[fitsbits] FIT Keyword data dictionaries

[Peter Bunclark]

Doug Mink wrote:

> Peter Bunclark wrote in response to Doug Mink's citation of an image header
> with DATE-OBS= '1611-04-15':
>
> > The FITS standard tell us that DATE-OBS ``...value field shall contain a
> > character string giving the day on which the observations represented by
> > the array were made...''  so is it, in fact, an appropriate keyword to
> > represent the epoch of an artificially-generated map given that the
> > observations used were presumably made much later?
>
> In this case, the visible planets in the image were actually from the epoch
> of DATE-OBS.  If I had a catalog with proper motions deep enough to show
> the Milky Way well, I would have used it, and the stars in the image would
> have been of the appropriate epoch as well.  I might try it again with
> the Tycho-2 to see how it looks--I unsuccessfully tried the PPM, which was
> the biggest all-sky available when I did the project, just before the release
> of the Tycho catalog (whose faint star proper motions wouldn't have made the
> 17th century star positions any more accurate than the HST GSC anyway).
> Part of my project (without the FITS files) is online at
>
> http://tdc-www.harvard.edu/software/skypic/rome97.html
>
> -Doug Mink

Nice pictures, but I'm afraid I still disagree; DATE-OBS has to refer
to an actual observation; had you scanned in Galileo's notebook and
produced a FITS file you could legitimately have used DATE-OBS,
but in your case you mean DATE-MAP or something!

Cheers,
    Pete.