The Joint Space Operations Center or JSpOC:
...is the organization responsible for performing all of the orbit determination activity necessary to maintain the US space catalogue.
During announced space launches and satellite deployment (and of course unannounced ones as well!) various space monitoring technologies including ground based radar and optical telescopes make a series of observations of space objects for the purposes of orbit determination. At some point if an object crosses some threshold of potential to be a new object, it is assigned a NORAD Catalog Number. If an observed object is later determined to be a known object, the identity assigned to the catalog number is updated. Thus one object can easily appear under more than one catalog number.
International designators are applied to identifiable components of a space launch or a space mission. The naming is getting more complex as components attach and leave the international space station, or are shuttled there then deployed to space at a later time (e.g. cubesats getting tossed out of an airlock).
In this case 2017-003 is the third launch of 2017 to be assigned an international designator.
While observing the deployment, eleven objects were observed in addition to the Falcon 9 2nd stage, and assigned NORAD catalog numbers 41917 through 41927, and the letters A, B, C, D, E, F, G, H, J, K and M were assigned to them. It is not clear why the letter 'L' was skipped, but the letter 'I' is always skipped because of a potential ambiguity - it can be confused with the digit '1'. 41917 through 41926 likely matched the 10 Iridium satellites in character, and were later identified as IRIDIUM 102, 103, 104, 105, 106, 108, 109, 111, 112, 114 (not in that order, see below) the satellites known to be in the F9 2nd stage payload.
However, the eleventh object, number 41927 was in question. It seems to have been given the international designator 2017-003M, but was also called 2017-003L. From here:
I've been in email contact with Dr T.S. Kelso, who runs Celestrak.com. The first thing he did is got the orbital information from Iridium and matched them up, so the Celestrak database now has the satellites named. He has changed Object M to Object L in his database, as he can see no reason, other than an input error, for the letter L to be missed. If it was left for the stage, then there would have also been a spare. Catalog ID as well. (emphasis added)
As for the 11th object:
- In all likelihood, OBJECT L is some kind of upper stage, unless there were deployment mechanisms expected to be large enough to track. We don’t have any RCS data yet (and even that can be unreliable), so there is no way to know if this is something large like a rocket body or something else. All we can do is continue to monitor at this point.
Today I've done a search on the celestrak website: https://www.celestrak.com/satcat/search.asp and by playing around with the search input, including the text in the un-selected data field, I could get the chunk of catalog numbers that included the ROI. You can see that number 41927 has been identified, and there are mysteriously un-named catalog numbers immediately following.