This same situation could have existed in measurement of wave length, i.e., 
determining length by the distance between crest "A" and crest "B" rather 
than between "A" and "A", It is also possible that heights for one set of 
waves were measured and the periods and lengths for another set and the 
results of all three grouped together as characteristics of the same wave. 
As a result of the above inconsistencies, it was decided, after a talk 
with the Navy observer, to neglect the local wind wave data and to concen- 
trate solely on the comparison of observed and hindcast swell characteristics. 
The steepness ratio of the waves in the observed swell group is more 
reasonable, none of these exceeding H/E ratios of 1/7. However, it was noticed 
in this group that the observed wave lengths were as a whole much smaller than 
the calculated lengths based on the usual formula lL = 5.12T¢tanh 2M da/L, ob- 
tained from the observed period T. There appears to be an inconsistency here 
as the observed wave lengths should approximate the calculated wave lengths 
based on the observed period. This again might indicate measurements in- 
volving dissimilar wave period crests in determining the wave length as shown 
in Figure ). A curve was plotted (Figure 5) from all the data in the observed 
swell group and it was ascertained that about 60 per cent of all the recorded 
waves had a ratio of length measured to calculated of 0.5 or less, and that 
only about 5 per cent of the waves approximated a unity ratio of measured to 
calculated length. 
Similar discrepancies have been noted in past tests; for example, 
wave characteristic data taken by M. P. O'Brien* at the mouth of the Columbia 
River in the early 1930's showed that the measured wave length was approx- 
imately 75 percent of the calculated wave length. Dr. Willard J. Pierson, 
Jr.** of New York University in a recent paper expressed the belief that the 
wave Length in deep water is not correctly given by the equation L, = Bellet, 
except for a condition of the sea having but one simple wave train. For a 
"turbulent" sea consisting of various wave trains cf many period components 
from several directions, he determines that the actual "average" length should 
theoretically be about six-tenths of this value determined from this "average 
period". This might to some extent explain the predominance of measured 
wave lengths shorter than calculated wave lengths. The average value of this 
ratio for these data was 0.93. 
An analysis of observed deep water wave directions showed that, the 
greatest frequencies were from the West and South with percentages of total 
7 
* M. P. O'Brien "Wave Action and Salinity Currents at the Mouth of the 
Columbia River" presented before the Oceano. Soc. of the Pacific, Denver, 
Mun Zk USB) 
x W. J. Pierson, Jr. "An Interpretation of the Observable Properties of Sea 
Waves in Terms of the Power Spectrum of the Gaussian Record" presented at 
34th Annual Meeting of AGU at Wash., D. C., May 4-6, 1953 (to be publ.) 
