do not require historical information for their estimation. They can be based 

 on a current engineering and property valuation assessment of the facilities 

 to be protected. Historical information relating specific levels of economic 

 loss (in dollars) to the measured or hindcast wave height of the associated 

 storm can then be used to derive a function of the form, as follows: 



$L( 



r A(H -H, )'l 



H ) = $L 1 - e ^ ^° 

 s maxL J 



(31) 



where A is a coefficient determined by regression. This function, illus- 

 trated in Figure 16, can then be used to estimate the expected annual economic 

 losses, or user costs without project, according to Equations 16 and 17 by 



SI ■./.".(¥}«. 



(32) 



where F(H ) is the cumulative probability distribution of significant wave 

 heights derived in Step 1. A joint distribution F(Hg , T ) should be ap- 

 plied where operations and facilities are particularly sensitive to a certain 

 range of periods. Software has been developed to estimate both $L(Hg) and 



E{$L/yr} given an estimate of $L 



H 



Lo 



, Extremal Type I F(H ) coeffi- 



cients, and at least one historical data point [h , $L(H )) . The program 



$U 



MAXIMUM CONCEIVABLE LOSSES FROM WAVE ATTACK 



■$L(H^)=$L^^^[l-eA'^s-^Lo>] 



A HISTORICAL Hj AND $L (Hj) VALUES 

 • H. (WHERE $L (HI = 0.0) 



2 3 4 5 6 7 



SIGNIFICANT WAVE HEIGHT, H M 



Figure 16. Economic loss function versus incident wave height 



59 



