surprising. Owen's unverified roughness correction factor may explain some of 

 the difference between his estimate and Aaen's data. 



63. Another explanation is that significant scale effects could have 

 been present in Owen's 1:25-scale laboratory tests. Aaen (1977) found that 

 the scale effect depends on the magnitude of overtopping; the model overesti- 

 mates the prototype for very small amounts of overtopping. For three of 

 Aaen's storms, the relatively high crest elevation F/H caused F* to be 

 outside of the range tested by Owen. For storms 5 and 6, F# was at its 

 upper limit. In fact, the high crest height allowed very little overtopping 

 during any of the storms. Storm 4 had the most overtopping because it had the 

 highest water level and largest wave height. Still, Aaen's overtopping rate 

 of 4 x 10~3 ft^/sec/ft of breakwater is so small that it would take 30 sec to 

 fill a gallon jug along each foot of breakwater. Further information is 

 needed to ascertain the cause of the discrepancy between Aaen's data and 

 Owen's method. 



64. Fukuda, Uno, and Irie (1974) measured actual overtopping rates at a 

 seawall fronted by artificial concrete blocks. They found that Goda's curves 

 for seawalls covered with artificial blocks overpredicted their data by 

 between one and two orders-of-magnitude. Fukuda, Uno, and Irie believe this 

 drastic difference is caused by different offshore slopes. While Goda's 

 method is derived for offshore slopes of 1:10 to 1:30, Fukuda's seawall had an 

 offshore slope of 1:80. Fukuda, Uno, and Irie believe that their 1:80 slope 

 caused significantly more energy loss than Goda's offshore slopes. Since the 

 artificial- block seawall is not described in detail, these data are not 

 compared with estimates from any of the other methods. 



34 



