THE NEW CALACS!
Pixel-based CTE Correction of ACS/WFC:!
Extended Sources!
Ray A. LUCAS1, Norman A. GROGIN1, Marco CHIABERGE1,2,3, and the ACS
Team1!
1 Space
Telescope Science Institute, Baltimore MD,
Hopkins University, 3 INAF-IRA, Bologna, Italy !
ABSTRACT!
2
Center for Astrophysical Sciences, Johns
In the presence of a high electric field, the dark current of a single pixel can be
greatly enhanced. These hot pixels accumulate as a function of time on orbit;
however, the reduction of the operating temperature of the WFC CCDs has
dramatically reduced the dark current of the hot pixels. ACS devices undergo a
worse with time due to radiation damage to on-orbit
CCD
detectors.
affects
objects
thetheworst
in the
y-direction,
monthly
annealing
processItwhich
greatly
reduces
population
of hot
pixels and along columns, though
greatest when farthest from the relevant amplifier.
It affects
across
does
not affectobjects
the normal
pixels.!a wide range of magnitudes, but the effects are worst in fainter
Putting the electrons back where they belong!
The effect of CTE grows
also to a much lesser degree across rows in the xdirection, and its effect is
sources, and affect photometry, astrometry, and the
morphological shapes of sources. Using our new pixel-based CTE corrections, we perform an initial exploration of the effect primarily as seen along columns in the y-direction of the WFC at a recent epoch. We analyze
extended-source photometry of ~1800 sources from our recent re-observation of a portion of the GOODS-North deep field. These observations were tailored to allow (1) determination of extended-source CTE losses via
differential photometry of bright sources placed at different positions on the WFC detectors in different exposures, as well as (2) determination of CTE losses for extremely faint sources, not significantly detectable on
individual exposures, whose locations we know a priori from the much deeper GOODS observations back when CTE losses were negligible. !
INTRODUCTION!
!  The new Advanced Camera for Surveys (ACS) calibration pipeline (CALACS) contains the
pixel-based charge transfer efficiency (CTE) correction developed according to the Anderson &
Bedin (2010) code.!
!!  CTE affects both the photometry and morphologies of extended objects such as galaxies. As a
result, charge can be lost from objects, and shapes can be changed. In this initial part of the study,
we concentrate on the photometry or recovery of the charge within a given region around these
extended objects.!
!
Pos 1 FLT!
Pos 2 FLT!
Pos 1 FLC!
Pos 2 FLC!
Pos 2!
METHOD!
!  We observed two positions in a well-studied area: the GOODS-North field,
displacing the two positions in the y-direction by one full chip of the ACS/
WFC, so that there is a central region in which objects which were near an
amp in one position are far from an amp in the other.!
Figure 2: Cutouts of Pos1 (blue) and Pos2 (red) images from FLT (top) and new CALACS FLC images (bottom). Note y-direction CTE trails in FLT images at
top, especially from star. These new 550-second z-band images have modest background levels of ~15.8 e- (Pos1) and ~9.9e- (Pos2). These somewhat mitigate
the CTE loss. For reference, long blank-field WFC exposures in broadband filters can have background levels > 100e-. Since Pos1 and Pos2 have different
background levels, the CTE in them will not be exactly the same: Pos1 should have better CTE than Pos2.
RESULTS!
FLT (left) vs FLC (right) Relative Photometry
FLT (left) vs FLC (right) Absolute Photometry
Figures 3a, b: FLT (uncorrected for CTE) vs FLC (corrected for CTE) relative (delta) photometry between Pos1 and Pos2 (Figure 3a, left) and absolute photometry offset
from original GOODS “truth” image (Figure 3b, right).
CONCLUSIONS!
!  We processed and combined the Position 1 and Position 2 *_flt.fits images from
the two positions, aligning them to a deeper original GOODS ACS/WFC “truth”
image cut-out which we use as a reference image for both alignment and
verification of existence of real objects in our shallower post-SM4 data. Using
the new CALACS, we also processed and combined the new Position 1 and
Position 2 *_flc.fits images, again using the same original GOODS “truth” image
cut-out as reference., as we did for the corresponding *_flt.fits images.!
!  We ran dual-mode SExtractor on the “truth” image and on our combined
*_flt.fits images, and then on the “truth” image and our combined *_flc.fits (new
CALACS-produced) images.!
!  We then compared the photometry for data with the CTE formula of Chiaberge
et al. 2009 with that of the combined *_flt.fits images and the combined
*_flc.fits (new CALACS) images, and the original GOODS “truth” image.!
Pos 1!
Figure 1: The original circa ~2002 GOODS “truth” image with footprint overlay of new Pos1 (blue) & Pos2 (red), plus objects
detected, color-coded by S/N. About 1800 objects were detected in the original GOODS image. S/N color code is in Fig. 3a,b.
!  Although what we can say is somewhat limited by the lower S/N
of our new images, we found that the CTE effect did not seem as
great for faint extended sources in our data as we had expected. !
!  We also found that our new CTE correction seems to have little
effect on extended sources at the faintest magnitudes. !
!  Since the effect itself seems smaller than we had expected, the
fact that the new CTE correction algorithm doesn’t appear to do
much at faint magnitudes is probably a good thing. The cleaner
background in the “new CALACS”-produced FLC images may
enable fainter detections. We plan to study this further. !
Figure 4. Residual magnitude loss for y=2000 versus Log flux in electrons.
We compare the residual magnitude loss for different bins of source flux,
before (FLT, black) and after (FLC, blue) the pixel-based CTE correction is
performed. In red (CTE formula), we show the residual loss obtained if the
CTE correction formula (optimized for post-SM4 data, Chiaberge, M., private
communication) is applied to the photometry from the FLT image. In this
case, the extended sources are considered as point sources. Details on the
photometric correction formula, which is formally derived for stars only, can
be found in Chiaberge et al., ACS-ISR 2009-01.
REFERENCES!
!
!
Anderson, J. & Bedin, L. R. 2010, PASP, 122, 1035–1064!
Chiaberge et al. ACS ISR 2009-01 !