The Advanced Camera for Surveys Wide Field Camera (ACS WFC) is one of the currently operating cameras on Hubble. The camera consists of two CCDs, each with 2048 rows and 4096 columns. Due to damage to the camera’s CCD its images now show streaks or trails coming off of sources in the image. The streaks are caused by charge getting trapped and later released as the CCD is being read out. The technical term for this behavior is “charge transfer inefficiency” (CTI), though often the complementary term “charge transfer efficiency” (CTE) is used instead.
Correcting for this smearing effect is important for both scientific and aesthetic uses of ACS WFC data. This software is designed to perform that correction. A description of the correction can be found here: http://blogs.stsci.edu/newsletter/2011/02/07/correcting-pixels-for-cti-in-acs%E2%80%99s-wfc/
One interesting aspect of CTI is that in terms of correcting the streaks, any given column of data is completely independent of the others and all of the 8192 columns can be processed separate from each other, making the process conceptually easy to parallelize. Each column of data is corrected by a computationally intensive algorithm that requires a large number of iterations.
The data we’ve included here are in FITS format, a file type common for astronomy data. You can read more about FITS here: http://en.wikipedia.org/wiki/FITS
ACS WFC data files have a primary header and six image data extensions. The first three extensions are related to one ACS WFC CCD and the last three are for the other CCD. Of primary interest are extensions 1 and 4 since they hold the actual image data.
We’ve included two data files because the CTE correction algorithm is sensitive to the overall level of signal in the images. Images with high average levels of signal require more iterations, and more time, to correct than lower signal images. The image of spiral galaxy NGC 4258 has a high background level and requires significantly more processing than the image of globular cluster NGC 104 (also known as 47 Tucanae).
The data files can be found in the directories ngc104 and ngc4258, respectively.
The data have file names consisting of 9 characters/numbers followed by an underscore and a suffix, plus the .fits extension. For example: jb1f89eoq_flt.fits. Different file types are distinguishable by their suffixes.
The flt suffix is for data which has been processed through our calibration pipeline but has not been corrected for CTE. These will be your inputs to the acstools correction software.
We also have included files with the suffix cte_ref which are files we have processed that you can use as test comparison data.
DS9 is the tool most commonly used to view FITS images. It is available at http://hea-www.harvard.edu/RD/ds9/.
Acstools is a Python package containing the module PixCteCorr. The task YCte in PixCteCorr is the primary function used to perform the CTE correction. The actual work of calculating the corrections is done in C extensions called by the Python routine. Python performs the file IO operations using pyfits.
For information about installing acstools see Installation.
For information about running acstools see Running CTE Correction.
Because each column of data can be processed independent of the others it makes sense to parallelize the correction so that it can be applied to many columns at a time. The software currently applies the correction to each column one at a time in a loop in the function FixYCte in the source file src/FixYCte.c. The outer-most loop in the function loops over columns of data.