|
Dear Jay,
thanks for posting this interesting
screenshot.
The fine structure in the
spectrum is superposition of the three sources, transmitting
coherently on the same frequency but occupying different time slots 0.6 s
apart. The scheme of the 3.6 s repeating frames is currently:
0 3.6
s
+-----------+
|K E
N N*. .| 14.881 kHz
|. N E K . .| 12.649
kHz
|N . K E . .| 11.905
kHz
+-----------+
K = Krasnodar (West)
N = Novosibirsk (Center)
E = Khabarovsk (Far East)
* second dash from N on 14.881 has
alternating phase
The mechanism is similar
to selective fading, but the delay involved is not from propagation
but from the (much longer) transmitter delay. If you were for example
getting only N and K, the two dashes 1.2 seconds apart would result in a
0.83 Hz periodic ripple in the spectral envelope.
As the transmitters are laid out on a ~ 6000 km
baseline, the components will have different diurnal phase variations, resulting
in a slow movement of the minima in the spectrogram. If the later component
is lagging more, the ripple pattern moves down in frequency (ie. contracts
towards zero).
The plot appended here shows the three
signals separated in time domain, as observed here on Mar.
20. The idea behind this was to get more data for
a simple model for propagation loss, and I was expecting something
along the lines of 1/sqrt(r) and -2 to -3 dB/Mm.
But the pattern appears to be much more
complex. For example, at 11.9 and 12.6 kHz,
nighttime fieldstrength from Novosibirsk at 5 Mm actually exceeds that from
Krasnodar at 2 Mm (ie. the red and green thin lines go above the thick lines).
This is not the case at 14.8 kHz (blue). Also, the 8 Mm path from Khabarovsk
exhibits an unexpectedly deep daytime minimum. There are more plots in http://df6nm.bplaced.net/VLF/Alpha/ , showing that
level and phase patterns are essentially repeating every day.
The results seem to indicate that a fieldstrength
prediction for long ranges would seem feasible in principle, but one would have
to include interference between groundwave and multiple waveguide modes.
There may be favourable and unfavourable combinations of distance and
frequency, which (unlike at HF) are rather fixed and will not be averaged out by
slight ionospheric movements. This is also pointed out by an
older dual-frequency observation of DHO at 475 km:
Kind regards,
Markus (DF6NM)
----- Original Message -----
Sent: Wednesday, April 14, 2010 2:04
PM
Subject: LF: VLF...alpha long selective
fade?
Been monitoring / plotting the
alpha stations over the past couple weeks while working to improve
the VLF
receiving setup. The following very slow 'selective fade' has been seen every
night that
I've looked (4 times) so it's not a one off occurrence. Anyone
care to speculate on what would cause
this?
Full screen here: http://www.w1vd.com/capture/alphafade1.jpg
. Condensed version as an attachment.
Jay W1VD WD2XNS
WE2XGR/2
| 
