HIP numbers

[Malenfant]

I'm a bit confused about HIP 65011 (not 60511, sorry!):

http://www.ari.uni-heidelberg.de/aricns ... c01037.htm

ARICNS says it's Gliese 507A, about 14pc from Sol.
The HIPPARCOS catalogue says that HIP 65011 is about 42 ly from Sol, which roughly tallies with ARICNS.

But Celestia says HIP 65011 is 212.63 ly from Sol?! Is Celestia wrong on this one? Or has the parallax for this star been updated since HIPPARCOS and ARICNS hasn't noticed?

EDIT: clarified number. Sorry!

[Malenfant]

BTW, Gliese 507 supposedly has a companion that shows up on ARICNS as Gliese 507 B, or HIP 65012. This companion star isn't in Celestia at all (even when I type in the HIP number for it, nothing shows up).

[Malenfant]

A google search found this paper from 2004 which says that the parallax of HIP 65011/65012 is 75.96 mas, or 42.92 ly.

So at least up to two years ago, it seems to have been a lot closer and had a companion... (and one companion might even be a close double itself actually).


EDIT: corrected HIP number.

[ajtribick]

Does the Hipparcos catalog say that? When I searched using the Hipparcos query form the parallax given for HIP 60511 is 0.72 mas, which corresponds to a distance of over 4500 light years, which gives a suspiciously high luminosity for a G9V star.

I suspect this is one of the cases where Hipparcos measured a faulty parallax. How we should go about correcting this in Celestia is another matter, because there are two camps: those who go for greatest accuracy and those who prefer purity of data.

As far as I can tell, at the moment the data purity camp seems to be in charge, so we are currently using the Hipparcos catalog as is, preserving any errors and quirks (e.g. the use of the spectral types of the dimmer components of binary stars, check out Capella)

How to correct this for greatest accuracy while still preserving some semblance of data purity is a matter up for debate I think.

[t00fri]

...
How we should go about correcting this in Celestia is another matter, because there are two camps: those who go for greatest accuracy and those who prefer purity of data.
...

While this is true in principle, there emerged a satisfactory solution to that "controversy" recently: We agreed that we make my Perl scripts part of the Celestia source code distribution! These Perl scripts serve to "mass" extract objects (stars, dso's ...) from published, professional catalogs, are human readable and extensively commented by me.

So we now have a concise documentation of any changes and additional assumptions entering the Celestia data bases relative to the /published/ catalogs.

Neat, isn't it?

The typical Windows user of a /binary/ Celestia distribution, will presumably never notice

Bye Fridger

[ajtribick]

Ah, glad to see that this has been resolved!

I'm one of those typical Windows binary users, however if I needed those scripts (not sure if I would, but you never know), it would be available on the CVS, am I right?

[t00fri]

Ah, glad to see that this has been resolved!

I'm one of those typical Windows binary users, however if I needed those scripts (not sure if I would, but you never know), it would be available on the CVS, am I right?

Indeed and Perl is --of course-- available also for typical Windows binary users . Imagine to extract (and debug!) all those many DSO's within seconds YOURSELF....

The extensive Perl script I wrote for my deepsky.dsc file is in src/tools/galaxies in the CVS archive. It will be replaced this weekend by a much improved one (together with a much improved deepsky.dsc).

Bye Fridger

[ajtribick]

Ignore my comment about Capella... that was a consequence of me thinking I was updating stars.dat, but actually downloading stars.txt

Silly me.

[Malenfant]

Does the Hipparcos catalog say that? When I searched using the Hipparcos query form the parallax given for HIP 60511 is 0.72 mas, which corresponds to a distance of over 4500 light years, which gives a suspiciously high luminosity for a G9V star.

Erm, the HIPPARCOS Vizier search form (top button there) gives the parallax of 65011 as 75.96 mas, or 0.07596 as. Distance is 3.26/parallax, so that's 42.9 ly.


EDIT: Sorry folks, I must be numerically dyslexic or something, I always end up muddling up numbers when I'm transcribing them. The star i'm looking for is HIP 65011, not 60511. Sorry for the confusion.

So yeah, 60511 is at 0.72 mas, but that was the wrong star I pointed you to .

[symaski62]

cel://Follow/HIP%2060511/2005-08-27T20: ... 05911&lm=4

hip 60511 => (1.3889 Kpc)

[ajtribick]

I see what is causing the problem - HIP 65011 is located very close in the sky to HIP 64979, a star which has a parallax of 15.34 mas (=212 light years). The code that generates stars.dat is designed to avoid splitting up binary stars, so works on the assumption that two stars located very close together in the sky are in fact a true binary, so the parallax of one of the stars overrides the catalogue value.

This is what has happened in the case of HIP 65011 - the generating software has detected the apparent binary, and has assigned HIP 65011 the parallax of HIP 64979.

[Malenfant]

I see what is causing the problem - HIP 65011 is located very close in the sky to HIP 64979, a star which has a parallax of 15.34 mas (=212 light years). The code that generates stars.dat is designed to avoid splitting up binary stars, so works on the assumption that two stars located very close together in the sky are in fact a true binary, so the parallax of one of the stars overrides the catalogue value.

If that's true, that sounds like a really bad way of doing it. Surely the code should link stars that have the same parallax and RA/Dec, not that just have a similar RA/Dec?