| 摘要: |
| 【目的】寻求用于陕北地区洪水频率分析的适宜方法，为该区水利防洪工程的设计建设提供支持。【方法】在介绍期望矩原理的基础上，选取陕北交口河、赵石窖、绥德、刘家河、张村驿及志丹6个水文站的洪峰流量资料，将期望矩法应用于陕北地区洪水序列参数估计中，计算设计洪水值，并拟合洪水频率曲线，利用累积相对偏差平方和评价所得洪水频率曲线对洪峰序列拟合效果，并与普通矩法拟合结果进行比较分析。【结果】利用期望矩法和普通矩法对陕北6个水文站洪水频率进行分析表明，用期望矩法估计的交口河、赵石窖、绥德、刘家河、张村驿及志丹6个水文测站洪峰流量序列的设计值与实测值累积相对偏差平方和分别为2.952，7.221，17.413，1.838，0.478和6.281，而用普通矩法估计的6个水文站设计值与实测值累积相对偏差平方和分别为1.217，15.830，42.497，2.008，0.755和70.253。由此可得，除交口河站以外，其他5个站的设计值误差均较普通矩法小，期望矩法能充分利用历史洪水资料，降低设计值偏差，可更好地拟合洪水序列。【结论】在考虑历史洪水资料时，期望矩法是一种简单、稳定和有效的参数估计方法，可用于陕北地区的洪水频率计算。 |
| 关键词: 洪水频率 参数估计 期望矩法 P-Ⅲ型分布 陕北地区 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目（51179160，50879070，51079037）；高等学校博士学科点专项科研基金项目（20110204110017） |
|
| Using expected moments algorithm for flood frequency analysis in northern Shaanxi |
|
WANG Jun-zhen,SONG Song-bai
|
| Abstract: |
| 【Objective】This study aimed to obtain a reasonable method for flood frequency analysis in northern Shaanxi to support flood management and hydraulic engineering in this area.【Method】Based on the introduction to principles of expected moments algorithm (EMA),this paper applied EMA in flood frequency analysis in northern Shaanxi.Annual maximum flow information was obtained from hydrologic stations including Jiaokou River,Zhaoshiyao,Sudie,Liujia River,Zhangjiashan and Zhidan.Design flood rate was calculated and the flood frequency curve was fitted.Then,the cumulative squared error was used to evaluate the fitting results.The results were also compared with that of the traditional moments method.【Result】The cumulative sums of squared relative error of EMA method for Jiaokou River,Zhaoshiyao,Sudie,Liujia river,Zhangjiashan and Zhidan stations were 2.952,7.221,17.413,1.838,0.478,6.281,respectively,while those of traditional method were 1.217,15.830,42.497,2.008,0.755,and 70.253,respectively.EMA method gave more accurate results at stations except Jiaokou River.EMA method could efficiently use historical flood data and better fit the flood series with higher precision.【Conclusion】EMA is a simple,stable and effective method for flood frequency analysis when historical flood information is available,and it can be used in northern Shaanxi. |
| Key words: flood frequency parameters estimation expected moments algorithm P-Ⅲ distribution Northern Shaanxi |
