The propagation losses for the summer profile, 500 m depth, Straits of Sicily 

 (see Figure 4.25) are quite high. Although the cause for this is not immediately evident, 

 high bottom reflection losses are suspected. Thus, calculations of propagation loss and 

 bottom reflection loss were made for several frequencies using both the low and high 

 attenuation models. The resulting propagation losses (low attenuation) at 100, 200, 400 

 and 800 Hz are plotted in Figure 4.25. They are represented by the open circles. The 

 differences between the computed losses are high, 12 dB at 100 Hz, 18 dB at 200 Hz, 

 26 dB at 400 Hz and over 40 dB at 800 Hz. 



Figures 4.26 and 4.27 show bottom reflection loss versus grazing angle for the 

 low and high attenuation models and for the frequencies of 100, 200, 400 and 800 Hz. 

 There are fluctuations in the data which are indicative of the layering in the sediments. 

 The fluctuations are greater in magnitude as the frequency is increased; the layers become 

 a major bottom interaction mechanism. 



The differences in bottom reflection loss calculated for the two sediment models 

 are large even at 100 Hz: 1.5 dB at 10°, etc. These data are one output of the normal mode 

 model and thus the largest angle shown is the grazing angle associated with the last mode 

 used in the calculation. 



4.3 Comparison With Cylindrical Spreading 



The normal mode model used here assumes a perfectly reflecting sea surface, so the 

 only sources of loss are bottom loss and volume attenuation. This can be used to classify 

 different propagation loss curves. Figure 4.28 shows curves for cylindrical spreading loss 

 plus volume attenuation which we computed by Thorp's equation (Reference 8) for the 

 three ranges 50, 100 and 150 km. Actual propagation loss cannot fit these curves in 

 absolute value since the propagation must fall off initially with spherical spreading loss. 

 However, the shape should be the same if absorption is the only frequency dependent 

 mechanism at work. The winter profile results of Figure 4.20 are found to fit the 50 m 

 range quite well above 200 Hz. The loss for the summer profiles (Figures 4.21 and 4.25) 

 increases more rapidly with frequency than does the absorption only curve (Figure 4.28). 

 This is to be expected since bottom loss is a factor in the summer (downward refracting) 

 profiles and bottom loss generally increases with frequency. 



The increase in propagation loss at low frequencies over that of the absorption only 

 curve occurs because at sufficiently low frequencies no modes are trapped by the positive 

 gradients. Then all modes are equivalent to bottom reflected paths. With decreasing frequen- 

 cy, these paths become steeper resulting in both greater reflection loss and more reflections, 

 since the distance between reflections are closer. This is most evident in the East of Singapore 

 case and can be easily seen in the propagation loss curves for the Korea Strait (Figure 4.24). 



The three curves of Figure 4.28 can be used also to indicate ducted propagation by 

 noting the distance between them at a given frequency and comparing this distance with 

 any propagation loss plots set for the same three ranges. If the distances are the same, then 

 only absorption and cylindrical spreading are present. (The computational model used here 

 has no surface reflection loss.) By such comparisons, we find that the propagation is ducted 

 in a nearly lossless duct in the following: 



• Winter Profile Cases 



North Sea above 400 Hz (Figure 4. 1 ) 

 Lands End above 300 Hz (Figure 4.1 1) 



61 



