re OBJECT keyword

[Rob Seaman]

I agree with the group that argues for a new keyword (or keywords) to
contain an official IAU designation for an object.  The OBJECT keyword
was never defined to contain only official names, and no attempt has
ever been made by large segments of the community to enforce any
particular naming policy.  It seems more than a little late to start now.

However, I also agree that the data acquisition systems should ideally
be responsible for creating the OBJNAME or IAUDESIG keywords (preferably
only one or the other, not both).  Unfortunately, I doubt that I'm the
only one who can comment from experience on the unlikelihood of this
happening for legacy instrumentation.

Note that not all observations are associated with a particular IAU
designated object.  Does the Hubble Deep Field have an official name?
The DSS fields will have no IAUDESIG, for instance.  It would be
impossible to assign a OBJNAME to a "serendipity mode" observation
at some random telescope pointing.  What about names for individual
exposures from whole-sky or other telescope surveys?  The precise
intent of a survey is to provide fodder for only later generating
names, after all.

Note also that the IAU may not be the sole body that we need to consider
as being responsible for granting names.  Do we need to reserve other
xxxDESIG keywords?  (Might I suggest that a shorter abbreviation be
found for "designation" so that more than three characters remain for
the naming organization acronym?)  Alternately, is OBJNAME appropriate
for other than official IAU nomenclature?  Without reserving multiple
keywords to allow keeping multiple namespaces separate, a single keyword
may not be able to resolve naming conflicts.  Is an ambiguous official
name better than an unambiguous ad hoc name?

I'm also leary of relying entirely on positional identifications for
resolving names.  In crowded fields it is quite possible for the
boresight of the telescope to lie nearer to a happenstance object than
to the intended object.  On the other hand, an observation of, say,
halo objects of a nearby galaxy may result in a telescope pointing
that is far from the official center of the galaxy...while narrow
band observations of distant objects may use red shifted filters such
that otherwise centered foreground objects aren't really present in
the field of view in any appropriate scientific sense.  (I'm sure
others can come up with even more convincing examples.)

In general, assigning a single object name (an official IAU name or
otherwise) to each observation of a particular field-of-view relies
intrinsically on the intent of the observer.  And searching an archive
for all observations of NGC1234 (or is that ngc-1234?) is only one
desirable type of database query.

What about observations containing multiple objects?  What about
observations of specific regions of larger objects?  I may want to
search not only for observations of Jupiter, but specifically for
observations of the Jovian Great Red Spot.  Is our software to be
responsible for determining not only whether the GRS is in the
field-of-view, but also whether it is on the proper side of the
planet?  I presume the IAU has a lot to say about languages - but
does FITS have any specific requirements about the language used
to specify a name?  Abbreviations?  Is case significant?  ...

The nature of the problem has not been sufficiently well defined yet.
We shouldn't hurry to a solution until it is.

Rob Seaman
-- 
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