# Is TESS really lost? What does Celestrak mean exactly?

From this answer to the question What artificial satellite has the farthest orbit around the Earth?:

I found the following "far out" spacecraft:

• TESS (Transiting Exoplanet Survey Satellite) recently launched, not in final orbit yet
• Spektr-R
• IBEX or Interstellar Boundary Explorer
• Geotail

Here are there IDs:

TESS       43435    2018-038A
Spektr-R   37755    2011-037A
IBEX       33401    2008-051A
Geotail    22049    1992-044A


I downloaded both TLEs and orbit data from JPL Horizons in order to piece together this qualitative data.

Question: So why does Celestrak show three of them as lost? What does "lost" mean in Celestrak-speak?

I'm not looking for guesses or free-form interpretations, I'd like to know how Celestrak uses the term "lost" in this case, and what it means specifically.

GEOTAIL
1 22049U 92044A   19100.47047473 -.00001329  00000-0  00000+0 0  9998
2 22049  16.9903  61.2002 5647056 148.0163 356.4514  0.19163824 10715

• celestrak.com/satcat/lost.php – Chris Apr 16 '19 at 2:03
• @Chris If "Objects which should have TLEs but for which no TLEs were found in the last 30 days." is the answer, then it's worth posting as the answer. Thanks! However, the devil may be in the details; what does "should" mean? – uhoh Apr 16 '19 at 2:04
• I checked all documentation for instances of the word "Lost", over 15 PDF documents yielded no results (for such a common word that seems odd). – Magic Octopus Urn Apr 17 '19 at 17:45

You're right; these objects aren't lost. However, Space-Track doesn't actually track a lot of higher-orbiting objects. They just don't seem particularly interested in anything with a multi-day orbit; I don't know why.

Of the four objects you mention, TESS and IBEX are tracked by the people operating those missions. I know of no public data for IBEX, but JPL's Horizons system provides excellent ephemerides for TESS :

https://ssd.jpl.nasa.gov/horizons.cgi

One of the near-earth asteroid surveys observed it, briefly causing us to think we'd found an interesting rock until we identified it. An observer in the UK imaged it as well; the positions they measured are at

https://www.projectpluto.com/pluto/mpecs/18038a.htm

and match the JPL ephemerides well. I compute TLEs from the JPL ephemerides. Space-Track doesn't seem to bother.

For GEOTAIL, Spektr-R, and Spektr-R's booster (the last isn't even a catalogued object!), the asteroid surveys have provided plenty of observational data, and I use these to provide TLEs which are, in turn, used by the asteroid surveys to identify junk they happen to find. (You can imagine that it would be embarrassing to announce an incoming "rock", only to find out that it was some bit of junk. Not a very likely scenario, but when the surveys find junk, they like to identify it as such and move on. Satellite observing is not part of their purpose.)

Links to pages giving the observational data for these objects, and many other bits of high-flying junk, are at

https://www.projectpluto.com/pluto/mpecs/pseudo.htm

The TLEs for TESS, Spektr-R, GEOTAIL, and other objects not listed by Space-Track are at

https://www.github.com/Bill-Gray/tles

In re "objects which should have TLEs but for which no TLEs were found in the last 30 days": Space-Track's coverage of some high-fliers is episodic. I think they only emit new TLEs if they have new observations. For some Velas, for example, they have data on some months and not in others. (For many, they don't have anything at all. We do have TLEs for eight of the twelve Velas at present.)

I found this out the hard way when the surveys found some "new" artsats that I couldn't identify. I eventually figured out that it was just that Space-Track had temporarily or permanently stopped providing TLEs for them.

• +1 Thank you for the thoughtful and really informative answer! – uhoh Apr 18 '19 at 2:05
• "Not a very likely scenario"? Someone should have told the discoverers of J002E3 and the potentially-impacting 2007 VN84. – Mark Apr 20 '19 at 0:46
• – uhoh Apr 28 '19 at 0:40