基于年最大抽样序列的暴雨频率估计值的修正研究  

Revision of Extreme Precipitation Frequency Estimation Based on Annual Maximum Series

作　　者：邵月红[1] 刘玲[1] 吴俊梅 刘永和[3] SHAO Yuehong;LIU Ling;WU Junmei;LIU Yonghe(College of Hydrometeorology,Nanjing University of Information Science and Technology,Nanjing 210044,China;Kunshan Meteorological Bureau,Kunshan 215300,China;Institute of Resources and Environment,Henan Polytechnic University,Jiaozuo 454000,China)

机构地区：[1]南京信息工程大学水文气象学院,江苏南京210044 [2]昆山气象局,江苏昆山215300 [3]河南理工大学资源与环境学院,河南焦作454000

出　　处：《水文》2019年第1期7-14,共8页Journal of China Hydrology

基　　金：江苏省青年基金项目(BK20141001);淮河流域气象中心开放项目暨安徽省大气科学与卫星遥感重点实验室开放基金项目(HRM201702)

摘　　要：以江苏省为研究区,利用地区线性矩法和周文德频率转换关系式对年最大抽样序列估算的频率估计值进行修正,推求合理可靠的频率估计值。研究结果表明:基于气象成因和水文统计特性相结合的方法将研究区划分为5个水文气象一致区。通过蒙特卡罗模拟、均方根误差判断各一致区的最优分布;年最大抽样序列的最优分布以GEV为主,年超大抽样序列的最优分布以GPA为主。基于两个实测序列计算的频率估计值比值和基于年最大抽样与周文德转换后的年超大抽样计算的估计值比值进行比较,研究表明:周文德频率转换关系式适用于该研究区。在此基础上,对年最大抽样序列估计的频率估计值进行修正。站点的频率估计值分布特征表现为:随着重现期的增加而增加,不同重现期下的暴雨频率估计值空间分布趋势基本一致,从南向北渐增,暴雨中心分布在苏北的宿迁和连云港一带;站点的频率估计值与同频率的观测值吻合较好,相对误差较小、相关性很高。基于年最大抽样序列和周文德公式频率转换能够推求计算合理可靠的暴雨频率估计值。The extreme precipitation frequency estimation based on the annual maximum series(AMS) was revised by using hydrometeorological regional L-moments method and Chow’s equation for Jiangsu Province, China. Firstly, the study area was divided into five homogeneous regions by a coupled methods of hydrologic statistics and climatological factors. The optimum distribution for each region was determined by the Monte Carlo simulation and root mean square error. For AMS, Generalized Extreme Value(GEV) is the more appropriate distribution in most homogeneous regions, and Generalized Pareto distribution(GPA) is the optimum curve for annual exceedance series(AES). In the next step, the frequency estimation ratios were respectively computed by using the real AMS and AES, the real AMS and the generated AES by Chow’s equation. The results show that Chow’s formula is applicable for the study area. A set of rational and reliable frequency estimation can be obtained by using hydrometeorological regional L-moments method based on AMS and Chow’s equation. The spatial distribution of quantiles show that the northern area are greater than the south area, and all of them increase with the increase of return period. The highest frequency estimates were observed near the Suqian and Lianyungang stations in the northern area of Jiangsu Province. The frequency estimates are in good agreement with the observed values at the same frequency, which can provide important basis for flood control and water resource management.

关 键 词：地区线性矩法 年最大抽样序列 年超大抽样序列 频率估计值 

分 类 号：P333[天文地球—水文科学]