2007年主汛期武汉区域四种数值模式定量降水预报评估  

Evaluation of Quantitative Precipitation Estimation of Numerical Weather Prediction Models in Wuhan Region During Main Flood Season of 2007

作　　者：孟英杰[1] 吴洪宝[1] 王丽[2] 张萍萍[2] 

机构地区：[1]南京信息工程大学大气科学学院,南京210044 [2]武汉中心气象台,武汉430074

出　　处：《暴雨灾害》2008年第3期273-277,共5页Torrential Rain and Disasters

基　　金：湖北省科技厅"城市防灾减灾应急气象服务系统开发研究"项目(2006AA301B45)资助

摘　　要：利用武汉中心气象台研发的"武汉区域气象中心天气轨道业务产品检验与评估平台,"对武汉区域气象中心在业务中使用的T213、AREM、日本数值预报模式和德国数值预报模式在2007年主汛期的降水预报进行了分级降水检验以及时空分布演变综合评估。结果表明,日本数值预报模式的综合预报性能最好,AREM次好,各模式均存在对强降水预报漏报率偏大的问题;AREM模式对降水带分布和中心强度的预报与实况最接近,表现出对降水带分布较强的预报能力,其它模式对强降水中心位置及强度的预报均有一定偏差;四种数值模式对区域强降水过程的发展趋势具有较强的预报能力,但降水量预报与实况有一定的差距。By using the evaluation and analysis system of weather products in wuhan regional meteorological center, four numerical weather prediction models are verified and assessed for their QPE using in Wnhan region during main flood season of 2007. Firstly, the integrated forecast ability of the Japanese global model is the best, AREM the second best, and the fail rates of four models are all bigger than observations. Secondly, the distribution and intension of precipitation of AREM furthest closing to the observations shows a good spatial forecast capability for the quantitative precipitation forecasting. Thirdly, Four models have good performs on forecasting the developing trends of rainfall processes, but their QPE need to be improved for some errors comparing with the rain gauge observations.

关 键 词：数值模式 TS评分 漏报率 空报率 预报偏差 

分 类 号：P456.7[天文地球—大气科学及气象学]