Spitzer Documentation & Tools
MIPS Instrument Handbook

8.2.9        Example: 160 micron bright extended source

In this section, we describe how to reduce 160 micron data for bright extended sources. The standard 160 micron default, non-filtered reduction for extended sources may be good enough for most purposes; see Figure 8.15.  You may be able to improve on the standard reduction by making a corrected illumination correction file from the data themselves and applying the correction before mosaicking.  If the DCE after the stim is significantly brighter than following BCDs, you may not want to include this DCE in the mosaic, depending on the level of redundancy in the data set. 


The use of median filtered data is not recommended for extended sources at 160 microns; see Figure 8.15.


A second example of a galactic object (the cluster L1228) on a bright background appears in Figure 8.16 through Figure 8.18.  Again, constructing a mosaic from the filtered BCDs is not recommended.




Figure 8.15: Mosaic of unfiltered (left) and filtered (right; DO NOT DO THIS!) 160 micron scan BCDs of NGC 300.  On the left, the dark and bright ''dotted lines'' are due to bad stim-subtracted solutions for some pixels due to the bright object.  Note on the right among other things the dark sidelobes introduced by the filtering.  See text.


Figure 8.16: Scan map of the molecular cloud L1228 at 160 microns from the Galactic First Look Survey (2 degrees long).  The image was created using MOPEX and the unfiltered BCDs, and illustrates how well the 160 micron array behaves on a relatively bright background.


Figure 8.17: As above, but using the filtered BCDs; the linear stretch is the same in both images.  In this case the 'damage' on the intensity of the mosaic is not horrendous, but the flux level has been modified (see next figure below).


Figure 8.18 :  The integrated flux of the L1228 mosaic on the cross-scan direction along the scanning direction for the unfiltered (top line) and filtered (bottom line) BCDs.  The flux density using filtered BCDs has dropped by nearly a factor of 3.