Signal variation between day and night

Could someone explain to me why there is too much difference between day and night signals?

Captura de pantalla de 2021-12-30 13-09-08

Looks like Thermal Inversion to me. Have issues with it on my longer links (little water by me, but really wild temperature swings cause the same problem)

Basically, the mirages you see in the visible spectrum also effect radio signals. Frequently worse then visible spectrum.

Note, different portions of the 5 GHz spectrum will be effected differently and at different times, that is one of the advantages of Mimosa’s Auto Channel selection in that it “could” work around these issues somewhat. (With it’s own issues being introduced unfortunately)

Mimosa has a really cool post about their own solution to thermal inversion. Deploying a Diversity Link | Mimosa by Airspan

Does this link go over water?

Are there AM towers near to where you are? I think (think) most AM stations put out a stronger signal at night so more noise - but your graph seems to indicate the opposite - better at night than day - are there any other sources of noise that you know of, air conditioners compressors etc?

Thank you very much for the reply.
There are some lakes, but they are not very big.
Noise sources do not exist near the towers.

Thank you @William5 .
I’m going to read the article that you recommend.
And try the mimosa channel auto select

Yea then prolly what @William5 was saying thermal ducting -

Let us know how you mitigate / deal with this issue.