PART V: ANALYSIS 



28. Both the maximum overtopping rate and the total overtopped volume 

 are important in evaluating damages associated with storms. Determining the 

 rate (a function of wave height and swl) and the volume (a function of the 

 overtopping rate through time) is a necessary exercise for evaluating poten- 

 tial damages and developing a flood routing plan. The data base collected 

 during the physical model study represents overtopping rates for a relatively 

 narrow set of conditions. Therefore, to evaluate overtopping associated with 

 various design situations, two methods which extend the physical model data 

 base were used: the Storm Time-History Method (STHM) and the Relative 

 Freeboard Method (RFM) . 



Storm Time-History Method 



29. The STHM groups the physical model study data into a series of 

 smaller data sets which are then used to predict overtopping rates throughout 

 the storm as surge and wave conditions evolve. Overtopping tests (presented 

 in Appendix A) were separated into subsets based on storm type (hurricane or 

 extratropical) and by percent of wave height (or gain) produced at the wave 

 paddle in 10 percent increments, essentially separating the storm wave heights 

 by a percent of the fully developed storm conditions. 



30. Subsets, which range in number of samples or overtopping values 

 between 8 and 14, are presented in Appendix B. The subset for the hurricane 

 model test with 100 percent gain is shown in Table 4. The swl's are in feet 

 above NGVD, and the overtopping rates are in cubic feet per second per linear 

 foot of seawall. This case will be further developed as an example of the 

 calculations. Only final results from the other percent gains for the extra- 

 tropical storms and hurricanes will be presented in the main text. 



31. Linear regression techniques were employed to produce graphs which 

 relate the overtopping rate to swl for each 10 percent increment of the gain. 

 Regression estimates for overtopping rate were made, and curves were plotted 

 for each storm type and each 10 percent increment of gain. Figure 7 shows the 

 regression curve for the data points in Table 4. The solid line (noted by Q ) 

 represents the regression estimates with the dashed lines being upper and 

 lower (Q and Q , respectively) limits for symmetric prediction intervals at a 



20 



