基于EC集合预报产品的降水预报检验

Precipitation Forecasting Verification Based on EC Ensemble Prediction Products

为了对EC集合预报产品的应用性能进行检验,笔者基于EC集合预报的概率匹配平均降水量、融合降水量以及中位数、众数和控制预报降水量等5个产品,通过TS评分方法对内蒙古2014年汛期降水过程预报进行检验。结果表明,各检验成员产品对过程的把握能力差异较大,集合预报对降水偏大的降水过程评分比较高,对暴雨以上量级降水的预报存在着较大的偏差,预报效果较差。平均TS评分结果表明,集合预报产品尤其是概率预报产品具有一定的参考价值,其次是控制预报和融合产品。

The paper aims to test the application performance of EC ensemble prediction products. Based onthe 5 products： probability of EC ensemble forecast to match average precipitation, fusion precipitation,median, mode and control forecast precipitation, and with the TS score method, the authors tested theprecipitation forecast in flood season in Inner Mongolia in 2014. The results showed that differences existedamong product＇s ability to grasp the process, there was a high score on ensemble forecast of the precipitationprocess, for heavy rain and above, there was a large scale precipitation forecast deviation, the forecast effectwas poorer. Average TS score results showed that the ensemble forecast products, especially the probabilityforecast, had certain reference value, the second was control forecast and fusion products.