摘要
应用2008年奥运期间(6—10月)所使用的多种中尺度数值预报业务运行模式GRAPES_MESO15km(GRAPES)、华北中尺度模式(MM5)和北京区域气象中心的快速更新循环同化预报系统(RUC),对北京市24小时定量降水预报产品采用多个统计量进行了较详细的统计检验,并应用多元回归方法对多种模式的降水预报产品进行了集成试验。检验结果表明:GRAPES_MESO15km模式对奥运期间北京市晴雨及小雨预报具有较好的预报能力,而RUC产品对中雨及以上量级的降水预报则表现出较好的预报效果,这对今后北京市精细化降水预报具有一定参考价值。
The GRAPES-MESO15km,MM5 and BJ-RUC（rapid assimilation forecast update cycle system of Beijing）were used in the meteorological services for the Beijing 2008 Olympic Games.The prediction qualities of 24-h quantitative precipitation forecast products of these models were examined in detail.Firstly,some statistics were calculated,such as the correlation coefficient,TS,FAR,RMSE etc.Then,by use of the multiple regression method,the integration test to the precipitation forecast products of multiple models was taken.The results show that GRAPES-MESO15km model has a good forecasting performance in the aspect of forecasts for light rain and raining or not,while RUC has a good forecasting capability to the moderate and greater rain.The verification results have a good reference value for the future refinement precipitation forecast in Beijing.
出处
《气象》
CSCD
北大核心
2009年第11期3-8,共6页
Meteorological Monthly
基金
973"我国南方致洪暴雨监测与预测的理论和方法研究项目-高分辨
非静力中尺度暴雨数值预报模式的发展及其预报系统的建立"课题(2004CB418306)
中国气象局北京城市气象研究所城市气象科学研究基金项目(UMRF200706)
关键词
多模式
降水产品
检验
集成试验
multiple models precipitation forecast product performance examination integration test
