Hello! I would appreciae guidance on how to handle air mass extinction over wide fields, especially with the new CMOS small pixel chips coming out.
My calculations may be in error, so please let me know if so.
With full size chips (35mm) and a scope of about 2000mm focal length, the field of view is about 1 degree. Using air mass equation, and assuming 0.25 mag extinction per airmass, at 30 degrees from the horizon, the magnitude difference between top and bottom is nearly 0.02 magnitudes.
For shorter focal lengths, this difference becomes much more significant and can impede the goal of millimag measurement.
I suppose the comp mags can be adjusted by hand using trigonometry to calculate the air mass difference and then mannually adjusting the ADUs. Can such a process be automated? This would be especially important for short period variables where a hundred images might be obtained through the course of an evening
Thank you and best regards.
Mike
Back when eps Aur was having its eclipse, I did some time series with 50 mm focal length. At airmass=4 at the start of a run, eps was more than 0.1 mag fainter than comp star eta. At the end of a run, at airmass near 1, eps eta was about 0.05 mags brighter than eps -- before correcting for differential extinction. So, an extreme example of the effect you have identified.
Certainly the process of correcting can be automated, but it depends so much on how you process your data. My present processing flow inserts an airmass in the FITS image headers using the target star in the image. It would be relatively easy to compute an airmass for each star in the image, and carry that airmass along with the raw instrumental magnitude derived for each star in each image, thereby facilitating correction.
My data processing flow is an evolution of what I've been using for 15+ years, so it is a "dog's breakfast" of scripts calling IRAF, Starlink, and wcstools programs. Things like airmass are presently calculated using python/astropy routines. My observing is mostly all-night time series of eclipsing binaries.
Gary Billings
Thank you!
Mike