Several of the MIPS-920 24µm standard observations (HD 58142 from campaign H and HD 159330 from campaign Q), as well as one of the scan maps from Q, were analyzed for this report. The data were processed as usual with mips_sloper and mips_caler (v2.41). The distortion correction was then applied to the individual calibrated DCEs using mips_enhancer. Aperture photometry was done on the star in each DCE from the datasets both with and without the distortion correction, using the usual 6-pixel aperture and pixel 12-16 sky annulus.
Figure 1 shows the resultant fluxes from the aperture photometry of the DCEs WITHOUT distortion correction, as a function of DCE, for two different standard photometry mode observations. Note the systematic offset between DCEs, at the 2-3% level; these correspond to alternating top-bottom array positions. There is also a slight increase in the mean flux between positions on the left and right sides of the array (DCEs 1-15 vs. 16-30). The effect has been shown previously in the MIPS-920 reports.
Scan maps also show this effect. Photometry of a point source at each of its positions on the array during the scan is shown in Figure 2. The top-bottom offset seen in photometry mode observations is seen here as a gradient, with a larger total magnitude (~8%, decreasing from the bottom to the top of the array) since a larger range of array positions are sampled in scan mode. Note that since some of the positions nearly overlap, the presence of source latents within the photometry apertures causes an artificial increase in the measured flux (e.g. the position near row 32).
The preliminary explanation that this photometry gradient was due to the read-2 offset effect has proven false - all the data shown here have been corrected for that effect. Since it is flux-dependent, the read-2 offset turns out to have almost no effect on point sources, and is only responsible for the gradient seen in the background at shorter exposure times. However, another possible source for the photometry gradient is the effect of image distortion. Figure 3 shows photometry plots identical to those in Figures 1 and 2, except that the data have now been corrected for distortion. The systematic offset in source flux between the top/bottom and left/right array positions in the photometry mode observations is now gone, as is the row-dependent photometric gradient in the scan map observations. The overall dispersion is also considerably improved, with 1 sigma now less than 1.5% for the flux standards. The scatter in the scan map measurements is larger, although much of it is caused by the latent contamination.
Conclusion: the distortion correction appears to improve the data, removing all systematics in the point-source photometry as a function of source position on the array.