Abstract

Axial sound speed and depth of the Sofar channel in the north and south Atlantic have been estimated and the seasonal variations that affect Sofar propagation examined. The estimates are derived from applying averaging techniques to archived oceanographic data. The computed values appear to agree within one part per thousand with the measured data from long‐range propagation experiments. A more accurate sound speed mapping method is recommended, as is the incorporation of measured values of speed and depth in any Sofar computer program.

Axial sound speed and depth of the Sofar channel in the north and south Atlantic have been estimated and the seasonal variations that affect Solar propagation examined. The estimates are derived from applying averaging techniques to archived oceanographic data. The computed values appear to agree within oni: part per thousand with the measured data from long-range propagation experiments. A more accurate sound speed mapping method is recommended, as is the incorporation of measured values of speed and depth in any Solar computer program. Oceanographers and underwater acousticians have become interested in Sofar and its applications si•nceWorld War II [Ewing et al., 1946]. For example, deep 6•an currents have been measured by neutrally buoyant floats equipped with sound sources [Rossby and Webb, 1970], and Johnson [1966] has used Sofar methods to locate the epicenters of submarine earthquakes. The U.S. Navy has based virtually its entire undersea surveillance capability on the ability to locate the position of submerged submarines by sound fixing and ranging techniques. In all of these applications, in addition to the original one of accurately locating the position of downed aircraft and/or sinking ships equipped with Sofar bombs, the accuracy of Sofar fixes is of paramount importance to the investigator. Because the accuracy of the fix is primarily dependent upon the use Of correct sound speed in the Sofar channel, measurement and compilation of axial sound speed has been an ongoing program in the oceanographic community since World War II. Some of the early work done at the Woods Hole Oceanographic Institution by G. P. Woollard is still largely classified. This paper summarizes some of the more recent results in the Atlantic that have been achieved since the advent of computer oriented oceanographic data banks of the National Oceanographic Data Centd• (NODC) and data files established at the Fleet Numerical Weather Center, Monterey, and Naval Undersea Center, San Diego. Accuracy of Sofar fixes is determined in part by ability to read arrival time accurately, -knowledge of hydrophone position, and knowledge of sound speed in the deep ocean sound channel. Because the arrival times read in Sofar work are those of the peak signal level (corresponding to the arrival of the axial ray) [Urick, 1963], sound speeds at the Solar axis must be determined before a reliable Sofar fix can be obtained [Johnson, 1966, 1969; Northrop, 1972; Munk, 1974]. Institution and the Nhval Oceanographic Office [Fenner and Bucca, 1969, 1971]. Very little Sofar axis analysis has been published for the south Atlantic [Bryan et al., 1963]. METHOD Sound speeds were computed by means of Wilson's [1960] equation from hydrographic (Nansen bottle) data on file in the NODC data bank. The sound speed Co is calculated, in meters per second, in the following way [Wilson, 1960]:

No abstract