Bill Gray Oct 28 1:53 PM
On 10/27/2015 06:05 AM, Marco Langbroek wrote:
> Since May this year I have put 2010-050B and 2013-070B (Chang'e
> 2 & 3 rb) on my regular observation schedule, using the 'remote'
> telescopes Q65 Warrumbungle in Australia and G68 SSON in Arizona...
Which has been _very_ helpful. I've not updated the pseudo-MPECs
for a while (need to do that!), but I've been getting fresh ephemerides
in for the MPC's DASO Page. That also helps to ensure that when these
objects get stumbled across, they can go directly to "not a minor
planet", rather than get everybody briefly excited. There's another
possible use for the data, which I'll get to below.
> Combination of MPC 461 obs with Q65 obs allowed me to get a good
> value for the current tumble periodicity of 2013-070B:
>
> http://sattrackcam.blogspot.nl/2015/09/the-tumble-periodicity-of-change-3.html
Now, _that's_ interesting...
As I mentioned previously, I've wanted to come up with a better
physical model for the forces acting on bits of junk such as this.
However, I've been a little reluctant to do so, thinking it might well
be a wild goose chase. If the object is still outgassing, and sometimes
randomly re-orienting itself and accelerating a bit, how the heck can I
hope to model that?
But your results indicate that the spin rate, at least, is quite
stable. So if I model the way solar radiation pressure pushes a tin
can around (these objects are basically big tin cans, with highly reflective
sides and not very reflective ends), I might conceivably be able to get
decent results. It _would_ involve fitting a lot of additional parameters,
such as the spin axis of the object. Fitting that many parameters is a
little iffy; you run into a famous comment of John von Neumann's about
being able to model an elephant if you have enough parameters, and being
able to predict the elephant will fly if you add one more parameter. But
it might be worth trying.
If I do that, determining all the extra parameters will require an
unusual amount of data. At which point, I'll be glad to have a few test
cases where the objects have been observed at occasional intervals
(instead of the common case where they get lots of observations at
only a few points).
> This led two (radio-) observers to inquire whether the Chang'e 5 r/b
> was ever seen by asteroid survey telescopes.
I'd wondered about that object... the other Chang'e boosters were found
quickly, as were the DSCOVR and LCROSS boosters, and I expected the
Chang'e 5 booster to show up.
Your elements suggest a possible answer. The other objects are all in
high orbits, or very elliptical orbits, or both. Either way, they
have times when they aren't moving all that fast. The fast-movers get
thrown out by the surveys as "probably just junk". But you've got
the Chang'e 5 booster with eccentricity 0.614 and in a 13.6-day orbit.
The inclination is about 62 degrees, which might help keep it in
places surveys don't look very often.
Still, I'm glad to hear of this; if we come up with another
unidentified piece of junk and can backtrack it a ways, it's possible
we can link it up to this information.
> But TLE's are not very adequate for objects like these anyway).
True. I've found that fitting TLEs to the SGP4 model (the one usually
used for short-period objects) works reasonably well on these high objects;
you can get accuracy of a few dozen kilometers, usually, for a TLE
covering a one-day period.
But the default is that TLEs for these objects must be fitted to SDP4,
a model that works wonderfully with 24-hour and 12-hour orbits, but isn't
great for longer-period artsats. (At least, not with my code. I'm trying
to figure out if it's just something I'm doing wrong. But I've not seen good
TLEs using SDP4 anywhere else, either.)
-- Bill