach for the shallow convection scheme (Han and Pan 2011, hereafter HAN) was applied to the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) model, and it showed a better performance on climate simulations than an effective eddy diffusivity approach (Tiedtke 1983, hereafter TDK). The HAN scheme successfully reduced unrealistic moisture accumulation below the inversion and stratocumulus clouds near the coast of the eastern Pacific and Atlantic Oceans. However, the Global/Regional Integrated Model system (GRIMs) and Weather Research and Forecasting (WRF) model had not included the mass flux shallow convection scheme yet. The purpose of this study is to describe the implementation of the shallow convection parameterization based on the mass flux approach into the GRIMs and the WRF model and compared the impacts of two different shallow convection schemes on global and regional climate simulations. The experiments are performed to examine their general features of the seasonal simulations under the global model framework and their impacts on a summer monsoon precipitation in East Asia under the regional model framework. Sensitivity of the simulated thermodynamic characteristics to two kinds of mass-flux deep convection schemes, Simplified Arakawa-Schubert (SAS) deep convection scheme (Pan and Wu 1995) and modified SAS scheme (Han and Pan 2011), is also investigated.

The TDK scheme simulates more vertical mixing of moisture and heat in the same way as the vertical diffusion with prescribed values of eddy diffusivity within clouds, whereas the HAN scheme is based on the SAS deep convection scheme, which represents entrainment, detrainment, and compensating subsidence to an updraft. Thus, the vertical mixing in the HAN scheme is reduced by the stronger cumulus convection, because the error is decreased by the deep convection scheme and the role of the shallow convection scheme is found to be limited in the global model simulations. Similarly, the HAN scheme has little effect on the simulated monsoonal summer rainfall climate over East Asia by the WRF model, whereas the experiments with TDK scheme show the similar changes in precipitation, almost regardless of the ability of the deep convection scheme.