Table 6, Correlation between spectral 



-peakedness parar.eter and 



other wave parameters. 



Location 



Cases 



Correlat 



ions 



Cases 



Correlations 

 Tp vs. Qp 



Qp vs. 

 No. of peaks 



Kg vs. (fc 



(single 



-peaked cases) 



Atlantic City, N.J. 



1,037 



0.22 



0.08 



322 



0.27 



Virginia Beach, Va. 



982 



0.13 



0.02 



292 



0.20 



Nags Head, N.C. 



945 



0.31 



0.20 



286 



0.20 



Lake Worth, Fla. 



982 



0.05 



0.08 



332 



0.29 



Naples, Fla. 



866 



0.06 



0.16 



351 



0.31 



Michigan City, Ind. (buoy) 



246 



0.09 



0.27 



128 



0.34 



Michigan City, Ind. (pressure) 



127 



0.06 



0.61 



39 



0.80 



Presque Isle, Pa. (buoy) 



110 



0.10 



0.14 



54 



0.25 



Presque Isle, Pa. (pressure) 



119 



0.06 



0.36 



25 



0.43 



Huntington Beach, Calif. 



746 



0.34 



0.24 



274 



0.05 



Pt. Mugu, Calif. 



799 



0.22 



0.25 



34 



0.73 



weakly correlated with relative water depth as evidenced by a small cor- 

 relation between peak period and % for single-peaked spectra (Table 6), 

 This may also be a result of the definition of Qp. A low correlation 

 between significant height and Qp is also shown in the table. 



The spectral-peakedness parameter appears to be related to signif- 

 icant wave height and peak spectral period in a complicated way (see 

 Table 7 for Nags Head and Hiontington Beach) . High mean values of Qp 

 occur for significant wave heights and peak periods near the annual mean 

 for most locations (Table 7). Mean Qp values decrease fairly system- 

 atically for Tp increasingly higher than the annual mean Tp in most 

 cases (Table 7) . Qp values have a tendency to increase with Hg for 

 Tp shorter than the mean and decrease with Hs for Hg and Tp 

 greater than the annual means. High values of Hg are often associ- 

 ated with relatively high mean Qp for short Tp, but low mean Qp 

 for long Tp. The latter is mostly a result of the nonsinusoidal pro- 

 file for long waves in shallow water. 



On the basis of evidence in this report and Goda (1976), it seems 

 reasonable to hypothesize that high peakedness parameters are indicative 

 of strong wave grouping, but low peakedness parameters in shallow water 

 are not necessarily indicative of weak grouping, especially when the 

 relative water depth is small. 



c. Skewness and Kurtosis of Sea-Surface Elevation Distribution 

 Function . The distribution function for sea-surface elevation also pro- 

 vides useful insight on shallow-water spectral statistics. Mean values 

 of skewness and kurtosis (defined in eq. 6) for the selected locations 

 are given in Table 8. It is evident that, except at the Great Lakes 

 sites, the shallow-water distribution functions tend to have more high 

 than low extremes (high skewness) and that they tend to be more sharply 

 focused than the normalized Gaussian distribution (high kurtosis). These 

 findings are consistent with cnoidal wave profiles shown in Figure 2. 



Cumulative distribution curves for skewness and kurtosis are shown 

 in Figures 24, 25, and 26. The Atlantic coast locations have a tendency 

 for higher skewness and kurtosis values than the gulf and Pacific coast 



50 



