Author: Brian Skiff Date: 951030 Revision: #0 951030 Key Words: astrometry, photometry, filters, asteroids, catalogs
For images involving any reasonable sky coverage (i.e. more than a square degree or so), the PPM is the astrometric catalogue of choice. If the 90,000 star supplement is included, it has nearly twice as many stars as the nearest competitor, viz. the ACRS. Both have similarly-small errors at the current epoch (around 0".3), and so at the image scales involved with TASS, errors are essentially zero compared to the scatter that will derive from having 3-4" pixels. Because you won't be getting high precision, even a few stars (like four or five) will be enough for an astrometric solution. The PPM is statistically complete to perhaps V ~8.5, never mind that two-thirds of the stars are fainter than that. Plenty of stars in the PPM at V ~10. If you really need fainter stars (i.e. higher density) there is no serious problem with using the GSC, again especially at the image scales involved. The GSC is being used by both professionals and amateurs for narrow-field asteroid and comet astrometry, and there seems to be no serious problems with results at the 1" level.
Soon enough we will have the Hipparcos and Tycho "output" catalogues. I really don't think these are going to solve any problems faced by people doing practical astrometry or photometry. For the 100,000 or so stars in the Hipparcos list, we'll have these terrific positions and motions, but already the PPM/ACRS are better than most of us can measure at ordinary image scales. The photometry won't give us much either, since already at V=8.0 the precision has fallen to 0.1 mag. The extended Tycho mission results promise positions for one or two million stars to mag. 12 or so, but the photometry will be in the half-mag. or worse range, so not any better than what's in the PPM, say.
The discussion of photometric standards seemed to be limited to the equatorial Landolt stars and the raw GSC magnitudes. The early TASS systems won't be scanning at the Equator and the GSC sucks as a photometric source, patently obvious in the Lasker et al. AJ papers describing the GSC in the first place. The GSC photometry was not intended to be used as a reference system, and it is unsuitable, sure enough. The GSC is also not an inventory of the sky to _any_ magnitude level, although claimed by many to be so. It is an engineering document, that's all. The deepest complete all-sky stellar inventory we have is the BD/CoD/CPD, published between 80 and 150 years ago!
As far as photometric standards, the obvious solution for TASS is to use the Guide Star Photometric Catalogue. This is the set of about six stars in the range 9 < V < 15 measured in V and B near the center of each sky survey plate used for the GSC itself (6-deg centers in the north, 5 deg in the south), encompassing some 9000 stars. This catalogue was published by Lasker et al. in ApJSuppl 68,1: it takes up the entire Sept 1988 issue of the ApJ Supplements. To save time in observing, sequences were adopted from the literature as available (e.g. clusters along the galactic plane). Evidence from my own observations and from a paper published by the South African Astronomical Observatory indicates these sequences are quite adequate for work at the couple-percent level. The big pixels of TASS (again) will mean you'll be lucky to get better than a few percent precision, even from good photometric sites, so sequences such as these will be just fine (Landolt stars will be overkill). Note also another bundle of much fainter stars in Humphreys et al. 1991 AJ 102,395, again structured around the POSS-I field centers. I think the goal of whatever photometry is done with TASS should be to maintain reasonable _accuracy_ on some standard system (R probably) rather than to worry much at all about internal _precision_, which you simply won't get with the big pixels. Since the GSPC stars are scattered all over the sky, a TASS system will scan across one or more of these fields on a regular basis (in the north they're roughly half an hour of sidereal time apart).
You and Alain Maury (perhaps others) gave the pros and cons about using filters. Alain is certainly right about the magnitude loss, which will cut quite severly into the discovery rate of anything. You are right that being able to report a magnitude that's more-or-less on a standard system is also valuable. Given that the systems will likely be deployed at crummy sites (suburban backyards), I would suggest using a weak "orange" long-pass filter that cuts on just redward of the sodium lines. This kills most of the worst effects of light pollution, and combined with the natural sensitivity of CCDs, gives you a photometric passband transformable to standard Cousins R while maintaining throughput. Something equivalent to a Wratten 23A filter should do the trick, which you can get in standard filter mounts for the 135mm lenses at the local camera store. A Wr25 probably is too dark, not reaching full transmission until about 6200A.
Someone mentioned asteroid orbital-element catalogues, citing various secondary sources. Might as well use the original stuff that's complete and revised daily: see Ted Bowell's Web page, where he has about 13,000 orbits (and another 15,000 to come!) already integrated and ready for use. The present selection includes all numbered objects plus everything with at least a 40-day arc. The elements include perturbations by all the major planets (Earth and Moon separately) plus Ceres, Pallas, and Vesta. Many of these are unpublished results deriving from our astrometric work at Lowell on Palomar and UK Schmidt survey plates, prints, and films, as well as on 18-inch Schmidt films taken by the Shoemakers (their 25,000+ film collection is now here) and other astrograph collections in our possession. Ted's homepage is at the URL:
http://www.lowell.edu/users/elgb/
...and the element file is at:
ftp://ftp.lowell.edu/pub/elgb/astorb.html
...which is linked from his Welcome page. Note that despite being an ftp address, an html-style page comes up. TASS people may also be interested in the LONEOS program, which will involve scan-mode observations for near-Earth asteroids with a CCD array on a Schmidt. A description of this is also given a page at Ted's address. Once this gets running (next spring?), we'll certainly need help following up on the many hundreds of objects we'll be finding each month (we're talking 10Mb/minute data rates here for the full-up system). The LONEOS page has some comments about how amateur observers might contribute.