摘要
选取2013年1月至2018年4月的青岛潮高数据,建立广义极值分布(GEV)模型分析青岛潮高月最大值序列,采用极大似然估计法估计模型中的参数,利用轮廓似然估计确定精确的参数置信区间,推断出未来10个月可能出现的涨潮最大高度.通过建立青岛潮高的Gumbel模型,与GEV模型作出比较,进一步说明GEV模型的优越性.
Based on the Qingdao tide height data from January 2013 to April 2018, a generalized extreme value distribution(GEV) model is established to analyze the monthly maximum value series of Qingdao tidal height. The maximum likelihood estimation method is used to estimate the parameters of the model. The accurate parameter confidence interval is determined by using the contour likelihood estimation and the maximum height of the high tide that may occur in the next 10 months is inferred. The superiority of the GEV model is further illustrated by establishing the Gumbel model of Qingdao tidal height and comparing with the GEV model.
引文
[1] 唐少华,暴景春.用统计分析方法优化潮高表示模型[J].海洋测绘,2010(2):14-19.
[2] 刘灿权.利用Microsort Excel求任意时刻潮高和任意潮高的潮时[J].舰船电子工程,2012,32(1):72-73.
[3] 曹兵,邬惠明.警戒潮位核定规范的几个问题探讨[J].海洋预报,2016,33(5):70-77.
[4] Coles S.An Introduction to Statistical Modeling of Extreme Values[M].New York:Springer,2001.
[5] 史道济.实用极值统计方法[M].天津:天津科学技术出版社,2006.
[6] Eliane C,Pinheiro,Silvia L,Ferrari P.A comparative review of generalizations of the Gumbel extreme value distribution with an application to wind speed data[J].Journal of Statistical Computation and Simulation,2016,86(11):2241-2261.