Reference 5 show remarkable spatial and temporal persistence of the channel 

 in the mid- Atlantic. In a survey lasting twelve days and extending over 

 some 1000 kilometers, the velocity minimum changed by less than 2 meters per 

 second and remained at practically constant depth. At all the measurement 

 stations the velocity below the minimum increased at the pressure rate. 

 However, a different study (Reference 6) carried out south of Bermuda 

 showed considerable fine structure in the sound channel region there, 

 with sound velocity changes of about one meter per second in a few hours at 

 the same depth, raising the question whether a single sound velocity 

 profile is sufficient to describe the deep channel in a large area for any 

 length of time. Probably extreme temporal variations in the deep sound 

 channel are due to fronts and may be of a local and predictable nature. 



More data on SOFAR studies in the South Atlantic are given in 

 Reference 7. Reference 8 discusses long range sound propagation in the 

 Southern Ocean or Antarctic Ocean (Project Neptune). The Antarctic 

 Ocean appears to have a different velocity structure from the Atlantic or 

 Pacific. The depth of the SOFAR channel decreases rapidly between 40°S 

 and 60°S and no true SOFAR channel exists further south. 



To make use of this propagation there is a requirement for 

 absolute rather than relative accuracy since for accurate location an 

 exact value of V is required instead of a knowledge of the velocity profile 

 which suffices for predicting the ray paths in the channel. Possible errors 

 in SOFAR location are similar to errors in location by LORAN due to timing 

 errors. These are discussed in detail in Reference 9. The location of the 

 source is determined from the intersections of hyperbolae which describe 

 positions having fixed differences in distance from the listening stations. 



10 



arthur ZDAittleJnr. 



