RE: FW: Forced reentry of Cosmos 2495

I am now confident that the fireball seen over the western U.S. on 2014 Sep 03 near 04:33 UTC was not Cosmos 2495
(14025A / 39732). That became improbable when it was recognized that the slow-moving fireball seen more than 10 h
earlier over Western Kazakhstan and Orenburg region (Russia) must have been Cosmos 2495. For it to have been the cause
of both fireballs would have required an incomplete de-orbit burn.

I tested that hypothesis by creating a TLE at the approximate time of the de-orbit burn (2014 Sep 02 17:40 UTC), that
when it reached Kazakhstan ~45 min later, would have just reached the altitude when re-entering objects become
self-luminous, which I normally take to be 96 km. I then fit a decay term by trial and error using Satevo, that would
have resulted in re-entry on Sep 03 near 04:40 UTC, several minutes after the U.S. fireball sighting.

The resulting de-orbit TLE, shown below, passes over Western Kazakhstan at about 18:14 UTC, at about 96 km altitude:

1 70000U          14245.73611112  .17600000  00000-0  80777-4 0    05
2 70000  81.3770 299.0950 0128000  52.4336 214.0000 16.37300000    00

This is the final TLE before re-entry of the above orbit, propagated by Satevo:

1 70000U          14246.17706951 1.15185463  27284 2  16568-3 0 90009
2 70000  81.3675 298.4863 0025344  50.6326 309.5915 16.65049125    77

It passes over the area of the fireball sighting in the U.S. on Sep 03 near 04:24 UTC, about 10 min earlier than the
fireball was seen. This represents the best-case scenario, and it depends on favourable, but unrealistic assumptions. 

No effort was made to estimate the decay term from the object's actual ballistic coefficient. Had this been done, I
suspect the decay term would have been significantly greater, causing the object to decay well before it reached the
U.S. Based on the videos of the Kazakhstan fireball, the re-entry probably had progressed well below 96 km. Had a more
realistic altitude been chosen, the case for survival to reach the U.S. would be further weakened.

My preliminary analysis of the possibility that the U.S. fireball was due to uncatalogued debris jettisoned just prior
to the de-orbit burn tends to support, but does not prove that hypothesis. 

I propagated the last known TLE to the approximate time of the de-orbit burn, and assumed that the orbit of the
jettisoned debris would have differed only in rate of decay. I then fit a decay term by trial and error using Satevo,
that would have resulted in re-entry on Sep 03 near 04:40 UTC, several minutes after the U.S. fireball sighting:

1 70001U          14245.73611113  .13400000  00000-0  78682-2 0    01
2 70001  81.3770 299.0950 0034344 130.9033 134.4989 16.17169265    09

This is the final TLE before re-entry of the above orbit, propagated by Satevo:

1 70001U          14246.18294601 2.42452645  43066 2  41010-2 0 90000
2 70001  81.3681 298.4965 0008820 129.1327 230.9457 16.47601197    72

This orbit passed over the region of the U.S. fireball sightings about one minute too early. Since the underlying model
of TLEs (SGP4) is not designed to propagate the final descent accurately, I made lower altitude versions of the above,
closer to the likely altitude of the fireball. I was able to obtain a reasonably good agreement with the trajectory data
reported by the Cloudbait observatory, albeit about one minute early. The result is sufficiently close to warrant a more
rigorous analysis. The meteor explanation is not ruled out.

It is necessary to review the previous eight Kobal't-M launches to learn as much as possible about the debris jettisoned
prior to the de-orbit burn. Of particular interest is the typical time of jettison relative the de-orbit burn, the
delta-V of separation, and the ballistic coefficient. The number of pieces of debris catalogued has ranged between zero
and three. The zeros may have been the result of decay before cataloguing could occur.

Ted Molczan


_______________________________________________
Seesat-l mailing list
http://mailman.satobs.org/mailman/listinfo/seesat-l