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



determined by waves which pass over all bottom topography without 

 suffering bottom reflections. 



The next four Figures, 13 to 16, will show a comparison of normal- 

 mode theory with the field measurements. 



On Figure 13 observed group-velocity dispersion is compared with 

 dispersion computed from a multilayered model which will be described 

 later. The agreement is excellent for the first three modes. 



Figure 14 shows the close agreement between theory and experi- 

 ment of the depth variation of pressure for the first normal mode 

 at two frequencies of 20 and 40 Hz. 



Figure 15 shows a computed oscillogram for the first two modes. 

 If we compare peak intensity as a function of range for waves of the 

 first normal mode, we get the type of agreement shown on Figure 16. 

 In these computations the effect of wave scattering from the rough 

 ice boundaries was taken into account using a modified formula of 

 Mellen and Marsh (1965) . An RMS ice roughness of 3 m fits the data 

 for Profile 1, while an RMS roughness of 4 m agrees reasonably well 

 with Profile 2. These experiments were performed in 1968 (Kutschale, 

 1969) and provided solid evidence that indeed there is a close corres- 

 pondence between surface ice roughness and wave amplitude . We shall 

 return to this matter shortly in connection with propagation loss. 



Attention will now be turned to computational aspects of deriving 

 estimates of propagation loss as a function of range. Wave theory 

 will be emphasized because of the close agreement of field data 

 previously shown with computations derived from wave theory. Further- 

 more, propagation loss of ice vibrations detected by vertical and 

 horizontal component geophones is included in the wave theory and 

 the method is useful at extremely low frequencies or in shallow water. 



702 



