Application of Taylor diagram in the evaluation of joint environmental distributions' performances

• Published in: Marine systems & ocean technology : journal of SOBENA--Sociedade Brasileira de Engenharia Naval 2020-09, Vol.15 (3), p.151-159
• Main Authors: Simão, Marina Leivas, Videiro, Paulo Mauricio, Silva, Plínio Bueno Andrade, de Freitas Assad, Luiz Paulo, Sagrilo, Luis Volnei Sudati
• Format: Article
• Language: English
• Subjects: Distribution; Distribution functions; Engineering; Environmental assessment; Environmental factors; Evaluation; Extreme values; Fittings; Offshore Engineering; Offshore structures; Original Paper; Parameters; Peak periods; Probability theory; Significant waves; Spectra; Statistical analysis; Statistical models; Wave data; Wave height
• ISSN: 1679-396X; 2199-4749
• DOI: 10.1007/s40868-020-00081-5

Probability-based design of offshore structures usually require a joint distribution of significant wave heights and zero-up-crossing periods, or spectral peak periods, in order to describe the long-term wave scenario, and thus predict long-term extreme responses. Several joint statistical models have been proposed in the literature, often comprising environmental parameters aside from significant wave heights and zero-up-crossing or spectral peak periods, such as wind and current velocities, among others. However, since there is no theoretical environmental distribution function, the assessment of different models is done on an empirical basis for each available data set. The fitting of a joint distribution model on a set of long-term wave data, given the extensive number of available joint statistical models and the inherent extrapolations required for extreme conditions, can be extremely cumbersome. This paper proposes the application of Taylor diagram in the assessment of the fit of joint models. The aim is to find the models with the best performances through an easy and straightforward visual comparison. For simplification purposes, the joint statistical model considered comprises significant wave heights and spectral peak periods only. Results show that Taylor diagram is a useful tool for primarily evaluations of adjusted probability models. The indicatives for better model performances identified in the diagram were shown to correspond to other common identifiers, which demonstrates that the diagram can be a reliable indicator for good models. Some study cases are presented to demonstrate Taylor diagram application in joint probability distributions fittings for wave environmental parameters.