Evaluating regional cloud-permitting simulations of the WRF model for the Tropical Warm Pool International Cloud Experiment (TWP-ICE, Darwin 2006)

Data from the Tropical Warm Pool International Cloud Experiment (TWP-ICE) were used to evaluate Weather Research and Forecasting (WRF) model simulations with foci on the performance of three six-class bulk microphysical parameterizations (BMPs). Before the comparison with data from TWP-ICE, a suite of WRF simulations were carried out under an idealized condition, in which the other physical parameterizations were turned off. The idealized simulations were intended to examine the interaction of BMP at a "cloud-resolving" scale (250 m) with the nonhydrostatic dynamic core of the WRF model. The other suite of nested WRF simulations was targeted on the objective analysis of TWP-ICE at a "cloud-permitting" scale (quasi-convective resolving, 4 km). Wide ranges of discrepancies exist among the three BMPs when compared with ground-based and satellite remote sensing retrievals for TWP-ICE. Although many processes and associated parameters may influence clouds, it is strongly believed that atmospheric processes fundamentally govern the cloud feedbacks through the interactions between the atmospheric circulations, cloudiness, and the radiative and latent heating of the atmosphere. Based on the idealized experiments, we suggest that the discrepancy is a result of the different treatment of ice-phase microphysical processes (e.g., cloud ice, snow, and graupel). Because of the turn-off of the radiation and other physical parameterizations, the cloud radiation feedback is not studied in idealized experiments. On the other hand, the "cloud-permitting" experiments engage all physical parameterizations in the WRF model so that the radiative heating processes are considered together with other physical processes. Common features between these two experiment suites indicate that the major discrepancies among the three BMPs are similar. This strongly suggests the importance of ice-phase microphysics. To isolate the influence of cloud radiation feedback, we further carried out an additional suite of simulations, which turns off the interactions between cloud and radiation schemes. It is found that the cloud radiation feedback plays a secondary, but nonnegligible role in contributing to the wide range of discrepancies among the three BMPs.