PORTER: SOFAR PROPAGATION OF WIDE-BAND SIGNALS TO LONG RANGES 



Dr. H. W. Kutschale {Lamont-Doherty Geological Observatory of 

 Columbia University): In the Arctic, dispersion is characteristic 

 of the signal propagation. We have been computing pressure pulses 

 for a number of years using normal mode theory. Currently, in 

 connection with propagation loss, we have been using the FFP with 

 considerable success. Recently Fred DiNapoli has extended the FFP 

 to transient signals and I think his results would be of considerable 

 interest to the community. 



And one other point — the initial normal mode computations made 

 by Pekeris for the two-layer case were subsequently done by him in 

 the time domain using a ray theory, and this type of approach has 

 been extended to multilayer systems in seismology. I think some of 

 these techniques might be quite useful for the higher frequency 

 acoustics community. 



Dr. Hersey: I think one note I have to introduce is based on 

 such things as Ira Dyer's question last night about the requirement 

 that we really do contribute to solving the problems of people who 

 pay us. Unfortunately, unless we can make a compelling story for 

 any such thing as the thing we have been discussing in the last few 

 minutes — namely, working out impulse response type methods of 

 analyzing propagation — we will very quickly be accused with the 

 refrain that I keep hearing in Washington, "Oh, those fellows are 

 just doing science for science's sake." That is, they say, we have 

 abandoned the idea of making science useful. 



This simply won't be supported unless we can make a very, very 

 cogent case for it supporting a practical end. I'm sorry, but I 

 believe that is true in the present climate. 



665 



