基于动态极值理论和Copula函数的极端海平面高度预测建模  

Prediction modeling of extreme sea level based on dynamic extreme value theory and Copula function

作　　者：汪杨骏 侯太平[1] 张韧[1,2] 钱龙霞[1] 王锋 龙强 

机构地区：[1]解放军理工大学气象海洋学院,江苏南京211101 [2]南京信息工程大学气象灾害预报预警与评估协同创新中心,江苏南京210044 [3]河北省唐山市曹妃甸工业区气象局,河北唐山063000

出　　处：《海洋工程》2016年第4期62-70,共9页The Ocean Engineering

基　　金：气象水文预先研究项目(407010602);唐山市曹妃甸工业区专项(CQZ-2014001)

摘　　要：全球气候变化背景下,海平面上升是一个潜在的重大风险,为防范气候灾害,应对极端气象海洋事件,需客观、定量地对未来极端海平面变化进行科学预测。为此,基于Copula函数和动态极值分析理论,综合考虑平均海平面变化(包括垂直陆地运动和基准的局地变化)与潮、涌、浪等其他气候变化的增水对极端海平面高度的影响,采用DREAM方法改进Bayes推断对动态极值模型的参数空间估计问题,提出一种新的模型对未来极端海平面高度变化进行预测,旨在改进传统模型存在的不确定性问题,并运用该模型对气候变化背景下厦门地区未来35年的海平面变化情景进行了模型应用和实验模拟。In the context of global climate change,sea level rise is a potentially major risk. In order to prevent the climate disasters,and to deal with extreme weather events,it is necessary to objectively and quantitatively predict the future extreme sea level changes.To this end,based on the theory of Copula function and that the dynamic extreme value analysis,the authers of this paper have made comprehensive consideration of the influence of the mean sea level changes（ including the vertical land movement and the baseline）and the changes of the water level of the tide,surge and wave,and used Bayes method to improve the parameter estimation of DREAM.A new method is proposed to predict the future extreme sea level changes,and the proposed method is used to improve the existing problem of the traditional method,and a model is used to simulate the sea level changes in Xiamen area in 35 years.

关 键 词：极端海平面高度 动态极值模型 DREAM方法 BAYES推断 COPULA函数 

分 类 号：P731.23[天文地球—海洋科学]