[fitswcs] WCS for co-moving maps

David Berry davidstuartberry at gmail.com
Mon Sep 14 03:49:12 EDT 2015


On 13 September 2015 at 13:45, Dirk Petry <dpetry at eso.org> wrote:
> Hi,
>
> OFLN/OFLT sounds interesting also for our purposes at ALMA.
> I will have a look at the documentation.
> Is anyone working towards making this part of the FITS standard?

Not that I know of.

> How widely is this adopted, yet?

It was news to me that this specific facility of AST was used by
anyone outside JCMT, although the AST library itself has a much wider
audience including DS9 (which does all its WCS using AST).  Here's an
example header that represents polar offsets centred on M31 (i.e. the
north pole of the primary WCS lon/lat system is centred on M31, and
the secondary WCS gives the corresponding absolute RA/Dec values):

CTYPE1  = 'OFLN-TAN'
CTYPE2  = 'OFLT-TAN'
CRVAL1  =     5.78640635151322
CRVAL2  =     89.9138940173094
CRPIX1  =                150.5
CRPIX2  =                150.5
CDELT1  =                 0.01
CDELT2  =                -0.01
EQUINOX =               2000.0
PC1_1   =    0.994891253923403
PC1_2   =   -0.100952428731157
PC2_1   =    0.100952428731156
PC2_2   =    0.994891253923403
MJD-OBS =     51544.4992571308
RADESYS = 'FK5     '
CRPIX1A =                150.5
CRPIX2A =                150.5
CRVAL1A =              10.6847
CRVAL2A =               41.269
CTYPE1A = 'RA---TAN'
CTYPE2A = 'DEC--TAN'
CDELT1A =                -0.01
CDELT2A =                 0.01
RADESYSA= 'FK5     '
EQUINOXA=               2000.0

and here's the equivalent 'Cartesian' offset systemr (i.e. this time
the (0,0) origin of the primary WCS lon/lat system is centred on M31)

CTYPE1  = 'OFLN-TAN'           / RA offset from 0:42:47.1, 41:11:00
CTYPE2  = 'OFLT-TAN'           / Dec offset from 0:42:47.1, 41:11:00
CRVAL1  = -0.00868123360008796 / Value at ref. pixel on axis 1
CRVAL2  =   0.0856672428320622 / Value at ref. pixel on axis 2
CRPIX1  =                150.5 / Reference pixel
CRPIX2  =                150.5 / Reference pixel
CDELT1  =                -0.01 / Degrees/pixel
CDELT2  =                 0.01 / Degrees/pixel
EQUINOX =               2000.0 / Equinox of coordinates
PC1_1   =                  1.0 / Transformation matrix element
PC1_2   =  -1.3273828411248E-4 / Transformation matrix element
PC2_1   = 0.000132738284112092 / Transformation matrix element
PC2_2   =                  1.0 / Transformation matrix element
MJD-OBS =     51544.4992571308 / Modified Julian Date of observation
RADESYS = 'FK5     '           / Reference frame for RA/DEC values
CRPIX1A =                150.5 / Reference pixel on axis 1
CRPIX2A =                150.5 / Reference pixel on axis 2
CRVAL1A =              10.6847 / Value at ref. pixel on axis 1
CRVAL2A =               41.269 / Value at ref. pixel on axis 2
CTYPE1A = 'RA---TAN'           / Type of co-ordinate on axis 1
CTYPE2A = 'DEC--TAN'           / Type of co-ordinate on axis 2
CDELT1A =                -0.01 / Pixel size on axis 1
CDELT2A =                 0.01 / Pixel size on axis 2
RADESYSA= 'FK5     '           / Reference frame for RA/DEC values
EQUINOXA=               2000.0 / [yr] Epoch of reference equinox

In the case of a moving object such as a planet, the secondary axis
descriptions are usually just used to define the orientation of the
offset axes within an absolute coordinate system

David



More information about the fitswcs mailing list