SUPERSONIC TRANSMISSION 



143 



laboratory groups. Plots of these runs were analyzed 

 with respect to three factors : refraction pattern, 

 depth of the receiving hydrophone and bottom 

 character. Two otlier parameters which are also 

 significant, the water depth and wind force, were not 

 taken into account in order not to spht the .s! mple 

 into too many small divisions. 



No analy.sis is at present available concerning the 

 effect of water depth on transmission. It is known 

 that in very shallow water (5 fathoms) transmission 

 is poorer in the presence of downward refraction than 

 it is over the same type of bottom in deeper water.' 

 For the purposes of the analysis reported below, runs 

 in water of less than 10 fathoms and runs in water of 

 more than 200 fathoms have been omitted. As for 

 wind force, a separate analysis has been made at 

 UCDWR, which will be reported at the end of this 

 section. 



The following types of bottoms have been treated 

 separately: SAND, SAND-AND-MUD (including 

 SAND-MUD and MUD-SAND), MUD (including 

 only the soft muds), CLAY (the plastic muds), 

 ROCK (including ROCK and CORAL), and STONY 

 (including gravel, cobbles, and similar notations on 

 the original sheets). Sand-and-shells was treated as 

 SAND. 



The depths of the receiving hydrophone were sub- 

 divided into three classes: shallow (0 to 16 ft), inter- 

 mediate (17 to 100 ft), and deep (more than 100 ft). 

 These classes were chosen for convenience and uni- 

 formity. A division of hydrophone depths into depths 

 above and below the thermocline might have been 

 preferable from a theoretical point of view; but on 

 many bathythermograph traces the location of the 

 thermocline is not uniquely determined. Therefore, a 

 more mechanical division on the basis of hydrophone 

 depth in feet was decided on. 



The bathythermograph patterns were divided into 

 the usual classes, described in Section 5. 1.4 as NAN, 

 CHARLIE, MIKE, and PETER. All patterns were 

 classified as in deep water, that is, the classification 

 BAKER (used for most conditions in shallow water) 

 was never used. The MIKE patterns were subdivided 

 into two classes, DEEP MIKE, consisting of all pat- 

 terns in which the water was isothermal to at least 

 100 ft below the surface, and SHALLOW MIKE, in- 

 cluding all other MIKE cases."= In the case of certain 



well-reflecting bottoms, NAN and CHARLIE were 

 combined into one class. 



As a preliminary step, median and quartile R^ 

 ranges'* were determined for all combinations of the 

 three parameters considered in this analysis ; quartiles 

 were omitted wherever the number of runs was 7 or 

 less. Table 1 lists the results obtained for the 

 UCDWR runs in shallow water. Table 2 lists the 

 results obtained for the WHOI runs available at the 

 time of the analysis. 



For each class of runs, two figures are .supplied in 

 the upper right-hand corner of the box for median 

 values of Rw in order to indicate the size and extent 

 of the sample. The first of these two figures is the 

 total number of runs making up the sample. The 

 second number, which shall be called the "adjusted 

 number of days" and is separated from the total 

 number of runs by a slant line, indicates how widely 

 distributed the sample is in time. The latter figure is 

 supplied because it has been found that the acoustic 

 data obtained on a particular day and at a particular 

 location resemble each other more closely than data 

 which have been obtained on different days, even 

 though the oceanographic conditions are closely 

 similar. Instead of simply noting the number of dif- 

 ferent days on which the various runs making up the 

 sample were obtained, it was decided to give an "ad- 

 justed" number, computed as follows. If the number 

 of runs made on k different days are denoted by ni, 

 n2,- ■ • ,nk, then the adjusted number of days K is 

 defined as the expression 



K = 





(3) 



K equals the number of days k if all the rii are very 

 nearly equal; in other words, if the sample is evenly 

 distributed over the various days on which runs in 

 this classification were obtained. But if some of the 

 days furnish only one or two runs with other days 

 contributing large numbers of runs, the days with 

 very few runs will not be counted fully. To give a 



' This division of the MIKE patterns was made for this 

 analysis only. The designations DEEP MIKE and SHALLOW 

 MIKE have no official standing in Navy doctrine. 



■^ These ranges represent that range at which the trans- 

 mitted sound level is 40 db below the level at 100 yd. In some 

 cases, the level at 100 yd was ascertained by extrapolating in 

 from several hundred yards. In the case of WHOI data, Rm is 

 determined with reference to the sound level at 100 yd at the 

 depth of the hydrophone in question, and Rm is thus nothing 

 but a measure of the slope of the transmission anomaly vs 

 range; while for UCDWR data, reference is made to the sound 

 level at 100 yd at a depth of 16 ft below the sea surface. 



