OCEAN WAVES AS A METEOROLOGICAL TOOL 
nates the effect of tides and other very long waves. The 
pressure fluctuations are converted into electrical modu- 
lations, which are transmitted through a cable to a shore 
recorder. Finally the records are subjected to a har- 
monic analysis by a special instrument [2]. The natural 
wave spectrum is therefore modified in three stages: 
(1) waves of short period are removed by hydrodynamic 
filtering; (2) waves of very long period are removed by 
the slow leak in the underwater unit; (3) the remaining 
record is broken down further by harmonic analysis. 
Hydrodynamic filtering is particularly adapted for 
extracting the low forerunners from the complex pat- 
tern of the sea surface. The response characteristics of 
this filter can be expressed by the ratio AP/AP) of the 
amplitude of the pressure fluctuation at the bottom to 
SHORT IRREGULARITIES 
FILTERED OUT BY 
HYDRODYNAMIC ACTION 
WATER PRESSURE 
ACTS ON OUTER BELLOWS 
UNDERWATER 
UNIT 
OUTER 
BELLOWS 
INNER 
BELLOWS 
AIRTIGHT. 
CASING 
a: 
z. 
A 
= H-POTENTIOMETER 
i 
THREE-LEAD CABLE 
1093 
equations (8) and (9) are in error by approximately 20 
per cent.® Ewing and Press [4] have suggested that this 
discrepancy might be related to the nonrigidity of the 
sea bottom. Folsom [5] obtains about half this dis- 
crepancy in laboratory experiments employing metal 
wave tanks, thus indicating that the discrepancy might 
also be related to other factors. 
Two methods for determining wave direction are being 
investigated. At the Department of Engineering, Uni- 
versity of California, Berkeley, J. D. Isaacs has meas- 
ured wave direction by means of a Rayleigh disk at- 
tached to the underwater unit. The orientation of this 
disk is recorded ashore by means of a Selsyn motor. In 
the case of a simple wave train the disk will align itself 
normal to the plane of orbital motion. In the case of an 
SHORE 
“RECORDER 
Fic. 4—Recording mechanism. The pressure is transmitted through an outer bellows into a second bellows inside the in- 
strument casing, so that the pressure of the air inside the two bellows always equals the pressure in the water outside. The 
air inside the bellows can pass through a slow leak into the instrument casing, and the average pressure inside the casing 
equals the average pressure in the water, that is, the hydrostatic pressure. The leak is so slow that the pressure inside the in- 
strument does not change appreciably during one wave period, yet it permits pressure equalization related to tides. The 
total displacement of the inner bellows depends on the difference in pressure between the inside and the outside of the bel- 
lows and therefore measures the deviations of pressure (amplitude AP) from the hydrostatic mean. 
that just beneath the surface. Let h designate the water 
depth, L the wave length, 7 the wave period, and g 
the acceleration due to gravity. The length Z can be 
eliminated between the equations 
AP 1 
AP, cosh (Qnh/L)’ (8) 
ei _ 21g 
Tye yi 
so that the hydrodynamic filtering effect AP/AP) can 
be computed as a function of depth and wave period. 
For instrument depths (on the bottom) of 40, 150, and 
600 ft, the pressure fluctuations for wave periods of 4, 
8, and 16 sec, respectively, are reduced to 10 per cent 
of their surface value (AP/AP, = 0.10). 
Measurements by Seiwell [13] would indicate that 
2h 
tanh Tr? (9) 
interference pattern the records are difficult to inter- 
pret. At the Scripps Institution, R. S. Arthur is investi- 
gating the determination of wave direction from a 
comparison of the phase relationships of records from 
two underwater instruments placed roughly parallel to 
shore. The development of a reliable method for measur- 
ing wave direction is an essential requirement in mak- 
ing use of wave records for tracking storms. 
Interpretation of Records. In order to overcome the 
almost prohibitive labor involved in carrying out har- 
monic analysis by numerical computations, a very 
effective instrument for frequency analysis has been 
developed at the Admiralty Research Laboratory in 
Teddington, England [2]. Figure 5 shows a set of period 
6. Seiwell uses the deep-water formula (27/7)? = 2rg/L 
rather than equation (9) (see discussion by Ewing and Press 
[4]). 
