• Tom Johnson

Reverse Engineering SSN Practices and Performance

Just wrapped up the 3rd International Conference on Space Situational Awareness (ICSSA) in Tres Cantos, Madrid. We presented our paper "Reverse Engineering Space Surveillance Network Practices and Performance from Two Line Element Sets." We can learn a lot from looking at TLE epoch times. The full paper is available at

It turns out OneWeb and Starlink TLEs are often epoched on the integer hour. Mostly when still doing a low-thrust orbit raise. The graphs below show the semi-major axis rate (via a linear fit to each RSOs TLEs) as a function of the launch date. RSOs on the same launch date are "jittered" to try to provide more visibility (otherwise the markers stomp on each other). The plots reflect the status as of Dec 21, 2021 (yes, already out of date!).

OneWeb Constellation

Starlink Constellation

Here's what you get when you plot a lot of TLEs at epoch ...

Here are some histograms where we've binned the points in az/el space and rendered in polar coordinates. Azimuth on the angle axis, elevation on the radial (with zenith in the center at 90 degrees). Bins are color coded on a logarithmic scale.

We'll start with the big radar at Eglin AFB. Not sure what we mean? It's this thing Note the nice "fences" on the south side at low elevation angle.

Eglin phased array radar histogram.

Here's the same info, but now we've looked at the whether each TLE was ascending (positive elevation angle rate) or negative (descending elevation angle rate). If you add 1 for the ascending passes and subtract 1 for the descending passes, then each bin value indicates the predominate trend. Now we see the low elevation fences showing that the lowest el "blue" fence is where the RSOs are "exiting the fence" as they descend. Or conversely, rising through the fence as indicated by the red band directly above it.

Here's the new Space Fence radar at Kwajalein. Also a phased array, but the faces are pointing straight up. You can see it at

Space Fence radar histogram.

More details in the paper ...

109 views0 comments

Recent Posts

See All