Affiliation
American Association of Variable Star Observers (AAVSO)
Tue, 12/16/2014 - 23:33
Over the past year I have been selecting first-order extinction stars for PEP observing from the "northern" list in Appendix A of Henden and Kaitchuck. There are a number of stars in the list which SIMBAD considers to be variable (eg: 21 LMi, 60 Leo, rho Vir, 24 CVn, sig And, 32 Per, B Per). Whether these stars were considered variable at the time of second publication of H&K (1990), I do not know. My question is: how much variability is sufficient to disqualify a star as a reliable reference for the Hardie extinction measurement?
Tom
Hi Tom,
You should not accept Simbad's classification of these stars as "variable"; always go to VSX. At best, they are listed in VSX as suspected variables, with unknown ranges. At the time the book was written, these were considered standard stars, and I see no reason to classify them otherwise.
Even if they are slightly variable, they won't affect extinction calculations unless they vary significantly over a few hours. Extinction is on the order of 0.25mag/airmass, and almost all of the "weight" on extinction is when the star is at high airmass, where it is near the horizon and changing airmass rapidly. The star would have to vary several tenths of a magnitude in an hour or so to throw the extinction calculation off much, and it will be very obvious when looking at the graph of magnitude vs. airmass (which should be linear).
Arne
Arne:
I think you are describing the consequences of variability when measuring a single star over the course of hours. What about the Hardie method where one is measuring multiple stars at different airmass in quick succession? My concern is that this method depends upon knowing the "true" magnitudes of these stars, and if the stars are varying, how much variation from the standard magnitude can be tolerated?
Hi Tom,
The same goes for the Hardie method. Remember that it is the airmass difference that is important - you are trying to measure stars near the zenith (X=1) and close to the horizon (X=2-3) if possible, to get the best fit. The extinction coefficient is 0.25. So if one star is at X=1 and one star is at X=2, the high airmass star should be extincted by 0.25magnitudes. You are also doing this for several pairs, not just one. So slight variability just isn't going to affect the result. Usually you can measure extinction to about 10 percent since it isn't a strong function. For a single pair, and assuming both stars are variable and randomly so, that means variability in the 0.1mag range for each star. Such a variability would have immediately identified these stars as true variables, and they just aren't in VSX that way. Don't worry. Even if one star is variable, that high/low pair will be in obvious discrepancy and can be discarded.
Arne