Hi Alan  My experiments with  skywave propagation from years ago was made with a
Loran -C receiver.  This is very usefull because Loran-C is a pulsed system.
Because the pulses are long ( on the order of 200 microseconds ) and ground
waves and skywaves or first hop and second hop skywaves tend to overlap. You can
still glean valuable information about skywaves.
       I observed signals at a range of say 800miles.In the daytime the
groundwave would be just barely visible.  The first hop skywave would have a
delay of about 35 microsec, and would be typically be 6 db or so stronger.  As
night approached , the skywave amplitude would double or triple.  Then its delay
would increase to around 60 microseconds.  Then the amplitude would triple again

    Note if the daytime delay were 35 microseconds the two pulses would
subtract each other .  If  the delay were 40 microsec the pulses would add.  The
observed signal is further complicated because the pulses have a rise time of 6
cycles of the carrier freq or 60 microsec.
      Anyhow,  the slow change in delay of the first and second hop skywaves
causes the pulses to add up   with  each other and subtract from each other 4 or
5 times as the delay gradually increases  The same thing will also happen with
cw signals.
       On occaision one could observe a ground wave signal followed by a
skywave of 500 microseconds delay .   The transmitter range would be on the
order of 400 miles.  From this one can calculate the height of the reflecting
layer.
      For people seriously  interested in getting an understanding of LF
propagation I would highly recommend the use of Loran-c receiver an oscilloscope
and a scope camera.
Taking data during a solar eclipse can be very interesting
      73 de John VE1ZJ

Alan Melia wrote:

Hi David, well no, its too early for sundown effects. The twilight shadow
reaches the west coast of Greece about 1625z with full darkness about 1650z.
(old GeoClock) That was the puzzling thing. There is a slight 'wriggle' in
the plot at dusk with two slight dips about an hour apart but nothing as
consistent and deep as the 1550z dip. Most odd.

Cheers de Alan G3NYK
[email protected]