Moisture plays a key role in the atmospheric circulation through the impact on radiative fluxes, temperature changes, and hydrological cycle, and its representation is one of the most uncertain components in the weather and climate models. Because moisture is highly variable in space, its variability within a grid box should be represented appropriately in the model, especially in global circulation models (GCMs). In GCMs, sub-grid scale variability of moisture and temperature is considered to determine cloud macrophysical properties including cloud fraction and amount of cloud condensate. Since cloud information is important for the complex microphysical processes of warm and ice phase clouds, precipitation is sensitive to the macrophysical cloud condensate amount. In this study, we attempt to understand the role of the variability of moisture within a grid box in the global feature of precipitation. We modify the parameter controlling the sub-grid cover and condensate amount of liquid and ice clouds in the National Center for Atmospheric Research/Community Atmospheric Model version 5 (NCAR/CAM5) and examine its impact on the global distribution and intensity of convective and large-scale precipitations and convective activity in multi-year simulations. We also compare the model results with reanalysis data. Detailed results will be presented at conference.