[Amps] article in Nature, modified Aurora via HF

[Bill Fuqua]

       I think the real issue is being missed there. The real test of 
non-linearity is the production of harmonic and intermodulation distortion. 
This means allows measurements of nonlinearity as small as a very small 
fraction of a percent. Even to a part per  million with very specialized 
equipment.
       If  there were nonlinear propagation the RF spectra would be 
unusable.  The apparent increase in signal strength could be  in 
perception.  Which has to do with your reference.  On receive, a signal to 
noise ratio of 0 db would be unintelligible but a 3 or 6 db increase in 
signal would provide lots of improvement.  And are you measuring signal or 
signal+noise?  It comes down to this. If no harmonics are produced and no 
intermodulation products produced between the transmitting and receiving 
sites the propagation path is linear.
      Now, if you should be unfortunate enough to have a thermonuclear 
blast between the transmitter and receiver you would have lots of 
nonlinearity.  This is due to the greatly ionized region (large volume of 
plasma) due to the blast. This scrambles signals and presents all sorts of 
problems for those trying to use RF communications.
      If you want to do such an experiment arrange for two stations near by 
to transmit a KW signal and then look for the intermodulation products at 
the far end of the path.  From that you can calculate the degree of 
nonlinearity.  Don't use stations are close enough that there may be mutual 
coupling between their antennas and sending a 2 tone test will not work 
either because the degree of nonlinearity that you are looking for will be 
very very small. But don't forget. There is also nonlinearity in your 
receiver as well. But all of those sources of nonlinearity are  small 
compared to the nonlinearity that would produce several db increase in 
receive signal strength over the change transmitter output power.

73
Bill wa4lav