Figure 1: Coadded 24 micron darks: 3-second exposures (top left), 4-second exposures (top right), 10-second exposures (bottom left), 30-second exposures (bottom left).
Principal:
Deputy:
Analyst:
AORKEYS:
Last Updated:
All data were processed with DAT version 2.15 in order to derive calibrated images. The 24 micron images were coadded using mips_enhancer with the dark flag turned on.
Figure 1 shows the coadded dark images at each exposure time. The illumination pattern and mean counts are nearly identical to darks obtained on the ground. The changing illumination pattern with exposure time is probably due to the low-flux nonlinearity in readout 4 (see the jailbar report). The mean counts are higher in the shorter exposures, as was seen on the ground. Examination of RAW data ramps (Fig. 2) shows that the second read for pixels near the bottom of the array is noticeably lower than the other reads, which artificially inflates the fitted slopes when the total number of reads is small. Thus, what appears to be amplifier glow on the bottom of the 3- and 4-second exposures is not real. The read 2 offset should not affect the calibration since it will be washed out in sky images where even the lowest backgrounds are a factor of ~100 larger than the dark current. Using slopes in the top half of the array only, the dark counts are nearly the same for all exposure times (1.8, 1.8, 1.4, and 1.4 DN/s for 3, 4, 10, and 30 seconds, respectively).
There is virtually no scattered light visible in the dark images; Figure 3 shows the residual image from subtraction of a ground dark from the on-orbit dark with the same exposure time. There is no illumination gradient, as was seen when the thermal background was high, and the residual counts are consistent with typical dark current variations (+/- 0.5 DN/s) (see also the results from MIPS-950D2, which are similar).
Figure 1: Coadded 24 micron darks: 3-second exposures (top left),
4-second exposures (top right), 10-second exposures (bottom left), 30-second
exposures (bottom left).
Figure 2: RAW data ramps for pixel (55,2) (left) and pixel (55,61) (right).
Note the offset at read #2 for pixel (55,2), which is typical of pixels at
the bottom of the array. The offset slowly decreases relative to the rest of
the ramp for pixels further from the bottom of the array.
Figure 3: Residual image of the on-orbit 3s dark subtracted by
a ground dark of the same exposure time.
Due to the lack of scattered light, and the similarity to ground data, suitable dark calibration images can be obtained in orbit. Taking into account the read-2 bias in the slope calculations at short exposure times, the dark current is roughly 1.5 DN/s at all exposure times; thus, we can proceed with the planned use of only 10-second darks for routine calibration.