Stefan schrieb:
> The most enjoyable finding (within ham experiments)is, that the
> E field decay relly seems to be rather ~ 1/sqrt(d) than ~1/d :-)

1/sqrt(d) is doing better than 1/d, so far with just one data point
to go by.

> Does that mean that if the ERP is 3dB higher (21kV instead of the
> 15kV i had yesterday), the maximum distance will be 4 times the
> old distance, ideally?

There is also an attenuation factor.  Approx 3dB per 1000km in
daylight, and 2dB/1000km at night, at 10kHz.   Combine that with
1/sqrt(d) and we are getting a good match with your measured flux
density.

The D layer is quite absorbing to higher frequencies, but at VLF
it is more of a mirror. It is appropriate then to look at propagation
in terms of cavity modes, as compared to following rays.

9kHz is a bad frequency - background noise is highest there because
lightning energy couples well to the Earth-ionosphere cavity at
that frequency, giving a broad peak in the noise spectrum.

Better might be say 2kHz, just above the waveguide cut-off frequency,
where noise is low.   Obviously lower ERP but maybe the lower
noise more than compensates?   Must do some arithmetic...

(Always there is more arithmetic.   Flying kites sounds nice.)

Earth 'loops'?   Could work in some situations.  Limestone country
with underground stream to complete the loop?   Top to bottom of
a long rocky escarpment, forcing current to go the long way round?
Oxbow bend in a stream over rock or clay?  Hard to model.  Easy and
fun to test.
--
Paul Nicholson
--