###
气象:2011,37(5):540-546
本文二维码信息
码上扫一扫!
定量分析数值模式日降水预报结果的不确定性
(1.中国气象局武汉暴雨研究所, 武汉 430074;2.武汉大学水利水电学院, 武汉 430072)
Quantitative Analysis of the Uncertainty of the Daily Precipitation Predicted by Numerical Model
(1.Institute of Heavy Rain, China Meteorological Administration, Wuhan 430074;2.Department of Hydrology and Water Resources, Wuhan University, Wuhan 430072)
摘要
图/表
参考文献
相似文献
本文已被:浏览 1323次   下载 2076
投稿时间:2010-08-13    修订日期:2011-01-28
中文摘要: 集对数值预报模式日降水预报结果与观测资料,根据联合熵处于最大、最小联合熵之间的相对位置设定不确定性指标U,以便定量分析NWM降水预报结果的不确定性。初步计算结果表明:U与模式预报随机偏差有着紧密联系,但与系统偏差无关;NWM的降水预报不确定性指标超过60%,即其既具有参考性,又具有不确定性,特别是对强降水的预报,不确定性更大;对我国大陆范围24小时降水预报而言,日本模式比T213不确定性小,特别是日本谱模式(RSM)预报在出现40~80 mm降水时比T213具有更高的参考价值。
Abstract:Collecting the pairs of the numerical weather prediction model daily precipitation forecast results and observed data, the uncertainty index U is set, according to the relative position that the joint entropy exists between maximum joint entropy and minimum joint entropy, so as to analyze the uncertainty of NWM precipitation prediction results quantitatively. Preliminary results show that: U is closely linked with the random deviation error of the model forecast, but has nothing to do with the system deviation error; the uncertainty index of the NWM precipitation forecast is over 60%; it has both reference significance and uncertainty; there is bigger uncertainty to forecast the heavy precipitation. As for the continent wide 24 h precipitation forecast in China, the uncertainty of the Japan model is smaller than the T213. When a 40 to 80 mm precipitation appears in the forecast of Japan Regional Spectral Model (RSM), it has the higher reference value than the T213.
文章编号:     中图分类号:    文献标志码:
基金项目:国家自然科学基金资助项目 (40675070),湖北省科技攻关计划 (2007AA301B57)和公益性行业(气象)专项(GYHY200806002)共同资助
引用文本:
沈铁元,廖移山,彭涛,崔春光,殷志远,宋星原,张利平,2011.定量分析数值模式日降水预报结果的不确定性[J].气象,37(5):540-546.
SHEN Tieyuan,LIAO Yishan,PENG Tao,CUI Chunguang,YIN Zhiyuan,SONG Xingyuan,ZHANG Lipin,2011.Quantitative Analysis of the Uncertainty of the Daily Precipitation Predicted by Numerical Model[J].Meteor Mon,37(5):540-546.