Abstract:

In order to test forecast performance of different wet processes on a torrential rain, three cumulus parameterization schemes and seven cloud microphysics schemes in WRF with horizontal resolution of 9 km were selected to simulate a torrential rain occurring in Qingdao during 9-12 May 2014, and methods of Threat Sore (TS), rainfall distribution and time series of hourly rainfall were used to evaluate the forecast performance. The results show that the cumulus parameterization schemes of KF, GD and BMJ used in WRF with horizontal resolution of 9 km could improve precipitation TS with different degrees, and GD scheme did the best in the aspect of simulated precipitation area and intensity. For the simulation of heavy rainfall, there were little differences among seven cloud microphysics schemes with averaged TS of 0.64, in which KESSLER scheme did the best with TS of 0.73 and then WSM6, LIN and WSM5. They also took on the features of bigger precipitation area and stronger rainfall. The cumulus parameterization scheme had greater influence on the process of precipitation than that of cloud microphysics schemes, generally.

Key words: cumulus and cloud microphysics parameterization schemes, torrential rain, forecast performance

摘要：

基于9 km分辨率的中尺度数值模式WRF，通过TS评分、降水空间分布和降水强度检验评估3种积云对流和7种云微物理参数化方案对2014年5月10日青岛地区的一次暴雨天气的预报性能。结果表明：在分辨率为9 km的模式中考虑KF、GD和BMJ积云对流参数化方案时，能够不同程度提高大雨和暴雨的TS评分，且GD方案模拟的降雨落区和强度更接近实况。7种云微物理方案对暴雨模拟效果相差不大，平均TS评分达到0.64，其中KESSLER方案预报性能最好，TS评分达到0.73，其次是WSM6、LIN和WSM5方案，但也大都表现出暴雨范围偏大、雨量偏强的特点。对于此次降雨过程，积云对流参数化方案的预报性能优于云微物理过程方案的表现。

关键词: 积云和云微物理参数化方案, 暴雨, 预报性能