189. The point of this section is that the obseirved wave climate is 

 consistent with existing knowledge of wave behavior as defined by the conven- 

 tional parameters of characteristic height, period, and direction. It is 

 noted that bulk peak direction ^p,iDS ^^ cases of large directional spread 

 may have a large uncertainty of estimation, as mentioned above. The range of 

 observed peak directions is from 60 to -60 deg in azimuth relative to shore 

 normal. Most of the observations cluster around the shore-normal direction 

 because all observations were used in the analysis and the sampling is 

 naturally biased toward the longer waves of low- energy, nonstorm conditions. 



Directional Spread 



190. A primary objective in the present study is to characterize 

 directional spread. Two measures of bulk directional spread were introduced 

 above: (a) the arc containing the central half of total energy in the 

 integrated direction spectrum A^j^g , and (b) the spectrally weighted sum of 

 the arcs containing the central half of the energy at each frequency ASg^ . 

 The distributions of these parameters, their correlation with each other, and 

 their correlation with height, period, and direction parameters are considered 

 here . 



191. Figure 16 shows the distribution of directional spreads determined 

 from the integrated direction spectrum. The distribution is rather smooth and 

 distinctly unimodal indicating that wave -field energy (in this data set) 

 always has a finite directional spread. There are very few cases with a 

 spread of less than 20 deg and very few with a spread greater than 60 deg. 



The bulk of cases lies in the range 35 to 45 deg. 



192. What is striking about this result is that there is no indication 

 of a unidirectional wave field. Such a field would appear in Figure 16 at 



deg on the abscissa. There are none. In light of the experiments by 

 Vincent and Briggs (1989) and Kaihatu and Briggs (in preparation), mentioned 

 previously, the most probable spread from Figure 16 is nearer the high- spread 

 end of their range of cases. For purposes of guidance for model studies in 

 engineering design, this result suggests strongly that finite directional 

 spread be included as a test parameter. The range indicated in Figure 16 can 

 be considered representative of sites adjacent to broad, shallow continental 

 shelves . 



76 



