DATE-OBS='31/12/99'

[Rob Seaman]

Lucio Chiappetti writes:

> Personally I believe that a date without a time is no use. In fact
> I've never used something like DATE-OBS (which is just a reserved
> keyword, not a mandatory one ...). I think there is no interest in
> knowing that a given measurement was done on a given date, if one
> cannot know the time !

There have been several other comments similar to this.

Clearly, a time resolution of a day is rarely scientifically useful.
The justification for including MJD (or JD) in FITS headers is just as
clear, although the associated complexities such as supporting multiple
systems of time may themselves be only rarely necessary.

But FITS headers are not only scientific documents, they are political
documents, and also often serve as the primary log sheet for an
astronomer's observing run.  The date is usually the easiest way to
select one proprietary set of data from another - at least I find this
true of data from general-purpose, public, ground-based, optical-IR
observatories.

More to the point, a date and a time don't just represent different
sampling intervals from an inertial continuum of time - our telescopes
are anchored to a moving Earth.  This is true even for the space-based
telescopes - at least the ones in Earth orbit.  All our data contain
diurnal (and annual) signatures of one kind or another.  Keeping the
date separate is the same thing as keeping the time separate as a
searchable keyword.

A not insignificant bookkeeping issue for both MJD and ISO 8601 is the
simple length of the keyword value.  Measuring the time of an observation
to millisecond accuracy requires 13 significant figures for MJD and 17
significant "figures" (ignoring punctuation) for a fully qualified ISO
string.  Not only may a particular series of observations not require
that much absolute precision, they may also be better served by times
recorded as relative offsets.

No single date/time mechanism suffices for all astronomical data.
FITS should remain flexible on this issue.

Rob
-- 
seaman at noao.edu, http://iraf.noao.edu/~seaman
NOAO, 950 N Cherry Ave, Tucson AZ 85719, 520-318-8248
PGP: 98 8D 8B 49 74 9A 41 88  3A 43 87 54 51 BF 30 4B