2. A sensitivity analysis is recommended on those input parameters that 

 are poorly known. For example, if there is some uncertainty in the value of 

 the design water level, predictions should be made over the range of expected 

 water levels keeping all other factors fixed. Comparison between the predic- 

 tions at different levels will indicate the importance of water level. 



3. Estimate reflection coefficients from model results. 



4. If possible, final breakwater design should be made with the use of 

 physical models. The program MADSEN can be used to assist in designing and 

 interpreting physical laboratory models and results for permeable breakwaters. 



Copies of the program decks for the program MADSEN and OVER described in 

 Appendixes F and G may be obtained from the Automatic Data Processing Coordina- 

 tor, Coastal Engineering Research Center, Fort Belvoir, Virginia 22060. 



63 



