I have been plotting the difference between individual star magnitudes in calspec exposures in the PROD.SE_OBJECT database and the coadded photometry for the stars in the Y1A1_COADD database. I've been noticing nearly every exposure has two distinct populations of stars separated by 1.3-1.5 mag and that some of the stars are duplicates with one copy in each of the two populations.

During the Friday Calibration phonecon, Robert Gruendl told me that this is the same size as a 'gain correction' that was applied at some point and that I might be picking up multiple steps of reprocessing. We rewrote the query(below) where I select the objects from the PROD.SE_OBJECT and I also began manually removing duplicates, but that still doesn't completely eliminate the problem. 


One of the sets of exposure numbers for which I am having problems is 

g = 383546, r =383549 i = 383531, and z = 383534

These are exposures of WD0308-565 taken on Dec 4, 2014. 

I've attached a plot from the g band exposure so you can see what I mean:

The y-axis is g_Y1A1_COADD - g_exposure (corrected so that the median value is zero) and the x-axis is g_exposure. Ignore the line with errorbars in the middle. 

One more fact that might help you figure out where this is going wrong: The two population problem only happens on exposures taken during Y2 and Y3, and happens during every standard exposure that I've seen during Y2 and Y3.



easyaccess -q -s -db-desoper -c "SELECT e.expnum, c.filename, e.nite, e.object, e.band, i.ccdnum, i.airmass,
e.mjd_obs, e.exptime, i.skybrite, i.skysigma, i.elliptic as image_ellipt,
0.27*i.fwhm as image_fwhm_arcsec, i.saturate as image_sat_level, c.filetype,
i.crpix1, i.crpix2, i.naxis1, i.naxis2 ,rasicam.gskyphot, rasicam.gskyvar,
rasicam.lskyphot, rasicam.lskyvar, rasicam.source,
s.object_number, s.x_image, s.y_image, s.ra, s.dec,
s.flux_aper_8, s.fluxerr_aper_8, 3600.*s.fwhm_world as fwhm_arcsec,
s.class_star, s.spread_model, s.spreaderr_model, s.flags
LEFT OUTER JOIN PROD.RASICAM_DECAM rasicam on e.expnum = rasicam.expnum,
AND t.pfw_attempt_id=c.pfw_attempt_id
AND c.filetype='cat_finalcut'
AND t.pfw_attempt_id=i.pfw_attempt_id
AND i.filetype='red_immask'
AND i.ccdnum=c.ccdnum
AND e.expnum=c.expnum
AND e.expnum=$expnum
AND c.filename=s.filename
AND ((s.class_star > 0.8) OR ((s.spread_model + 3.*spreaderr_model) between -0.003 AND 0.003))
AND s.flux_aper_8 > 2000. AND s.flags < 3
ORDER BY s.filename"

    CommentAdd your comment...