From Figure 10. 1, one can see that the energy for frequencies 

 less than Z-nl^(:i is negligible and is probably due to such effects 

 as a slow drift of the recording instrument and a tilting back and 

 forth of the wave pole due to the varying pressures of the wind act- 

 ing on it. The total E value for the spectrum (E equals twice 

 the variance of the wave record, and it also equals the sum of 

 the squares of the amplitudes of the spectral cornponents) for fre- 

 quencies equal to or greater than 2it9/96 is 4.94 ft . When the 

 upper and lower confidence bounds are taken into consideration, 

 as will be explained shortly, one can conclude that the true value 

 probably lies between 5.28 ft^ and 4.59 ft^. (See also Table 10.1.) 



Since 



(10.1) E = 0.242 (-^)^ 



as given in Pierson, Neumann and James [1955], where E is in 



2 2 



ft and V is in knots, an E value of 4.94 ft implies a wind speed 



of 18.25 knots, and E value of 5.28 ft^ implies a wind of about 



18.5 knots, and an E value of 4.59 ft implies a wind speed of about 



18.0 knots. 



The Neumann spectra for 19.00. 18.5. 18.25, 18.00 and 17.5 



knots were computed in order to cover the above range, and a little 



extra, and plotted against the observed spectruin. The results are 



137 



