[vsnet-chat 2953] Re: Survey bookkeeping

[Brian Skiff <bas@lowell.edu>]

Dr. Paczynski,
     Thank you for your interesting note.  I agree that the avalanche of new
published variables with lightcurves will overwhelm the data centers, to say
nothing of the GCVS group in Moscow.  My immediate efforts as regards the
surveys with small instruments has been in the hopes of precluding the addition
of new entries into SIMBAD of stars that already have some fairly well-
established designation; my opinion is that ASAS, ROTSE, et al. should not
invent new proprietary names for these stars unless absolutely necessary.  The
problem is that at present the SIMBAD folks do not make cross-identifications
unless they are given explcitly in a paper, so it is easy to find objects with
as many as six separate entries in their database (ones with two and three
entries may constitute some 20-30% of the database).  At the survey end, 
authors are clearly misinterpreting their results by making no search at all
for the stars they find (the ROTSE list is an egregious example).  The OGLE and
MACHO searches are using large apertures, so most of the stars they find are
uncatalogued, or appear only in USNO-A2.0 for instance, which is a transient
catalogue whose designations should not be promulgated.  For these it is
reasonable to create a new name for objects of interest.  The surveys being
done with telephoto lenses might well want a running number for internal
bookkeeping, but they shouldn't publish those names since practically all the
stars appear in Tycho-2 and/or the GSC, or whatever is coming next in the
"catalogues of precision" arena.  (Tycho-2 is a transient catalogue, too, but
the adoption of the GSC numbers by the Tycho consortium has ironically made the
GSC numbers more permanent.)
     You ask about a comprehensive list of "things".  To a great extent, and
to the extent that the literature provides accurate coordinates, that list is
essentially the SIMBAD database.  Most of its limitations are those of the
literature, although certainly much work needs to be done linking objects and
obtaining accurate positions for everything already there.  (The situation for
extragalactic objects is far superior thanks to the efforts of the NED team.)
     Sometimes I have wondered whether there needs to be something analogous
to the Minor Planet Center, but for variable stars.  This would collect
observations and identifications, make linkages, lightcurves, etc., and act
as the "official" (IAU-sanctioned) clearinghouse for assigning designations
(and priority of discovery, I suppose, as does the MPC for asteroids/comets).
This is essentially the task of the GCVS group at the Sternberg Institute, but
they are used to working on the timescale of a year or two, although this is
getting better.  Thus we have the current problem of knowing whether an object
is a new variable or not, which Taichi Kato has tried to ameliorate by making
a big list of reported but undesignated variables---like a list of asteroids
having only preliminary designations and insecure orbits.  Perhaps there will
be a big rush of new discoveries for two or three years, after which a lull
follow during which the GCVS and SIMBAD can catch up.  But when one sees the
marvelous lists and lightcurves occupying entire issues of 'Acta Astronomica',
it is reasonable to wonder whether something two or three orders of magnitude
more efficient is required to organize things.
     The Minor Planet Center operates with only _two_ people doing the vast
majority of the work.  Their observational database is growing at the rate of
a few hundred thousand positions per month; five years ago the total archive
was less than one million.  The rate of asteroid "numbering" (analogous to the
GCVS assigning a permanent variable-star name), formerly fairly steady at
150-200 per year, now increases at ~300 per _month_.  It works only because the
MPC insists that data be submitted in a specific format, and much of the work
is highly automated, but still large numbers of individual checks are made
because of the probabilistic vargaries involved in orbit determination (read:
lightcurve classification).  Batches of improved orbits are published each day.
     Part of the problem of setting up such an operation is finding people
who have the right mental outlook.  Most astronomers would regard it as just
bookeeping/bottle-washing/button-sorting, and not "Real Science" and thus
unimportant---which is why both SIMBAD and NED are ridiculously understaffed:
astro-accountants are hard to find, and review committees think spending money
on it is a waste.  The committee members still expect it to happen, but 
rarely think of doing that tedious work on their own projects, or even
designing the products so that they are amenable to being worked on by the
database people.
     Well, this would deal only in one dimension with a specific class of
stellar objects (no proper motions, radial-velocities, etc.).  The big
extragalactic surveys are a whole 'nother deal.  And how this ties in with
notions of a worldwide "virtual observatory" where raw/reduced data are
available on-line is something people smarter than me will have to ponder.

\Brian

P.S.:  The designated LMC variables are contained in volume 5 of the GCVS,
which covers extragalactic variables.  Most of the stars there have accurate
(~1") positions, and so are readily identifiable in new surveys.  See:
http://vizier.u-strasbg.fr/viz-bin/Cat?II/214A   for some details and access.

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp