BUCK: ARCTIC ENVIRONMENTAL LF ACOUSTICS MEASUREMENTS, 

 MODELS AND PLANS 



We took all of the data for the case of a 400-foot source, 



100-foot hydrophone and water depths greater than 3,000 meters and 



fitted them to a cascaded filter model made up of 1-pole (i.e., 



-6 dB/octave slope) filters. This empirical model is shown in 



Figure 7, which is best explained by way of an example. At 200 



nautical miles, the filter is made up of three 1-pole filters in 



series (-18 dB/octave slope) each with a 3 dB break frequency of 



48 Hz and a total insertion loss of 54 dB. The model can be used 



at any range to derive loss without using an integral number of 



filters by determining the insertion loss, 3 dB break frequency and 



slope and applying a Bodie plot (- 10 n log [1 + f/f ] ) . 



3dB 



In Figure 8 are plotted transmission loss measurements as a 

 function of frequency at ranges of 23, 72, 200, and 480 nautical 

 miles, which correspond, respectively, to one, two, three, and four 

 1-pole stages of the model which are plotted in solid lines. The 

 empirical fit is seen to be quite good. 



Figure 9 is taken from one of Hank Kutschale's figures where he 

 compares early measurements by the author and Bob Mellen with his 

 computed FFP and ray theory models. We have added the much more 

 extensive 1970 data and the filter model of the previous two figures. 

 Over 100 km, our controlled measurements and filter model agree 

 almost exactly with Kutschale's computed FFP. The difference at 

 short range is to be expected since RRR rays and, hence, local bottom 

 topography plays an important role. 



It was previously mentioned that the Arctic's stable vertical 

 velocity profile was a steep positive gradient from the surface down 

 to about 1,200 feet and a less-steep positive gradient from there 

 to the bottom, and that the concentrated rays traveling in the uppermost 



738 



