PART IV: SUMMARY AND RECOMMENDATIONS 



56. An intercomparison of empirical design methods and predictive 

 relationships such as presented in Part III may be useful in lending validity 

 to certain design tools. However, it is difficult to make broad generaliza- 

 tions about the reliability of empirical methods. The inherent simplicity of 

 the empirical methods evaluated herein and the lack of a large prototype 

 database tended to result in widely varying predictions for most design 

 relationships. As with any research study, more prototype, numerical, and 

 physical model data sets would greatly increase knowledge and predictive 

 abilities . 



57. However, several design methods and relationships were shown to be 

 useful predictors, if their limitations are realized throughout the design 

 process. Suh and Dalrymple's (1987) relationship predicting salient length in 

 the on- offshore direction tended to overpredict the salient excursion 



(R = 0.67), but was a very good predictor for pocket-beach type projects with 

 tombolo formation. Dividing Harris and Herbich's (1986) volume relationship 

 by Nir's (1982) salient/tombolo thickness relationship gave a prediction of 

 the salient/tombolo area. The correlation between measured (assumed to have a 

 triangular shape) and predicted area was weak (R = 0.18), but a trend was 

 observed. Seiji, Uda, and Tanaka's limit for "no gap erosion" was shown to be 

 a fairly good predictor. Berenguer and Enriquez ' (1988) and Ahren's rela- 

 tionships to predict gap erosion for pocket beaches, developed from pocket 

 beach data in Spain, were poor indicators of actual gap erosion for US 

 headland projects (J? = -0.99, R = -0.24, respectively). Hallermeier 's (1983) 

 recommended depth for breakwater projects was shown to correlate well to 

 actual project depth for all but one project (overall R = 0.55). 



58. The success of the Lakeview Park project lends validity to the 

 method used to design it, the Diffraction Energy Method. A comparison of 

 Lakeview Park as -constructed project parameters to those recommended by the 

 JMC method and Toyoshima's median- depth system was conducted. Both the latter 

 two methods resulted in smaller segment lengths and gap distances than the as- 

 constructed project. The structures were also positioned closer to shore than 

 indicated by the Diffraction Energy Method. 



59. As with any design procedure, limitations of any empirical method 



38 



