Impact of precipitating ice on the simulation of a heavy rainfall event with advanced research WRF using two bulk microphysical schemes

In this study, the Weather Research and Forecasting (WRF) model version 3.2 is used to examine the impact of precipitating ice and especially snow-graupel partitioning in the simulation of a heavy rainfall event over Chalkidiki peninsula in Northern Greece. This major precipitation event, associated with a case of cyclogenesis over the Aegean Sea, occurred on the 8th of October 2006 causing severe flooding and damage. Two widely used microphysical parameterizations, the Purdue Lin (PLIN) and WRF Single-Moment 6-class scheme (WSM6) are compared with available raingauge measurements over the complex topography of Chalkidiki. To further investigate the importance of snow and graupel relative mass content and the treatment of precipitating ice sedimentation velocity, two older versions of the WSM6 scheme were compiled and run with the current model. The verification results indicate that all simulations were found to match raingauge data more closely over the eastern mountainous Chalkidiki peninsula where maximum accumulations were observed. In other stations all schemes overestimate 24h accumulated rainfall except a station situated at the western part of the peninsula, where none of the simulations was able to reproduce observed rainfall. Graupel dominance in PLIN generates rapid precipitation fallout at the point of maximum predicted 24h accumulation. Similar behavior is shown in WSM6 from WRF version 2, but with significant less rainfall. Increasing snow amounts aloft, due to the unified treatment of precipitating ice in WSM6 from WRF version 3, modifies rain dynamics which decrease rainfall rates, but increases 24h accumulations. A sensitivity experiment where PLIN is used with snow accretion by graupel turned off, indicated that this process seems to be the most important factor controlling the differences in surface precipitation between PLIN and WSM6 from WRF version 3, determining the spatial and temporal distribution of this heavy precipitation event. The results also revealed that snow overestimation can lead to high rainfall accumulations, even though rain is more evenly distributed over the 24h period, deteriorating precipitation forecast.