relatively steep, making the breaker ratio correspondingly high. Beach 

 profiles from the FRJ data set showed more gentle slopes than the CRIEPI 

 experiment, causing the predicted breaker ratio to become lower, and the waves 

 to break farther offshore. As a result, the relationship derived from the LWT 

 data produced a slope dependence which appears not to apply to the more gently 

 sloping bars found in the field. 



496. The energy-based significant wave height was used in the numerical 

 model to determine the wave height distribution across shore. On a field 

 beach, the break point constantly moves back and forth due to random variation 

 in wave parameters . A problem is to find a measure of the wave height that 

 will on the average reproduce properties of the random breaking waves. As an 

 alternative to the significant wave height, the mean wave height H , deter- 

 mined by assuming a Rayleigh distribution, was used in some simulations. 

 Since H is smaller than H^^ , the waves broke farther inshore but moved 

 less sand. However, better agreement was not achieved using H , in contrast 

 to what was reported by Mimura, Otsuka, and Watanabe (1987) based on their 

 small tank experiments. 



497. The nonlinear shoaling law derived by Shuto (1974) was also tested 

 in some field data simulations. It seemed to overestimate shoaling just 

 before breaking, as was the case for the LWT experiments. Longer period waves 

 calculated by the nonlinear theory markedly increased in height in shallow 

 water, creating a bar too far offshore. Consequently, linear wave theory was 

 judged to be more satisfactory and was used throughout. 



498. Median grain size probably varied across the beach profile (see 

 Figure 77) with a notably larger grain size on the foreshore. To represent 

 this variation in the model, two different grain sizes were used along the 

 profile. A larger grain size (2.0 mm) was specified on the foreshore to a 

 distance approximately 130 m from the baseline, and a finer grain size 

 (0.15 mm) was employed from this point and seaward. The larger grain size 

 requires larger equilibrium energy dissipation with correspondingly more wave 

 energy needed to move material. As for the LWT simulations, the equilibrium 

 energy dissipation design curve of Moore (1982) was reduced by a factor of 

 0.75. Additional variation in median grain size across shore somewhat 

 improved the fit of the model in trial simulations but was considered to be 



208 



