interference below carrier

C/N 2

The measurement you are making above is strictly (C+N)/N   In your example this makes negligible difference. See https://www.satsig.net/lnb/ebno-calculator.htm  If you consider the interference "carrier", its (C+N)/N = 2 dB and its C/N is -2.3 dB. The interference is actually 2.3 dB below your noise floor.  Think: If the interference was the same level as your noise floor you would see a 3 dB high hump.

Leave the spectrum analyser in max hold for a while to see if there are brief bursts of higher level interference that might cause intermittent error bursts in your transmission.

Why not find the source(s) of the interference and try to get it removed ?  Check your own receive system first. Is your polarisation correctly adjusted ?  Do you have receive sidelobes aimed at the adjacent satellies ?  Is the interference visible at other receive sites ? Does the interference change exactly as you would expect if you mispoint your dish slightly ?

Anyone with an operational carrier should exploit any outages by closely examining the noise floor under the carrier to see if interference exists.

Anyone with TDMA return link can see interference in real time by just watching the return link spectrum with its intermittent wanted bursts. Try using normal freq sweep display and max hold as well as zero span to look for TDMA burst interference.
wxw
Best regards, Eric.

In addition to all the excellent advices of Eric, I would like to add one grain more. I think you may contact the satellite operator you purchase the segment from, and ask them for screenshot of the space taken from their earth station, as well in max hold and normal sweep. It will allow you not only to find out whether your earth station has some problem (With polarity, side lobes, local interferences etc.), but will also allow you to request to replace the segment in the future, in case there is an interference that can not be eliminated.

Kind Regards,
Nimrod