Observing Operations | Reviews | Survey Management

The Monitor Telescope Pipeline (MTPIPE)

Presented by Douglas L. Tucker


General Summary

The original 0.6m Monitor Telescope (a.k.a., "MT") was assigned three main tasks:
  • to set up the SDSS standard star network, calibrating 100+ primary standard stars against the three fundamental primaries, BD+17 4708, BD+21 607, and BD+26 2606;
  • to calibrate roughly 2500 27x27 arcmin^2 secondary transfer fields, which in turn would be used to calibrate the main photometric survey; and
  • to monitor the transparency and extinction of the sky at APO in real-time, in order to forewarn the 2.5-m operators of any important changes in the night's photometric quality.
In January 1998, it was determined that the MT was not up to these tasks, so they were split up between two other telescopes. Observations for setting up the standard star network were to be made with the USNO 40in telescope (Flagstaff, AZ); these observations started in March 1998 and continued through January 2000. The other two tasks -- calibration of secondary transfer fields and APO sky monitoring -- were to be done by a replacement telescope for the MT. This replacement telescope, a 0.5m DFM known as the Photometric Telescope (or "PT"), now resides in the old MT dome. (Due to the larger field-of-view of the PT, now only about 1500 40x40arcmin^2 secondary transfer fields need to be calibrated.)

The Monitor Telescope Pipeline (``MTPIPE'') is charged with the duty of reducing the data for the first two of the above tasks -- that of setting up the primary standard star network and that of calibrating the secondary transfer fields to this photometric system. (Incidentally, many of the tools developed for these offline reductions are also being used for the third directive, that of realtime monitoring of the sky conditions, although this is not an official requirement for MTPIPE.)

The main trunk of MTPIPE is composed of four parts:

  • preMtFrames, which is basically a "pre-burner" to the other three packages. It
    • creates a directory structure for a night's reductions
    • copies night log from the spool directory into the reductions directory
    • copies the appropriate input parameter files into the reductions directory
    • runs checks on the contents of the observing log
    • produces quality assurance histograms for the raw bias, dome flat, and twilight frames

  • mtFrames, which
    • creates master bias frames, dome flats, and twilight flats for the night's data
    • bias subtracts and flattens target images
    • detects objects in the flattened frames
    • performs aperture photometry on these objects
    • outputs the results for each target to a file
    • produces quality assurance plots

  • excal, which
    • takes the output of mtFrames (object positions, aperture photometry) for the "primary standard star" fields
    • determines the scale and orientation of the images
    • identifies all primary standard stars in these fields,
    • invokes a least squares routine to solve the SDSS photometric equations for the night (see Appendix below for details)
    • if desired, the least squares routine can also calibrate any uncalibrated primaries observed during this night
    • outputs the photometric solution for each night and observed magnitudes of each primary standard star to a file
    • produces quality assurance plots

  • kali, which
    • takes the output of mtFrames (object positions, aperture photometry) for the "secondary patch" transfer fields
    • determines the astrometric solution (row,col on CCD --> RA,DEC) for each field
    • applies the astrometric solution to each field
    • applies the photometric solution from excal on each field
    • outputs the positions and photometry for each secondary patch target to a file
    • produces quality assurance plots

These secondary patches are later scanned over by the imaging camera on the SDSS 2.5m telescope. They thus act as calibration transfer fields to set the photometric zeropoints for the imaging camera data. This calibration is performed by the FCALIB package, which will be discussed by Brian Lee.

A flowchart of the mtpipe reduction process can be found in the figure below.



Appendix: The SDSS Photometric Equations

Specifically, the SDSS photometric equations which are solved by the least squares routine are of the form:

g'(instr) = g' + a_g + b_g*(g'-r') 
                     + c_g*[(g'-r') - (g'-r')_zp]*[X - X_zp] 
                     + k_g*X 
where we have used the g' band equation for the purpose of illustration.

The variables are defined thusly:

  • g'(instr) is the instrumental magnitude; i.e., -2.5log(counts/sec)
  • g' is the known magnitude for this standard star
  • a_g is the zeropoint
  • b_g is the instrumental color term coefficient (b_g is defined to be 0 for the USNO40 data and is approximately zero for the PT data)
  • c_g is the atmospheric color term coefficient
  • k_g is the first-order extinction
  • X is the airmass
  • (g'-r') is the known color of this standard star
  • (g'-r')_zp is a zeropoint color (defined to be a cosmic average color)
  • X_zp is a zeropoint airmass (defined to be 1.3 for the SDSS)
Note that the above equation converts known magnitudes and colors into an instrumental magnitude. After solving this equation, excal also determines the coeffients for the inverse equation, which converts instrumental magnitudes and colors into a calibrated magnitude:

g' = g'(instr) + a_g' + b_g'*(g'(instr)-r'(instr)) 
               + c_g'*[(g'(instr)-r'(instr)) - (g'(instr)-r'(instr)_zp]*[X-X_zp] 
               + k_g'*X 

where 

     a_g' = a_g - b_g*(a_g - a_r)
     b_g' = b_g
     c_g' = c_g
     k_g' = k_g - b_g*(k_g - k_g)
     (g'(instr)-r'(instr))_zp = (g'-r')_zp + (a_g - a_r) + (k_g - k_r)*X
This equation is used by kali to calibrate the stars in the secondary patch transfer fields.


Last updated: 5 July 2000
Douglas Tucker (dtucker@fnal.gov)