Journal of Arid Meteorology

Previous Articles     Next Articles

Application Test of Ensemble Forecast Methodology to Rainstorm Forecast in Shanxi Province

  

  1. Shanxi Meteorological Observatory, Taiyuan 030006, China
  • Online:2017-04-30 Published:2017-04-30

集合预报方法在山西暴雨预报中的应用试验

  

  1. 山西省气象台,山西 太原 030006
  • 作者简介:闫慧,女,硕士,工程师,主要从事数值模拟和天气预报技术研究. E-mail: yanhui8796@163.com
  • 基金资助:

    国家自然科学基金“黄河中游地区突发性大暴雨MCC结构特征研究”(41475050)、中国气象局小型基建项目“全国集合预报业务系统建设(一期)”和山西省气象局科学技术研究课题“山西暴雨集合概率预报方法研究”(SXKMSTQ20165210)共同资助

Abstract:

Based on the products of ECMWF high resolution deterministic forecast and ensemble forecast, the rainfall forecasts of 9 rainstorm processes in 2014 in Shanxi Province were tested and analyzed by using the local statistics quantities fusion and joint probability forecast method. The results indicated that the major influence systems of the 9 rainstorm processes were significantly different. The strong precipitations were concentrated in the central and southern regions of Shanxi Province, which presented regional, local or large-range dispersion characteristics. The heavy rainfall areas forecasted by the local statistics quantities fusion method were more closed to the observation, while the greater probability range of joint probability forecast had a good indication to the rainstorm area. TS scores showed that for the regional and local rainstorm weather, we should be more referred to the statistics quantities fusion forecast, while for the large-range dispersion rainstorm process, ECMWF deterministic forecast was more significant than that of the local statistics quantities fusion forecast with higher false prediction rate.

Key words: ensemble forecast, rainstorm, statistics quantities fusion, joint probability

摘要:

利用ECMWF高分辨率确定性预报和大气模式集合预报产品,采用本地统计量融合方法和联合概率方法,对2014年山西省9次暴雨天气过程的降水预报进行检验分析。结果表明:(1)2014年山西省9次暴雨过程影响系统差异较大,强降水多集中在山西中南部地区,暴雨落区表现为区域性、局地性或大范围分散性特征;(2)本地统计量融合方法对于大雨落区的预报与实况较为接近,而联合概率预报方法的大概率范围对于暴雨落区预报具有较好的指示意义;(3)TS评分检验表明,对于区域性或局地性暴雨天气过程,可更多参考本地统计量融合方法,而对于大范围分散性的暴雨天气过程,本地统计量融合方法空报较多,ECMWF确定性预报更具参考价值。

关键词: 集合预报, 暴雨, 统计量融合, 联合概率

CLC Number: