This paper employs the Peak Over Threshold/Generalized Pareto Distribution (POT/GPD) method to assess the extreme value of significant wave height in the South China Sea, based on the tropical cyclone wave hindcast data from 1972 to 2011. The POT/GPD method is compared with the Annual Maxima/Pearson-III (AM/P-III) method, which is recommended by the Code of Hydrology for Sea Harbour in China. Comparison results show that, these two methods have merits and limitations associated with their sample methods. The AM method is a special form of the POT method and the POT method is the more nature sample method.
Reasonable estimation of extreme wave heights and associated return periods is crucial for the offshore and coastal engineering. It affects the design, construction, installation and operation activities of structures in the nearshore or offshore zone (López-Ruiz et al., 2016). In order to assess the extreme wave height of specified return period, the suitable sample method and the appropriate probability distribution based on the accurate wave data are needed.
In this paper, the study area is the South China Sea, which is affected by the tropical cyclone frequently. The total number of tropical cyclone in this sea area from 1972 to 2011 is 942 and the mean number of tropical cyclone of this period is 23.55. Therefore, the tropical cyclone wave hindcast data are suitable for the extreme wave height assessment in the South China Sea.
In the field of ocean and costal engineering, a Peak Over Threshold (POT) method has been commonly employed to extract independent peak storm wave heights from time-series wave data for the extreme wave analysis (Goda, 1988; Davison and Smith, 1990; Mathiesen et al., 1994; Caires and Sterl, 2005). Then, the Generalized Pareto Distribution (GPD) model (Coles, 2001) is applied to fit the sample obtained by the POT method. Considering the superiority of POT/GPD method, it is used to assess extreme wave heights of different return periods in the South China Sea, which is recommended (Hawkes et al., 2008) and widely used in the assessment of extreme wave heights in the nearshore and offshore zones (Mendez et al., 2006; Tancredi et al., 2006; Thompson et al., 2009; Mazas and Hamm, 2011; Solari and Losada, 2012).
