always the maximum in the shoaling transformation. Because a steep decay in 

 wave height occurred between Gages 4 and 5 for tests involving a bar, the 

 height at Gage 4 was used as the breaking wave height in analysis of both the 

 barred and the plane -sloped cases. 



161. Statistical wave heights derived from data from Gage 4 were com- 

 pared with Equation 34, the breaker height expression for irregular waves 

 (Goda 1975) . Goda assumed the breaker height would vary and gave a range to 

 the coefficient A to predict this variation. In the present study, it 

 seemed reasonable to compare the higher value of A with the maximum wave 

 height measured at Gage 4. Figure 69 shows the predicted wave height given by 

 Equation 34, in which the coefficient A was the maximum specified by Goda 

 plotted versus H,,,^ normalized by (H s ) . The line of perfect prediction is 

 also plotted in Figure 69. The Goda equation predicts the lower measured 

 values well, but for tests with steep-faced bars, resulting in higher 



£ — values , the equation gives an overprediction. 



162. Because Equation 34 (Goda 1975) predicted H,^ well using the 

 upper limit of A , the lower limit of A was used to estimate a statisti- 

 cally averaged wave height. Goda arbitrarily selected the lower limit to be 

 two-thirds of the upper limit; therefore, significant wave height was used 

 since it is the average of the highest one- third wave heights. Predicted 

 values of (fib)s using A = 0.12 in Equation 34 were plotted against the 

 significant wave height normalized by (H s ) at Gage 4 in Figure 70. The 

 predicted values of (flb)s agree well with the measurements for 



fa < 10 deg , but are too high for fa = 15 deg. 



163. With the reasonable success in associating A-values with H,,,^ 

 and H s , a procedure was developed to determine the A-value required to 

 predict H rms /(H S ) . Because the Rayleigh distribution gives H s = Jl H rms , 

 the minimum value of the empirical coefficient recommended by Goda (1975) was 

 reduced by Jl, which yielded A = 0.085 . Predictions with this value of A 

 gave reasonable results for all seaward angles, but it underestimated the 

 measured values for fa < 5 deg. Therefore, the coefficient was rounded to 



A = 0.09 . The predicted versus measured values using A = 0.09 are shown in 

 Figure 71. 



164. Equation 34 predicts the measured wave heights well for the plane - 

 slope cases, and for the barred slope cases for fa < 10 deg. The equation 

 overpredicted wave heights in the tests with the steeper bar angle 



129 



