Search for: All records

Abstract Previous observational and modeling studies have suggested that moisture plays a dominant role in Madden–Julian oscillation (MJO) evolution. Using a realistic MJO simulation by incorporating the role of mesoscale stratiform heating in the Zhang–McFarlane deep convection scheme in the National Center for Atmospheric Research Community Atmosphere Model, version 5.3 (NCAR CAM5.3), this study investigates the factors responsible for the improved MJO simulation by examining moisture variations during different MJO phases. The results of column moist static energy (MSE) and moisture budgets show that during the suppressed phases of MJO, vertical advection acts to increase MSE anomalies for the development of deep convection while radiative heating and surface heat flux decrease MSE. The opposite holds true at the MJO mature phase. However, their roles largely cancel each other, leaving horizontal advection to play a major role in the low-level MSE increase during the suppressed phase of the MJO and MSE decrease after the MJO mature phase. A further analysis combining moisture and temperature budget equations is performed to demonstrate the effects of vertical advection and cloud processes within the column at each level. The vertical profiles of column-confined moisture tendency show that large-scale vertical advection induced by latent heat release and evaporation within shallow convective clouds is also important to the lower-tropospheric moistening during suppressed phases. This confirms the role of shallow convection in low-level moistening ahead of MJO deep convection. Radiative heating is vital across all MJO phases, and its warming effects keep the column humidity anomaly maintained in mature phases. None of these features are reproduced by the standard CAM5.3. 

Abstract With the recent advances in data science, machine learning has been increasingly applied to convection and cloud parameterizations in global climate models (GCMs). This study extends the work of Han et al. (2020,https://doi.org/10.1029/2020MS002076) and uses an ensemble of 32‐layer deep convolutional residual neural networks, referred to as ResCu‐en, to emulate convection and cloud processes simulated by a superparameterized GCM, SPCAM. ResCu‐en predicts GCM grid‐scale temperature and moisture tendencies, and cloud liquid and ice water contents from moist physics processes. The surface rainfall is derived from the column‐integrated moisture tendency. The prediction uncertainty inherent in deep learning algorithms in emulating the moist physics is reduced by ensemble averaging. Results in 1‐year independent offline validation show that ResCu‐en has high prediction accuracy for all output variables, both in the current climate and in a warmer climate with +4K sea surface temperature. The analysis of different neural net configurations shows that the success to generalize in a warmer climate is attributed to convective memory and the 1‐dimensional convolution layers incorporated into ResCu‐en. We further implement a member of ResCu‐en into CAM5 with real world geography and run the neural‐network‐enabled CAM5 (NCAM) for 5 years without encountering any numerical integration instability. The simulation generally captures the global distribution of the mean precipitation, with a better simulation of precipitation intensity and diurnal cycle. However, there are large biases in temperature and moisture in high latitudes. These results highlight the importance of convective memory and demonstrate the potential for machine learning to enhance climate modeling.