>It seems like Space-Track is the official source of TLEs. Is that correct? It seems like CelesTrak and AMSAT are

> secondary sources. Is that correct?


This is almost always the case, but there are some situations that arise that require us to get element sets from other sources, usually the satellite operator.    

>It seems like the process is this:

  1. >2. At some point an official ID number is generated. Who does this? I’m assuming NORAD?

The "official" number is the Satellite Catalog number generated by the United State Space Force's 18th Space Defense Squadron (18SDS).   This number is generated as soon as the analysts at Vandenburg Space Force Base get tracking data (radar echoes or images) of a new object detected in orbit.  

      >3. At some point, the satellite is given a name like “AO-123” versus “2024-199AY” or “OBJECT AY”. Who does this? Is that an “official” name that all TLEs should use?


Naming is essentially up to the organization that is using the name.   For example, object 29661 was known as RAFT, RAFT1, Navy-OSCAR 60, NO-60 or in my house "Daddy's Satellite", although to my knowledge neither of my kids  issued any TLEs for that object.


  1. >4. At some point it seems like the TLE sources align with the same name. Who drives that?


18th SDS works with the satellite owner to assign a name to an object.   It can take some time for an owner to be sure their spacecraft matches a particular Satellite Catalog number while a flock of cubesats are all clumped together after launch.    It is only after the objects drift apart that it is possible to make a match.   In the case of amateur satellites people rely heavily on the work of Nico, PA0DLO who uses careful analysis of Doppler shift to match a downlink with TLEs for a particular object number.


>When will Space-Track and CelesTrak list it as AO-123? Who decides that?

Given the current geopolitical situation, the satellite owner may not be in a big rush to work with the US military.   TS Kelso and his team decide what happens on CelesTrak.    


>Also, why is AMSAT’s TLEs so far behind Space-Track and CelesTrak? Notice AMSAT lists revolution 49614 versus

> the others that list revolution 49716 so it’s quite a bit behind.


AMSAT issues a TLE bulletin in the first hour of the UTC day using the most recent information available,     Note that the last digit of each line of a TLE is a checksum, so in your example AMSAT TLEs are only 10 orbits behind, i.e less than one day.    


>Lastly, sometimes the TLE sources never seem to align on a name. <snip>

>Why don’t they use the same name after all these years? Who decides that?


AMSAT's bulletins are prepared by a team of students at the New England Sci Tech STEM education center under my guidance.    If a satellite has an OSCAR number, we will use that in the form of *O-##.  If no OSCAR number has been assigned, a judgement is made to use the name most familiar to the amateur satellite community.     It is of course up to other organizations to use names most appropriate for their audience, and it is not unusual to see more than one, e.g  ASRTU-1(AO-123).   


Joe Fitzgerald (KM1P)

AMSAT Orbital Elements Manager