CFH12K Field Bright Stars Mapping

Telescope Field Mapping (TFM) Documentation Page

Table of contents:
Use TFM Now


   The ``Telescope Field Mapping'' (TFM) tool maps the bright stars in and around the field of view of CFHT's wide-field CCD imager CFH12K. Bright stars are the prime source of image contamination in CCD wide-field imaging: large halos, scattered lights and blooming cause far more pixels to be lost for science than bad cosmetics in particular. Statistics show that there is a least one 8th magnitude star per square degree in the sky. It is obvious that any CFH12K field of view will include several fairly bright stars.
   Any CFH12K observer should check his pointings using this tool to ensure that no major bright stars related problems (see items 1, 2 & 3 below) will badly obliterate the data. Of course, there are times when the object of interest can not be moved significantly enough on the CFH12K field of view to avoid contamination, but there are ways to at least minimize the impact of these bright stars.
   Note that the CFHT prime focus bonnette can not be rotated, hence dodging bright stars can only be accomplished through translations along the alpha and delta axis.
   TFM respects the size and gaps between the CCDs as well as the off-centered position of the telescope pointing (the point of the sky corresponding to the telescope pointing coordinates lies on the top corner of CCD02, not exactly at the center of the mosaic. See this page for further information.)
   However be aware that the gaps between the CCDs are not perfectly uniform, nor are the CCDs perfectly aligned in respect to each other (typical alignment angle is 0.3 degree), and can play a role at the scale of 10 arcseconds (see the "Mosaic Geometry" section on this page for further information.). So don't count on TFM to position an object precisely in a gap for example. Worse, the rotation of the camera on the sky is defined by the alignment of the telescope prime focus top-end at installation at the beginning of the observing run. Due to innacuracies in the process, the alignment of the instrument changes by up to 0.5 degree from an observing run to another. TFM maps the sky on the CFH12K field for a rotation angle of 0 degree, hence a small rotation induced by the mounting process will shift all the positions by as much as 20 arcseconds in the outer parts of the field. Also, the telescope pointing accuracy is roughly 10 arcseconds unless a SAO star is used to reset the pointing just before. A SAO star pre-pointing allows a pointing accuracy on a nearby field of 1 arcsecond but the process of calibrating on the SAO stars eats up 2 to 3 minutes. So, all in all, TFM can not (so could not any other tool) be used to prepare CFH12K pointings more accurate than 10 arcseconds.
   A sinus projection is used to map the Guide Star Catalog objects on a two dimensional plane. Moreover, the CFHT prime focus wide-field corrector optical distortion is included in the star positions mapping. A common effect for such optics, this radial distortion increases the actual distance of objects to the center of the field (see Cuillandre et al. PASP, 1996, 108, 1120).


1: On-field stars

This slide illustrates how bright stars falling within or near the CFH12K field of view can badly degrade the quality of the data. Here are described the four typical cases of contamination caused by bright stars: reflection halos, blooming and scattered light from the edge of the focal plane. The letters pointing the various phenomena on the slide are described:
      • A: The transmission of the optics (dewar window, filters, wide-field corrector lenses) are less than 100% (usually 96% with coatings). The surface of the CCD itself is reflective (specially in the blue) and light bounces back from the CCD to the dewar window, then back to the CCD. This is what can be seen as the "A" effect: a very large halo (7 arcminutes diameter). This halo does not unfortunately produce a uniform illumination and it can be difficult to salvage objects lying in that area. This effect is critical for stars of magnitude less or equal to 6 (V-band). Note that the flux in this halo is no more than 1% of the overall flux from the star. For fainter objects, this halo is totally dominated by the sky background photon noise and does not affect the signal. Check out the "tips" section below about the limits one should set on the on-field stars.

      • B: The pixel full well capacity is about 100,000 electrons. Beyond that, the electrons start overflowing on the surrounding pixels along the columns. A bright star can easily contaminate entire columns of a CCD, resulting in dead scientific area far away from the star itself. This effect kills less data than the halos.

      • C: When a star lies near outside the edge of the CCD focal plane there is still some reflections effects happening with the light baffle inside the CCD dewar. The contamination is much lower (factor of 5) than in the case A though.

      • D: When a star is a bit more further out in the field, its light hits the edge of the filter (which have baffles that reduces the effect but don't eliminate it) and results in an injection light beam on the CCD. This can spray light onto the CCD focal plane up to 7 arcminutes from the edge. This is the less damaging effect.

2: Within 1 degree radius stars

Baffling and surfaces blackening have been greatly improved at CFHT's prime focus prior CFH12K first light in 1999. While UH8K used to suffer a lot from stars within the field of view of the wide-field corrector in particular (0.5 degree radius), the effect on the CFH12K has been strongly reduced (factor of 10). The tests shown on this slide illustrate the effect caused by the star Sirius, the brightest star in the sky (the star was positioned at various locations around the field of view of CFH12K as illustrated by the small images distribution). For all the other stars in the sky, the effect will be lower of course (scaled by their relative brightness), often dominated by the sky photon noise, but still present.
3: Beyond 1 degree radius stars

Stars beyond that limit are not a concern anymore. This is illustrated by this slide.


  1: Look at the TFM output for the wished pointing. If there is a bright star of magnitude less than 4 within the 1 degree radius, try to compromise the pointing to put the star as close as possible from, or outside, the 1 degree radius area. Iterate using the "Display GIF Image" option.
  2: Look at the TFM output for the wished pointing. If there is a bright star of magnitude less than 7 inside the CFH12K field of view (the mosaic), try to compromise the pointing to put the star outside the focal plane. However, you have to expect the following item 3. Iterate using the "Display GIF Image" option.
  3: Look at the TFM output for the wished pointing. If there is a bright star of magnitude less than 6 outside and very near (less than 1 arcminute) the CFH12K field of view, try to compromise the pointing to put the star further away (like 2 arcminutes). However, you have to expect the following item 4. Iterate using the "Display GIF Image" option.
  4: Look at the TFM output for the wished pointing. If there is a bright star of magnitude less than 8 outside and near (more than 1 arcminute) the CFH12K field of view, try to compromise the pointing to put the star further away (3 arcminutes). Iterate using the "Display GIF Image" option.


   To properly select the correct pointing on your field, iterate a few times until the brightest stars don't represent a serious danger anymore. For angular scaling reference, the individual detector size is 7'x14'. The center of the field is given in the text area on top of the window while the axis are labelled in arcminutes with the reference (0,0) set at the center of the mosaic.
   A radius of 1.0 degree is used for checking for bright stars contamination. As shown above, even extremely bright stars are not a problem beyond the 1 degree radius.
   TFM can also be used to produce CFH12K finding charts (use a of radius of 0.4 degree). Access this fonction with this link.
   TFM also runs in ``monitoring'' real-time mode during observations giving the map of the field currently pointed by the telescope (1.0 degree radius, no input from the user).
   Known problem: TFM fails if a pole (+90.0 or -90.0) is included in the field (projection problem).



   Comments on the CFH12K pages to Jean-Charles Cuillandre:
   ``Telescope Field Mapping'' was developed by Frédéric Magnard (CFHT) and Jean-Charles Cuillandre (CFHT).
   CFH12K is the CFHT CCD wide-field imaging mosaic.