HD000319 Miles

[Andrew Smith]

Hi, I noticed when doing the "instrumental response" for my echelle that the Miles Library A1V star HD000319 has a significantly different continuum level to other A1V stars in the Miles Library and to HD198001 in the NOAO library which is also ~ A1V. While the others show small differences the difference with HD00319 reaches 20 at 420nm when normalised in the standard way in ISIS.

I may be mistaken or doing something wrong but of I am correct it is best to avoid using this star as a template for Instrumental response.

Regards Andrew

[Robin Leadbeater]

Hi Andrew,

Possible causes are extinction and metalicity, neither of which are taken into account in the generic Pickles spectra.

As long as the MILES spectrum is a true representation of the actual star spectrum though (which it should be of course) then it should be ok to use it. (This shows the advantage of using actual spectra over the generic Pickles spectra ) As a cross check you could generate a calibrated spectrum of it using another MILES star for calibration to make sure it has not changed since it was measured for the MILES database.

Concerning the NOAO database, http://www.noao.edu/cflib/
I have been looking in more detail into how the NOAO spectra were generated using the paper referenced there
http://www.noao.edu/cflib/Valdes.ps
and I believe there is a problem using them for instrument response calibration. It appears that the continua of these stars was generated using the synthetic spectra from the Pickles library rather than the actual continua of the measured spectrum and so are not representative of the true spectrum of the star. See page 10 of the paper
As far as I can see this means using NOAO spectra for calibration is no better than using Pickles generic spectra. Perhaps others can check my logic here.

Cheers
Robin

[Andrew Smith]

Thanks for the reply Robin, but these are both in the Miles Library (unless I have gone completely mad). I agree there may be good astrophysical reasons why they differ but unless you observe a Miles Library star and then use the library spectra of that star to calculate the "instrument response" you could well be in error! It seems to challenge the very basis of using a generic A star with a Miles Library A star of the same spectral type.


Regards Andrew

[Robin Leadbeater]

Hi Andrew,

There is indeed no point using a MILES star if you are then going to compare it with a Pickles spectrum -why would you do that ?

Just because two stars are classified as the same spectral type does not mean they will have identical spectra. The MK classification process makes no reference to the shape of the continuum, only the relative strength and width of various lines. Metalicity and extinction both significantly affect the continuum shape for a given classification and classification itself is a pretty inexact science. Here is an example of two MILES stars of the same spectral type but different metalicity which shows exactly what you are seeing
http://www.spectro-aras.com/forum/viewt ... city#p1703
Also where do the classifications for the MILES stars actually come from? Are they reliable ?

Any use of calibration stars with a published spectral classification compared with Pickles stars as the reference spectrum will be approximate at best and it is generally a good idea to do at least some rough checks on any star used in this way to make sure it is at least likely to be "typical" of the spectral type and does not show high IS extinction or abnormal metalicity. Using the MILES spectra gets round all this because there we have actual measured (and as far as we can tell reliable) spectra for the star so the spectral classification assigned to it is irrelevant. I suspect if you measure the spectra of these two stars, You will find they are actually different as the MILES spectra suggest even though they have been classified as being the same spectra type in the MK system

Robin

[Martin Dubs]

Hi Andrew,

the Miles catalog is a collection of quite different stars. The same classification does not mean that all spectra look alike. there are other parameters such as metallicity, interstellar absorption and effective temperature which may vary quite a lot. Notice tha the spectral class is defined by the appearance of the spectral lines and not by their effective temperature. I include a list of these parameters, from which you can see by sorting the catalog after spectral class, that the effective temperature of HD000319 is particularly low for an A1V star, which explains the reddish appearance. At first I suspected interstellar reddening, but this seems to be of minor importance in this case (E(B-V) = 0.012). That is why it is preferable to use a actual star from the Miles database for calibration and not a synthetic spectrum such as Pickles for spectrum calibration. I agree with Robin that the NOAO database is not very reliable for calibration, there are several examples of spectra quite different from the corresponding Miles spectra. However the NOAO spectra may be useful for spectra calibration requiring a larger wavelength coverage. Notice also that the reddening has to be included in the spectra for calibration (as is done in ISIS, but not in the original Miles database). There has be an intense discussion about this in this forum about two years ago (search for dereddening or similar)

Regards, Martin

[Robin Leadbeater]

unless you observe a Miles Library star and then use the library spectra of that star to calculate the "instrument response" you could well be in error! It seems to challenge the very basis of using a generic A star with a Miles Library A star of the same spectral type.

That is exactly how the MILES stars should be used (ie dont mix MILES and Pickles spectra). If you have to use a star which does not have a reliable actually measured spectrum then the best you can do is assume it is typical (having done some precautionary checks eg Francois T published a list of low extinction A/B stars) and compare it with a generic eg Pickles spectrum, not an arbitrary MILES spectrum. What we need are more actual "MILES" quaility star spectra to give better coverage, something I tried to get AAVSO interested in doing, like they do for comparison stars for photometry

Robin

[Andrew Smith]

Thanks Robin & Martin.

I had forgotten the previous debate about calibration stars and the need for more "standards".

I am not sure where Pickles came in as I was not proposing using them at all. I was seduced however, by the simple "choose any A or B star and then compare it with an similar star from the Miles Library" approach to "instrumental response". I will have to redo mine.

Thanks for taking the time to respond. Your list will be most helpful Martin.

Regards Andrew