UBVb - uvby - MK - Radial Velocity Database For Stephenson-Sanduleak Southern Luminous Stars B. Cameron Reed Department of Physics Alma College Alma, Michigan 48801 e-mail: reed@alma.edu (517) 463-7266 CURRENT STATISTICS (before substitution of 1997 file for radvel.dat) Date of last modification: 14 Sept 1996 Stars in LS Catalog 5132 Total number of references 359 UBVb References - observations 250 9736 MK References - observations 150 6442 uvby References - observations 45 1215 RV References - observations 49 1072 Total Observations 18,465 Stars with UBV 3663 Stars with b 2356 Stars with MK 2598 Stars with UBV + b 2166 Stars with uvby 830 Stars w/ UBV + b + MK 1666 UBV + b + MK + uvby 703 Stars with radial velocities 864 UBV + b + MK + uvby + RV 371 Stars with any data at all 4084 1. Introduction Since its publication over two decades ago, Stephenson and Sanduleak's (1971) Luminous Stars in the Southern Milky Way (Pub. Warner and Swasey Obs., vol. 1, no. 1; hereafter, "LS" catalog) has served as a starting point for numerous observing campaigns directed at mapping the distribution of spiral features and reddening in the southern-hemisphere Milky Way. Consisting of positions, approximate photographic magnitudes, rough spectral classifications and finding charts for 5132 stars, the LS catalog is a fairly complete guide to the OB stars and A and F giants and supergiants within 10 degrees of the galactic plane between galactic longitudes l = 186deg and 381deg to a limiting photographic magnitude of about 12.5. A vast amount of information on the LS stars now exists distributed throughout the astronomical literature. Given the value of these stars to synoptic galactic structure programs, a compilation and uniform analysis of available information would appear to be worthwhile. This work presents the first step in such an effort: a compilation of published photoelectric UBVb, and Stromgren uvby photometry, MK classifications, and radial velocities (RV's) for the LS stars. 2. Strategy The primary criteria applied for incorporation of data into this compilation are originality and adherence to the UBVb, uvby, and MK systems. Homogeneity of radial velocities is of course a more difficult issue. This work began as a collection of papers in the author's possession reporting LS star photometry and spectroscopy. Stars not appearing in this initial collection (the vast majority) were queried on the Simbad database and a star-by-star index-card file of references and data was constructed. As each reference was acquired, it was checked as thoroughly as practicable for LS stars by cross-referencing finding charts and identifications such as HD, CPD, CD and HR numbers. In computerizing the data, a separate file is created for each reference. In practice, this means separate files for UBVb, MK, uvby, and RV data. This practice is carried out even if a paper reports, for example, uvbyb photometry: two separate files are generated (this is because the initial version of the database consisted only of UBVb), though the paper involved is assigned a single "master reference number". The rationale for this is to render straightforward future corrections and updating. Each reference was examined for any relevant notes (emission, variability, duplicity, etc.); these are recorded via numeric codes. In a compilation of this size, complete consistency is impossible to achieve: readers interested in any particular star are urged to treat this work as a sourcebook and consult original references whenever possible. 2.1 UBVb and uvby Photometry Excepting a very few values representing averaged observations, only photoelectric data reported in the original research literature have been incorporated. Care was taken to exclude any results adopted from or that derived from averaging or homogenizing data from different sources: what appears here is, as much as possible, unadulterated. V magnitudes deriving from uvby photometry have not been incorporated. A chronic problem area was the number of measurements reported for each star: while most authors state this explicitly, some give only a minimum number and others no details at all. If no information along these lines could be gleaned, the number of measurements was taken to be one for compilation purposes. A number of UBV observations (134 at this writing) appear in older references not easily available to this author. However, they do appear in their original form in the U.S. Naval Observatory photoelectric catalogue (Blanco, et. al., Pub. USNO, 2nd series, vol. XXI, 1968). These have been incorporated here as coming from a single reference, number 180. Future work aimed at homogenizing and analyzing the present compilation should therefore drop data attributed to reference 180 as they cannot be regarded as internally consistent. In a few cases older data for a subsequently much-observed star has not been tracked down, however, to my knowledge no LS star for which photoelectric (or spectroscopic) data are available has escaped compilation. Some LS stars are well-known variables with extensive light curves available; for these only an indication that the star is a named variable is given. The same note codes are used for both UBVb and uvby data. 2.2 Spectroscopy The general precepts used in assembling the photometric compilation are retained here: only original data from references directly available to the author have been incorporated. A three-stage process was followed in assembling the spectroscopic data. First, all photometric references which also contained original spectral classifications were incorporated; this proved to be some 70 papers. Second, classifications from "large-scale" sources such as the Stephenson-Sanduleak catalog itself, Houk's Michigan Spectral Survey, Hiltner, Garrison, and Schild (ApJ 157, 313, 1969 =1969ApJ...157..313H), and Garrison, Hiltner and Schild (ApJS 35, 111, 1977 =1977ApJS...35..111G) were incorporated. Lastly, the Simbad database was queried for MK classifications for every LS star and notes made as to references to be acquired and reviewed for suitability. Most classifications derive from grating, prism, or objective-prism spectra; in a few instances they derive from equivalent widths, microphotometer tracings or the like. Such "non-traditional" classifications are noted with corresponding numerical codes. Two complications arose when dealing with spectra that did appear with photometry. The first is that notations used to designate peculiarities (uncertain classification, emission, broadening, interstellar lines, double lines, etc.) differ from source to source and have evolved with time. Within the limits imposed in producing a printable collation, I have attempted to remain faithful to original notations. The second is that there is an element of subjectivity to classification: two credible sources may legitimately assign different classifications to the same star; this is particularly so in the case of luminosity classifications where low-dispersion spectra are involved. Except for resolving obvious cases of misidentification (a star classified as both B and K, for example), no attempt has been made to scan the compilation for such differences, hence no detailed spectroscopic notes have been compiled for any star. An important exception was made to the "original data only" precept in compiling spectral data: the catalog of L. F. Smith [reference 301; MNRAS 138, 109 (1968) =1968MNRAS.138..109S] was used in its entirety, even though not all data therein is original to her. Her paper introduced a new classification system for the Wolf-Rayet stars and included re-classification of previous material on the new system. Dispersion codes for all entries from this paper are given as "see comments in original reference". 2.3 Radial Velocities Here we began by examining all references we had on hand (> 300) from the construction of the UBVb, uvby, and MK databases for any that reported original radial velocities. Following this, the SIMBAD database was queried for RV data for each LS star. Again, each reference was examined for any relevant notes (emission, variability, duplicity, etc.) and information as to the spectral dispersion employed, if such was reported. RV papers are notoriously inconsistent in the way that results are reported: some record each RV (and possibly an associated uncertainty) obtained from each line utilized on a given spectrum, while others give only an overall direct or weighted average. In compiling the present database, we have tried to adhere to the following simple conventions for averaging different RV's reported for the same star from the same reference. If the reference records only an average RV and (possibly an) associated uncertainty sRV, we report both of these but give the number of measures as zero; if no uncertainty is reported, that column of the database will be blank. If individual observations were recorded, we computed averages and standard deviations (irrespective of any uncertainties attached to individual observations) and record these along with the number of contributing measures; if a value was indicated as uncertain, we gave it half weight. If two or more values were "bracketed together", we averaged them and treated that average as one entry into the calculation of that star's overall average RV from that reference. If an average value was reported along with a "quality code" quantified in terms of a range of probable errors, we took the upper limit of the range as the uncertainty, except in cases where the velocity is of such poor quality (typically "D" quality) that only a minimum uncertainty was specified; such cases are explicitly flagged with corresponding notes. Similarly, if an average RV and number of measures was recorded but no quality code was specified, we record the uncertainty as zero and flag the lack of a quality code by a note. These last two conditions apply in particular to papers published in the Publications de L'Observatoire de Haute-Provence (POHP). We deviate in one major way from our intent to incorporate only original data in this compilation. This is that, in view of the difficulty of securing them, we decided not to try to acquire radial velocity references dated earlier than 1951, instead adopting, with two minor exceptions, the average RV's given in Wilson's (1953) General Catalogue of Stellar Radial Velocities (GCRV). The two exceptions are a paper by Moore and Paddock (ApJ, 112, 48, 1950 =1950ApJ...112...48M) which was already had on hand wherein an RV for one star (LS 238) is reported, and Abt's (ApJS, 19, 387, 1970 =1970ApJS...19..387A) compilation of RV's measured at Mount Wilson between 1909 and 1951 for stars between zero and twelve hours of right ascension; this paper involves five stars (LS 41, 110, 140, 151, and 822). GCRV values for these six stars were therefore discarded. Finally, we searched Duflot, Figon, and Meyssonnier's recent Wilson-Evans-Batten (WEB) average RV compilation (AAS, 114, 269, 1995) for any LS stars that the above search strategies failed to turn up; this resulted in 42 stars, which we treat as coming from a single reference. 3. Acknowledgements Much of the spectroscopic data was checked and keypunched by Amy Beatty, and much of the radial velocity data by Kristine Kuhna; both served as summer research assistants. This research has made use of the Simbad database, operated at CDS, Strasbourg, France. I am grateful to the inter-library loan staff at Alma College for tracking down a number of references, and to the Office of the Provost of Alma College for support of publication costs. This research was supported by an award from Research Corporation.