雷达资料同化模式对降雨-径流预报的影响  

The Impact of Radar Data Assimilation Modes on Rainfall-Runoff Forecasting

作　　者：田济扬[1,2] 刘荣华[1,2] 丁留谦[1] 刘含影 TIAN Jiyang;LIU Ronghua;DING Liuqian;LIU Hanying(China Institute of Water Resources and Hydropower Research,Beijing 100038,China;Research Center on Flood and Drought Disaster Reduction of Ministry of Water Resources,Beijing 100038,China)

机构地区：[1]中国水利水电科学研究院,北京100038 [2]水利部防洪抗旱减灾工程技术研究中心,北京100038

出　　处：《水文》2022年第4期30-37,共8页Journal of China Hydrology

基　　金：国家重点研发计划资助项目(2019YFC1510605);国家自然科学基金资助项目(51909274);中国水科院五大人才计划项目(JZ0199A022021)。

摘　　要：雷达资料同化是提高局地数值降雨预报精度的重要手段。以2014年6月17日和2016年7月9日福建梅溪流域发生的2场典型降雨洪水事件为例,设计9种不同的雷达资料同化模式,并引入到以WRF模式和基于CNFF-HM的分布式水文模型为主体构建的陆气耦合模型中,开展降雨-径流预报试验;探究中小流域尺度合理的雷达资料同化模式;分析雷达资料同化对降雨-径流预报结果的影响。结果表明,雷达资料同化能够提高中小流域陆气耦合降雨-径流预报精度;当同化时间间隔为6 h时,同化雷达反射率的降雨预报效果最好,但随着同化时间间隔缩短,雷达径向风的同化效果得到明显改善;逐小时同化雷达径向风支持下的陆气耦合模型,其降雨-径流预报结果最佳,2场降雨的累积雨量误差分别仅为7.55%和5.80%,相应洪水的洪峰流量误差为-12.65%和-18.26%,峰现时间均提前1h,NSE为0.866和0.927,在中小流域暴雨洪水预报中有一定的应用前景。Radar data assimilation is an effective method to improve the numerical rainfall prediction.In order to investigate the reasonable radar data assimilation mode and analyse the impact of radar data assimilation modes on rainfall-runoff forecasting,two storms occurred in Meixi catchment on June 17th 2014 and July 9th 2016,were chosen to test the nine different radar data assimi⁃lation modes,which were integrated in the coupled atmospheric-hydrological modeling system.The coupled atmospheric-hydrologi⁃cal modeling system was constructed by WRF model and CNFF-HM.The results show that radar data assimilation could improve the rainfall-runoff forecasting accuracy at small and medium catchment scale.With a six-hour interval,assimilating radar reflectiv⁃ity is the best choice for improve rainfall forecast.However,assimilating radial velocity can significantly improve the rainfall fore⁃casting as the time interval shortens.With the support of hourly radial velocity assimilation,the coupled atmospheric-hydrological modeling system can obtain the best rainfall and runoff forecasting.The relative errors of accumulated rainfall(RE)for the two storms are only 7.55%and 5.8%,respectively.The corresponding errors of flood peak flow(Rf)are-12.65%and-18.26%,the errors of flood peak appearance time(△t)are both 1 h and Nash-Sutcliffe efficiency coefficient(NSE)can reach 0.866 and 0.927,which indicates that the coupled atmospheric-hydrological modeling system with hourly radial velocity assimilation is an effi⁃cient technology to forecast the heavy rainfall and severe flood at small and medium catchment scale.

关 键 词：WRF模式 CNFF-HM模型 雷达资料同化 数值降雨预报 洪水预报 

分 类 号：TV87[水利工程—水利水电工程] X43[环境科学与工程—灾害防治]