3

I'm interested in generating a "corrected TLE1" based on observation data. I would have observed azimuth and elevation, site location, and time. All of those data fields are observed with high quality (0.1 degree for pointing angles, GPS locked location and time). I would have an array of these observations for let say several minutes at 1 second intervals or less. Range would be unknown.

Is there any open source software to do this?

1Two Line Element set, a common orbital ephemeris format described here: https://celestrak.org/NORAD/documentation/tle-fmt.php

Community
  • 1
Judd
  • 91
  • 1
  • 2
  • What does TLE stand for? – Rodo Jul 21 '21 at 18:15
  • 1
    Two Line Element, a common orbital ephemeris format described here: https://celestrak.com/NORAD/documentation/tle-fmt.php – Judd Jul 21 '21 at 18:17
  • @Judd I've added that back into your question, but I don't know what a "corrected TLE" is. Corrected for what? Are there "uncorrected TLEs" as well? Also note that if one has only 60 seconds of visual observations the quality of the resulting TLE will likely be lower than one based on deployment state vectors provided by a launch provider. – uhoh Jul 21 '21 at 22:49
  • @Judd, probably you are interested in making corrections to a KNOWN TLE, using additional observations. But, do you really need TLE or your problem is rather refine the orbital elements (using additional observations)? – Ng Ph Jul 22 '21 at 09:04
  • @nhoh, Satellite operators are not 100% accurate in the TLEs provided to tracking stations. This results in loss of data from the spacecraft, especially when you are dealing with higher frequencies or larger dishes (both reduce the beamwidth of the antenna system). From the ground station we can measure a set of correction vectors, and apply that in real time. However, this can be a problem if the final correction is large and you may not acquire the satellite at all on the next pass. If you can use the measurements to generate a new TLE then you have a better chance of AOS on the next pass. – Judd Jul 22 '21 at 14:56
  • @NgPh, You are correct. We do need a TLE format as that is pretty much universal input to tracking ground stations. Newer orbital models would be better, but I would still have to convert to TLE to load to the tracking pedestal. The TLEs would be short-lived, however, not used more than a day or two. – Judd Jul 22 '21 at 14:59
  • @Judd, I suppose "a day or two" the time period between two passes( also a period of no observations?) If your Antenna Control Unit implements the SGP4 and you wish to treat this SW as a black box, then it looks like you need to launch into some Kalman filtering. Not a straightforward exercise, I may say. Let's see whether somebody else has a better idea (I am curious to know the answer too!). – Ng Ph Jul 24 '21 at 13:39
  • I know that Stackexchange does not like this, but some information can be found in these articles:
    • https://en.wikipedia.org/wiki/Orbit_determination
    • https://en.wikipedia.org/wiki/Gauss%27s_method
    • https://en.wikipedia.org/wiki/Lambert%27s_problem
     – axd Sep 25 '21 at 18:06

0 Answers0