network area is at a depth below 400 meters. 17 In 

 the Gulf of Alaska during winter the sound-channel 

 axis is at a depth of around 100 meters, but the 

 channel is weak. On the contrary, though compar- 

 atively shallow, the channel of the deep Bering Sea 

 is strong. The angle of the limiting ray (defined as 

 the angle of departure at the axis of the ray which 

 becomes horizontal at the top of the channel — see 

 reference 17 for detailed definitions) is between 8° 

 and 10°, which is comparable to the region between 

 35° and 45° N in the northeast Pacific. The width 

 of the channel (vertical distance between top and 

 bottom) is generally less than 900 meters. 



A similar sound channel occurring at shallow 

 depths has been reported in Canadian Atlantic 

 waters of the Bay of Fundy, Gulf of St. Lawrence 

 and the Scotian Shelf during the summer and autumn 

 months. 18 The depth of the axis in those waters is 

 50 to 75 meters, but the channel is generally bottom- 

 limited because of the shallow depths occurring in 

 that region. The Bering Sea, however, is much deeper 

 than most of the aforementioned Atlantic region and 

 the channel caused by refraction of sound is not 

 bottom-limited. The bottom of the channel is as 

 shallow as 600 meters at a few stations and, by 

 extrapolation, not deeper than 1200 meters at the 

 remainder, Recent unpublished observations in the 

 Gulf of Alaska taken on the Northern Holiday 

 cruise 1 " indicate a strong channel in that region with 

 the axis around 75 meters, and one also occurs in 

 the far northwest Pacific off the Okhotsk Sea and 

 the Kamchatkan Peninsula. Because of their shallow 

 depth, such sound channels are favorable for ship- 



17 E. R. Anderson, Preliminary Study of the Deep Sound 

 Channel in the Area Covered by the Eastern North Pacific 

 SOFAR Network, Naval Research Laboratory, U. S. Navy Journal 

 of Underwater Acoustics, vol. 1, no. 1 (RESTRICTED), January 

 1951, pp. 75-86. 



18 W. B. Bailey er a/., Sound Channels in Canadian Atlantic 

 Waters (RESTRICTED), Canada, Atlantic Oceanographic Group, 

 13 November 1950. 



!9 W. S. Wooster, Operation NORTHERN HOLIDAY, August- 

 September 1951; A Preliminary Report, Scripps Institution of 

 Oceanography, Reference no. 51-46, 15 November 1951. 



to-ship transmissions at long ranges. 20 Since a sub- 

 marine can dive to depths well within the upper part 

 of the sound channel, transmission between subma- 

 rines might be accomplished at even greater ranges. 



The sound conditions in the deep Bering Sea 

 would be particularly adaptable to a Sofar network. 

 Cables to hydrophones would be short because the 

 channel is shallow, the bottom slopes at the Aleutian 

 Islands are steep, and protected bays are available 

 for cable-landing sites. Stations could be located 

 along the Aleutian Chain, and it is possible that a 

 station could be located at the Pribilof Islands which 

 are relatively near the edge of the continental shelf. 

 The axis of the sound channel appears to rise across 

 the edge of the shelf although it approaches very 

 near the bottom, so that it is possible that reception 

 would be satisfactory at such a location. More de- 

 tailed oceanographic observations and actual acous- 

 tic tests would be required in order to verify this. 



Because of the lack of traffic in the region, 

 operational use of such a Sofar network would not 

 be practical except in case of war. However, the 

 region might profitably be used in peacetime for 

 research and experimentation on the propagation of 

 Sofar signals. The horizontal gradient in sound ve- 

 locity at the axis, of the order of 15 feet per second 

 per 100 nautical miles, offers an opportunity to study 

 horizontal refraction effects. Much interest lies also 

 in the effect of seamounts and ridges upon the prop- 

 agation of the signal. Northwest of Adak there is 

 a continuous ridge of some 200 miles long of varying 

 minimum depth around 400 meters over which such 

 tests as desired could be made. The fairly regular 

 continental slope and flat Bering Sea shelf provide 

 an excellent location for examining the change of 

 the sound channel and sound propagation entering 

 from deep into shallow water. For such investigations 

 the Bering Sea — though remote — should be consid- 

 ered as a test area because of the favorable sound 

 conditions and because Navy bases from which ships 

 could operate do exist in the region. 



20 K. V. Mackenzie, Long-Range Sound Transmission in the 

 Deep Bering Sea (CONFIDENTIAL), NEL Report 280 (in press). 

 * Computed according to reference 11. 



