36 



was probably due to the small range of the variable (14.0 to 19.5) and the 

 relatively short distance that the slope was used in the wave flume. 



This simplified model was used by CERC's Research Division for the 

 bore runup overtopping module. 



Task C 



Data from Task C (excluding C+, Tests 39 through 44) were analyzed 

 to determine a regression model for a broken wave overtopping module. 

 Input conditions for the regression analysis (in prototype scale) are given 

 in Table 10. 



Overtopping rate was nondimensionalized in the same manner in 

 Task B, that is, as Q' = Ql{g*f^)^'^- Other variables that were determined 

 to be significant in the regression analysis were: 



P/1 = swl// 



PI2 = Hlf 



PB = LIf 



0-' 



where L^ is deepwater wavelength defined as 



L^ = (g/(2ji)*7'2 



The model that gave the best results was weighted by wave height and 

 is given as 



Q' = 0.004162 - 0.007285*/'/! + 0.003252*^/1^ 



+ 0.001559*/'/22 - 0.000025997*P/3 + 0.0000002 17*P/32 



As in Task B, this model was selected based on the sum of squares of re- 

 siduals of the dimensional overtopping rates. Correlation coefficient for 

 the nondimensional model was 0.9865 {R^ = 0.9732). Sum of squares of 

 residuals for the dimensional overtopping was 0.0511 for 35 test runs, 

 yielding an average error of ± 0.038 cfs/ft. 



It should again be emphasized that the regression analysis should not 

 be used beyond the limits of the data set or for any other sites. Table 1 1 

 lists the ranges of variables used in the analysis. 



1978 Profile 



All tests conducted using the 1978 survey of Profile 2 were combmed 

 in a single data set for analysis. The data included Tasks A and C plus the 



Chapter 3 Research Tasks A, B, and C 



