Reference Star Question

[Keith Graham]

Hi All,

I have a question regarding reference stars.

I realize that the best reference stars to use are those of spectral classes OV, BV, or AV. But I also see that such spectral classes are not always in optimum position with respect to the target. I also see that spectral/luminosity classes other than the above are in the Miles library and other libraries. I even see spectral classes listed down through M.

I have taken spectra of stars of varying spectral/luminosity classes and I have found that their ir/extinction corrected profiles are very similar. I also realize that O,B,and A stars are best to use since their continua have fewer lines than other classes and that luminosity classes other than V may be variable. With this in mind, my question is:

If the closest Miles library listed reference star to a target is something other than OV,BV, or AV (i.e. a G2III star) can this star be used to generate the IR profile that will be used to correct the target even though the target might be of a totally different spectral/luminosity class?

I am asking this because it appears to me that only reference stars in close proximity to the target should be used for ir/extinction correction, but OV, BV, and AV stars are not always found close enough to the target to establish a reliable correction. I am assuming that those many other spectral/luminosity class reference stars are in the catalogs for a reason. So if only OV, BV, or AV stars should be used for ir/extinction correction, why are so many other classes represented in the catalogs?

Cheers,

Keith

[Francois Teyssier]

Dear Keith,

I use this Excel Speadsheet : http://www.astronomie-amateur.fr/Docume ... inder.xlsm
Using BSC
Usually I try to get a reference star at 5° max from the target

Cheers

François

[Christian Buil]

I confirm, this Excel Speadsheet from François is a must. A very good help for select A/B reference stars (choose low interstellar absorption objects) + Pickles for the theoritical profile.

Christian

[Keith Graham]

Hi Francois and Christian,

Thank you for the spreadsheet. I had already downloaded this spreadsheet a few days ago and have seen the many reference stars it contains. But I am still wondering why the Miles library contains so many stars other than OV,BV, and AV if only those classes are useful as reference stars?

Your response brought another question to mind. I was under the impression that it is best to use a Miles star as reference so one could apply the Miles spectrum to the same star taken by the spectroscopist. The Pickles library has only sample spectra from each class that will be applied to a reference star of the same class. I would think that the ir correction profile derived from an exact match (using the Miles spectrum) would give a more accurate correction than simply applying a Pickles same class spectrum to a reference star spectrum. Is this a correct assumption? If so, then if the Pickles spectrum is used, is the resulting correction profile “good enough” to give an accurate correction to the target?

Cheers,

Keith

[Robin Leadbeater]

Hi Francois and Christian,

Thank you for the spreadsheet. I had already downloaded this spreadsheet a few days ago and have seen the many reference stars it contains. But I am still wondering why the Miles library contains so many stars other than OV,BV, and AV if only those classes are useful as reference stars?

Hi Keith,

The MILES database http://miles.iac.es/ (and the Pickles library of standard spectra) was not produced as a database of reference spectra for spectral calibration. It was produced to generate synthetic spectra of stellar populations for comparison with measured spectra of galaxies for example. This is why it has a wide range of spectral types.
We are using it for a different purpose so only the (non de-reddened) hot star spectra are useful to us as reference spectra.

Cheers
Robin

[Robin Leadbeater]

Thank you for the spreadsheet. I had already downloaded this spreadsheet a few days ago and have seen the many reference stars it contains.

Hi Keith,

Note these are currently strictly only "Potential" reference stars as, unlike the MILES stars, we do not have calibrated spectra for them, only their published spectral type. This means we have to assume that they are identical to the standard (eg Pickles) spectrum for that spectral type, which may not always be true.
This is why I suggested a project to produce calibrated spectra of these stars using known standard eg MILES stars. This would extend the database of reliable reference spectra beyond those in MILES

Cheers
Robin

[Keith Graham]

Hi Robin,

Your explanation concurs with my original understanding of the Pickles spectra. That is why I am now a bit confused regarding the suggestion that I use the spreadsheet stars as reference stars and Pickles to obtain a profile to be used to correct the target. If I understand correctly, I should use Miles as a first choice. If a reference star close to the target is not available in Miles, then I would find a reference star on the spreadsheet that is close and use the Pickles spectrum to create the profile used to correct the target. Does this sound right?

If the above is correct, it appears to me that when a Miles star is not close enough to the target for a reliable correction profile, then a closer spreadsheet star should be used with the Pickles spectrum as this would be better for correction than a Miles star that is too far from the target . Does this sound right?

I am just trying to understand when it is better to use the Miles library and when it is better to use the spreadsheet/Pickles option.

Cheers,

Keith

[Robin Leadbeater]

I am just trying to understand when it is better to use the Miles library and when it is better to use the spreadsheet/Pickles option.

Keith

My view is that is an impossible question to answer as you are comparing apples and oranges ie comparing the (approximately known) potential error due to atmospheric extinction by using a MILES star at a different altitude, with the unknown (but hopefully small) error from using a star at the right altitude but with an only approximately known spectrum. Personally I would try to measure both a Miles star and one nearer. If they then give conflicting results you can always go back on another (photometric) night and measure the two reference stars relative to each other but at the same elevation to resolve the difference.

Robin

[Robin Leadbeater]

......you can always go back on another (photometric) night and measure the two reference stars relative to each other but at the same elevation to resolve the difference.

Having done this, you have then of course turned a star of uncertain quality chosen from the spreadsheet into another reliable reference star with a known spectrum, which you (or anyone else) can then use for any future measurements of the same or nearby target.

Cheers
Robin

[Keith Graham]

Hi Robin, Francois, and Christian,

Robin, I do understand everything you have said – it all makes perfect sense to me. Correcting a target with the same star in the Miles library is certainly the best way to go. But this brings me back to earlier emails from Francois and Christian regarding the use of the Excel spreadsheet with Pickles. So perhaps I should re-phrase my concern to Francois and Christian.

Francois and Christian, you mentioned in an earlier message that you use the spreadsheet and Pickles to easily find a reference star. The spreadsheet certainly does offer a multitude of reference stars that makes it easy to find one in the vicinity of the target. But how can one be certain that a Pickles spectrum will give an accurate correction of the reference star? Can you please explain how you use this combination? If the Pickles spectrum is potentially at an altitude quite distant from the reference star, how will this result in an accurate extinction correction profile? Are Pickles spectra corrected above the atmosphere as are the Miles spectra? Then again, why not use a Miles star of the same spectral/luminosity class as the reference star to get the correction profile instead of Pickles? We know that the Miles have been corrected for above the atmosphere, so would not the Miles stars create a potentially more accurate correction profile?

Along the same line, I noticed that the NOAO Indo-us library contains a number of OV,BV, and AV stars. Are these reliable stars to used for correction purposes?

Cheers,

Keith