THE VELOCITY OF SOUND 

 IN SEA WATER AT ZERO DEPTH 



INTRODUCTION 



An accurate knowledge of the velocity of sound in sea water as a function of tempera- 

 ture, chemical composition (salinity and dissolved gas content), and pressure (depth) is 

 essential to sonar-not only for ranging determinations but for calculation of possible 

 refraction which may operate either to channel the sound energy or to send this energy 

 into the ocean depths. 



Available tables for the calculation of soimd velocity in the ocean are based either on 

 theoretical computations or radio -acoustic ranging (1-5). These latter field measure- 

 ments, hoTX'ever, assume the existence of homogeneous water between two distant points. 

 The theoretical tables use compressibility and density determinations made over forty 

 years ago and assume the temperature dependence of specific heat for distilled water to 

 be the same as for sea water. 



Kuwahara's tables (3), calculated from empirical formulas for the compressibility 

 of sea water, are as accurate as any, and are generally used in this country. These 

 tables have been graphically presented (6), and used by the Naval Research Laboratory. 

 A comparison of velocity values obtained from various tables is made for "normal" sea 

 water of 35 parts per thousand (ppt or °/oo ) salinity in Table 1.* 



BACKGROUND 



Prelihiinary measurements of sound velocity in sea water with a less accurate 

 instrument (7) (i.e., ±1 m/sec) indicated that Kuwahara's tables yield velocities that are 

 too low by 3-4 m/sec. The Revised British Admirality Tables (4) yield velocities slightly 

 lower than Kuwahara's. Kuwahara claims an accuracy of 3 m/sec for his tables, but the 

 preliminary evidence seemed to indicate that his error is systematic and that his values 

 are consistently low. More information was desired to check the need for revision of 

 sound velocity tables. 



The ultrasonic interferometers in present use at the Naval Research Laboratory, 

 capable of measuring the velocity of sound in liquids to a precision of 1 part in 10,000, 

 were chosen as ideal instruments for checking the existing tables. 



EXPERIMENTAL PROCEDURE 



Samples of sea water were obtained from the Bermuda-Xey West area by various 

 field parties from the Sound Division, NRL. Of the samples received, three were selected 



Tables and figures are grouped after the tex;t. 



1 



