Final published version
Research output: Contribution to Journal/Magazine › Journal article › peer-review
Research output: Contribution to Journal/Magazine › Journal article › peer-review
}
TY - JOUR
T1 - On the spatial dependence of extreme ocean storm seas
AU - Ross, Emma
AU - Kereszturi, Monika
AU - van Nee, Mirrelijn
AU - Randell, David
AU - Jonathan, Philip
PY - 2017/11/15
Y1 - 2017/11/15
N2 - Contemporaneous occurrences of extreme seas at multiple locations in a neighbourhood can cause greater structural reliability and human safety concerns than extremes at a single location. Understanding spatial dependence of extreme seas is important therefore in metocean design, yet has received little rigorous attention in the offshore engineering literature. We characterise the spatial dependence of storm peak significant wave height using three models motivated by max-stable processes for locations in the northern North Sea. Models for marginal extremes per location, and dependence of extremes between locations, are estimated using Bayesian inference with composite spatial likelihoods. We show that, in addition to marginal directional non-stationarity of extreme seas per location, all three models indicate spatial anisotropy in extremal dependence quantified by the spatial covariance matrix of the corresponding max-stable process. Estimates suggest that extreme seas show greater extremal dependence from West to East than from North to South.
AB - Contemporaneous occurrences of extreme seas at multiple locations in a neighbourhood can cause greater structural reliability and human safety concerns than extremes at a single location. Understanding spatial dependence of extreme seas is important therefore in metocean design, yet has received little rigorous attention in the offshore engineering literature. We characterise the spatial dependence of storm peak significant wave height using three models motivated by max-stable processes for locations in the northern North Sea. Models for marginal extremes per location, and dependence of extremes between locations, are estimated using Bayesian inference with composite spatial likelihoods. We show that, in addition to marginal directional non-stationarity of extreme seas per location, all three models indicate spatial anisotropy in extremal dependence quantified by the spatial covariance matrix of the corresponding max-stable process. Estimates suggest that extreme seas show greater extremal dependence from West to East than from North to South.
KW - Extreme
KW - Spatial
KW - Dependence
KW - Max-stable process
KW - Composite likelihood
KW - Pooling
KW - North Sea
U2 - 10.1016/j.oceaneng.2017.08.051
DO - 10.1016/j.oceaneng.2017.08.051
M3 - Journal article
VL - 145
SP - 359
EP - 372
JO - Ocean Engineering
JF - Ocean Engineering
SN - 0029-8018
IS - Supplement C
ER -