Using the WRF model and GFS data, a torrential rain process resulted from a large-scale synoptic system is back-forecasted, and the impact of different precipitation schemes in the WRF model with three horizontal resolutions (45 km, 15 km and 5 km) on precipitation forecast effects is analyzed.The results show: (1) For a torrential rain attributed to large-scale forcing, especially that mainly resulted from stratus clouds, variations in the microphysical process parameterization schemes have much more influence on precipitation than that of cumulus parameterization schemes. (2) The forecast effect of different microphysical processes differs from light to torrential rain. Among them, the TS score of Kessler scheme is sharply down when the precipitation level is up. As for other six schemes, the TS scores of light and torrential rainfall are much higher than those of moderate and heavy rainfall. (3)For light rainfall, the Lin scheme performs best; for torrential rainfall, the WSM 3 scheme is the best; and for moderate and heavy rainfall, the WSM 5 scheme is the best. In general, WSM3 proves to be the best of 7 schemes, WSM5 the next, and Kessler the worst according to TS scores. (4)The forecast effects of the WRF model do not always keep step with the increase of horizontal resolution. Obviously, there is a threshold for the model horizontal resolution. The amplitude of the forecast improvement decreases when the horizontal resolution excesses the threshold.%利用WRF模式和GFS资料对一次大尺度天气系统作用下的暴雨过程进行了回报,分析了WRF模式不同降水方案和3种不同的水平分辨率(45 km,15 km和5 km)对降水预报效果的影响.结果表明:①对于大尺度强迫作用较强的暴雨,尤其是层状云降水为主的暴雨,云微物理过程方案对降水的影响远大于积云参数化方案对降水的影响.②WRF模式不同的微物理过程方案对各等级降水量的预报效果差别较大.其中Kessler方案的TS评分明显随降水量级的增加而减小,其他6个方案的TS评分都呈现“两头大,中间小”的特点,即小雨和暴雨的TS评分较高,而中雨和大雨的TS评分较低.③对于小雨量级的降水,Lin方案的预报效果最好；对于中雨和大雨量级的降水,WSM 3方案的预报效果最好；对于暴雨量级的降水,WSM 5方案的预报效果最好；整体预报效果最好的是WSM 3方案,其次是WSM 5方案,Kessler方案最差.④WRF模式的降水预报效果并不总是随水平分辨率的提高而提高.模式水平分辨率的提高存在明显的阈值(15 km左右),当模式的水平分辨率提高到超过这一阈值以后,预报效果开始转差.
展开▼
