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



Propagation loss estimates based on dispersive propagation have 

 been presented. A diffraction correction to WKB theory has been 

 necessary to prevent divergence of the energy calculations near the 

 ray turning points. Reasonably good agreement has been obtained be- 

 tween this theory and loss data obtained in the western Mediterranean. 



Future work should include both experimental and analytical 

 effort. Wide-band data, with sources and receivers on the channel 

 axis, are needed for other areas besides the Mediterranean. It is 

 important to know if simultaneous ray and mode resolution is possible 

 elsewhere. It would be very interesting to see wide-band data ob- 

 tained with a swept frequency source. The ability to alter the dis- 

 persion of the received signal has interesting ramifications for 

 future experiments. More detailed development of the propagation- 

 loss software package is required. It should be based on more exact 

 mode theory or a better diffraction correction. We should also study 

 the possible generalization of the parabolic-equation technique to 

 include wide-band signals. 



DISCUSSION 



Dr. P. W. Smith (Bolt, Beranek, and Newman, Inc.): Dr. Porter's 

 pictures are the most fascinating, stimulating ones I have seen for 

 a long while. They represent a frequency-time decomposition of the 

 energy transmitted from an explosive sound. I guess this is a situ- 

 ation where one can see what he looks for, in a certain sense, and 

 you were lucky to see both rays and modes at the same time. 



It seems to me that there is a certain fundamental limitation on 

 the product of resolution time and resolution bandwidth, but you can 

 arrange that area in any shape you want. You could, for example, try 

 to get more resolution in time, in which case you would have to have 



658 



