Figure 20(b) shows characteristics of secondary spectral peaks at Nags 

 Head when period of the dominant peak is 12 to 13 seconds. Secondary 

 peak periods of 4 to 9 seconds are common. In most cases these secondary 

 peaks represent locally generated wave energy or swell waves generated 

 near the Nags Head coast. The average energy associated with these peaks 

 is relatively high. Secondary peaks at periods longer than 10 seconds 

 are almost totally absent. 



At Huntington Beach, spectra with relatively short Tp have a very 

 strong tendency for a secondary peak at long period (see Fig. 21, a). 

 Locally generated waves at Huntington Beach are usually superimposed on 

 long-period swell waves. Huntington Beach spectra with long Tp have 

 a small tendency for secondary peaks with short periods (see Fig. 21,b). 

 They also have a small tendency for secondary peaks at long period, which 

 presumably correspond to secondary swell waves superimposed on the domi- 

 nant swell. However, most Huntington Beach spectra with long Tp appear 

 to be dominated by the main spectral peak. 



b. Spectral Peakedness . The spectral-peakedness parameter, , 

 defined by equation (4) was computed for each spectrum. The spectral- 

 peakedness parameter is shown on each plot in Appendix A. 



Cumulative distribution functions of Qp are shown in Figure 22 

 for the selected locations. The Qp values cover a wide range at most 

 locations-from less than one to greater than five. Qp values tend to 

 be smallest for the Atlantic coast sites. The highest C^ values tend 

 to occur for the Great Lakes gages, with the pressure gages having higher 

 Qp values than the buoy gages. Qp values for the gulf and Pacific coast 

 gages tend to be intermediate to values for the Great Lakes and Atlantic 

 coast gages except for the highest 5 percent of the Pt, Mugu Qp values. 

 The cumulative distribution function for Pt. Mugu shows a rapid increase 

 in Qp values associated with probabilities of less than 5 percent. 

 This feature is a result of occasional large overcompensation of the 

 pressure spectrum which creates a high spectral peak at high frequency. 

 Several cases in which Qp was greater than nine are omitted from Fig- 

 ure 22. These cases all correspond to very low wave conditions except 

 for several overcompensated Pt. Mugu cases. 



Goda (1976) indicated that high values of Qp are associated 

 with strong groining of high waves. Pen-and-ink record traces for sev- 

 eral cases with reasonably high waves and high Qp values are shown in 

 Figure 23. Some grouping of the high waves is evident in the records. 

 Records with low Qp values in Figures 13 and 18 show little evidence 

 of wave grouping; yet the record in Figure 11 for high, long waves shows 

 evidence of high waves occurring together, but the Qp value is only 

 1.35. 



The spectral-peakedness parameter might be expected to weakly corre- 

 late with the number of major spectral peaks so that single-peaked spec- 

 tra tend to have high peakedness. The data indicate a low correlation 

 at some locations, but not at others (Table 6). Qp also appears to be 



48 



