Representing Celestial Coordinate Systems - 10

Gnomonic projection: pixel grid with pixel, linear, native,
& celestial coordinates

Prev 1 2 3 4 5 6 7 8 9 10 11 Next

What aspects of the WCS proposal are specific to FITS, and what would we do differently if starting again from scratch (e.g. XML)?

  • All angles in degrees?

    OK, suits astronomy and software.

  • Left-handed (X, Y) coordinates?

    Required for but can be very confusing.

  • Define the Euler angles by specifying the celestial coordinates at the reference point (CRVALi)?

    Also necessary for first-order behaviour but complicates non-zenithal projections.

  • Require at all?

    Yes, it ties the celestial coordinate system to a particular point in the image.

  • Encode the algorithm type in CTYPEi?

    No, too restrictive.