2012年深圳前汛期降水EC与HAPS模式检验  

Verification of Precipitation of the Annually First Rain Season of Shenzhen in 2012 Between Numerical Models From ECMWF And HAPS

作　　者：徐文文[1] 陈申鹏 李晴岚 吴敏 

机构地区：[1]深圳市国家气候观象台,广东深圳518040 [2]深圳先进技术研究院,广东深圳518040 [3]慈溪市气象局,浙江慈溪315300

出　　处：《广东气象》2013年第5期6-9,共4页Guangdong Meteorology

基　　金：深圳市科技研发资金项目(JCYJ20120617115926138);中国气象局华南区域气象中心科技攻关项目(GRMC2012M15)共同资助

摘　　要：对2012年前汛期（4-6月）欧洲中心全球模式与HAPS中尺度模式的降雨预报场进行逐6h晴雨检验与24 h分量级降水检验,得出：EC逐6h降水预报空报多而漏报少,预报无降水时可信度更高,TS评分在0.3 ～0.4之间;24 h分级检验中预报能力随着降水量级的增大而减小,空报率、漏报率增大,TS评分减小;EC随着预报提前量的减小TS评分增加,起报时次对预报能力无明显影响;总体来说,EC能为深圳10 d预报提供较好的参考价值,尤其是6d预报,对于暴雨的预报有待预报员订正.HAPS逐6h晴雨预报TS评分在0.48～0.32之间,提前量6h内预报效果好,TS评分0.4以上,有较高的参考价值;24 h分级降水预报能力与EC相同,也是随着降水量级的提升预报能力降低;但是HAPS预报能力与提前量的相关性主要表现在提前量6h以内,提前量减小TS评分明显增大,对于提前量大于6h的预报能力无明显变化;总体来说,预报员可依靠HAPS在短期预报基础上做好短临预报的订正工作,开展24 h分区预报.Assessment is made of fine-sky versus rain forecast at 6h intervals and 24b graded rain forecast based on the output of a global ECMWF model and HAPS, a mesoscale model, for the annually first rain season of 2012 （April to June） in Shenzhen. The results are as follows. The 6h interval forecast by the EC1VIWF model has large number of false alarms but few misses, its reliability is higher when forecasting absence of precipitation, with the TS score between 0.3 and 0.4. For the 24h graded rain forecast, its forecasting capability decreases, the rates of false alarms and misses increase, and the TS score declines, with the increase of rain magnitude. With the decrease of advanced forecast time, the ECMWF modelg TS score in- creases and the time to initiate forecasting does not have apparent impact on its forecasting capability. In general, the ECMWF model can provide good basis for 10 d, especially 6 d, forecast in Shenzhen, while the forecast of heavy rain needs the forecaster to correct. For HAPS, the TS score of its 6 h forecast of fine-sky versus rain is between 0.48 and 0.32, having high value of reliability. Its capability is also the same as that of the ECMWF model in the 24h graded rain forecast ; it also decreases with the increase of the magnitude of rain. However, the correlativity between the forecasting capability of HAPS and the amount of advanced forecast time is mainly shown within the advanced time of 6 h. A decreased time of advancement is associat- ed with much increased TS scores while no apparent changes are found with the forecasting capability with the advanced duration of time larger than 6 h. In general, the forecaster can operate 24 h graded rain fore- cast using HAPS to correct nowcasting on the basis of short-term forecast.

关 键 词：天气学 前汛期降水 欧洲中心全球模式 HAPS中尺度模式 深圳 

分 类 号：P44[天文地球—大气科学及气象学]