MIPS 70-micron scan image mosaic

Summary:

We will demonstrate use of the software tool MOPEX to create a mosaic of 70-micron images obtained with MIPS. This mosaic can be compared to the mosaic produced with MOPEX as part of the standard data processing pipeline.

Background Preparation:

In order to follow this demonstration, you will have to do the following in preparation:

• Download and read: the MIPS Data Handbook, the MIPS chapter of the Spitzer Observer's Manual, the Spot and Leopard User's Guides, the Post-BCD Software User's Guides (specifically the overview and mosaicker guides), and the MIPS Data Products and Filenames documentation.


• Install the necessary software on your computer: Leopard (and Spot) and MOPEX.

Sample Data:

For this demonstration, we will be using MIPS 70-micron scan mode data obtained as part of the Galactic First Look Survey (GFLS). Specifically, we will use the AOR galscan105.6p3 from program 104 (AOR reqkey 4961280). Use Leopard to download this data (e.g., see the detailed instructions for downloading data in the Retrieve Data with Leopard). Unpacking the data will create a directory called r4961280/. The 70-micron data are stored in the ch2/ sub-directory.

Create a Mosaic with MOPEX:

1. If you wish, you can examine the mosaic produced automatically by the data processing pipeline. This is the file r4961280/ch2/pbcd/SPITZER_M2_4961280_0_1_E165918_msaic.fits. As of this writing, the GFLS data available in the archive were processed with the S10 pipeline, whereas the current version of MOPEX (and the various namelists we will use - see below) have been optimized for version S11. Thus, we may expect that our mosaic will look slightly (but not significantly) different from the current version of the pipeline-generated mosaic. Reprocessing of the archive data is underway, and the S11 version of the GFLS data will be available in the near future.
   
   You can use any FITS viewer to examine the mosaic image; an example using SAOIMAGE-DS9 is shown here (note that the mosaic covers a strip several degrees in length - only a small portion is seen in the magnified main viewing window):
   
   


2. Obtain the appropriate namelist and pmask files to use with these data. The former is a parameter file for running MOPEX, while the latter is a mask of permanently bad pixels in the 70-micron array. These files are available on the post-BCD website or directly from these links: mosaic_M2.nl and mips70_pmask.fits.


3. Use your favorite text editor to ensure that the value of the PMASK_FILE_NAME parameter in the namelist is set to mips70_pmask.fits.


4. To create the mosaic with MOPEX, we will use the Basic Calibrated Data (BCD) files located in r4961280/ch2/bcd/. The first step is to create input lists for the BCD files (*_bcd.fits), uncertainty images (*_bunc.fits), and mask files (*_bmask.fits). The input order for the BCD files does not matter, but the bunc and bmask files must be listed in the same order. Sample input lists are available here:
   
   
   • bcdfiles
   • buncfiles
   • bmaskfiles
   
   Store these files in the r4961280/ch2/bcd/ directory (along with the namelist and pmask files). Note that in order to streamline the data processing for the purposes of this demonstration, we have only included half of the available images in the sample input lists. If you wish, you may repeat the demontration using all of the available data.


5. We are now ready to run MOPEX. (Don't forget to source the MOPEX set-up file, mopex.csh.)

unix% mosaic.pl -I bcdfiles -S buncfiles -d bmaskfiles -O bcd_out -n mosaic_M2.nl

This will take several minutes to complete (e.g., about 15 minutes on a "standard" SunBlade 100). MOPEX creates a sub-directory called bcd_out/; the output image is located at bcd_out/Combine/mosaic.fits. Again, the mosaic image can be examined using your favorite FITS file viewer:




6. The "striping" of the mosaic image (which is visible in the output of both the automatic pipeline and MOPEX) is a data processing artifact caused by differences in detector response that are not removed by the pipeline at the BCD level. The striping in the GFLS data is particularly strong because the data were obtained prior to a bias change in the MIPS germanium arrays that was implemented specifically to address this artifact. This effect can be mitigated in the GFLS data by running MOPEX on the filtered BCD images (*_fbcd.fits) produced by the data processing pipeline, which are also located in the r4961280/ch2/bcd/ directory.

   NOTE, however, that although the resultant mosaic will look better (see below), the filtering process renders the flux calibration of any extended emission in the image unreliable. Point source photometry should still be fine for targets fainter than 0.5 Jy at 70-microns (brighter targets will have some flux removed; i.e., they will actually be brighter than reported by photometry on the filtered mosaic image).

   Construct a new BCD input list for the filtered files, fbcdfiles, and run MOPEX again:

unix% mosaic.pl -I fbcdfiles -S buncfiles -d bmaskfiles -O fbcd_out -n mosaic_M2.nl

You can use the same input file lists for the bmask and bunc files, as well as the same namelist and pmask files, as were used for running MOPEX on the unfiltered BCD data. MOPEX creates a sub-directory called fbcd_out/; the output image is located at fbcd_out/Combine/mosaic.fits. Again, the mosaic image can be examined using your favorite FITS file viewer:




7. The following image shows a side-by-side comparison of the three mosaic images used in this demonstration: automatic data processing pipeline (left), MOPEX with unfiltered BCD data (middle), MOPEX with filtered BCD data (right). All three panels show the same region of sky at the same intensity scaling (keep in mind that only half of the available BCD data were processed with MOPEX in this demonstration, so the left panel shows slightly more sky - in the upper left corner - than the other two panels).

Return to the Data Analysis Demos main page.

Last modified on