In late Oct2019 we hit a week or so of very clear weather. The seeing was generally good during this time. I colloected five nights of acm data starting of 20191018 (when the moon was just befor third quarter) to 20191027 when we were close to new moon and doing a lot of hetdex observations. I'll keep some stream consciousness botes here about the resuctions of these nights.

1. The five good nights.
2. Reduction location and initial reduction script.
3. My older wcsf methods
4. Overplot USNO sources.

These nights have some moon, but were extremely clear and mostly good seeing. performed with a script.

% cat BaseDir 
/media/sco/DataDisk1/sco/AD/HET_work/acm_nights
% acm_nights_markII list.dates none N  
% cat acm.nights_SUMMARY 
Basepath used  = /media/sco/DataDisk1/sco/AD/HET_work/acm_nights 
  Date       Ntot  Nopen Nbias Ndark Nnone Ntemps  NoB   Nog   Nor   Noi   Moon_illum
  20191018    587    355    20     0   202   587     0    99    23   232    84.100 
  20191020   1003    323     5     0   665  1003     0   208     3   112    66.400 
  20191022    722    609     5     0    98   722     0   437     0   172    44.900 
  20191026    659    541     5     0   103   659     0   384     0   157     6.900 
  20191027    422    361     5     0    46   422     0   240     0   121     2.100 

The ccd processing, image catalog preparation, and onsky sets derivation can be performed with a script.

% pwd; ls 
/home/sco/ACM_work_Oct2019/red_20191018/S
BaseDir  list.dates  RunDate*

To run a night (above the S subdirectory):
    ./S/RunDate 20191018

The goal now is to develop an automated wcsf system. In the onsky_set document, I show how to compile a list of images (for one sky position) for investigative work.

% onsky_set EX1 1 Y N  

I pick JUST one image: 
% cat List.FINAL_ISET 
/home/sco/ACM_work_Oct2019/red_20191018/local_red/FIXUP/20191018T024545.3_acm_sci.fits

Here are some properties of this image prior to WCS calibration 
RADESYS = 'FK5     '
CUNIT1  = 'DEGREES '
CUNIT2  = 'DEGREES '
CTYPE1  = 'RA---TAN'
CRPIX1  =                  185 / reference pixel
CRVAL1  =           290.697915
CTYPE2  = 'DEC--TAN'
CRPIX2  =                  399 / reference pixel
CRVAL2  =             53.34493
CD1_1   = 6.72660898999311E-05
CD1_2   = 3.37937381118536E-05
CD2_1   = -3.3793738111854E-05
CD2_2   = 6.72660898999311E-05

The fiducal pixle is at CRPIX1,CTYPE2 = 185,399 

This position, according to the header and xy2sky is: 
% xy2sky /home/sco/ACM_work_Oct2019/red_20191018/local_red/FIXUP/20191018T024545.3_acm_sci.fits 185 399
19:22:47.500 +53:20:41.75 J2000 185.000   399.000 

To use my older calibration method: 
Usage: wcsf /home/sco/ACM_work_Oct2019/red_20191018/local_red/FIXUP/20191018T024545.3_acm_sci.fits N 
arg1 - Name of file with list of FITS images (can be full path)
arg2 - run in debug mode (Y/N)

This is the first ds9 view of the test image discussed in the text above. The acm test image is shown on the left. I have labeled the Ra,Dec of one target identified as target 11 in the DSS image (on the right). I have also marked with a thick blacj circle the X,Y location used for the CRPIX1,CRPIX2 position in the original PAS version of the image. This is the value that must be corrected. As can be seen in the right panel, the Ra,Dec of target 14 is quite different from that predicted by the PAS header. The RA wiki reports that the IHMP position on the acm is X.Y=185.2,396.1. This is not exactly the same as the header CRVAL1,CRVAL2=185.399, but it is close. Hence, the WCS error that PAS makes is in the assignment of Ra,DEc (CRVAL1,CRVAL2) for this position. Re-deriving a correct value for CRVAL1,CRVAL2 is our current goal.

This is the second ds9 view of the test image discussed in the text above. After we have identified a target in the field (previous figure) we can simply use the X,Y (for CRPX1,CRPIX2) and the Ra,Dec (for CRVAL1,CRVAL2) of that position to fix up the PAS acm image header. This is performed by the rough_wcs routine and it gives us what we see above. We have simply collected an astrometric catalog from the USNOB-1.0 catalog, and overplotted those sources on the corrected image. Now, instaed of all the source being displaced by 20-30 arcseconds, most of the circles fall on top of a real source in our acm image. Those circles that are emtpy are probably associated with sources that were too faint for our acm image (or were just incorrect entries in the USNO catalog). In any case, we now have an image that is calirbated to better than 1 arcsecond of error, and henne were are able to cross-match to other astrometric catalogs and derive n improved WCS solution using many more stars.

% ls ./local_red/WCS 
20191018T024545.3_acm_sci.fits
# Here are some of the the header cards in our new image  
CTYPE1  = 'RA---TAN'           /  Gnomonic Projection
CTYPE2  = 'DEC--TAN'           /  Gnomonic Projection
CRPIX1  =              387.000 / Axis 1 Reference
CRPIX2  =              385.000 / Axis 2 Reference
CRVAL1  =        290.722656250 / Ra  at Frame, J2000 (deg)
CRVAL2  =         53.324096680 / Dec at Frame, J2000 (deg)
CROTA2  =     333.147003173828 / Image Twist +AXIS2 W of N, (deg)
CDELT1  =      0.0000752099586 / Axis 1 Pixel Size (degs)
CDELT2  =      0.0000752099586 / Axis 2 Pixel Size (degs)
RMSRA   =               0.1360 / RA_rms in arcsec
RMSDEC  =               0.1810 / DEC_rms in arcsec
NUMWCS  =                   19 / Number of points in WCS fit
AZHET   =             321.2751 / HET structure AZ (deg)
HETQ    =             295.6245 / HET parallactic angle (deg)
POSANG  =              26.8530 / 360-CROTA2 (tzcal.sh,old)
TZ      =              91.2290 / PA-HETQ (tzcal.sh,old)

The WDS derivation in the previous section was succesful. In preparation for deriving an autometed version of this prcedure, I want to be able to overplot USNO targets on my calibrated image using the curcle markers that are scaled in size to reflect magnitude (bright sources large, faint sources small). I have to build this into the cdfp2reg.shroutine, and I discuss this in a document about overplotting USNO sources with ds9. Basically, the goal is to have a display like that belwo give me overplotted points that convey magnitude as well as position.

% usno_look_wcs_run /home/sco/ACM_work_Oct2019/red_20191018/Test1/local_red/WCS/20191018T024545.3_acm_sci.fits N