Estimation of Storm Peak and Intrastorm Directional–Seasonal Design Conditions in the North Sea

Abstract

Specification of realistic environmental design conditions for marine structures is of fun damental importance to their reliability over time. Design conditions for extreme waves and storm severities are typically estimated by extreme value analysis of time series of measured or hindcast significant wave height, Hs. This analysis is complicated by two effects. First, Hs exhibits temporal dependence. Second, the characteristics of Hf are nonstationary with respect to multiple covariates, particularly wave direction, and sea son. We develop directional-seasonal design values for storm peak significant wave height (Hf) by estimation of, and simulation under a nonstationary extreme value model for Hf. Design values for significant wave height (Hs) are estimated by simulating storm trajectories of Hs consistent with the simulated storm peak events. Design distributions for individual maximum wave height (H^fi are estimated by marginalization using the known conditional distribution for Hm ax given Hs. Particular attention is paid to the assess ment of model bias and quantification of model parameter and design value uncertainty using bootstrap resampling. We also outline existing work on extension to estimation of maximum crest elevation and total extreme water level.

Spatial and Seasonal Variability of Metocean Design Criteria in the Southern South China Sea From Covariate Extreme Value Analysis

“…The CEVA method has been described in detail from a methodology perspective by [18], and from a user perspective by [1]. For this reason, we provide just an overview of the methodology in this section.…”

“…Recently, new deductive methods for data extrapolation were developed. These rely on non-stationary extreme value analysis, hereafter referred to as CEVA ("Covariate Extreme Value Analysis") and are described in ( [1], [18], [19], [20]). Briefly, CEVA performs a non-stationary directional-seasonal analysis of storm peak significant wave height using a penalized maximum likelihood approach.…”

Temporal evolution of the extreme excursions of multivariate $k$th order Markov processes with application to oceanographic data

“…We remark that their work cannot be considered to model the extremes of bivariate Markov processes since the associated directional covariate does not take on extreme values. Feld et al (2015) use a sophisticated covariate model for the most extreme observation (the most extreme value of the dominant variable) in an excursion, combined with a historical matching approach for the intra-excursion trajectory; in Section 3.4 we adopt a version of this methodology as a benchmark for our case study. Finally, we mention well-established literature on multivariate time series, e.g., Tiao and Tsay (1989), which is not directly applicable to modelling environmental extremes because such models are only designed to model typical behaviours.…”

Waves and associated currents–experiences from 5 years metocean measurements in the northern North Sea

“…The semi-parametric conditional extremes model introduced by Heffernan and Tawn [11] has been strongly recommended for estimation of joint distributions of metocean parameters through a series of studies where this model has been adopted and applied in different ocean basins, including the northern North Sea [12][13][14][15][16]. Bivariate modelling of different combinations of environmental parameters have been performed based on a wide range of bivariate parametric probability distributions, see for instance [17][18][19][20][21][22].…”