“7．18”山东暴雨过程分析II：WRF模式边界层参数化方案对降水的影响

目的]分析不同行星边界层参数化方案对此次暴雨过程中降水模拟效果的影响。方法]利用WRF3.4模式,选用不同边界层参数化方案对"7.18"山东暴雨过程进行敏感性试验,研究不同行星边界层参数化方案对此次暴雨过程中的降水分布、降水极值出现时间以及BS、ETS评分的影响。结果]不论是否使用边界层参数化方案,均能模拟出雨带的基本走向,但不同方案对于降水中心强度及其位置的模拟存在一定差异。不同边界层方案得到的模拟降水峰值的出现时间普遍比实际降水落后2~3 h,且其降水峰值之间存在较大差异,其中YSU方案对降水峰值的模拟效果最好。综合ETS评分和BS评分结果可知,对于降水量较小(25 mm)的模拟,不使用边界层方案反而效果最好;对于中等降水量级(25~50 mm)的模拟,使用Boulac方案比较合适;对于暴雨或特大暴雨(50 mm)的模拟,使用YSU方案效果最好。结论]WRF模式中不同边界层参数化方案对降水模拟效果的影响较大,选择合适的边界层参数化方案能显著提高降水的预报准确率。

Analysis of “7.18” Shandong Heavy Rainfall II:the Impact of PBL Parameterizations to Precipitation in WRF Model

Objective] The aim was to analyze effects of different boundary layer parameterization schemes on precipitation simulation results in the heavy rainfall process.Method] Using WRF mode 3.4 version,sensitivity test was conducted on “7.18” Shandong heavy rainfall process, effects of different boundary layer parameterization schemes on precipitation distribution,extreme occurrence time,BS and ETS scoring results were analyzed.Result] The results showed that the rain band’s general trend can be well simulated no matter the PBL parameterization schemes are used or not,but simulation of the intensity and location of the rainfall center differ to some extent .The showing up time of the rain peak ob-tained by different PBL schemes is generally 2-3 hours later than the actual result,and their rain peaks greatly differ with each other,among which the result of YSU scheme’s simulation is the best.Synthesizing the result of ETS (Equitable Threat Score) and BS (Brier Score) show that no use of PBL schemes is the best for the simulation in the case of small amount of precipitation (i.e.,less than 25 mm),while Boulac scheme is more appropriate in the case of medium rainfall-class (i.e.,25-50 mm).As for rainstorm or extreme precipitation (i.e.,more than 50 mm), YSU scheme gets the best result.Conclusion] Different boundary layer parameterization schemes in WRF mode have significant influence on simulation results of rainstorm ,using appropriate boundary layer parameterization scheme can improve the accuracy of precipitation forecast signif-icantly.