RE: Question about the age of available TLEs

From: Chris Peat via Seesat-l <>
Date: Tue, 13 Feb 2018 08:51:56 +0100
Hi Gavin,

My apologies - it seems I overlooked another source which explains the conundrum. In addition to Space-Track and the Mike McCants file, we also use the following source for amateur radio satellites because they are not all listed on Space-Track (presumably because they were launched with one or more classified payloads);


As you can see from the file name, this appears to come from NASA originally. What I did not realise, is that the ISS is included in this list because it is of great interest to amateur radio enthusiasts. The epoch of the ISS elements is often ahead of those on Space-Track. Our import script always favours elements with the latest epoch, so these elements overwrite the Space-Track ones.


I hope this clarifies things for you.


Best wishes,

Chris Peat



From: Dr. T.S. Kelso [] 
Sent: 12 February 2018 22:05
To: Gavin Eadie via Seesat-l
Cc: Thomas Goodey via Seesat-l; Chris Peat
Subject: Re: Question about the age of available TLEs




CelesTrak does get TLEs from Space Track and we check every two hours for new data, so it does seem unusual to see that kind of gap. I regularly check to ensure all data is getting updated, as well, but perhaps there was a hiccup at the time of your comparisons.


With regard to the NASA TLEs, where they fit a TLE to their planning ephemeris, we actually provide that data on CelesTrak, as well. It is listed under the Supplemental TLE section, where the data is produced from operator data, rather than by JSpOC using SSN observations. In most cases, we fit the operator data using SGP4 as implemented in STK (which uses Dave Vallado’s code), but in some (e.g., ISS and Orbcomm), we get that data from the operator. You can find the supplemental TLEs at


For the ISS data, we simply parse out the TLE data from the NASA ISS Trajectory Data. One would expect NASA’s predicted locations to be much more accurate than propagating a current TLE over the same period, since they use high-fidelity models of the area and mass of the ISS, attitude of the ISS and orientation of the solar panels, along with high-fidelity atmospheric drag models. And they would also incorporate any planned maneuvers, too. So, these supplemental TLEs should be a more reliable source of information for planning, whether that is for researchers having experiments onboard or someone wanting to go out and observe it.


Please let me know if you have any questions on the supplemental TLEs.  - TS


Dr. T.S. Kelso


E-Mail: <> 

On 2018 Feb 11, at 23:30, Chris Peat via Seesat-l <> wrote:


Hi all,

After following this discussion, I just want to clarify where Heavens-Above
gets the TLEs used for predictions.

For the classified objects, we use Mike McCants file.

For all other objects, we use Space-Track as the source, using a script
which checks for updates every 4 hours. If the TLEs for the ISS are the same
as on the NASA page, this is pure coincidence.

Best wishes,


Seesat-l mailing list


Seesat-l mailing list
Received on Tue Feb 13 2018 - 01:53:09 UTC

This archive was generated by hypermail 2.3.0 : Tue Feb 13 2018 - 07:53:09 UTC