欢迎访问《中国农学通报》,

中国农学通报 ›› 2015, Vol. 31 ›› Issue (10): 232-239.doi: 10.11924/j.issn.1000-6850.casb14100074

所属专题: 农业气象

• 资源 环境 生态 土壤 气象 • 上一篇    下一篇

乌江流域面雨量的WRF预报检验

唐延婧1,彭 芳2,尚媛媛1,胡跃文3   

  1. (1贵州省气象服务中心,贵阳 550002;2贵州省气象台,贵阳 550002;3贵州省气象信息中心,贵阳 550002)
  • 收稿日期:2014-10-21 修回日期:2015-03-18 接受日期:2015-02-25 出版日期:2015-05-05 发布日期:2015-05-05
  • 通讯作者: 唐延婧
  • 基金资助:
    公益性行业(气象)专项“乌江流域精细化水文气象服务系统的研究与开发”(GYHY201306059)。

WRF Area Rainfall Forecast Verification of Wujiang River Watershed

Tang Yanjing1, Peng Fang2, Shang Yuanyuan1, Hu Yuewen3   

  1. (1Meteorological Service Center of Guizhou Province, Guiyang 550002;2Meteorological Station of Guizhou Province, Guiyang 550002;3Meteorological Information Center of Guizhou Porvince, Guiyang 550002)
  • Received:2014-10-21 Revised:2015-03-18 Accepted:2015-02-25 Online:2015-05-05 Published:2015-05-05

摘要: 为了研究WRF对乌江流域面雨量的预报能力,认识WRF降水预报的特点,提出释用思路。笔者运用晴雨预报、等级预报检验和预报偏差分析方法,对乌江流域2010—2013年汛期面雨量的WRF降水预报进行检验。结果表明,乌江流域的晴雨预报具有相当的可靠性。等级预报准确率在53%左右。按降水等级检验发现:WRF对无雨和小雨的预报,尤其小雨的预报有较好的参考性;但对中雨以上的降水预报能力较差。中雨以上随降水量增大,预报偏差增大;预报偏差大多偏小,接近正常的较少,偏差100%的较多。WRF对大降水的预报能力较差,对中雨的预报有一定可靠性。根据检验结果提出简单的集合预报方法,其与实况相关性更好,预报接近正常的比率提高,预报偏小的情况也有所改进。

关键词: 暴雨, 暴雨, 时空分布特征, M-K突变检验方法

Abstract: The paper aims to verify the accuracy of WRF area rainfall product of Wujiang River Watershed, understand the character of WRF rainfall forecast, and propose interpretation method. WRF area rainfall product of Wujiang River Watershed in flood seasons from 2010 to 2013 was verified through methods of clear-rain verification, classified verification and forecast departure. It was found that the clear-rain forecast was quite reliable. Classified verification showed that WRF forecast was better in no-rain and especially small-rain level, but the result was not so good for over moderate-rain level. In levels above moderate-rain, forecast departures were mostly smaller than the real, and rarely close to the real. As rainfall became heavier, the departures became lager. WRF performed badly in heavy precipitation and passably in moderate-rain level. The ensemble forecast was related better to the real, which was more close to the normal.