Yet another comment.
Today a have find some web page
(www-star.stanford.edu/~vlf/south_pole/south%20pole.htm)
where presented waterpole picture with some signal  E ~ 4uV/m near 10kHz.
It is dificult to estimate FFT-bandwith with the  picture. But full screen
is only few seconds. Signals are nice detectable. If this is right
then much weaker signal to be detectable in long mode.

On Fri, 26 Feb 2010, James Moritz wrote:
> 
> Dear Markus, LF Group,
> 
> I checked the noise level again during the daytime, and the noise level was 
> about 6dB lower than last night, i.e. about 13uV/m per sqrtHz
> 
> One possibility for the discrepancy between results is that the effective 
> height of your antenna is reduced at VLF compared to the LF calibration 
> point. In my transmitting vertical field strength measuring sessions on 
> 500k, 136k and earlier 73k, the Heff of the same antenna works out 
> consistently less as the frequency is reduced. I think this is due to 
> increased "site loss" due to surrounding trees at the lower frequencies. 
> H-field sensing loop antennas can be expected to be less affected. Of 
> course, if there are no trees near your antenna, that theory is blown out of 
> the water...
> 
> Cheers, Jim Moritz
> 73 de M0BMU
> 


Regards,
Alexander