Affiliation
American Association of Variable Star Observers (AAVSO)
Mon, 12/15/2014 - 21:26
Many of the stars I am gathering data on don't report a spectral type or color index in the VSX. As that can be an important element in classifying a star I would like to be able to determine or approximate a B-V or other color index. However, in some cases the measurements one gets can vary materially depending upon where in the cycle the magnitude measurements are taken. Is there a convention as to where, in a variable's cycle (peak, mean or bottom), one should draw upon to determine a color index value?
Hi Jim,
The usual color index for a variable is the mean value. This of course only approximates the star, and there are many classes of variables where a mean doesn't do the star justice. For example, a CV might be A0-like in outburst, but in quiescence, the red donor star might contribute. Or, an eclipsing binary with unequal temperature stars will look different in/out of eclipse. This is why you usually observe variables in two passbands and determine the color of the target at the exact moment of the observations whenever possible. However, the mean value gets you most of the way towards the right solution.
Arne