Simulations of Stratus Clouds over Eastern China in CAM5: Sources of Errors

A previous study by Zhang et al. suggested two biases of the high-resolution configured Community Atmosphere Model, version 5 (CAM5), in simulating stratus clouds over eastern China, including an underestimation of stratus occurrence frequency and a spurious low stratus amount when present (AWP) value center over the Sichuan basin. In this study, the causes for these two problems are further explored. The underestimate of stratus occurrence frequency in the model is attributed to the bias in large-scale ambient environmental fields. This is confirmed by investigating the differences between two climate counterparts. Results suggest that when the environmental fields in the climate ensemble become more realistic, the simulations of stratus cloud radiative forcing and cloud fraction are enhanced, mainly caused by a corresponding increase in the stratus occurrence frequency. The specific sources of the cloud changes between these two ambient climates are then investigated. The presence of a low stratus AWP value center is found to be sensitive to the choice of dynamical core. This is confirmed by comparing the simulations from two dynamical core counterparts: a default finite-volume core and an alternative Eulerian spectral transform core. Experiments with these two cores suggest that the spectral CAM5 is able to alleviate this problem. Correspondingly, the subsiding motions when stratus clouds occur in the default core are largely suppressed in the spectral core. As a result, the spectral CAM5 has more midtopped nimbostratus cloud fraction than the default configuration over the Sichuan basin, especially in the lower levels of the cloud profiles.