KUTSCHALE: LOW-FREQUENCY PROPAGATION IN THE ICE-COVERED ARCTIC OCEAN 



The first question which may be raised is whether topside rough- 

 ness is related to bottomside roughness. Experience shows that, 

 speaking in broad terms, the ice is compensated at depth. Figure 22 

 shows a set of coincident topside and bottomside measurements illus- 

 trating the above point. However, it should be stated that one is 

 not really concerned with the exact nature of the correspondence if 

 a consistent rule is found relating topside roughness to propagation 

 loss. 



That such a rule exists is strongly suggested by data of the type 

 shown in Figure 23 (Diachok, 1973). On this figure, dB loss per bounce 

 appears to vary smoothly as a function of the ridge index, n, defined 

 as the ntomber of pressure ridges per km of path. Similar plots are 

 obtained at other frequencies but, of course, the dB loss per bounce 

 for a given n is strongly frequency dependent, being far smaller at 

 20 Hz on the figure. Typical values of dB loss per bounce for n = 3 

 are: 5 Hz, 0; 10 Hz, 0; 20 Hz, 0.15; 50 Hz, 0.7; 100 Hz, 1.3; 160 Hz, 

 2.5. 



Another very important fact is that n appears to vary in a regular 

 pattern geographically and seasonally. This is indicated in Figure 24. 

 Note that the greatest ice roughnesses are north of Ellesmere Island. 



Figure 24 sets the stage for describing the recent experiments 

 conducted in cooperation with O. I. Diachok of the Naval Oceanographic 

 Office. We were fortunate that Ice Island T-3 was located in some 

 of the roughest ice in the central Arctic Ocean north of Ellesmere 

 Island. By setting flight lines in various directions, we could 

 sample considerable variations of ice roughness along propagation 

 paths. I should mention that this was not the first airborne experi- 

 ment designed to relate roughness to propagation loss over long 

 ranges but, during the other two experiments in 1968 and 1970, a 



715 



