How many hot subdwarf stars were rejected from the PG survey

The Palomar-Green (PG) survey for UV-excess objects selected objects with U-B<-0.46 for spectroscopic follow-up. The color selection was done from photographic photometry, with typical error sigma_(U-B)=0.38. Spectroscopic detection of the Ca II K line in color-selected candidates was thought to indicate that a metal-weak cool star (sdF-sdG) had entered the list owing to photometric errors. About 1100 such "K-line" stars were rejected on this basis from the final published PG catalogue, as not having "genuine" UV excesses. However, another possibility is that some of these objects are composite (binary) stars, consisting of a hot subdwarf (sdB or sdO) and a cool companion (F-G-K). Recent interest in binary-star formation channels for sdB stars in particular has brought renewed attention to the issue of completeness of lists of known hot subdwarfs, especially ones with cool companions. We have studied the nature of the "rejected" PG candidate stars, by assembling available information for a subset of 173 stars between r magnitudes 14.0 and 16.0 that have photometric data from both Data Release 2 of the Sloan Digital Sky Survey (SDSS) and the Two-Micron All Sky Survey (2MASS). Two of the stars have SDSS spectra. Both from their location in (u-g,g-r) and (g-r,r-K_s) two-color diagrams and from detailed fitting of single-star models to the spectral energy distributions, we conclude that the vast majority of these stars can be interpreted as metal-poor F and G subdwarfs, consistent with the original interpretation by the authors of the PG survey. We discuss the seven outliers individually; these may plausibly be binary systems that include a hot subdwarf star as a member, or they are hot stars that entered the list of rejected stars by accident.