摘要
采用包括广义极值分布、威布尔分布、广义逻辑斯特分布和皮尔逊3型分布在内的4种三参数概率分布模式推算广东省85个气象站历年日最大降水不同重现水平的分位数。主要结果如下:1)参数估计结果表明不存在普遍适用的极端降水概率分布模式;2)广东极端降水以珠江三角洲和沿海较为突出。其中,清远、汕尾、阳江为广东三大极端降水中心区;3)采用4种概率分布模式推算的极端降水重现水平差别很大。
Four probability distribution models,generalized extreme value distribution,Weibull distribution,log-logistic distribution and Pearson type-III distribution,were used to calculate the different return levels of maximum daily precipitation for 85 stations in Guangdong province.The main results are as follows: 1) The results of parameter estimation indicate that there is no universally applicable extreme precipitation probability distribution model;2) The extreme precipitation occurs more frequently in the Pearl River Delta and the coastal areas of Guangdong among which Qingyuan,Shanwei and Yangjiang are three extreme precipitation centers;3) The return levels of extreme precipitation differ greatly between the calculations of four probability distribution models.
出处
《中山大学学报(自然科学版)》
CAS
CSCD
北大核心
2012年第1期102-106,共5页
Acta Scientiarum Naturalium Universitatis Sunyatseni
基金
广东水利创新资助项目(2011370004209292)
关键词
三参数概率分布模式
极端降水
拟合优度检验
重现水平
广东
three-parameter probability distribution models
extreme precipitation
goodness-of-fit test
return levels
Guangdong