摘要
利用误差分析和累计降水检验的方法,对日本、德国、T213三种数值预报模式在强降水过程中的预报进行了统计和比较,并分析了强降水过程中日本模式的预报定量误差。结果表明:三模式产品在强降水过程预报中预报值基本上小于实况,而且随着实况量级的增加,预报偏小的次数增多。其中日本模式的预报最为稳定,且结果较接近实况,为解释应用提供了依据和数据参考。
The precipitation products of JMA,GRE and T213 in strong precipitation courses were statisticed and compared by using the means of error analysis and accumulative verification, and the quantitative error of JMA＇s precipitation forecasting in heavy rainfall processes was analyzed. The results show that the forecasting value of these three models are all less than the observed value in heavy rainfall forecast, and with the precipitation increasing, the times of underestimating increases. JMA＇s forecasting is the most steady one among JMA,GRE,T213,which has the consistent character and is more near to the reality. It provides the gist and reference for JMA forecasting＇s explaining and applying.
出处
《暴雨灾害》
2007年第4期301-305,共5页
Torrential Rain and Disasters
基金
中国气象局新技术重点推广项目"长江中游气象水文预报与服务系统研究"(CMATG2006Z08)资助
关键词
数值模式
降水预报
定量误差
统计检验
NWP model
Pprecipitation forecast
Quantitative error
statistical lest
