DATE-OBS='31/12/99'

[Steve Allen]

In article <4qrl92$r5q at noao.edu>, Rob Seaman <seaman at noao.edu> wrote:
>Question:  Are there any examples (digitized plates, for instance)
>of nineteenth century FITS data?

Give me a moment to run over to the archives and I'll make one :-)

More seriously, doing this right in an archival sense involves
significant historical research into the (often unwritten)
conventional assumptions of the era.  Without doing that it is pretty
likely that a modern interpretation of the logbooks will presume
something wrong.  And in that case the FITS file should contain
something that will prevent naive machine interpretation of the data.

I've run into quite a bit of that as I dug through the archives
finding the history of the published positions of Lick Observatory.
The adopted longitude value is directly related to the time standards
at Lick, but there's also the issues of why GMT did not equal UT and
other international conventions which were initially used with
suboptimal inputs.  Having the history of the adopted positions in
hand I *might* be able to disentangle the TIMESYS assumptions inherent
in the 19th century Lick plates -- at least for some of the observers.

>                                  Any keyword changes should handle
>pre-twentieth century data also.  (Someone else can suggest a solution
>for dates BC...)

Already available.  1 AD follows 1BC, but astronomically 1BC is year 0.
If this is an issue for FITS, however, then we would first need to
consider a CALENSYS keyword with values such as 'JULIAN', 'GREGORIAN',
and 'ORTHODOX' so that we can encode the DATE-OBS information from
19th century Russian plates.  With this in hand it should be possible
to scan Galileo's notebooks and convert them to FITS images with
correct WCS information.

I am being facetious with the CALENSYS suggestion, but interpretation
of the archival record of planetary observations demands the TIMESYS
keyword.  Values for the TIMESYS keyword in archival material could
include locally relevant tags such as the name of the observer at the
Meridian Circle who was then setting the clocks.  The value of TIMESYS
not being one of a few certified-machine-interpretable choices would
serve both to document historically and to prevent automated
misinterpretation.

--
Steve Allen          UCO/Lick Observatory       Santa Cruz, CA 95064
sla at ucolick.org      Voice: +1 408 459 3046     FAX: +1 408 454 9863
WWW: http://www.ucolick.org/~sla                PGP public keys:  see WWW