I think the Aricebo HF transmitter is 0.6 MW, not 6 MW
6 MW is EIRP, I guess.
The Aricebo ERP is 220 MW at 8175 kHz and just under 100 MW at 5200 kHz.
They are almost certainly noise from some production technique, RF welding, etc.
The HF generators of such industrial plants must have frequency inside specially allocated ISM bands which are mostly the same in various countries: 6765 - 6795 kHz, 13553 - 13567 kHz and 26957 - 27283 kHz for SW. And even taking into account the fact of presence of non-standardized plants of such kind or some spurious emissions, the most activity of the "sweepers" must be observed on these ISM bands. However, it is not obviously on the posted screenshots: they are active on 24 and 28 MHz instead.
Yes, and those facts are why I did not say these signals ARE from things like RF welding, and rather said "almost certainly".
However, I have seen visually identical signals from a plastic welder, and it was slightly outside the ISM bands (in that case 25 - 26 MHZ).
Combine that with the fact I see the majority of these signals (outside the ISM bands) when propagation is in from Asia, and China is well known for ignoring the rules and doing their own thing, and I think the majority of these that I see are quite possibly, almost certainly (but not quite certain) are some similar industrial technique.
Of course, that is why this is still a mystery. A list of possible sources has been ruled out, the most probable source should not be happening on these frequencies, and at this point we are still guessing.
With the signals I was looking at locally there was a definite reduction in number visible on any given Sunday. I plotted this over a period of 6 weeks and
every Sunday the number was less than one tenth of what was seen on Saturday and Monday. This is a strong indicator, in my opinion, that these signals are man made.
The waveforms are unstable and varied, especially in phase. And there are seemingly many different sources up at any time (based on propagation and antenna bearings). These would argue, to me, against things like research facilities, which should be few in number and probably using more traditional or predictable waveforms. Also, research tends not to avoid one specific day of the week, you generally do research when you can get the time on the facility or when the conditions are right for your tests.
Just my opinion of course.
T!