大通站枯水流量特征及概率拟合分析  

Analysis of Low Flow and Probability Fitting Model at Datong Station

作　　者：徐亚男[1] 荣艳淑[1] 贺晓婧[1] 王文[2] 

机构地区：[1]河海大学水文水资源学院,江苏南京210098 [2]河海大学水文水资源与水利工程科学国家重点实验室,江苏南京210098

出　　处：《水电能源科学》2015年第12期32-36,共5页Water Resources and Power

基　　金：国家自然科学基金项目(41371050)

摘　　要：以大通水文站为代表站，选取1950～2012年共63年的逐日流量资料，分析了最近60多年来长江中下游地区枯水流量的特征。用P—Ⅲ型曲线、广义极值分布模型（GEV）和广义帕累托分布模型（GPD）分别拟合汛期和非汛期的日流量极小值，通过柯尔莫哥洛夫检验（K—S检验）、概率点据相关系数检验（PPCC检验）及均方误差三种方法，检验分布模型的拟合精度。结果表明，长江中下游汛期低流量事件增多、强度增大，非汛期低流量事件减少、强度减弱。三种模型对大通站汛期和非汛期的枯水流量都有较好的拟合精度，但以P-Ⅲ分布模型模拟更优。用P-Ⅲ分布模型模拟的历史低流量事件显示，汛期最强的枯水流量强度达到68年一遇，非汛期最强的低流量事件强度达到38年一遇。在汛期，长江中下游防汛的同时，也要注意防旱。This paper presented the analysis of low flow for the middle and lower reaches of the Yangtze River, based on the daily flow data of Datong hydrological station from 1950 to 2012. Three probability models including Pearson-Ⅲ curve model （P-Ⅲ）, generalized extreme value distribution （GEV） and generalized Pareto distribution （GPD） were chose to fit the low flow series of flood season and non-flood season in the Datong hydrological station. Three methods including Kolmogorov-Smirnov test, PPCC test and mean square error were used to estimate the fitting ability of three probability models. The results indicate that the frequency and intensity of low flow in flood season increase, while those in the non flood season decrease. Three models all have well ability to fit the low flow series. However, the P-Ⅲ model is the best one. According to the fitting result of P-Ⅲ , the minimum low flow event in flood season appeared in 2009 with the return period of 68 year, while the minimum low flow event in the non-flood season happened in 1963 with the return period of 38-year. Therefore, we must pay attention to the drought in the flood season when the flood hazard is concerned.

关 键 词：枯水流量 P-Ⅲ分布 广义极值分布 广义帕累托分布 重现期 

分 类 号：TV121.5[水利工程—水文学及水资源] P343[天文地球—水文科学]