FITZGERALD: CONVERGENCE ZONE DEPENDENCE ON FREQUENCY 



and RSR type modes. This effect is again produced by the slow decay 

 of the lower-frequency eigenfunction in the shadow region of the 

 associated ray. The 31-Hz SOFAR modes with phase velocities slightly 

 less than the speed of sound at the surface decay slowly into the 

 upper thermocline and sense the pressure-release surface . The 

 corresponding 200-Hz modes decay rapidly and do not sense the surface. 

 The same effect occurs for modes with phase velocities slightly less 

 than the speed of sound at the bottom. The higher 31-Hz RSR modes 

 sense the profile's steep negative gradient in the bottom. The 200-Hz 

 modes do not. 



DISCUSSION 



Dr. Gordon Raisbeck (Arthur D. Little, Inc.): You showed that 

 the convergence-zone length when plotted against a certain mode para- 

 meter was different for two different frequencies only over part of 

 the range of the argument. 



You also showed that the amplitudes of the modes were far from 

 uniform. It happens in the examples you showed that the amplitudes 

 are large just in the same range of the parameter where they diverge. 



You failed to say that there was a common cause. Was this 

 accidental or is this a general rule that the amplitudes will be high 

 just in the range of parameter where there is a difference between 

 the two frequencies? 



Dr. Fitzgerald: The plot of interference wavelength versus 

 average phase velocity is independent of mode excitation. That plot 

 comes only from the calculation of the separation constants, the 

 eigenvalues. So it's totally independent of the excitation. The 

 excitation is determined by the location of the source and receiver. 



681 



