The Astronomical Journal, 126:2209–2229, 2003 November E # 2003. The American Astronomical Society. All rights reserved. Printed in U.S.A. A LARGE, UNIFORM SAMPLE OF X-RAY–EMITTING AGNs: SELECTION APPROACH AND AN INITIAL CATALOG FROM THE ROSAT ALL-SKY AND SLOAN DIGITAL SKY SURVEYS Scott F. Anderson,1 Wolfgang Voges,2 Bruce Margon,3 Joachim Tru¨ mper,2 Marcel A. Agu¨ eros,1 Thomas Boller,2 Matthew J. Collinge,4 L. Homer,1 Gregory Stinson,1 Michael A. Strauss,4 James Annis,5 Percy Go´mez,6 Patrick B. Hall,4,7 Robert C. Nichol,6 Gordon T. Richards,4 Donald P. Schneider,8 Daniel E. Vanden Berk,9 Xiaohui Fan,10 Zˇ eljko Ivezic´,4 Jeffrey A. Munn,11 Heidi Jo Newberg,12 Michael W. Richmond,13 David H. Weinberg,14 Brian Yanny,5 Neta A. Bahcall,4 J. Brinkmann,15 Masataka Fukugita,16 and Donald G. York17 Received 2003 May 5; accepted 2003 August 12 ABSTRACT Many open questions in X-ray astronomy are limited by the relatively small number of objects in uniform optically identified and observed samples, especially when rare subclasses are considered or when subsets are isolated to search for evolution or correlations between wavebands. We describe the initial results of a new program aimed to ultimately yield 104 fully characterized X-ray source identifications—a sample about an order of magnitude larger than earlier efforts. The technique is detailed and employs X-ray data from the ROSAT All-Sky Survey (RASS) and optical imaging and spectroscopic follow-up from the Sloan Digital Sky Survey (SDSS); these two surveys prove to be serendipitously very well matched in sensitivity. As part of the SDSS software pipelines, optical objects in the SDSS photometric catalogs are automatically positionally cross-correlated with RASS X-ray sources. Then priorities for follow-on SDSS optical spectra of candidate counterparts are automatically assigned using an algorithm based on the known ratios of fx/fopt for various classes of X-ray emitters at typical RASS fluxes of 10À13 ergs cmÀ2 sÀ1. SDSS photometric parameters for optical morphology, magnitude, and colors, plus FIRST radio information, serve as proxies for object class. Initial application of this approach to RASS/SDSS data from 1400 deg2 of sky provides a catalog of more than 1200 spectroscopically confirmed quasars and other AGNs that are probable RASS identifications. Most of these are new identifications, and only a few percent of the AGN counterparts are likely to be random superpo- sitions. The magnitude and redshift ranges of the counterparts are very broad, extending over 15 < m < 21 and 0:03 < z < 3:6, respectively. Although most identifications are quasars and Seyfert 1 galaxies, a variety of other AGN subclasses are also sampled. Substantial numbers of rare AGN types are found, including more than 130 narrow-line Seyfert 1 galaxies and 45 BL Lac candidates. These early results already provide a very sizable set of source identifications, demonstrate the utility of the sample in multiwaveband investigations, and show the capability of the joint RASS/SDSS approach to efficiently proceed toward the largest homogeneously selected/observed sample of X-ray–emitting quasars and other kinds of AGNs. Key words: catalogs — quasars: general — surveys — X-rays On-line material: machine-readable tables 1. INTRODUCTION All-Sky Survey (RASS; Voges et al. 1999, 2000) covers the entire celestial sphere in the 0.1–2.4 keV range with the Although extrasolar X-ray sources were first observed four Position Sensitive Proportional Counter (PSPC; Pfeffermann decades ago (Giacconi et al. 1962), the first all-sky imaging et al. 1988) to a typical limiting sensitivity of 10À13 ergs X-ray survey has only recently been achieved. The ROSAT cmÀ2 sÀ1, although because of the scanning protocol, the 1 Department of Astronomy, University of Washington, Box 351580, Seattle, WA 98195; [email protected]. 2 Max-Planck-Institut fu¨r extraterrestrische Physik, Geissenbachstrasse 1, D-85741 Garching, Germany; [email protected]. 3 Space Science Telescope Institute, 3700 San Martin Drive, Baltimore, MD 21218; [email protected]. 4 Princeton University Observatory, Peyton Hall, Princeton, NJ 08544. 5 Fermi National Accelerator Laboratory, P.O. Box 500, Batavia, IL 60510. 6 Department of Physics, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, PA 15232. 7 Departamento de Astronomı´a y Astrofı´sica, Facultad de Fı´sica, Pontificia Universidad Cato´lica de Chile, Casilla 306, Santiago 22, Chile. 8 Department of Astronomy and Astrophysics, Pennsylvania State University, University Park, PA 16802. 9 Department of Physics and Astronomy, University of Pittsburgh, Pittsburgh, PA 15260. 10 Steward Observatory, University of Arizona, 933 North Cherry Avenue, Tucson, AZ 85732. 11 US Naval Observatory, Flagstaff Station, P.O. Box 1149, Flagstaff, AZ 86002. 12 Department of Physics, Applied Physics, and Astronomy, Rensselaer Polytechnic Institute, 110 Eighth Street, Troy, NY 12180. 13 Department of Physics, Rochester Institute of Technology, 85 Lomb Memorial Drive, Rochester, NY 14623-5603. 14 Department of Astronomy, Ohio State University, 140 West 18th Avenue, Columbus, OH 43210. 15 Apache Point Observatory, P.O. Box 59, Sunspot, NM 88349. 16 Institute for Cosmic Ray Research, University of Tokyo, Midori, Tanashi, Tokyo 188-8588, Japan. 17 Astronomy and Astrophysics Center, University of Chicago, 5640 South Ellis Avenue, Chicago, IL 60637. 2209 2210 ANDERSON ET AL. Vol. 126 exposure times and thus sensitivity limits vary markedly from rate redshifts, magnitudes, optical continuum slopes and the ecliptic pole to the equator. Depending on the level of emission-line identifications, intensities, and equivalent statistical significance discussed, 104–105 X-ray sources are widths are largely unavailable. contained in the RASS Bright and Faint Source Catalogs, Although the 103 total identifications from 103 deg2 of with more than 124,000 sources in the catalog versions con- sky provided by the EMSS (and similar later ROSAT sidered here. Positional accuracies derivable for the sources surveys) is an extremely respectable number, there remain vary with count rate, but moderately well-exposed sources scientific problems where the size of an identified subclass of typically have a positional uncertainty of 1000–3000.Many X-ray emitters of interest is uncomfortably small. Notable important inferences can and have been made from RASS examples include studies of the more rare subclasses of data based directly on the X-ray observations alone, with AGNs, such as narrow-line Seyfert 1’s (NLSy1’s) and BL minimal need for new correlative observations at other Lac objects. For example, among the large sample of more wavelengths. For example, the RASS point source catalogs than 400 AGNs from the EMSS, only about 40 BL Lac provide an exceptional measure of a portion of the X-ray objects were found, and yet, even this small subsample sug- log N/log S diagram (e.g., Voges et al. 1999). gested remarkable ‘‘ negative evolution ’’ (e.g., Morris et al. However, the scale of the effort involved in identifying a 1991) that demands further study. If such rare subclasses of large fraction of the abundant RASS sources, especially AGNs evolve in their properties with redshift and/or exhibit those in the Faint Source Catalog, poses an unusual analysis complex dependencies of X-ray luminosity with optical and/ problem. From ROSAT and also previous generations of or radio luminosity, as more typical quasars are known to do X-ray observatories, especially the extensive observations (e.g., Avni & Tananbaum 1986; Wilkes et al. 1994; Avni, of the Einstein Observatory (e.g., Gioia et al. 1984), such Worrall, & Morgan 1995), then an already small subsample X-ray source counterparts are known to include a highly must be further subdivided into yet smaller bins for analysis, heterogeneous mix of objects, ranging from nearby M leaving literally a handful of objects per bin. Another less dwarfs to distant quasars. In many cases, the X-ray data obvious example is that although modest- to high-redshift taken alone cannot unambiguously determine whether an quasars are not usually thought of as ‘‘ rare,’’ X-ray selection X-ray source is Galactic or extragalactic, much less finer dis- at the Einstein and RASS flux levels strongly favors the dis- tinctions about its nature. It has thus long been realized that covery of nearby and low-luminosity AGNs; in fact, the identification of optical counterparts is an essential entire EMSS sample of greater than 400 AGNs includes only companion study to large X-ray surveys. a small handful of X-ray–emitting quasars with z > 2, thus The most complete optical identification effort attempted limiting its utility for studies of the quasar X-ray luminosity on Einstein X-ray data was that of the Extended Medium function at moderate to high redshift. Sensitivity Survey (EMSS; e.g., Stocke et al. 1991). The In order to effectively build and expand on such earlier comparison with RASS is not inappropriate, since although large-scale counterpart identification programs, any new Einstein surveyed less than 10% of the celestial sphere, the effort (at comparable X-ray depth) must survey substan- flux levels and positional accuracy of source determinations tially in excess of the 103 deg2 covered by the EMSS and from the Einstein Imaging Proportional Counter detector must ultimately provide a sample numbering substantially were not greatly dissimilar to those from the ROSAT PSPC in excess of 103 optical identifications, obtained in a uni- (though, for example, the latter provides better positional form fashion. Moreover, if many more than 103 counter- information). Optical identification of sources in the EMSS parts are to be newly found on a reasonable timescale, more have been made in an exceptionally ambitious, decade-long efficient optical identification techniques than previously program (e.g., Stocke et al. 1991 and references therein), available must be used.