38.5 



14.3 13.9 



27.8 25 

 FIGURE 12 Computed wave directions as a function of period — 3 June 1961, 0215 — 0345Z 



12.2 11.6 



Since all other values appear fairly continuous in azimuth, it has been 

 assumed that some other mechanism or mode of propagation is responsible 

 for its existence. If it were due to some instrument characteristics or 

 a fault of the digitizing process, one might expect to find the peak in 

 both records; however, it appears only in the 0300 to 03^52 record. If 

 the peak is real and not extraneous, it may well have been caused by an 

 earthquake. The records show an earthquake of magnitude 5^ "to 52 occur- 

 ring at 0100 hrs, 13 min, 25.^ sec at a depth of about 29km near Kamchatka 

 (53»3°W*l6^«8'^) on this date. The Wg, second arrival (via the antipodes) 

 of seismic surface waves of the "G" type, would fall within the 0300 to 

 03^52 record interval. Such waves have periods of from 1 to if minutes, and 

 multiple arrivals are often observed (Richter, 1958). Such energy could be 

 imparted to the shorefast ice cover at many points along its periphery and 

 conceivably produce the observed response of 167-second period at the array. 



c. Beam Width 



As shown in Munk et si, (1963), the value for coherence may be 

 used as an indication of the beam width of the incoming waves. They give 

 the following equation for determining the beam width. 



Aa = (1 - R2)l/2y(Z-£ cos a ) 



(17) 



where Aa = beam width 



R = coherence 



. j/(27rD) 

 sina = <p/ —r—' 



= phase angle 



Values of A a for periods of 62.5, 38.5, and 27.8 seconds are 

 about 30°, 20°, and 15°, respectively. These beam widths seem quite 

 believable when the proximity and width of the storm are taken into account. 



25 



