1. 广州地理研究所广东省地理空间信息技术与应用公共实验室,广东,广州,510070
2.
3. 中国科学院地理科学与资源研究所中国科学院陆地水循环及地表过程重点实验室,北京,100101
4. 中山大学地理科学与规划学院,广东,广州,510275
纸质出版日期:2019,
网络出版日期:2019-5-25,
扫 描 看 全 文
赵玲玲, 陈子燊, 刘昌明, 等. 基于广义Pareto分布的洪水序列频率分析[J]. 中山大学学报(自然科学版)(中英文), 2019,58(3):32-39.
ZHAO Lingling, CHEN Zishen, LIU Changming, et al. Flood sequence frequency analysis based on generalized Pareto distribution[J]. Acta Scientiarum Naturalium Universitatis SunYatseni, 2019,58(3):32-39.
赵玲玲, 陈子燊, 刘昌明, 等. 基于广义Pareto分布的洪水序列频率分析[J]. 中山大学学报(自然科学版)(中英文), 2019,58(3):32-39. DOI: 10.13471/j.cnki.acta.snus.2019.03.004.
ZHAO Lingling, CHEN Zishen, LIU Changming, et al. Flood sequence frequency analysis based on generalized Pareto distribution[J]. Acta Scientiarum Naturalium Universitatis SunYatseni, 2019,58(3):32-39. DOI: 10.13471/j.cnki.acta.snus.2019.03.004.
广义Pareto分布是超定量洪水频率分析的常用线型,其阈值的合理选择是关键。对洪水序列超阈值分布的阈值选择和超定量样本的适用性检验两个方面深入讨论,为设计洪水提供更优的频率分析方法。基于广东省曹江流域大拜水文站1968~2013年日流量的洪水序列数据,采用不同的抽样方法,检验不同洪水超定量的泊松分布、广义极值分布和超定量样本分布的拟合优度指标。将择优的GPD、GEV和P-III型模型推算的洪水重现水平做了对比分析。获得以下结论:大拜站洪水序列GPD模型属于重尾分布;洪水GPD阈值的选择可以参考经验平均超过函数图,而最佳阈值应采用多种指标综合确定;三种概率分布的拟合优度结果显示,构建的大拜站洪水GPD模型精度与GEV和P-III型相比较高;GPD的参数估计方法对洪水重现水平的计算结果有较大影响。
The threshold selections
excess value sample fitness test of peaks over threshold model of the flood series were deeply approached to provide a better frequency analysis method for flood designing in this paper. As a flood POT model example with the daily discharges measured at Dabai hydrologic station located in Caojiang River basin of Guangdong Province. The threshold selections were used by empirical mean excess function chart
and the Poisson distribution and the fitness indicators of excess values of peaks above different high level were tested. And then
a contrastive analysis was made among the flood models of optimal GPDandPearson Type III distribution. The main findings were as follows: 1) The Caojiang River flood extreme value distributions were heavy tail type; 2) The empirical mean excess function chart can be used as an important selection reference of Dabai flood threshold
but the optimal threshold determination should be investigated with multiindex ; 3) The good-fitness test indicated that the GPD model was better than Pearson Type III; 4) The parameter estimate methods have higher impact on the calculation results of flood return levels.
广义Pareto分布P-III型分布超定量洪水阈值AndersonDarling检验
Generalized Pareto distributionPearson III type distributionpeak over flood thresholdsAnderson-Darling test
0
浏览量
184
下载量
1
CSCD
关联资源
相关文章
相关作者
相关机构