A selection of minor hacks for space wizards, comprising:
- range_fit.py: Take an imperfect TLE and massage it to fit radio ranging data
- hpredict.py: Compute a schedule of ground station passes, suitable for feeding the satellite or certain ground stations
- azelrange.py: Annotate log files with the angle and distance to the sat.
- Orbit differences.ipynb: An IPython notebook illustrating how you can quantify the difference between two TLEs (illustrating the volatility of Space Track's issued orbits)
- tle_util.py: TLE parser library used by the above.
- scrape_track.sh: Pull the latest TLE from JSpOC.
Sorry, no distutils yet. 'Orbit differences' is an ipython notebook, so install ipython to use it.
You'll want (probably via pip): numpy, scipy, matplotlib, pyephem
Also download https://pypi.python.org/packages/source/s/sgp4/sgp4-1.1.tar.gz and extract it into this directory.
If you want to use scrape_track, edit it to include your Space Track credentials.
Obtain a TLE that you wish to improve. This could be a prelaunch prediction from a launch provider, an out-of-date Space Track TLE, etc. Put it in a file. If it's in 3-line format, great. Otherwise prepend a line with the name of the spacecraft. If you want, you can have more than one satellite in the file, but they must have unique names. e.g.
./scrape_track.sh (...) cat norad.tle 0 DOVE 3 1 39429U 13066P 14036.66224386 .00001774 00000-0 34640-3 0 646 2 39429 097.7753 109.2910 0154894 310.7235 048.0617 14.56764061 11098
Obtain some LST ranging data. This can be obtained via lst_ctl -R. Save the output to a file. Remember which ground station the ranging was run from.
You can collect one or several passes worth of data, from one or several ground stations. Put each set in a separate file. Each pass should have at least 20 points to be useful.
One pass is sufficient to fit the mean anomaly; 2 to fit the mean motion; 3 or 4 passes are suggested if trying to fit all the orbital elements.
Run range_fit:
./range_fit.py <gs=rangefile> [gs=rangefile] [optional args] ...
TLEfile is the catalog containing one or more TLEs. You can only improve one at a time even if there are several in the file. The file will not be modified by range_fit.
Each set of ranging data is specified as gs=rangefile where 'gs' is the name of the ground station (their names and locations are hardcoded in the source) and 'rangefile' is the name of a file containing the range data as output by lst_ctl.
If the TLE catalog contains more than one satellite, you must tell range_fit which satellite you're interested in with the '-s' argument.
By default, range_fit will only try to fit the mean anomaly. You can tell it to fit all the parameters with the '-f' flag, or tell it exactly which ones to fit with other flags (range_fit.py -h for full info)
The output should be pretty self-explanatory. If you can get the RMS error down to < 5 km, that's pretty good (LST ranging RMS noise is around 3 km). But beware the ill-conditioned problem - if you fit all parameters with only one pass' worth of ranging data, the RMS error will be small but the orbit will probably be bogus.
e.g.
./hpredict.py -c norad.tle monterey,spokane,timbuktu=25
Lists passes for 3 ground stations with Monterey and Spokane having default elevation mask of 8 degrees, Timbuktu a higher mask of 25 degrees