Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/hst/wfpc2/documents/isr/wfpc2_isr0108.ps Дата изменения: Unknown Дата индексирования: Tue Feb 5 15:30:15 2013 Кодировка: Поисковые слова: р п р п р р п

Copyright© 1999 The Association of Universities for Research in Astronomy, Inc. All Rights Reserved.
Instrument Science Report WFPC2 2001­008
Creating WFPC2 Dark
Reference Files: Addendum
J. Mack, S. Baggett, J. Biretta
September 5, 2001
ABSTRACT
In this report, we describe recent updates to the procedures already in place for making
WFPC2 dark reference files. These updates include the ability to create dark reference
files by using either weekly or daily dark images as input. They also allow the user to spec­
ify the year of the WFCP2 superdark image which is compared to the output dark file dur­
ing processing. The new scripts are currently available for FTP from the WFPC2 website
to users who wish to create dark reference files which are tailored to closely match the
date of their HST observations. This ISR is written as a supplement to Instrument Science
Report WFPC2 2001­001 and is not intended to serve as a stand­alone document.
Obtaining the IRAF scripts
All the tools necessary for making dark reference files are available for FTP from the
WFPC2 Dark Reference File Website:
http://www.stsci.edu/instruments/wfpc2/Wfpc2_memos/wfpc2_darks_tab.html
The FTP file linked by this website contains several cl scripts, text files, and images and
requires about 50 MB of disk space to download. Among these files are the superdark
images for various years: 1999, 2000, and 2001.
Selecting Input Darks
When retrieving a set of dark images from the archive, keep in mind that there are separate
dark monitoring programs. The first of these takes darks at roughly weekly intervals and
the second takes darks on a daily basis. ``Weekly'' dark images are taken in sets of five and

Instrument Science Report WFPC2 2001­008
2
have exposure times of 1800 sec each. ``Daily'' dark images are taken in sets of three and
are only 1000 sec each. The dark images you select as input will be combined to make the
output dark reference file.
To achieve the same exposure time as a weekly dark reference file, you would thus need
9*1000 sec exposures to equal 5*1800 sec exposures. (The readnoise will be slightly
higher when using 9 images, though is not a significant increase compared with the dark
count rate.).
These scripts are designed to handle only dark images of 1800 or 1000 sec, though any
number of input files may be used. The specific number of files used is left to the discre­
tion of the user. When selecting the input images, try to keep them as close as possible to
the date of your observations and be sure that the images you select are all taken within the
same decontamination cycle. Decontamination cycle boundaries (decon dates) are listed at
http://www.stsci.edu/instruments/wfpc2/Wfpc2_memos/wfpc2_decon_dates.html.
Making the Darks
The wkdark task links together the following 4 tasks: superimg, superdqf, doheader,
calcheck, each of which are described in detail in ISR WFPC2 2001­001. Before running
this task, four input parameters must be defined. (The first two parameters are described in
the previous ISR.)
mk> wkdark
Rootname of Previous Dark (kau1324lu):
Useafter Date (Jan 25 2000 05:05:05):
Superdark Year (2001):
Are infiles daily or weekly? (weekly):
The ``Superdark Year'' should be chosen to match the date of the HST observations you
wish to recalibrate. Superdark images are currently available for 1999, 2000, and 2001. As
future superdark images are created, the FTP files from the web will be updated.
The last parameter indicates whether the input darks were taken from a weekly or a daily
dark monitoring program. This allows the correct normalization: by 1843.6 seconds if the
input images are weekly darks or by 1003.6 seconds if the input images are daily darks.
These normalizations are slightly higher than the image exposure times because they
include overhead times while the detector is waiting to be read out.
We note a minor clarification in the previous ISR's description of how the scripts create
the data quality files using the task superdqf. When comparing the current dark DQF with

Instrument Science Report WFPC2 2001­008
3
the superdark DQF, the data quality value is assigned a value of `2' if the pixel value came
from the current dark and if crrej used less than half of the total input frames, indicating
that the pixel value was noisy or questionable. The previous ISR was worded ``less than 3
of 5 of the input frames...''. The scripts are now written to count the number of input files
and to determine if less than half were rejected for a given pixel.
We also give caution when masking ``bad regions'' in the input dark frames. It is important
not to ``overmask'' the effects of CTE (i.e. masking too many columns, too large a region,
or the same region in all the dark images). If the same pixel is masked as bad in every
input file, the output dark reference file will be set to zero at that pixel and the dark signal
will not be properly subtracted when you recalibrate your observations.