Open fliker09 opened 4 years ago
Thanks for adding your first issue to Stellarium. If you have questions, please do not hesitate to contact us.
I wonder how accurate our satellite position models are, to start with. Has anybody made a thorough analysis and comparison, using the same TLE, between Stellarium's positions and some dedicated service? I assume prediction of satellite occultations require 0.1 arcsec positional accuracy or better.
I've spent on this last few weeks... It's indeed a complicated subject! And yes, Stellarium is capable to calculate it precisely, but you need to provide it proper TLE data for that. Data downloaded from Celestrak doesn't cut it - you need to provide data with epoch close to the event! The only source for that, for ISS, is NASA, they are providing TLE for epochs ahead. Good lord Stellarium allows to use custom URLs, so I upload the tailored TLE on my server and point Stellarium to it. Simulations have shown good agreement! Though there is another issue... Orbital track is not aligned with ISS itself, opening a bug report on this right now.
See also #1130
(@fliker09 can you provide the URL of the NASA TLE future epochs? thx!)
@axd1967 regarding #1130 - interesting! Similar situation with ISS transits... As for the TLE - there is an interesting story about it... NASA has retired the wenbsite where it was publishing the future epoch TLEs for ISS :-| They provide instead CCDS OEM records. I started digging into how to convert this stuff into TLE but I am not enough smart for this :-D Luckily the author of ISS Transit Prediction app accomplished that! But he did even more - he discovered somehow that Celestrak has its own ISS future epochs TLE source! Here it is: https://celestrak.com/NORAD/elements/supplemental/iss.txt It requires editing of the headers, but that's easy (hopefully will push it soon to my personal server).
Hi, what about this sub-second time steps? I'm using stellarium to catch ISS too and sub-second time steps can help. A different approach should be to draw the path of an object (Eg: ISS) in a given time interval, still exist that feature? Thanks
This feature would allow to properly visualize ISS transits (Sun/Moon/planets), as the events of this type are very fast, most of the time below a second. If you want you can hide this feature behind a switch in the configuration. Currently, the only planetarium capable to do it is KStars (0.1 seconds step is possible), but it has its own set of issues. Also, 0.1 is not enough for planetary transits, would be cool to get 0.001 time step! Let me know if any additional details are required!