[fitswcs] Status of WCS negotiations

[Don Wells]

Mark Calabretta writes:
 > 
 > On Mon 1998/07/20 11:41:18 -0400, Don Wells wrote
 > >.., multiple combinations of linear mappings and nonlinear
 > >projections can be used to represent a piece of data..
 > >... For example, the geometry of a 2000-square image might be
 > >represented well by a TAN projection with radial terms, but if a
 > >500-square subimage is extracted from it the same accuracy might be
 > >achieved without the radial terms. ..
 > 
 > Do I understand you correctly that the subimage is of reduced resolution and
 > that there are three descriptions?  Before:
 > 
 >    1) The 2000x2000 image with the CDij matrix and a TAN+ projection (this
 >       CDij matrix being inapplicable to the subimage).
 > 
 > After:
 > 
 >    2) The 500x500 offset subimage with the same CRVALn but with the CRPIXn and
 >       CDij reset so that the value of (x,y) in the plane of projection is
 >       maintained, plus the original TAN+ projection.  This describes exactly
 >       the same coordinate system as (1), or
 > 
 >    3) The 500x500 subimage with truncated TAN+ projection and with CRVALn,
 >       CRPIXn and CDij subtly modifed to account for the discarded radial
 >       terms - effectively a new plate solution.

Yes, you have understood my somewhat contrived example. Version (3) is
what is done every day in production solutions for optical WCS. As I
pointed out in another posting today, every optical WCS solution
produces a different combination of all of the parameters. Note that
in many systems the CRPIXn are poorly determined; if the CRPIXn
change, the CRVALn must change too, and you will find that the CDij
(or equivalent linear PROJPk terms) will also change.

 > The simple answer is not to throw away the radial terms - why would you?

Over a smaller field the radial distortion terms may be negligible and
a simple TAN projection may be sufficient. This may appeal to some people.

FOOTNOTE: today there are multiple catalogs of calibrator sources. It
would be reasonable to reduce the same image against multiple
catalogs, producing multiple WCS solutions. Support for multiple sets
of WCS keywords would be nice in this situation---at data analysis
time the user could switch WCS versions and remeasure program sources
to see how much the catalog differences propagate into the positions
of the program sources.

 > ...  You can have
 > multiple.. descriptions without inflicting multiple CD matrices on people
 > who aren't interested in them.

``The [man] doth protest too much, methinks.''

If we decide to support multiple versions of sets of WCS keywords, the
incremental cost of support for multiple linear transformations will
be lost in the general cost of parsing multiple sets of keywords. I
think that the real question is whether the community really wants to
have support for multiple sets of WCS keywords---I have inferred from
discussions with various people (e.g. Doug Tody) over the years that
this is a requirement for several applications.  OK, readers of
'fitswcs', it is time to declare your requirements:

    Who among you need multiple WCS solutions? 
    Do you need multiple entire sets of WCS keywords, including CDij?

-Don
-- 
  Donald C. Wells         Associate Scientist         dwells at nrao.edu
		    http://www.cv.nrao.edu/~dwells
  National Radio Astronomy Observatory                +1-804-296-0277
  520 Edgemont Road,   Charlottesville, Virginia       22903-2475 USA