I have assembled 107 reduced acm images in /home/sco/acm_SBSKY. These are from 16 nights during Feb,Mar,Apr,May of 2019. I have included a few nights where the moon illumination was around 50% so that we can begin to look at the sky brightness function associated with moon phase and angle. In addition, in late April and May, I began to notice the presence of Gunn r filter images taken with the acm. Hence, in this analysis I'll change up the assembly of the acm data table file. I'll look into generalizing the use of some of the table manipulation and plotting tools.

Here I want to concentrate on building a script that sets up the initial g,r,i tables. Then I want to make plots and compute stats witj various table/plotter tools in oder to investigate systematics in the g image data.

 

Two new major scripts: 
  acm_pupil_data == assemble a table file of PUPILLUM values from image headers. 
    Uses acm_pup to convert (X_STRT,Y_STRT) to RSTRT. 

  acm_pupil_cal  == assemble puil illumination values predicted from ZPSEC values.                       
    This script uses fits2table to gathe WAVELEN and ZPSEC values from the 
    FITS headers of images listed in a an input file. 
    For each image values of   con.ZPSEC,WAVELN are written to ZP.data  
    The ZP.data file is then processes by acm_pupcal.sh to compute the 
    final pupil illumination values from ZPSEC. The values are written
    to "Pupil.Values" and this file should have the same number of lines 
    as are in the input list.  

The images are in:     /home/sco/acm_SBSKY   (see list.03) 
I do analysis 03 in:   /home/sco/acm_reds_sco2019/Analyses/Analysis03/data 

% ls -1 /home/sco/acm_SBSKY/*fits > list.03 

I make a table file with fits2table. Here is the PARAMS The PARAMSfile is 
what determine what will be assembled i the table files.
 
% cat PARAMS
WAVELEN    Filter wavelength (angs) 
UTHOURS    Hours since 0hUT
UTDATE     UT date (YYYYMMDD)
RSTRT      Radius position of tracker at start (mm)
AZHET      HET structure AZ (deg)
MILLUM     percentage moon illumination
PHIMOON    angle of separation to moon (deg)
ZPSEC      ZP for a 1-sec exposure
ZPERR      mean error of ZPSEC
SKYSB      sky surface brightness (mags per sq.arcsec)
SKYSBERR   mean error of SKYSB

To get tables: 
% ACM_ANALYSIS_TABLES list.03 PARAMS N

I get the A1 table, and found 5 bad points out of the 107. I removed these 
and ran again. 
/home/sco/acm_SBSKY/20190512T073909.7_acm_sci.fits
/home/sco/acm_SBSKY/20190205T121618.8_acm_sci.fits
/home/sco/acm_SBSKY/20190307T081109.3_acm_sci.fits
/home/sco/acm_SBSKY/20190207T111542.6_acm_sci.fits
/home/sco/acm_SBSKY/20190409T100822.5_acm_sci.fits

Make a nice plot: 
% Generic_Points N                        # edit "xyplotter_auto.pars" 
% xyplotter_auto datg RSTRT ZPSEC 20 N        # edir List.20 , Axes.20 hereafter 
% xyplotter_auto dati RSTRT ZPSEC 20 N        # edir List.20 , Axes.20 hereafter 

Then I added the PUPILLUM data and made a final plot (List.20, Axes.20) 
This plot was made in: /home/sco/scohtm/scocodes/Night_of_acm/Analysis/analysis03/data
All of the compoent files are stored in: 
/home/sco/scohtm/scocodes/Night_of_acm/Analysis/analysis03/data.Fig_Pupil_1
To re-run the plot:     
   % xyplotter List.20 Axes.20 N

Note: I currently leave out the PUPILLUM values fro the 16 nights. These 
are in:  ACMPUPIL0.table 
The ACMPUPIL.table has more points but is only from the first 9 days or 
so of Jan2019. The acm_pupil_data script runs very slowly on mcs and I 
could not get any more pupil data. 

 

The model HET pupil illumination points (from the PUPILLUM cards in acm images) are shown as small blue dots. Observed pupil values, computed from the photometric zeropoints derived with PANSTARRS photometry, are plotted for g, r, and i acm images. This plot uses 102 images drawn from 16 (mostly) clear nights in the period Feb-May 2019. Some of the i data were from acm images taken outside of 18 degree twilight during HPF setups and hence may be noisier due to increased background levels.

Most of the data reduced thus far were near moonless nights. I have some g data when the moon was 50% illuminated, and I can see a trend in sky syrface brightness and moon angle (see figure below). I need more data from nights with moonlight. One immediate stratgey would be to reduce a lot of nights where the moon phase is close to 50%. At some stage I may be able to correct for moon phase and sky color (i.e. normalize all filters to on system) and derive a single set of data points for sky brightness verses anular sparation, but we'll need a lot more acm images to validate this approach.

The moon properties for times when the moon was up in the sky.

The sky surface brightness measured on acm images for times when the moon was up in the sky. The g points are mostly from a night when the moon phase was around 50% and we see for these points a clear trend with angular spearation.