I notice that some of the star positions in GCVS and to a low precision. eg Z,UMi Here is a list of positions for Z UMi from various catalogues (some are related):
Catalog
RA
DEC
GCVS
225.504167
83.0636111
USNO-A2
225.506448
83.0634420
GSC1.2
225.50561
83.0634800
Tycho 2
225.506156
83.0635164
GSC 2.2
225.506155
83.0635160
USNO B1
225.506156
83.0635140
UCAC 2
225.506156
83.0635164
UCAC 4
225.505689
83.0635000
APAS
225.505513
83.0634970
VSX
225.506167
83.0635278
Does anyone know the origin of the stellar positions in GCVS? How can I get a copy of GCVS based on something like UCAC 4? I have written some software that refers to GCVS to return magnitudes for any GCVS star in my field but it is failing to locate some stars because of position error.
Any ideas?
Thanks,
Clive
Hi Clive,
I've run into a similar issue when writing Python code to query astronomical catalogs. In a nutshell, I wanted to combine information from multiple catalogs (e.g., an identifier from one catalog and coordinates from another), and I found a couple of ways to do this. I think that this is what you are trying to achieve. Are you using Python?
Colin
Hi Colin,
Yes, that is what I am trying to achieve. For each frame I; Perform astrometry using GSC 1.2. Extract position of up to 1000 stars to text file. Using Excel (at the moment) match the stars to photometric reference stars from AAVSO VSP (I have run VSP on many fields and have collected a good database of references in BVR &I). Return ensemble photometry for any GCVS objects in the field. So, three catalogs used, GSC 1.2, AAVSO, GCVS. My Excel spreadsheet automatically builds the WebObs report file. My plan is to take my spreadsheet and make a more processor efficient version - Python seems like a good choice. But method fails when the object is not where it is supposed to be. Your thoughts would be appreciated.
Clive
Hi Clive,
The positions in the GCVS come from a variety of sources. In most cases the positions (and other information) in VSX will be more up to date and I suggest using those. In addition there will be more variable stars to measure in your field of view. So I wonder why you prefer GCVS over VSX?
Note that in some cases you will "miss" an object with large proper motion (unfortunately VSX does not yet include this, but a future Gaia release may come to the rescue).
Patrick
Hi Patrick,
Thank you for your interesting suggestions on this. I do use VSX when I want to look up a specific star. I use GCVS because it is a self contained catalog. So, I can search it to find stars in my image and match to stars in GCVS. I have attached an image that shows my spreadsheet and how it is supposed to (and usually) works. This one field (EM Cyg) has returned data for several other GCVS stars in the field but not EM Cyg. It will reject any stars that are below S/N 30 or are too bright (near saturating) so that might be why EM is missing - or it may be out of expected GCVS position. In the case of Z UMi i have attached a chart that shows the reduced position of Z UMi in over 220 images from a single night - it also shows the GCVS position for Z UMi. I should explain that although I like Z UMi as a variable I am actually studying some other stars that I suspect are variable in this field.
Regarding proper motion - that is interesting and I had wondered about that.
Regarding VSX, is there a catalog version of this in a text file that I can use instead of GCVS?
Hi Clive,
You can download a text version of VSX from VizieR. It is updated weekly.
Patrick
Thanks Patrick,
that is exactly what I will do. I didn't realise it was available from there.
Clive
While the GCVS is a good variable-star catalog (VSX is better IMHO), there are no guarantees as to the accuracy of the positions, or even if a specified object is actually variable. In addition, I don't think it provides proper motion information, so even if the position was good in 1950, the object may be many arcsec from that position today.
What you are trying to do is really neat, but really hard. :-) For example, some GCVS stars are infrared objects and not visible. Others are 18th magnitude in quiescence, so trying to find them in a galactic plane field is difficult. Still others were novae that might have been bright in 1970, but have faded below detectable limits today.
My suggestion is to use the VSX positions for all GCVS objects. Then, if you don't have a match within your astrometric precision, then don't report anything. For the Z UMi example you gave, the positions from all of those catalogs are within an arcsec of each other, so I'm not quite sure why you used this to indicate the problem! Perhaps the first thing to check is how accurate are your positions, and how large (in arcsec) are your pixels? For the positional accuracy, do the astrometry on ~5 images of the same field, preferably on different nights. Obtain the mean position, and compare it with a good catalog, such as GAIA or UCAC4. Then graph the position error as a function of magnitude. This will tell you how accurately you can place an isolated star. You may also have to add to that something like 2x pixel size, as that determines how close you can resolve blends. These numbers will be important for your matching.
Arne
Thank you Arne for your thoughts,
I often refer to VSX. I have attached a chart that shows what my spreadsheet does. My goal is to get as much out of each image as I can rather than just the one target. I have also attached a chart that shown where my astrometry reduction places Z UMi for >220 images from a time series over one night. Each image is 120 sec. My field of view is large, ~1.4x1.1degrees.I have attached one image - I know it is not orientated as one would expect but I don't worry so long as I can extract a table of stars. I should explain that I am not obsessed with Z UMi but I am studying some other stars in the field that I suspect are variable - hence the time series. In the second chart I have included the GCVS position for Z UMi. I think the problem with GCVS is that it reports RA in hh mm ss and no greater precision than that. If it were ss.ss that would fix it. You will see that my position error (stdev) is about 10x greater in RA than Dec (as might be expected). My telescope is only an 80mm refractor so I am really only working stars brighter than about mag 14.5. Also, the mount is not great at tracking.
I will do as you suggest and do a critical review of the tolerances that I am applying. Perhaps I need a new approach. I know that I get problems with close pairs.
Thank you very much for your thoughts.
Clive
Clive
What I have done is download the newer GCVS4 database and I use it in The SkyX Pro which will calc all the changing variables based on epoch etc and it give me excellent pointing.
I can send you the text file for this database if you would like, it is fairly large though.
Drop me a email if you are interested
John R
Hi John,
Tha sounds interesting and may do exactly what I want. I am not familiar with it. I will see if I can download it and if not then I'll get back to you. Thanks you very much for replying to my post.
Clive