scamp & wcs fits header

[mvdb]

Hi,
I'm new to scamp, so I apologize in advance if there is an obvious solution to the
following question. I tried to run scamp on an image that contains an approximate
astrometric solution in the fits header. However, it appears that the keywords
are not recognized or misread. This is part of the output:
Group 1: 1 field at 19:54:27.61 -27:02:41.1 with radius 7.367'
bls12l.cat A1 P1 19:54:27.61 -27:02:41.1 7.367' 0.3051"

In fact, the field center is approximately at:
RA = '17:48:58.88' / right ascension (telescope)
DEC = '-29:31:50.2' / declination (telescope)

It is not obvious, to me, if the header is wrong or if I am running scamp with the wrong
parameter file. FYI, I attached an .ahead file with the relevant keywords and the config file.
I tried running scamp on another image (taken with another instrument) and in that case, the wcs info is correctly read in.

Any help or suggestions are welcome.
Thanks,
Maureen

[ioannis]

Maureen,

Can you provide a link to an example FITS image that contains the full header? I can try to reproduce the error.

Cheers,
John

[mvdb]

Hi John,

you can get the image by going to this anonymous ftp site:
ftp://cfa-ftp.harvard.edu/outgoing/maureen/scamp

Maureen

[ioannis]

Maureen,

The problem is that your header contains both WCS-compliant FITS keywords like CRVAL?, CRPIX?, etc., and IRAF/TNX astrometry header keywords like WAT?_???, etc. Since SE understands both systems, the root of the seemingly bogus output astrometry from SCAMP is likely originating in the SE catalogs, which you can check by looking at the RA,DEC values of one of your catalogs. (I came to this conclusion with one of my datasets by trial-and-error, not by reading the source code; in any case, a nice feature for SE might be to have it check for multiple, conflicting astrometry keywords in the header before processing.)

Since I've never managed to figure out precisely which IRAF/TNX keywords interfere with SE+SCAMP, what I usually do is generate a stripped-down header of each of my FITS images that contains the minimum amount of information that I need like AIRMASS, EXPTIME, etc., and the WCS FITS keywords: CTYPE?, CD?_?, CRPIX?, and CRVAL?. If you use IDL, there are some handy tools in the Goddard IDL library (e.g., extast, putast) that do most of this in the right format for you.

Good luck!
John

[mvdb]

Thanks John,
however, even with a minimal header it doesn't seem to work, i.e. scamp still reports the wrong field center
(below, I pasted in the new header). Also, I checked the RA & DEC values in the SExtractor output, and they look
fine.
Maureen
================================================================================​===========
Kclin_strip[2048,2048][real]: BLSS12
No bad pixels, min=0., max=0. (old)
Line storage mode, physdim [2048,2048], length of user area 4131 s.u.
Created Tue 00:00:00 01-Jan-1980, Last modified Wed 14:16:18 03-Dec-2008
Pixel file "Kclin_strip.fits" [ok]
EXTEND = T / File may contain extensions
ORIGIN = 'NOAO-IRAF FITS Image Kernel December 2001' / FITS file originator
DATE = '2008-12-03T19:04:53'
IRAF-TLM= '14:16:18 (03/12/2008)'
COMMENT FITS (Flexible Image Transport System) format defined in Astronomy and
COMMENT Astrophysics Supplement Series v44/p363, v44/p371, v73/p359, v73/p365.
COMMENT Contact the NASA Science Office of Standards and Technology for the
COMMENT FITS Definition document #100 and other FITS information.
OBSID = 'ct4m.ispi.20080520T101841' / Observation Identifier
OBJECT = 'BLSS12 ' / object title
UTSHUT = '2008-05-20T10:18:28.604' / UT shutter open
COADDS = 1 / number of coadded images
FSAMPLE = 1 / number of Fowler samples
EXPTIME = 5. / exposure time
DATE-OBS= '2008-05-20' / Date (dd/mm/yy) of observation
UT = '10:18:40.3' / UT of TCS coordinates
EQUINOX = '2000. ' / epoch of RA & DEC
AIRMASS = '1.469 ' / airmass
ZD = '47.2 ' / zenith distance (degrees)
FILTNAME= 'KA ' / Filter name
CTYPE1 = 'RA---TAN'
CTYPE2 = 'DEC--TAN'
CRVAL1 = 267.210730518875
CRVAL2 = -29.524603594309
CRPIX1 = 1189.27986099391
CRPIX2 = 991.451377142334
CD1_1 = -8.4754050878614E-5
CD1_2 = -3.4477302929100E-7
CD2_1 = 3.44843485514198E-7
CD2_2 = -8.4736734471095E-5

[ioannis]

Hi Maureen,

I did not explicitly test on your header -- the only thing I can see that might interfere is that the EQUINOX keyword should be type float, not a string (see the SCAMP thread entitled 'Scamp getting stuck on "Astrometric Matching" step').

To verify, after "repairing" your FITS header, I was able to run SE+SCAMP on your image successfully and get the right field center. Specifically, all I did was to remove all *LTV*, *WAT*, *RADECSYS*, *LTM*, and *WCSDIM* header keywords, and change EQUINOX to type float. The IDL code I used (maureen.pro) and input/output files are here:
http://sdss.physics.nyu.edu/ioannis/tmp/maureen/

Cheers,
John

[mvdb]

Hi John,
Thanks much for your help; with the repaired fits header scamp works as expected now.

Maureen