Grant,
thanks for explaining ...
granthutchison wrote:Fridger:
...
RECONS contains no orbit data.
YES and NO
. Sorry I should have been more explicit.
The primary/secondary mass ratios are among the most crucial parameters for the (elliptical) orbit geometry, I guess you agree
. These are gotten, however, in RECONS via a quantitatively most dubious
mass-luminosity relation, using the
absolute magnitude values from the Gliese catalog as an input! Of course, RECONS/Gliese does not contain any standard orbit parameters...
In the past, I have prepared
dozens of such mass-luminosity plots, trying to reduce (in vain) the number of undetermined entries in my binary data. I suppose, I don't have to emphasize the extremely large uncertainties that such relations imply for
individual star systems. In the sense of
statistical averages such relations may have some use, but NOT for individual stars (being often not even of main sequence type...). Here, fluctuations may be immense, depending on parameters, like metallicity etc. My binary orbit sets are largely of
astrometric type, which is quite a different class of reliability, as you know.
Because I am aware of these inherent uncertainties since a long time, I have preferred to leave some of the respective entries with an attached '?'.
Other orbits have been checked against existing orbit plots just as you describe. We had some discussion about this years ago, when nearstars.stc was first created, and I recall offering examples from both Sirius and 61 Cyg. I made other checks at the time, orbit by orbit. Why would you imagine that I wouldn't have done this?
Oh good you did it. Sorry this must have slipped my mind... Such tests are so important (and fun, too) that I would have expected you to show the results here in the forum.
What would you like in order to be reassured about hierarchical systems? A check that the orbital definitions of "level 2" stay well within the Hill sphere defined by "level 1"? Or do you want a full dynamical simulation before accepting any multiple systems that have defined orbits from different sources?
I am happy, if the
a posteriori fitted relative distances are shown to satisfy a hierarchical setup in all cases... Users are simply confronted with a set of numbers in nearstars, whithout any comments about such important underlying prerequisites. Actually I am aware that most multiple star systems tend to be pretty hierarchical, so it's really a matter of presenting explicit cross checks to be really sure.
Let me finally point out that quite a few of your alternative names are
incompatible with the syntax accepted by SIMBAD (e.g Luyten xyz-w, Lacaille 9352, ...). Achieving total SIMBAD naming consistency would be another goal I would be very happy to see realized...It is very straightforward to do this using the SIMBAD scripting service. All you got to do is to produce a listing of your objects ( 1 column), to define a desired output format and then to submit. The resulting lists can easily be reduced to a few name classes.
Andrew has also accepted this suggestion of mine for his HIP star analysis with Perl. My Perl-based DSO database (galaxies, globulars) is SIMBAD compatible. My binary star database is in the process of being
completely revised and improved considerably, including direct Perl-based ftp downloading and merging of different standard catalogs. Of course SIMBAD compatibility holds there, throughout. This is however all part of an ambitious ongoing project of mine and probably not relevant for Celestia proper, due to quite different views between ChrisL and myself.
Fridger
