Spitzer Data Reductions

Spitzer NEP Data Reduction

Final mosaics of Epoch-1 observations of the WISE CVZ (NEP):

IRAC-1 (fov ~ 47 arcmin); note NGC6543 ("cats eye") and NGC552 (SB galaxy) near center
IRAC-2 (fov ~ 47 arcmin)
IRAC-3 (fov ~ 47 arcmin) with delta-dark subtraction
IRAC-4 (fov ~ 47 arcmin)

Below are the steps taken to create cleaned IRAC mosaics and photometry of sources extracted from the images.

IRAC AOR Observations

READOUT_MODE: FULL_ARRAY
ARRAY: 3.6_5.8u=YES, 4.5_8.0u=NO
HI_DYNAMIC: YES
STELLAR_MODE: NO
FRAME_TIME: 12.0
DITHER_PATTERN: TYPE=none
N_FRAMES_PER_POINTING: 1
MAP: TYPE=RECTANGULAR, ROWS=18, COLS=18, ROW_STEP=156.0, COL_STEP=156.0, 
RIENT=ARRAY, ROW_OFFSET=0.0,COL_OFFSET=0.0,N_CYCLE=1
SPECIAL: IMPACT = none, LATE_EPHEMERIS = NO,SECOND_LOOK = NO
RESOURCE_EST: TOTAL_DURATION=8908.7, SLEW_TIME=1808.8, 
SETTLE_TIME=0.0, SLEW_OVERHEAD=215.0, SPECIAL_OVERHEAD=0.0, UPLINK_VOLUME=15960, DOWNLINK_VOLUME=195950664
VERSION=S16.0.0
INTEGRATION_TIME: IRAC_3_6=12.0,IRAC_4_5=12.0,IRAC_5_8=12.0,IRAC_8_0=12.0

Steps

1. Starting with the standard pipeline BCDs, run a set of IDL tools to mitigate pixel artifacts:

IRAC Artifact Mitigation

Example: IRAC-1 BCD
Example: IRAC-1 BCD after artifact correction; note the column pulldown is now absent and the muxbleed is greatly improved.

These tools may also create overlap-corrected images, but this mode was not used in this case (instead, MOPEX was used; see below).

2. Additional corrections are needed to IRAC-3 BCDs to correct for first-frame residual gradients across the focal-plane.

Example: IRAC-3 BCD with residual dark current (note the background gradient).
After delta-dark correction

Here is an example of an IRAC-3 mosaic that looks like a patch-work quilt due to this phenomenon. And here is the same mosaic after carrying out delta-dark subtraction.

To create "delta dark" images to subtract from the BCDs, you do this by building optimal median combinations of the BCDs themselves. To quote from the IRAC handbook:

"Noticeable bias patterns in the remainder of the data can be mitigated by removing a delta-bias frame from the data. As the pattern is a function of image cadence (time between subsequent frames), only the data at a particular cadence should be stacked together in creating the delta-bias frames. The delta-bias frame can be constructed by forming a robust (outlier-rejected) median stack of the data at a particular cadence." Here is the procedure that I used: (1) for each BCD, identify a set of BCDs that were adjacent in time and median-combine the set. The sample size is set to +- 13 images, for a total of 26 images in the stack. For example, the set for BCD #250 ranged from 237 to 263. The maximum pixel value (per pixel stack) was rejected (to mitigate cosmic-rays and stars). Care was taken to exclude BCDs with very bright sources (since they tend to leave residuals in the median stack). The sample size was selected to balance the changing delta-dark (it is time-varient) with a large enough sample size to minimize noise and residuals it the resulting median-combined image. This image was subtracted from the BCD to created the corrected image (see example above). Note that only IRAC-3 BCDs were corrected in this way. For the other channels, the time-varient bias is a small effect and is not worth the risk of adding noise and residuals to the BCDs.

3. Overlap correction: using MOPEX, run the overlap correction module for all channels.

4. Combine the overlap-corrected BCDs into one large mosaic (per channel). The following date sets are used:

All other settings are from the defaults, including the "outlier rejection".

5. The last step is to derive the array-location dependent photometric corrections. This step is needed if you intend to measure stellar fluxes from the mosaics. The corrections range +- few %.

The procedure is this: create a set of "fake" BCDs that have the actual WCS coordinate info from each BCD, but the pixel values correspond to the array-location correction images. With these "fake" BCDs, combine together with MOPEX with the usual file lists, *except* use the previously derived "R-masks" (found in the directory "Rmask-mosaic" where the MOPEX results are collected). These R-masks contain the rejected pixels (usually due to cosmic-rays), and so you will be simulating the actual pixel coverage of your mosaics (Step 4, above).

Be sure to turn off the outlier rejection steps; namely

Examples of the resulting array-location dependent photometric correction mosaics are given below:

IRAC-1 array-location correction mosaic; values range +- 1%.
IRAC-2 array-location correction mosaic; values range +- 1%.
IRAC-3 array-location correction mosaic; values range +- 5%.
IRAC-4 array-location correction mosaic; values range +- 4%.

Special thanks to Frank Masci and Betsy Hall for help with the artifact mitigation step (#1).

Last update - [2 Nov 2007]
T.H. Jarrett (IPAC/SSC)