Fri, 04/08/2022 - 17:52
I'm working with one of the folks I mentor and want to make sure I'm providing correct information.
He did measurements on a variable star and provided additional data for two comp stars. Looking at the normalized relative flux plots of the comp stars over the measurement period (many hours), one trends up and the other trends down with a similar but opposite slope. The comp stars are reasonably color matched (B-V = 0.21 and 0.41). Am I correct in saying that some unknown systematic (likely airmass) is affecting the comp stars differently? Further, if the unknown systematic was affecting the comp stars in the same way, then the normalized relative flux plots for each comp star would be flat?
Best regards,
Gary
If it hasn't already been done, I suggest checking that neither of the comp stars is saturated by plotting the raw ADUs for both stars for the entire run. If there is no saturation, I have no other explanation.
Roy
Not enough data given to give any advice. Over what airmass range were the observations taken; what filter was used; what were the peak counts in both comp stars; how far away from center were the comps, and in what direction; was dark subtraction and flat-fielding done?
Arne
The comp stars were not saturated, peak ADU counts were less than 40,000. The observations were taken over a considerable airmass variation (1.2 to > 3.5) as the observer was trying to get most of the light curve for an eclipsing binary in one session. He has no plans to submit the results as many were taken at an airmass greater than 2. Darks and flats were applied. I do not currently know how far the comps were from the target but significant angular distance could result in a systematic. The data was G from an RGB CMOS color camera - a UV/IR cut-filter was in the optical chain.
The systematic was likely airmass. I just want to confirm that results of the type I described likely result from an unknown systematic affecting the comp responses differently. Further, had the systematic affected both comp stars the same, both normalized relative flux curves would be flat.
Hi Gary,
That helps a bit. The high airmass range is usually a suspect with any systematic trends. The angular distance between the two comps is then a factor, for first order extinction. At airmass 3.5, for example, the star with the greater altitude will have less extinction than the lower one, the amount depends on the field of view. However, both stars would trend towards fainter values, just at different rates, whereas you indicate one increases in brightness and one fades. There are VERY few atmospheric effects that will brighten an object, so I'd have to see the data to really understand what is happening.
Second order extinction is dependent on star color, in the form k" X (B-V) for example. The difference in the star colors is small (you indicate about 0.2mag), but X varies from 1.2 to 3.5, so the size of the second order correction will change by a factor of 3 during the time series. However, the comp stars would still trend towards being fainter.
My guess, without seeing the data, is that flatfielding wasn't done correctly, or that poor tracking moved the star images across the Bayer mask and modified the G channel result.
Arne