It is possible that cross-pol interference is entering into the wanted signals on the uplink from an earth station with poor transmit isolation or adjustment. Even if the x-pol satellite output TWTA or transponder is muted it is possible that the uplink transponder frequency band is IN USE and being cross strapped, using on-board switch matrix, down into another downlink beam.
The satellite NOC can diagnose this by comparing the carriers on the other transponder with the interference spectrum observed on the wanted transponder.
Interference can also occur on the uplink due to earth stations pointed towards an adjacent satellite. Such interference will correspond to carriers on the other satellite, and not match a pattern of carriers on the wanted satellite. This is not cross-pol interference.
If you are watching a carrier that is clean at the satellite and can see it also at low level if you change your receive polarisation then the cross-pol isolation you are measuring is the combination the satellite downlink antenna isolation plus your own receive antenna isolation. The same applies if you observe a satellite beacon.
I suggest you record the interference you see and send it to the satellite NOC asking for help. You might usefully provide plots of what you see on your wanted polarisation noise floor when your own carrier is turned off, and what you see on the other polarisation.
If the interference is intermittent, possibly from an individual TDMA uplink site, use max hold for several minutes to get a clear record.
With multiple satellites in orbit and multiple beams there are very many sources of interference. This is inevitable. Some interference is always to be expected and this is acceptable. Acceptable levels of interference should always included in link budgets, for example you might design a link with 3 sources of interference with levels like -28, -21 and -28 dB for aggregated uplink interference, transponder intermodulation and aggregated downlink interference.