IRAC: Bright Star Test

T. Jarrett
Oct 13, 2004

Data

Bright Object Test AORs from campaign IRAC-VW (note: data provided by Bill Glaccum)

The star is found in 2MASS with parameters:

Note: these integrated fluxes represent the total flux of the star. They should be systematically brighter than what is measured for Spitzer because the Spitzer measurements are carried out on a much smaller region area (driven by the size of the PSF image).

Method

Use the 2MASS position to locate the initial position of the saturated star. Then deploy a walking "centroid" that avoids low pixel values to locate the intensity-weighted centroid of the "crater".a Fit the PSF to the wings of the saturated star, recover the lost pixels. See Bright Star Rectification for details.

Results

Channel 1

Channel 1, test location #3.
2MASS sources marked in red; the crater is located and marked in blue
Close-up look at the bright star. The pixels have been regrided to match the PSF pixel grid.
The PSF model fit to the wings of the star.
The fit residual: ratio between the model and the raw image. A radial plot showing the residuals is given here: residual radial plot.
The rectified image.

Channel 2

Channel 2, test location #3.
2MASS sources marked in red; the crater is located and marked in blue
Close-up look at the bright star. The pixels have been regrided to match the PSF pixel grid.
The PSF model fit to the wings of the star.
The fit residual: ratio between the model and the raw image. A radial plot showing the residuals is given here: residual radial plot.
The rectified image.

Channel 3

Channel 3, test location #3.
2MASS sources marked in red; the crater is located and marked in blue
Close-up look at the bright star. The pixels have been regrided to match the PSF pixel grid.
The PSF model fit to the wings of the star.
The fit residual: ratio between the model and the raw image. A radial plot showing the residuals is given here: residual radial plot.
The rectified image.

Channel 4

Channel 4, test location #3.
2MASS sources marked in red; the crater is located and marked in blue
Close-up look at the bright star. The pixels have been regrided to match the PSF pixel grid.
The PSF model fit to the wings of the star.
The fit residual: ratio between the model and the raw image. A radial plot showing the residuals is given here: residual radial plot.
The rectified image.

Masking the "Band Effect" Features in Channels 3/4

Channels 3 and 4 have very prominent artifacts that reside close to the center of the star. They ultimately derive from the read-out electronics, and hence do not represent a real signal. In the following examples, the artifact is masked according to: center of star (based on centroid or interactive pointing) plus 4 camera pixels, plus 8 camera pixels, plus 12 camera pixel, ... etc.

Channel 3

Close-up look at the bright star before masking.
Close-up look at the bright star after artifact masking.
The fit residual: ratio between the model and the raw image. A radial plot showing the residuals is given here: residual radial plot.
The rectified image.

Channel 4

Close-up look at the bright star before masking.
Close-up look at the bright star after masking.
The fit residual: ratio between the model and the raw image. A radial plot showing the residuals is given here: residual radial plot.
The rectified image.

Photometry

Photometry of six samples per band is given here. A circular aperture with 25 pixel radius is used to measure the integrated flux. The background is assumed to be zero.

A histogram showing the photometry results for six observations (per band) is given here:

Histogram of Rectified Photometry with artifact masking in Channels 3/4
where the white line represents the "raw" BCD photometry (including saturated pixels), red is the "model" photometry, and green is the rectified photometry. The yellow dashed line shows the flux measured for the sub-array (non-saturated) observations.
The non-saturated (sub-array measurements) photometry agrees reasonably well with rectified/model photometry. The exception is channel 3 where the results are quite discrepent.

Notes about photometry

Channel 1 photometry
The star is clearly saturated in channel 1. The raw integrated flux is about 4000 mJy. Both the model and rectified photometry is about ~20000 mJy, representing a flux ratio 0.20. In summary, the rectified photometry looks well behaved.
Channel 2 photometry
The star is clearly saturated in channel 2. The raw integrated flux is about 3000 mJy. Both the model and rectified photometry is about ~13000 mJy, representing a flux ratio 0.23. There is some evidence that the model photometry is slightly underestimating the true flux of the star.
Channel 3 photometry with artifact masking
The star is moderately saturated in channel 3, with the raw counts hovering around 7000 mJy and the rectified counts around 9500 mJy, or a ratio of 0.8. Both channels 3 and 4 are characterized by significantly bright (with respect to the total) "band effect" ghosts to the west of the star core (see the rectified image). This artifact light has been masked and rectified using the PSF model. One major mystery remains: why is the non-saturated measurement (from the sub-array observations) much fainter than the BCD and rectified measurements?
Channel 4 photometry with artifact masking
The star is moderately saturated in channel 4, with the raw counts hovering around 3500 mJy and the rectified counts around 5000 counts, or a ratio of 0.7. Both channels 3 and 4 are characterized by significantly bright (with respect to the total) "band effect" ghosts to the west of the star core (see the rectified image). This artifact light has been masked and rectified using the PSF model.