A Double-Moment Multiple-Phase Four-Class Bulk Ice Scheme. Part II: Simulations of Convective Storms in Different Large-Scale Environments and Comparisons with other Bulk Parameterizations

Abstract

Part I of this study described a detailed four-class bulk ice scheme (4ICE) developed to simulate the hydro-meteor profiles of convective and stratiform precipitation associated with mesoscale convective systems. In Part II, the 4ICE scheme is incorporated into the Goddard Cumulus Ensemble (GCE) model and applied without any ?tuning? to two squall lines occurring in widely different environments, namely, one over the ?Pica) ocean in the Global Atmospheric Research Program's (GARP) Atlantic Tropical Experiment (GATE) and the other over a midlatitude continent in the Cooperative Huntsville Meteorological Experiment (COHMEX). Comparisons were made both with earlier three-class ice formulations and with observations. In both cases, the 4ICE scheme interacted with the dynamics so as to resemble the observations much more closely than did the model runs with either of the three-class ice parameterizations. The following features were well simulated in the COHMEX case: a lack of stratiform rain at the surface ahead of the storm, reflectivity maxima near 60 dBZ in the vicinity of the melting level, and intense radar echoes up to near the tropopause. These features were in strong contrast with the GATE simulation, which showed extensive trailing stratiform precipitation containing a horizontally oriented radar bright band. Peak reflectivities were below the melting level, rarely exceeding 50 dBz, with a steady decrease in reflectivity with height above. With the other bulk formulations, the large stratiform rain areas were not reproduced in the GATE conditions. The microphysical structure of the model clouds in both environments were more realistic than that of earlier modeling efforts. Number concentrations of ice of O(100 L?1) occurred above 6 km in the GATE model clouds as a result of ice enhancement and rime splintering in the 4ICE runs. These processes were more effective in the GATE simulation, because near the freezing level the weaker updrafts were comparable in magnitude to the fall speeds of newly frozen drops. Many of the ice crystals initiated at relatively warm temperatures (above ?15°C) grew rapidly by deposition into sizes large enough to be converted to snow. In contrast, in the more intense COHMEX updrafts, very large numbers of small ice crystals were initiated at colder temperatures (below ?15°C) by nucleation and stochastic freezing of droplets, such that relatively few ice crystals grew by deposition to sizes large enough to be converted to snow. In addition, the large number of frozen drops of O(5 L?1) in the 4ICE run am consistent with airborne microphysical data in intense COHMEX updrafts. Numerous sensitivity experiments were made with the four-class and three-class ice schemes, varying fall speed relationships, particle characteristics, and ice collection efficiencies. These tests provide strong support to the conclusion that the 4ICE scheme gives improved resemblance to observations despite present uncertainties in a number of important microphysical parameters.