Hi Mike and the Group,
If you care to trawl Rik's web site there are several plots on 137kHz when
CFH has gone off that indicate the noise here is related to the oncomming
daylight. This is confirmed by observations on Spectrogram that Brian made
in Portugal. The noise seems to make a steady decline but the effect is most
noticable on a waterfall display about 1 hour before sunrise in the UK. The
reponse of a waterfall is different to that of my CFH plots, as the latter
are processed to remove sharp spikes of noise, such as static crashes, which
may play a big part in the nightime noise pattern.

Interesting plots are for following dates:-
Oct 26th
Nov 15th
Dec 14th
Dec 15th
Dec 18th if you look at the periods when the carrier was cut.

I also have a plot where CFH was off for the 16 Dec that I did not lodge
with Rik, there is a lot local static (rain?) noise at one stage and much
qrm around 137.0 in the morning (G8RW), but that shows a decline of
'non-local' noise of 4-6dB between 2400z and 0600z
These graphs may not be as dramatic as is seen on some waterfalls as they
are taken with a long time-constant and a form of 'averaging'  which only
records the lowest of 6 sequential measurements. I still have all the raw
data for these dates and could replot all the data 'averaged' instead of
'censored' which might give an indication of the 'subjective' noise levels
as well.

My explanation of this 6dB decline is the shutting down of the area of
received signals (including the storm areas)  as daylight approaches. Our
biggest area of noises in the UK is to the east. The effect may be different
in central Europe, it should be essentially symmetrical. This is based on
the opposite effect which occurs in Canada where the S/N is best in the
early evening, and degrades about 4 hours after sunset.

To this is added the fact that most local noise sources are generally quiet
for a hour or so after dawn.

Cheers de Alan G3NYK
[email protected]