/A/ = ;= -^nj2 ^^^^ 



H 



where T = 30.8 sec is a characteristic period provided by Datawell. The 

 manufacturer states that this sensitivity decrease results in amplitude errors 

 of less than 3 percent for oscillation (wave) periods less than 15 sec. Fig- 

 ures Al and A2 present curves for the DW = /A/ - 1 , the Datawell-predicted 

 sensitivity decrease error, and, as can be seen, the actual sensitivity does 

 not decrease with period according to the Datawell relationship (Equation Al) 

 given above . 



Difference Error (d) 



4. The difference error d is the difference between the Datawell- 

 predicted decrease in sensitivity error and that found from the actual buoy 

 calibrations. 



5. In Tables A2 and A3, DW and d are tabulated as a function of T 

 for each buoy. The best accuracy would be obtained by choosing the calibra- 

 tion values nearest in time to the date of the measurements. 



6. Since these error corrections are oscillation-period dependent, 

 their application requires that the wave data be decomposed into amplitude co- 

 efficients or variance-spectrum coefficients for each frequency or period. A 

 less accurate but also less complicated procedure would be to apply a single 



correction to the wave height H based on the peak-spectral wave period 



m 

 o 



T . For correction of amplitudes or derived parameters linearly related to 

 amplitude, a correction factor F(T) can be obtained from the sum of DW and 

 d by: 



^^T> = 1 + (DW + d) (^2) 



which can be applied by multiplying the uncorrected amplitude by F(T) for 

 T . For correction of parameters related to the square of the amplitude 



A2 



