[Kstars-devel] Wrong Epoch in Star Data (Re: KStars for astrophotography)

[Massimiliano Masserelli]

Il 23/10/2012 18:33, Akarsh Simha ha scritto:

> I'm very sorry I didn't follow up on this matter earlier. Your list
> mail, if I note correctly, asked for what stars to test.

Actually, IIRC, I mentioned some stars as evidence.

> It already helps that you know that stars.dat has problems.

Yep. Problem is: not all stars in stars.dat have recognizable catalog 
numbers. I _suppose_ (and from inline docs I think we may safely assume) 
that all of those data are from Hipparcos catalogue. I tested more than 
a couple of those with catalogue names, and their data are consistent 
with my theory.

> So there are three star catalogs merged together in KStars:
> 1. Some bright star catalog I don't know about -- possibly, Hipparchus
> 2. Tycho-2
> 3. USNO NOMAD's brightest 100 million.

Yep. The problem I noticed is for sure in HIP catalogue, since data from 
that mission are expressed as J1991.25 (from memory, anyway not J2000).

> 1. Stars from which catalogs have the problem? (all 3? only the bright
> stars?)

I tested the problem searching in kstars using names found on stars.dat 
(which I think is no more accessed directly from the application). So, 
brighter stars, for sure.

> 2. Could you name some stars that have the problem? (or give me their
> RA / Dec, so I can find them as well)

Please forgive me if I'm not able at the moment to produce a satisfying 
amount of data. I quote my apr 2011 mail:

In particular, Arcturus (alp Boo) has RA 14 15 40.35 Dec 19 11 14.17
J2000. If you have a look at simbad
(http://simbad.u-strasbg.fr/simbad/sim-basic?Ident=arcturus&submit=SIMBAD+search 
) you'll see it should have RA 14 15 39.6720 Dec +19 10 56.677. I then 
looked at Hipparcos via Vizier
+(http://vizier.u-strasbg.fr/viz-bin/VizieR-S?HIP%2069673) and found 
that it shows the same coords as kstars, only in J1991.25 epoch (last 
batch of data on page, as follows):

...
RAhms   14 15 40.35             Right ascension in h m s, ICRS 
(J1991.25) (H3)
(pos.eq.ra;meta.main)
DEdms   +19 11 14.2             Declination in deg ' ", ICRS (J1991.25) (H4)
(pos.eq.dec;meta.main)
...


Many others are in same condition (try with brighter ones), save a few 
such as Albireo. I think this depends on position and proper motion.

> 3. Are all stars from a given catalog affected?

 From what I've seen, yes.

> 4. Do you have any free GPL-licensable replacement catalogs (you
> mention something about libnova)

I have the whole stars.dat with RAs and Decs corrected to J2000 via 
novas-c library (I _did_ mention libnova, but my memory was faulting at 
the time. :) ).

You can find it here:

https://dl.dropbox.com/u/35150102/stars_j2000.dat.bz2

I verified a few with VizieR, and results seem accurate enough. 
Obviously, I wasn't able to test "unnamed" stars, and there are more 
than a couple (more than a million, actually).

So: if there's an easy way to convert textual stars.dat into new format 
used by kstars, I think those converted data are more realistic than 
those we have. If not, maybe it's time to think about integrating a new 
catalogue for stars fainter than 8.0, hopefully this time with J2000 
coords. :)
-- 
Massimiliano "Max Negro" Masserelli