The ATLAS All-Sky Stellar Reference Catalog ("ATLAS-REFCAT2")

Draft version September 26, 2018
Typeset using LATEX default style in AASTeX62
The ATLAS All-Sky Stellar Reference Catalog
J. L. Tonry,1 L. Denneau,1 H. Flewelling,1 A. N. Heinze,1 C. A. Onken,2 S.J. Smartt,3 B. Stalder,4

https://ui.adsabs.harvard.edu/abs/2018ApJ...867..105T/abstract

Ed, I originally was excited about this catalog, especially for it being a potential Sloan filter source as well as other filters.

After reading the above referenced paper I became somewhat discouraged for this catalog as a potential sequence source.

A key point is that the catalog presents gri data as either a GaiaDR2 source or a Pan-STARRS source…. The big however, herein is that the Pan-STARRS data, as I understand it, is a composite of Gaia, GMP, Pan-STARRS, SkyMapper, Pathfinder, APASS, APASS DR9and Tycho-2/BSC; all being placed on the Pan-STARRS bandpasses using relations explained in the above referenced paper.

Once I developed a better understanding of how this source was created it occurred to me that for sequence purposes it would probably be wiser to use a direct data source rather than this catalog’s composite data, IMO.

As a result, I closed my file notes and no longer consider this a viable source, at least for sequence work.

Tim Crawford, Sequence Team

     Coming rather late to this, but:  I think the refcat2 g,r,i photometry, specifically, is in fact quite reliable, since it is based on new observations from the Pan-STARRS and smaller ATLAS telescopes as well as massaged and transformed data in existing catalogues.  For instance, they re-reduced all the southern-hemisphere APASS data (northward to +20 Dec) from the original images to improve the flat-fielding and smooth out the frame-to-frame and zonal errors.  Other stuff was done to the southern SkyMapper survey data.  Read the paper to find out the details.  Be sure to look at the errors shown for each star.  (The catalogue can be queried in VizieR as item j/apj/867/105)  You'll find that for stars brighter than mag 12 or so, the errors are often large (like ~0.1 mag), meaning they are from derived composite data, and are not reliable.  This bright limit varies around the sky.  Ordinarily the errors are < ~0.015 mag or so in each color for reliable data.

     Much of the z-band photometry, which happens to interest me most, is from the Pan-STARRS telescopes, so is saturated brighter than about mag 13.5 (some areas are brighter).  New brighter (like mag 9-12) z-band standards (and the other colors for that matter) will have to be manufactured, probably using the fainter stars in images of Landolt fields, for instance, and doing the transformations.  Hic labor, hic opus est.

     I think that (for now) the adopted global zero-point of the Pan-STARRS version of the Sloan system will have to be taken as 'gospel', but it would not surprise me if further improvements will become available (the LSST/Rubin survey telescope etc).

     An aside:  note that the transformations from griz to BVRI provided in the Pan-STARRS papers were derived from stellar models, _not_ from actual stars, such as Landolt or Stetson fields.  The will limit the accuracy to a few percent, rather worse than the data themselves.  Probably needs to be re-done.

\Brian