Spitzer Documentation & Tools
MIPS Instrument Handbook

Chapter 8    Introduction to Data Analysis

8.1  24 Micron Data

8.1.1   Overview and recommended steps for looking at your data

Generally speaking, the 24 micron data is well behaved and the pipeline reduction produces a final product with few artifacts. This means that the default pipeline Level 2 post-BCD 24 micron mosaics can, in many cases, be used for detailed analysis, although some care should still be taken. For instance, the 24 micron pipeline does not correct for latent images, and although it flags cosmic rays, it does not correct for them. It is also common to find a gradient in the background, which in some cases is zodiacal light, while in others it appears to be due to scattered light changing as a function of scan mirror position. To completely remove all these artifacts requires further data reduction. The basic steps to begin data reduction of 24 micron data are the following:

1. Check the online default Level 2 post-BCD mosaics (*maic.fits).

 

a. If this maic.fits file is ok, then you are done and can begin your analysis, although it is still useful to check the results by remaking the mosaic offline. 

b. If this maic.fits file shows latents (dark or bright), low-level jailbars, or gradients then you most likely need to generate your own mosaic.

 

2. Self-calibrate your data (see section 8.1.2) and re-make the mosaic using MOPEX using Overlap correction. This step often solves most of the remaining artifact issues. Be sure to use the EBCDs  that became available in pipeline S18.12 for photometric observations, which should also remove most of the remaining mirror spots.

 

3. If a gradient still exists in your final image that you would like removed, you may need to do a sky background removal (see section 8.1.3). Re-make mosaic using MOPEX. 

There are certainly a handful observations of extreme complexity that could require a more sophisticated data reduction than presented here in this basic overview, like, for instance, a field with a great deal of extended saturation. However in most cases, these three steps will be all that is required to produce science quality data.