PART 10 

 ANALYSIS OF WAVE POLE DATA 



Determination of spectra and correction for wave pole motion 



Three wave pole records were taken at the times shown in Part 7. The 

 records had been read at an 0.2 second interval at Woods Hole for another pur- 

 pose, and the niimbers were made available to us by Harlow Farmer of Woods 

 Hole, Every fourth point in the sequence was used in the determination of the 

 spectrura, and the result ^was that the first series had 1,758 points, the second 

 had 1,686 points and the third had 1,764 points. Sixty points were estimated 

 for each of the spectra^ The formulas given by Tukey and described in Part 8 

 were used to compute the spectra. Since the points on the record were 0.8 

 seconds apart, and since 60 lags were used^ the result was that the AE values 

 of the spectrum for frequencies between |Ji - Z-rrfk --^)/96 and fx = 2TT(k+ -l)/96 

 would be estimated and plotted at the point |jl = 2T:(k)/96 as k ranges from 

 to 60. Frequencies above 2it/1.6 would be aliased. 



The values for the three spectra which were obtained were averaged after 

 a study of the individual values showed no statistically significant variation at 

 the 5 percent level from record to record„ Nevertheless, there may have been 

 fluctuations from record to record due to variations in the wind field which 

 would show up at a lower level of significance. The average was multiplied by 

 the correction factor derived in Part 8 in order to obtain the true spectrizm of 



the waves o 



The result is shown in figure 10.1. The solid curve is the estimate of 



134 



