Introduction to SDSS FITS binary tables and Calibrated Object Catalog files

  • Where to grab the files from: subscribe to the sdss-obs@astro.princeton.edu mailing list.

    watch for messages regarding successful processsing of a run. When a new run becomes available, ftp to FNAL (you will need an acct (your computer guru should arrange this), anonymous ftp won't have it), grab a tsObj file or 6.

    Have the SDSS infrastructure thru astrotools setup on your machine. (your computer guru can arrange this).

    	Dir: /data/dp3.b/data/94/calibChunks
    
    

    filenames of the form: cd /data/dp3.b/data/94/calibChunks ls -l -rw-r--r-- 1 yanny sdss 859236480 Nov 14 20:49 tsObj-000094-1-0-0011.fit -rw-r--r-- 1 yanny sdss 967752000 Nov 14 21:28 tsObj-000094-2-0-0011.fit -rw-r--r-- 1 yanny sdss 946077120 Nov 14 21:31 tsObj-000094-3-0-0011.fit -rw-r--r-- 1 yanny sdss 953064000 Nov 14 21:33 tsObj-000094-4-0-0011.fit -rw-r--r-- 1 yanny sdss 921188160 Nov 14 21:03 tsObj-000094-5-0-0011.fit -rw-r--r-- 1 yanny sdss 864783360 Nov 14 20:45 tsObj-000094-6-0-0011.fit Note size (1GB per scanline), note starting frame number (11) if you read the header, you can find out number of fields (or grab the tsChunk* file, it's in there too): setup astrotools astrotools astls>set hdr [hdrReadAsFits [hdrNew] tsObj-000094-1-0-0011.fit] astls>hdrPrint $hdr astls> hdrPrint h0 RUN = 94 CAMCOL = 1 FIELD0 = 11 NFIELDS = 535 FILTERS = 'u g r i z' COBJ_ID = '1998-11-14T14:31:16 11249'

    So you know there are 535 fields per scanline.
  • source the selection/plotting TCL script:

    (these scripts are available from FNAL at /data/dp2.d/data/calibChunks/sample-scripts)

    	astls>source findQSOs.tcl
    	astls> info args findQSOs  
      run col field0 startfield nfields rootdir ilimit zlimit trustRHL gmr rmi imz
    
    	Think of some good parameters, try it out:
    
    	astls>findQSOs 94 4 11 505 10 /data/dp3.b/data 21.0 20.0 1 2.5 1.5 0.8
    
    	94: run number
    	 4: camera column (1-6)
    	11: start field for this run
           505: starting field to examine/plot
            10: number of fields to examine/plot
      /data/dp3.b/data: root directory where the tsObj file is stored.
    	21.0: i' must be brighter than this
    	20.0: z' must be brighter than this
    	1: do you trust RHL's deblender? 1=yes, 0=no
    	2.5: g' - r' must be redder than this
    	1.5: r' - i' must be redder than this
    	0.8: i' - z' must be bluer than this
    
    
    	This will produce:
    
    astls> findQSOs 94 4 11 505 10 /data/dp3.b/data 21 20 1 2.5 1.5 0.8
    field 505 size is 704
    field 506 size is 860
    field 507 size is 724
    field 508 size is 672
    field 509 size is 635
    field 510 size is 656
    field 511 size is 620
    field 512 size is 684
    field 513 size is 691
    field 514 size is 774
    run: 94 camcol:4 size: 4830
              run c fld id    row    col      RA          DEC
    Candidate: 94 4 510 155 1417.49 758.30 +03:25:04.8 +00:17:56.1 
    
    	 u      g    r     i     z     parentid
    	25.66 23.78 21.27 19.74 19.17 153
    
    
    =========
  • So, there's a candidate(!), look at the CC diagram created:

    postscript image ghostview starQQ-r94-c4.ps

    But, I want to check the atlas image, perhaps make a finding chart.

    if you have the corrected frames, you can make a finding chart.

    	
    	astls>source fc.tcl	#another handy script....
    	astls>fc 11 94 4 510 155 1417.49 758.30
    
    	11: start of scanline field number
    	94: run number
    	 4: camcol
    	510: field
    	155: id within field
    	1417.49: row within field
    	758.30: col within field
    
    	[pops up finding chart]
    	postscript image
    
    =============
  • You may wish to examine it more carefully in photo to look for deblending problems or bad flags.

    (you will need to get the photo source code, and everything it depends on and this isn't available via UPD quite yet....).

    	photo> set tab [objfileOpen fpObjc-000094-4-0510.fit 
    		fpAtlas-000094-4-0510.fit]
    {OBJfd h0} {ATLfd h1} {TSTfd {}} {ncolor 5} {OBJnrow 656} {filters {u g r i z}} 
    {run 94} {camcol 4} {field 510} {ref_band 2}
    photo> mtv_objc tab 155
    h2
    photo> pflags
    OBJC  CHILD NOPETRO BINNED1
    0     CHILD NOPETRO NOPETRO_BIG DEBLENDED_AS_PSF ELLIPFAINT
    1     CHILD NOPETRO PETROFAINT DEBLENDED_AS_PSF
    2     CHILD DEBLENDED_AS_PSF BINNED1
    3     CHILD DEBLENDED_AS_PSF BINNED1
    4     CHILD BAD_RADIAL DEBLENDED_AS_PSF BINNED1
    photo> 
    
    	postscript image
    	Flags are ok...
    	Check parent...
    
    photo>mtv_objc tab 153
    photo> pflags
    OBJC  BLENDED MANYPETRO BINNED1
    0     BLENDED MANYPETRO BAD_RADIAL BINNED1
    1     BLENDED BAD_RADIAL BINNED1
    2     BLENDED BAD_RADIAL PETROFAINT BINNED1
    3     BLENDED MANYPETRO BAD_RADIAL BINNED1
    4     BLENDED MANYPETRO BAD_RADIAL ELLIPFAINT BINNED1
    photo> 
    	postscript image
    
    
    Though it is near that brighter galaxy, Parent seems ok too, so this would be a good one to get a spectrum of, maybe?
  • Also, you can use the stand-alone code on the CDROM to make individual atlas images of children and parents in the same fashion, (just a bit more tedious, as you have to do it 5filters*(parent+2children)=15 times).
    	rd_atlas -c 2 fpAtlas-000094-4-0510.fit 153 ai.fit	#[r-band, child]
    	astls>regReadAsFits [regNew] ai.fit
    	h0
    	astls>saoDisplay h0
    	rd_atlas -c 3 fpAtlas-000094-4-0510.fit 155 ai.fit	#[i-band, parent]
    	astls>regReadAsFits [regNew] ai.fit
    	h0
    	astls>saoDisplay h0
    	etc...
    
    Script findQSOs.tcl