General circulation models (GCMs) are widely used for global weather forecasting and climate modeling. In a GCM, the convective parameterization, which represents the bulk effects of convection, is typically regarded as a large source of model uncertainty (Arakawa, 2004;Rybka & Tost, 2014;Tost et al., 2006). Cumulus convection interacts with other processes in complex ways. Heat and moisture are pumped out of the planetary boundary layer (PBL) by subgrid-scale cumulus cells in response to surface solar heating. Detrainment and/or re-evaporation from convective condensate and precipitation moisten the grid-scale environment, favoring large-scale condensation. Convective parameterizations also strongly regulate the partition of convective and large-scale precipitation over the tropics with resultant effects on clouds and tropical transients (Kim et al., 2012;Lin et al., 2013). The interdependency between subgrid-and grid-scale processes further influences the hydrological processes, cloud types, and thus cloud radiative forcing and its feedback (Arakawa, 2004;Hourdin et al., 2006).Apart from the mean states, the impact of convective parameterization on tropical variabilities has also been reported in some previous studies (

We investigated the resolution sensitivity of the Global‐to‐Regional Integrated forecast SysTem global nonhydrostatic model characterized by explicit dynamics–microphysics coupling using varying uniform resolutions (120, 60, 30, 15, and 5 km). The experiments followed the DYnamics of the Atmospheric general circulation Modeled On Non‐hydrostatic Domains (DYAMOND) winter protocol, which covers a 40‐day integration. These simulations did not activate parameterized convection. One 120 km test with parameterized convection was performed as a coarse‐resolution reference. Other model configurations for different simulations were kept as consistent as possible. Our results showed that the model gradually improved its representation of the fine‐scale features as the resolution increased. The 5 km simulation was overall close to a 3.75 km simulation during the first 12 days of the DYAMOND winter. With respect to the mean climate, the 5 km simulation had a more realistic rainfall distribution than the lower resolution explicit convection simulations. Cloud water and the related physical fields (e.g., shortwave cloud radiative forcing) had a large resolution sensitivity. The tropical rainfall frequency–intensity spectra became more realistic in the 5 km explicit convection simulation, but the 120 km run with parameterized convection showed a more realistic mean climate. As the resolution increases, the mean bulk effect of finely resolved model convection gradually converges to that of parameterized convection. The mean climate of this storm‐resolving model has slightly higher rainfall biases than a parameterized convection coarse‐resolution model, highlighting the importance of balancing resolved‐ and under‐resolved model convection for developing a unified multiscale global model.

Abstract. As a unified weather-forecast–climate model system, Global-to-Regional Integrated forecast SysTem (GRIST-A22.7.28) currently employs two separate physics suites for weather forecast and typical long-term climate simulation, respectively. Previous AMIP-style experiments have suggested that the weather (PhysW) and climate (PhysC) physics suites, when coupled to a common dynamical core, lead to different behaviors in terms of modeling clouds and precipitation. To explore the source of their discrepancies, this study compares the two suites using a single-column model (SCM). The SCM simulations demonstrate significant differences in the simulated precipitation and low clouds. Convective parameterization is found to be a key factor responsible for these differences. Compared with PhysC, parameterized convection of PhysW plays a more important role in moisture transport and rainfall formation. The convective parameterization of PhysW also better captures the onset and retreat of rainfall events, but stronger upward moisture transport largely decreases the tropical low clouds in PhysW. These features are in tune with the previous 3D AMIP simulations. Over the typical stratus-to-stratocumulus transition regime such as the Californian coast, turbulence in PhysW is weaker than that in PhysC, and shallow convection is more prone to be triggered and leads to larger ventilation above the cloud layer, reducing stratocumulus clouds there. These two suites also have intrinsic differences in the interaction between cloud microphysics and other processes, resulting in different time step sensitivities. PhysC tends to generate more stratiform clouds with decreasing time step. This is caused by separate treatment of stratiform cloud condensation and other microphysical processes, leading to a tight interaction between macrophysics and boundary layer turbulence. In PhysW, all the microphysical processes are executed at the same temporal scale, and thus no such time step sensitivity was found.