ESTIMATION OF THE SEVEREST SEA 



As mentioned earlier, the sea severity is most commonly expressed in 

 terms of the significant wave height. For estimation of the severest sea 

 condition, it is necessary to prepare statistical information on signifi- 

 cant wave height in the are where the mooring system is located. The 

 severest sea condition (the extreme significant wave height) expected to 

 occur during a specified period of time can be estimated by applying order 

 statistics if the probability function which governs the significant wave 

 height is know precisely. Unfortunately, the probability function which 

 is uniquely applicable to significant wave height has not yet been found. 

 Figure 2 shows an example which indicates that neither the Weibull nor 

 the log-normal distribution represents the data obtained in the North 

 Sea. Hence, the extreme significant wave height may best be evaluated 

 by applying the concept of asymptotic distribution of extreme values 

 which is applicable for any probability distribution. 



Let us assume that the cumulative distribution function of significant 

 wave height can be expressed in the following form: 



R^) = 1 - e (1) 



where, q(x) is a monotomically increasing function. Then, it can be 

 proved that the probable extreme value expected to occur in n-observations, 

 denoted by y , is given for large n as, 



-'" ^ (2) 



3. = \ (^- ^) 



The recent method is to express q(x) as a combination of an exponen- 

 tial and power of the significant wave height, such as. 



The constants involved in the function q(x) are determined numerically 

 by applying a nonlinear least squares fitting method for representing the 

 data of significant wave height by the formulation q(x) (Ochi and Whalen, 

 1980) . The cumulative distribution function computed by using the method 



95 



