A new model of peaks over threshold for multivariate extremes

Autor:	Yao Luo, Liang-sheng Zhu
Rok vydání:	2014
Předmět:	Multivariate statistics South china Renewable Energy Sustainability and the Environment Mechanical Engineering Parametric model Statistics Univariate Natural distribution Ocean Engineering Oceanography Extreme value theory Sensitivity analyses Mathematics
Zdroj:	China Ocean Engineering. 28:761-776
ISSN:	2191-8945 0890-5487
DOI:	10.1007/s13344-014-0059-7
Popis:	The peaks over threshold (POT) methods are used for the univariate and multivariate extreme value analyses of the wave and wind records collected from a hydrometric station in the South China Sea. A new multivariate POT method: Multivariate GPD (MGPD) model is proposed, which can be built easily according to developed parametric models and is a natural distribution of multivariate POT methods. A joint threshold selection approach is used in the MGPD model well. Finally, sensitivity analyses are carried out to calculate the return values of the base shear, and two declustering schemes are compared in this study.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::287242d5b86cdb788597f5005be4f197 https://doi.org/10.1007/s13344-014-0059-7 Zobrazit plný text záznamu Full text from SpringerLink