Contributions to the Nearby Stars (NStars) Project: Spectroscopy of Stars Earlier than M0 within 40 parsecs: The Northern Sample I. R.O. Gray Department of Physics and Astronomy Appalachian State University, Boone, NC 28608 [email protected] C.J. Corbally Vatican Observatory Research Group, Steward Observatory Tucson, AZ 85721-0065 [email protected] R.F. Garrison David Dunlap Observatory, Richmond Hill, Ontario [email protected] M.T. McFadden Department of Physics and Astronomy Appalachian State University, Boone, NC 28608 [email protected] P.E. Robinson Department of Physics and Astronomy Appalachian State University, Boone, NC 28608 arXiv:astro-ph/0308182v1 11 Aug 2003 ABSTRACT We have embarked on a project, under the aegis of the Nearby Stars (NStars)/ Space Interfer- ometry Mission Preparatory Science Program to obtain spectra, spectral types, and, where feasi- ble, basic physical parameters for the 3600 dwarf and giant stars earlier than M0 within 40 parsecs of the sun. In this paper we report on the results of this project for the first 664 stars in the north- ern hemisphere. These results include precise, homogeneous spectral types, basic physical pa- rameters (including the effective temperature, surface gravity and the overall metallicity, [M/H]) and measures of the chromospheric activity of our program stars. Observed and derived data pre- sented in this paper are also available on the project’s website http://stellar.phys.appstate.edu/. Subject headings: astronomical databases: surveys — stars: abundances — stars: activity — stars: fundamental parameters — stars: statistics 1 1. Introduction two gratings with 600 grooves mm−1 and 1200 grooves mm−1 and a thinned, back-illuminated The three institutions represented by the au- 1024 × 1024 Tektronix CCD operating in the thorship of this paper are cooperating on a project multipinned-phase mode. The observations were under the NASA/JPL Nearby Stars / Space Inter- made with a 100µm slit, which corresponds to ap- ferometry Mission Preparatory Science program to proximately 2′′ at the focus of the telescope; aver- obtain spectroscopic observations of all 3600 main- age seeing at the Dark Sky Observatory (DSO) is sequence and giant stars with spectral types earlier about 3′′. The 100µm slit with the two gratings than M0 within a radius of 40 pc. We are obtain- yields 2-pixel resolutions of 3.6A˚ and 1.8A˚ and ing blue-violet spectra at classification resolution spectral ranges of 3800 – 5600A˚ and 3800 – 4600A˚ ˚ (1.5 – 3.6A) for all of these stars. These spec- respectively. The lower-resolution spectra were tra are being used to obtain homogeneous, pre- used primarily for the late-type stars in the sam- cise, MK spectral types. In addition, these spec- ple (later than G5) whereas the higher-resolution tra are being used in conjunction with synthetic spectra were used primarily for the earlier-type spectra and existing intermediate-band Str¨omgren stars. An iron-argon hollow-cathode comparison uvby and broad-band V RI photometry to derive lamp was used for the wavelength calibration, and the basic astrophysical parameters (the effective all the spectra were reduced with standard meth- temperature, gravity and overall metal abundance ods using IRAF1. [M/H]) for many of these stars. We are also using The 1.8A˚ resolution spectra were rectified us- these spectra, which include the Ca II K and H ing an X-windows program, xmk19, written by lines, to obtain measures of the chromospheric ac- one of us (ROG) and were used in that format tivity of the program stars on the Mount Wilson for both spectral classification and for the deter- system. The purpose of this project is to pro- mination of the basic physical parameters. For vide data which will permit an efficient choice of the late-type stars rectification is problematical as targets for both the Space Interferometry Mission no useful “continuum” points can be identified. (SIM) and the projected Terrestrial Planet Finder In addition, the energy distribution contains use- (TPF). In addition, combination of these new data ful information for both the spectral classification with kinematical data should enable the identifica- and the determination of the basic physical pa- tion and characterization of stellar subpopulations rameters. We have therefore made an attempt within the solar neighborhood. to approximately flux calibrate the 3.6A˚ resolu- Observations for this project are being carried tion spectra even though they were obtained with out on the 1.9-m telescope of the David Dun- a narrow slit. During the course of our observa- lap Observatory in a northern polar cap (DEC > ◦ tions for this project at DSO over the past three +50 ), the 0.8-m telescope of the Dark Sky Ob- years, we have, at intervals of a few months, made servatory (−10◦ ≤ DEC ≤ +50◦), the 2.3-m Bok ◦ observations of spectrophotometric standards at telescope of Steward Observatory (−30 ≤ DEC ≤ a variety of airmasses. These standard observa- −10◦) and the 1.5-m telescope at Cerro Tololo ◦ tions have been used to approximately remove Interamerican Observatory (DEC ≤ −30 ), al- the effects of atmospheric extinction and to cal- though there is considerable overlap between all ibrate the spectrograph throughput as a function of these samples. In this paper, we report on re- of wavelength. Except for observations made at sults for the first 664 stars, all observed at the high airmasses (> 1.8 airmasses) this procedure Dark Sky Observatory. yields calibrations of relative fluxes with accura- cies on the order of ±10%. While this is suffi- 2. Observations and Calibration cient for the purposes of accurate spectral classifi- cation, the determination of the basic physical pa- The observations reported in this paper were all made on the 0.8m telescope of the Dark Sky Ob- 1IRAF is distributed by the National Optical Astronomy servatory (Appalachian State University) situated Observatories (NOAO). NOAO is operated by the Associ- on the escarpment of the Blue Ridge Mountains ation of Universities for Research in Astronomy (AURA), in northwestern North Carolina. The Gray/Miller Inc. under cooperative agreement with the National Sci- ence Foundation classification spectrograph was employed with 2 rameters requires a more accurate flux calibration. The program stars were first classified indepen- This we have achieved by “photometrically cor- dently on the computer screen by eye (using the recting” the fluxes using Str¨omgren photometry. graphics program xmk19) by at least two of the The DSO 3.6A˚ resolution spectra have a spectral authors and then the spectral types were com- range which includes three of the four Str¨omgren pared and iterated until complete agreement was bands, v, b and y. We perform numerical photom- obtained. There are significant overlaps between etry on these spectra and use the absolute flux the samples observed with the four telescopes em- calibration of the Str¨omgren system (Gray 1998) ployed for this project to ensure homogeneity of to derive flux corrections at the effective wave- our spectral types over the entire sky. This homo- lengths of the three photometric bands. Interpo- geneity is further ensured by a significant overlap lation and extrapolation of these corrections yield in the MK standards used for each sample, and flux corrections over the entire observed spectrum. close scrutiny of the non-overlapping standards to From comparison with spectrophotometric obser- verify consistency. The spectral types for the first vations in the literature, we find that this pro- sample of 664 stars from the northern hemisphere cedure yields not only relative but also absolute observed at DSO are recorded in Table 1. These fluxes with accuracies of about ±3% over nearly spectral types are multi-dimensional, as they in- the entire spectral range. Fortunately, most of our clude not only the temperature and luminosity program stars with spectral types of K3 and ear- types, but also indices indicating abundance pecu- lier have Str¨omgren photometry. For the later- liarities and the degree of chromospheric activity. type stars without Str¨omgren photometry, other Chromospheric activity is evident in our spectra considerations (see §4) preclude the derivation of through emission reversals in the cores of the Ca II basic physical parameters from our spectra and so K & H lines, and in more extreme situations, infill- the lack of accurate fluxes is not otherwise limit- ing and emission in the hydrogen lines. We have ing. indicated these different levels of chromospheric All of the spectra obtained for this project are activity in the spectral types with the following available on the project’s website2. Rectified spec- notation: “(k)” indicates slight emission reversals tra from DSO have an extension of .r18 or .r36 or infilling of the Ca II K & H lines are visible; “k” depending on the resolution. The 3.6A˚ resolu- indicates emission reversals are clearly evident in tion flux spectra, which have not been photometri- the Ca II K & H lines, but these emission lines do cally corrected, are normalized at a common point not extend above the surrounding (pseudo) con- (4503A)˚ and have an extension of .nor, whereas tinuum; “ke” indicates emission in the Ca II K photometrically corrected spectra are available in & H lines above the surrounding (pseudo) contin- a normalized format (.nfx) and in terms of abso- uum, usually accompanied with infilling of the Hβ lute fluxes (.flx) in units of erg s−1 cm−2 A˚−1. line, and “kee” indicates strong emission in Ca II Spectra obtained at the other observatories (see K & H, Hβ and perhaps even Hγ and Hδ. Because §1) are also available on this website, and will be chromospherically active stars tend also to be vari- the subject of future papers. able, the chromospheric activity “type” will also vary. We have, therefore, noted the observation 3.