{"id":4854,"date":"2016-05-06T14:14:28","date_gmt":"2016-05-07T00:14:28","guid":{"rendered":"http:\/\/www.eaobservatory.org\/jcmt\/?page_id=4854"},"modified":"2020-11-23T17:50:03","modified_gmt":"2020-11-24T03:50:03","slug":"scuba-2-dr-tutorial-4","status":"publish","type":"page","link":"https:\/\/www.eaobservatory.org\/jcmt\/science\/reductionanalysis-tutorials\/scuba-2-dr-tutorial-4\/","title":{"rendered":"SCUBA-2 Data Reduction \u2013 Tutorial 4"},"content":{"rendered":"

Note:<\/strong> This tutorial assumes a pre-existing installation of the latest version of the Starlink software suite (available here<\/a>), and that it has been initialized. The steps outlined below are for the BASH<\/em> and ZSH shells; if the TCSH or CSH<\/em> shell is being used, please replace the export<\/em> command with the appropriate setenv<\/em> equivalent. See Step 4 of SCUBA-2 DR Tutorial 1<\/a> for an example of how to do this. Please also remember to run KAPPA and SMURF first.<\/em><\/p>\n

Matched Filters<\/h2>\n

Matched filters are a method of wavelet analysis used for optimal detection of point sources, by removing the large scale structure for a map and convolving the remaining information with an image of the telescope beam. Some ORAC-DR<\/em> recipes (e.g. REDUCE_SCAN_FAINT_POINT_SOURCES<\/em>) will automatically carry out matched filtering on the group coadds produced by the pipeline. This tutorial will demonstrate how to use PICARD<\/em> to perform the same filtering on an already-reduced map.<\/p>\n

The map used for this is a single observation with one visible faint point source present. Better examples of matched filtering can show a map going from appearing blank to having obviously detected point sources. When using matched filtering, care is needed in interpreting the flux of the output sources. See 6.1 of Starlink<\/em> document SUN\/264<\/em>, and section 8.7 and Appendix D of SC\/21<\/em> (The SCUBA-2<\/em> Data Reduction Cookbook) for more details on the SCUBA-2<\/em> Matched Filter.<\/p>\n

    \n
  1. Download the example data from here<\/a> and copy it into a working directory on your computer:\n

    JCMTTutorial-MatchFilter.tar.gz<\/code><\/p>\n<\/li>\n

  2. Gunzip and untar this file:\n

    tar xvfz JCMTTutorial-MatchFilter.tar.gz<\/code><\/p>\n

    You will now have a JCMTTutorial-MatchFilter<\/em> directory, containing a reduced SCUBA-2<\/em> observation in NDF format, and a README<\/em> file.<\/p>\n<\/li>\n

  3. Examine the files:\n

    cd JCMTTutorial-MatchFilter
    \nls
    \nREADME.txt jcmts20140729_00011_850_reduced001_obs_000.sdf<\/code><\/p>\n

    This observation is a reduced SCUBA-2 Pong<\/em> map taken by the JCMT<\/em> Cosmology Legacy Survey on 2014-07-29. The date and scan number may be deduced from the file name, and the file metadata can be examined with the KAPPA<\/em> command fitslist<\/em>:<\/p>\n

    fitslist jcmts20140729_00011_850_reduced001_obs_000.sdf<\/code><\/p>\n<\/li>\n

  4. Run the standard matched filter on it. Starlink<\/em> provides a PICARD<\/em> recipe for doing this, which can be run like so:\n

    export ORAC_DATA_OUT=. (setenv ORAC_DATA_OUT .\/ if you are using TC shell)
    \npicard -log sf SCUBA2_MATCHED_FILTER jcmts20140729_00011_850_reduced001_obs_000.sdf<\/code><\/p>\n

    Read the messages written to screen, which are also saved in a log file that begins with .picard<\/em>.<\/p>\n<\/li>\n

  5. Open the matched filter image (extension _mf<\/em>) in Gaia<\/em>, and compare with the original image. Various source finding and photometry methods can be used to compare the before and after cases. GAIA<\/em> enables the user to interactively perform photometry on a displayed file.<\/li>\n
  6. Look at the background structure of the matched filter image. There is a characteristic pattern visible in it, which is recognizable in comparison with a normal SCUBA-2<\/em> map. Note that when searching for data in the JCMT<\/em> Science archive, it may be possible for the user to recognize which observations were reduced with a matched filter just from this pattern visible in the preview image.\n
  7. The point spread function (PSF) used by the recipe is stored in the file with the _psf<\/em> extension.\n
    \"The<\/a>

    The standard Point Spread Function (PSF) used in SCUBA-2 Tutorial 4 (zoomed).<\/p><\/div><\/li>\n<\/ol>\n

    Advanced Tips<\/h3>\n

    This can be run on a standard JCMT calibrator source to see the effect on the beam – try downloading a reduced observation of (e.g.) Arp 220 from the JCMT<\/em> Science Archive here<\/a> and running this PICARD<\/em> recipe on it. Look at the difference in the source before and after running matched filtering. it is also possible to manually use the KAPPA beamfit<\/em> command to analyze the maps.<\/p>\n

    The SCUBA2_MATCHED_FILTER PICARD<\/em> recipe takes various optional recipe parameters. It is possible to set the following recpars<\/em> to change the defaults:<\/p>\n

      \n
    1. Change the point spread function (by default it will construct one based on the measured SCUBA-2 beam from Dempsey et al. 2013):\n

      PSF_MATCHFILTER<\/code><\/p>\n

      Name of an NDF file containing a suitable PSF. Must exist in the current working directory. If not specified, the recipe will calculate one itself for each input file.<\/p>\n<\/li>\n

    2. Change the normalization scheme used for the PSF:\n

      PSF_NORM<\/code><\/p>\n

      Normalization scheme used for the PSF created by this recipe if one is not specified using the above parameter. This may be PEAK<\/em> or SUM<\/em> to indicate whether the Gaussian PSF should have a peak of unity or a sum of unity. If not specified, the recipe assumes PEAK<\/em>.<\/p>\n<\/li>\n

    3. Whether to SMOOTH<\/em> the data (default is 1, indicating it should be smoothed):\n

      SMOOTH_DATA<\/code><\/p>\n

      Flag to denote whether or not the image and PSF should be smoothed and have the smoothed version subtracted from the original. If not specified, the recipe assumes a value of 1 (smooth and subtract).<\/p>\n<\/li>\n

    4. The FWHM of the Gaussian used to smooth the data:\n

      SMOOTH_FWHM<\/code><\/p>\n

      FWHM of Gaussian used to smooth data and PSF images before convolving with the PSF. If not specified the recipe assumes a value of 30 arcsec.<\/p>\n<\/li>\n<\/ol>\n

      \n

      Other JCMT data reduction\/analysis tutorials are available here.<\/a><\/strong><\/p>\n","protected":false},"excerpt":{"rendered":"

      Note: This tutorial assumes a pre-existing installation of the latest version of the Starlink software suite (available here), and that it has been initialized. The steps outlined below are for the BASH and ZSH shells; if the TCSH or CSH shell is being used, please replace the export command with\u2026 Continue reading<\/a><\/p>\n","protected":false},"author":41,"featured_media":0,"parent":4510,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":[],"_links":{"self":[{"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/pages\/4854"}],"collection":[{"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/users\/41"}],"replies":[{"embeddable":true,"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/comments?post=4854"}],"version-history":[{"count":54,"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/pages\/4854\/revisions"}],"predecessor-version":[{"id":11488,"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/pages\/4854\/revisions\/11488"}],"up":[{"embeddable":true,"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/pages\/4510"}],"wp:attachment":[{"href":"https:\/\/www.eaobservatory.org\/jcmt\/wp-json\/wp\/v2\/media?parent=4854"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}