transmission data to coalesce into one well-defined trend. A prediction equa- 

 tion was fit to the data shown in Figure 18, and the following relation was 

 obtained: 



K t= ''\ ,0.592 ^ 



H A \ 



HO t 



1.0 + 



for 



f->..0 



mo 



Equation 12 explains about 97 percent of the variance in K for the range 

 considered. It is apparent from the composition of Equation 12 why the rela- 

 tive freeboard F/H was not a good variable for explaining wave transmis- 

 sion through relatively high breakwaters. 



33. For conditions where transmission is not dominated by wave energy 



propagating through the reef, relative freeboard F/H is the most influen- 



mo 



tial variable. Part of the value of the variable is in being able to account 



for the changing influence of wave height as the dominant mode of transmission 



shifts between wave propagation over the crest to wave runup and overtopping. 



For submerged reefs the relative freeboard correctly indicates the interesting 



property of being able to dissipate energy of large waves more effectively 



than that of small waves. For reefs being overtopped, the relative freeboard 



correctly indicates that larger waves have higher transmission coefficients. 



In spite of these assets, wave transmission for low and submerged reefs is far 



too complicated to be formulated adequately in terms of F/H alone partly 



mo 



because wave energy is still propagating through low and submerged reefs even 

 though transmission may be dominated by either overtopping or propagation over 

 the crest. In addition, energy going over the reef is quite dependent on 

 crest width and bulk of the structure which introduces the influence of other 

 variables. Considering the multitude of confusing influences and the complex- 

 ity of the phenomenon, the following regression relation was fit to the 

 167 tests with relative freeboards less than one: 



31 



