ECMWF模式在岳阳市降水预报中的效果检验和订正分析

Effect verification and correction analysis of ECMWF model in precipitation forecast over Yueyang

应用TS评分、预报偏差BIAS、ETS以及TSS等方法,对2016—2019年ECMWF模式在岳阳市降水预报中的效果进行检验和分析。检验结果表明,ECMWF模式对岳阳市降水预报总体效果较好,呈逐年提升态势且年际变化幅度不大,但季节差异大;对小雨量级降水预报能力优于其他量级但空报严重,中雨次之,对大雨及以上量级预报能力较弱,尤其是对暴雨预报能力差且漏报严重。针对该模式小雨量级降水空报率高和暴雨漏报率高的现象,分别进行了消空和消漏订正处理后再进行检验,结果表明,将模式预报的0.1~0.3 mm降水进行消空订正后有明显的技巧提升;将60 h时效内模式预报的逐12 h雨量≥35mm降水进行消漏订正后,暴雨漏报率明显下降,TS评分也有提升,有正技巧,但无论是消空还是消漏,还需要结合降水发生的季节、天气系统、气候背景以及地形等多方面因素综合考虑。

The effect of ECMWF model in precipitation forecast over Yueyang from 2016 to 2019 has been tested and analyzed by TS score,forecast BIAS,ETS and TSS.The test results showed that the ECMWF model had a good overall effect on the precipitation forecast in Yueyang,which showed a year-by-year increasing trend and a small annual variation range,but the seasonal difference was large.The ECMWF model had the best forecast ability for light rain,which was overstated seriously,then moderate rain is next,and the ability to forecast heavy rain and above is weak,especially to forecast intense fall which is seriously understated.Aiming at the phenomenon of high false-alarm forecast rate of light rain and high missing forecast rate of intense fall by the model,it′s tested after making correction to eliminate the understatement and overstatement,and it turned out that the forecast ability of the model improved significantly after being corrected to eliminate of overstatement about 0.1~0.3 mm of precipitation.Besides,the rainfall greater than 35mm predicted by the model has been corrected by elimination of understating within 60 h lead time,which shows that the miss-report rate of intense fall decreased obviously,and the TS score also increased with positive skills.On the whole,whether it′s to eliminate the false-alarm forecast or missing forecast,it′s necessary to take into account the season of precipitation,weather system,climate background and topography.