I'll just summarise to bring Mike Smith at Forest VA into the loop.

Stefan DK7FC wrote:

> f = 17470.1000 Hz
> Start time: 05.Dec.2017  21:40:00 UTC
> Symbol period: 30 s
> Characters: 4
> CRC bits: 16
> Coding 16K21A

Measures Eb/N0 = -7.5dB with cross-correlation at Forest,
that's -38.55 dB in 1Hz.

Stefan wrote:

> f = 17470.1000 Hz
> Start time: 06.Dec.2017  22:00:00 UTC
> Symbol period: 40 s
> Characters: 3
> CRC bits: 10
> Coding 16K21A

Correlates at Eb/N0 = -4.8 dB at Forest, -37.30 dB in 1Hz.

It's getting on for 1 in 10,000 chance that these two
correlations of the EbNaut messages are from random noise,
so this is leading to the first east-to-west amateur VLF
detection.

This would a remarkable result at 6818 km, the east-to-west
path is much harder than west-to-east due to anisotropy
of the ionosphere.

Hopefully some further transmissions from Stefan will
strengthen this result.  Ideally a blind decode of a
short message.

--
Paul Nicholson
--