data shown in the respective plate. On some plates a second curve (nonre- 

 gression) has been added. The second curve has been added where the data 

 scatter suggests that for design purposes a trend more conservative than the 

 regression curve should be used. The second curves are not regression curves 

 but are curves that have been fit by eye on the basis of the judgment of the 

 principal investigators. Where both curves are present the nonregression 

 curve is the one that is recommended for use for design purposes. It should 

 be noted that various vertical scales have been used in Plates 1-10. The 

 vertical scales were chosen to help portray the observed data effectively, but 

 the scales make direct comparisons between these plates difficult. Compar- 

 isons between various configurations are made later in the text. 



14. All of the curves shown in Plates 1-10 have been fit to an equation 

 of the general form 



C F' 

 Q = Q o e 1 (2) 



where C, is a dimensionless coefficient, and Q Q is a coefficient with the 

 same units as Q (ft /sec). The coefficients have been determined either by 

 regression analysis or "fit by eye" as mentioned above. Equation 2 seems to 

 have the proper form to fit all of the data sets rather well and is the same 

 form as the overtopping equation developed by Owen (1982) in his laboratory 

 study of irregular wave overtopping of sea dikes. Coefficients Q and C.j , 

 for both regression and nonregression curves, are given in Table 1. 



15. Although the parameter F' given by Equation 1 and used as the in- 

 dependent variable for Plates 1-10 may seem a bit abstract at first, it is 

 effective in consolidating the data into well defined trends that can be 

 readily identified. Generally, there is a large change in Q in the range 



of F 1 between 0.3 and 0.5. For F' greater than 0.5 there is little wave 

 overtopping, while for F 1 less than 0.3 there is considerable overtopping 

 regardless of the seawall/revetment configuration. 



16. These large amounts of wave overtopping result from the effect of 

 large waves hitting the seawall or seawall/revetment at high water levels. 



The term high waves means those with crest elevations probably in the range of 

 70 to 80 percent of the freeboard. For these conditions it is difficult to 

 envision a strategy which would be effective. The wave just surges up at the 

 wall and inundates the recurve then spills over the crest of the seawall in 



16 



