MEMO TO SCIENTISTS DOWNLOADING FLS IRAC DATA

You should be aware of the following issues that affect the quality
of IRAC data. Some of these issues will be eventually corrected or
eliminated, but they are still present in the data that are currently
released.

1.Persistent images in channel 1. When a bright source (K=13 mag or
brighter) is stared at for a sufficiently long time, it will leave a
persistent image in channel 1 that decays very slowly (persists for
several hours or more). The persistent images have a PSF that is wider
than the true source PSF. Such artifacts may exist in the FLS IRAC data
releases, due to bright sources in the FLS or preceding observations.

2.Persistent images in channel 4. These are different in nature from
the channel 1 persistent images. A bright source leaves a persistent
image that can last for more than a week. Again, such artifacts may
be found in FLS IRAC data.

3.Diffuse stray light: All raw IRAC images contain a stray light pattern,
resembling a "butterfly" in channels 1 and 2, and a "tic-tac-toe"
board in channels 3 and 4. These artifacts are due to zodiacal light
scattered onto the arrays. The science and calibration pipelines
subtract the scattered zodiacal light by subtracting a scattered light
template image scaled by the COBE/DIRBE zodiacal light model. Scattered
diffuse Galactic emission, especially near the Galactic plane, remain.

4.Stray light from point sources. SPOT allows you to overlay stray
light boxes on any image; if a bright star is placed in those boxes
during an observation, a scattered light patch will appear on the
array. The new stray light boxes are included in SPOT now and are also
shown in Observer's Manual version 4.0. Channels 3 and 4 have less stray
light. In channels 3 and 4 the stray light arises when a star lands on
a thin region just outside the array (the same region that causes the
"tic-tac-toe" pattern from diffuse stray light in flat fields). More
details are given on this webpage:

http://ssc.spitzer.caltech.edu/irac/scatt.html

which shows the location and shape of spurious scattered light spots and
rays that could be mistaken for real astronomical objects.

5.Muxbleed. Muxbleed shows up as horizontal (in raw or BCD images) bright
rows on either side of a bright point source. The IRAC arrays are read out
with 4 independent amplifiers that sweep along rows, thus the muxbleed
pattern has a characteristic pattern with every fourth column having a
correlated effect. The coefficients of the correction algorithm for
muxbleed in the pipeline partially correct the effect. We have found that,
especially for bright sources, muxbleed does not scale predictably with
source brightness, so it appears that no less than a nonlinear, iterative
fitting algorithm (under development but not yet working) could remove
it.

6.Banding and column pulldown. Banding appears as horizontal and
vertical bright rows and columns around a bright point source in
channel 3 and 4 images. Column pulldown affects all 4 channels
and causes an offset of columns containing bright sources. The effects
are not understood, nor are they corrected by the pipeline.

An algorithm to cosmetically correct the images for column pulldown has
been developed and is being tested. It does not perfectly correct the
pulldown effect in the mosaic columns, and it sometimes corrects columns
that did not need correction. Thus, leftover effects are still visible
even after this post-BCD tool is used.

7.Full-frame bias. Very bright sources on the array tend to cause an
array wide bias shift. Some experiments at fitting are underway, but
the amplitude seems to be variable. Using overlapping portions of frames
to perform overlap correction partially corrects for the bias offset.
This correction can be applied with the SSC mosaicer, but it is not
routinely run by the on-line pipeline processing.

8.Ghost images are visible near very bright sources in channels 1 and
2. These ghost images are caused by internal reflections within the
tilted filters. Filter ghosts also appear for very bright (saturated)
sources in channels 3 and 4. These ghosts are cross-shaped ("+"). Both
types of ghost images are seen the same row and column distance
from the source. See SSC IRAC webpage and the IRAC chapter in the
Observer's Manual for more information.

9. Calibration: The absolute calibration is to point sources using a
10 pixel (12") radius aperture. Photometry performed the same way on
the BCDs and mosaics will be accurate. For extended emission, there is
a poorly-understood, but quantified, difference in calibration. To use
these data products for extended emission, we recommend using the
ratio of the point to extended source throughput in the SOM, Table
6.4, to obtain the approximately correct calibration. Specifically,
multiply the surface brightnesses in channels 3 and 4 by f_p/f_ex =
[0.45, 0.61] / [0.72, 0.88] = 0.63, 0.69, respectively.

10. Pipeline version: The pipeline version used to process these data
was S9.5 for the BCD and S10 for the post-BCD data.