1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
4:
5: \documentclass[12pt,preprint]{aastex}
6:
7: % manuscript produces a one-column, double-spaced document:
8: %\documentclass[manuscript]{aastex}
9: % preprint2 produces a double-column, single-spaced document:
10: %\documentclass[preprint2]{aastex}
11:
12: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
13: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
14: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
15:
16: \slugcomment{}
17:
18: \shorttitle{The Cool Subdwarf Investigation I}
19:
20: \shortauthors{Jao et al.}
21:
22: \received{2007 July 11}
23: \begin{document}
24:
25: \title{Cool Subdwarf Investigations (CSI) I: New Thoughts for the Spectral
26: Types of K and M Subdwarfs}
27:
28: \author{Wei-Chun Jao\altaffilmark{1}, Todd J. Henry\altaffilmark{1},
29: Thomas D. Beaulieu\altaffilmark{1}, John
30: P. Subasavage\altaffilmark{1}}
31:
32: \affil{Department of Physics and Astronomy, Georgia State University,
33: Atlanta, GA 30302-4106}
34:
35: \email{jao, thenry, beaulieu, subasavage@chara.gsu.edu}
36:
37: \altaffiltext{1}{Visiting Astronomer, Cerro Tololo Inter-American
38: Observatory. CTIO is operated by AURA, Inc.\ under contract to the
39: National Science Foundation.}
40:
41: \begin{abstract}
42:
43: Using new spectra of 88 K and M-type subdwarfs, we consider novel
44: methods for assigning their spectral types and take steps toward
45: developing a comprehensive spectral sequence for subdwarf types K3.0
46: to M6.0. The types are assigned based on the overall morphology of
47: spectra covering 6000\AA~to 9000\AA. The types and sequence presented
48: link the spectral types of cool subdwarfs to their main sequence
49: counterparts, with emphasis on the relatively opacity-free region from
50: 8200--9000\AA. When available, supporting abundance, kinematic, and
51: trigonometric parallax information is used to provide more complete
52: portraits of the observed subdwarfs. We find that the CaHn (n $=$
53: 1--3) and TiO5 indices often used for subdwarf spectral typing are
54: affected in complicated ways by combinations of subdwarfs'
55: temperatures, metallicities, and gravities, and we use model grids to
56: evaluate the trends in all three parameters. Because of the complex
57: interplay of these three characteristics, it is not possible to
58: identify a star as an ``extreme'' subdwarf simply based on very low
59: metallicity, and we suggest that the modifiers ``extreme'' or
60: ``ultra'' only outline locations on spectroscopic indices plots, and
61: should not be used to imply low or very low metallicity stars. In
62: addition, we propose that ``VI'' be used to identify a star as a
63: subdwarf, rather than the confusing ``sd'' prefix, which is also used
64: for hot O and B subdwarfs that are unrelated to the cool subdwarfs
65: discussed in this paper.
66:
67: \end{abstract}
68:
69: \keywords{stars: subdwarfs --- stars: abundances --- stars: late-type
70: --- stars: classification --- stars: fundamental parameters ---
71: stars: temperatures}
72:
73: \section{Introduction}
74:
75: The HR diagram is the most important map of stellar astronomy. It
76: provides a relatively straightforward method for separating different
77: stellar luminosity classes, e.g.~supergiants, bright giants, giants,
78: subgiants, main sequence dwarfs, and white dwarfs, using their colors
79: and luminosities. However, spectroscopic and trigonometric parallax
80: results have revealed a seventh distinct stellar luminosity class ---
81: the subdwarfs --- that lie below the main sequence dwarfs on the HR
82: diagram.
83:
84: The realm of the subdwarfs has been previously explored by
85: \cite{Sandage1959}, \cite{Hartwick1984}, \cite{Monet1992},
86: \cite{Gizis1997}, and \cite{Jao2005}, to name a few. Subdwarfs'
87: locations on the HR diagram are in part due to having metallicity
88: abundances lower than most field stars, which causes their opacities
89: to differ from those of regular dwarfs. Subdwarfs are sometimes
90: called low metallicity halo stars or Galactic thick disk stars based
91: on their spectroscopic features, kinematics, and/or ages
92: \citep{Digby2003, Lepine2003, Burgasser2003, Reid2005, Monteiro2006}.
93: Regardless of how they are described, subdwarfs are fundamentally
94: different from their main sequence cousins.
95:
96: Because of their generally high intrinsic velocities in the Galaxy,
97: many subdwarfs have been selected using high proper motion efforts,
98: such as the Lowell proper motion (\citealt{Giclas1971, Giclas1978}),
99: LHS (\citealt{LHS}), and L{\'e}pine-Shara Proper Motion-North (LSPM,
100: \citealt{Lepine2005a}) catalogs. Recently, subdwarfs have been
101: selected by colors and spectroscopic observations via the Sloan
102: Digital Sky Survey (SDSS, \citealt{West2004}). After initial flagging
103: as a potential subdwarf, spectroscopic and astrometric
104: (i.e.~trigonometric parallax work) followup efforts are typically
105: carried out to confirm or refute candidates as true subdwarfs. Past
106: subdwarf identification efforts include \cite{Ryan1991},
107: \cite{Monet1992}, \cite{Carney1994}, \cite{Gizis1997},
108: \cite{Lepine2003}, \cite{West2004}, \cite{Reid2005} and
109: \cite{Burgasser2006}. Most of these studies confirmed subdwarfs
110: spectroscopically, but only \cite{Monet1992} provides the crucial
111: trigonometric parallaxes that allow subdwarfs to be placed on the HR
112: diagram.
113:
114: \cite{Gizis1997} presented a pioneering effort to assign numeric
115: subtypes for cool subdwarfs of spectral types K and M. First, he used
116: the flux ratio of molecular band features, CaHn (n $=$ 1--3; we will
117: use ``CaH'' to indicate all three bands or index values throughout
118: this paper, unless otherwise specified) and TiO5 with pseudo-continuum
119: points to calculate spectroscopic indices
120: ($f_{bands}$/$f_{continuum}$). On TiO5 vs CaH1 or TiO5 vs CaH2$+$CaH3
121: plots, two high order polynomial lines can be used to separate a
122: continuous distribution of stars into three categories: regular
123: dwarfs, subdwarfs, and extreme subdwarfs. Numerical subclasses were
124: then assigned using two independent linear fits for subdwarfs and
125: extreme subdwarfs. For a decade, this methodology has been used to
126: assign spectral types for cool subdwarfs.
127:
128: However, the current method of assigning subdwarf spectral types is
129: not directly linked to either their main sequence or giant
130: counterparts, as is typically (but not always) the case with normal
131: dwarfs and giants. In addition, we have found that many subdwarfs are
132: assigned different subtypes even though their differences are limited
133: to CaH. If these limitations can be overcome, a well-defined spectral
134: sequence would benefit many research areas beyond classification
135: efforts, including attempts to estimate effective temperatures and
136: distances, as well as providing insight into understanding Galactic
137: structure.
138:
139: In this work, we first provide clarification and recommendations for
140: subdwarf terminology by addressing the usage of the confusing spectral
141: prefix ``sd''. We then discuss our spectroscopic observations of 88
142: subdwarfs, generally targeting high proper motion stars in the
143: southern sky. We next outline how synthetic spectra can assist us in
144: understanding cool subdwarf spectral features. The bulk of this work
145: describes a detailed effort to provide a subdwarf spectral sequence
146: for stars having spectral types K3.0 through M6.0. We then apply this
147: spectral typing method to those subdwarfs found in the SDSS database.
148: Once we have a detailed understanding of what makes a star a subdwarf,
149: we then discuss (1) why it is premature to assign precise parameters
150: to subdwarf spectra, (2) misunderstandings related to the terms
151: ``extreme'' and ``ultra'' subdwarfs, (3) why the old method works for
152: dwarfs, but not subdwarfs, and finally (4) the recent set of subdwarf
153: spectral standards from \cite{Lepine2007}.
154:
155: \section{``VI'' Subdwarfs are Different from ``sd'' Subdwarfs}
156: \label{sec.VI}
157:
158: The first subluminous objects fainter than main sequence stars were
159: reported by \cite{Adams1922} when they were trying to determine the
160: luminosities of A-type stars. In \cite{Adams1935} they called these
161: stars ``intermediate white dwarfs'' to separate them from typical
162: white dwarfs, and in the same paper reported the first six
163: ``intermediate white dwarfs'' --- now known as LHS 405 (sdF3), LHS 540
164: (F8IV), LHS 1501 (A4p), LHS 2194 (sdF5), HD 132475 (F5/F6V), and HIP
165: 68321 (A4). All six have either A or F types (spectral types are from
166: SIMBAD) in the modern MK spectral classification system. However, the
167: term ``subdwarfs'' was not suggested until \cite{Kuiper1939}. He
168: expected three classes of objects to be found in his spectroscopic
169: survey of high proper motion stars--- white dwarfs, stars of large
170: spectroscopic parallaxes, and a class that was 2 or 3 magnitudes less
171: luminous than main sequence stars of the same color\footnote{Kuiper
172: assumed all of these high proper motion stars had $V_{tan}$ less than
173: 474 km/sec. Under this assumption, he could then make crude estimates
174: of their absolute magnitudes without having trigonometric parallaxes.
175: The results placed some stars three magnitudes below the main
176: sequence.}. He suggested the name ``subdwarfs'' be used to represent
177: this final, independent, class of stars. This name paralleled the use
178: of the term ``subgiants'' to describe stars that fall below the giants
179: on the HR diagram. The name also eliminated the confusion with white
180: dwarfs, which are much less luminous than main sequence stars (and we
181: now know are a completely different type of object). A year later,
182: \cite{Kuiper1940} reported the first three M-type subdwarfs ---
183: Kapteyn's star, LHS 20, and LHS 64 (\citealt{Gizis1997} confirmed that
184: all are M subdwarfs). Although they were termed subdwarfs, he used
185: the same spectral classification as dwarfs (M0 and M2, see
186: \citealt{Kuiper1940} Table 1). The ``sd'' spectral classification
187: prefix for subdwarfs did not appear until \cite{Joy1947}, when he used
188: the strengthening of the Lindblad depression around 4226\AA.
189:
190: In the late 1940's through 60's, the term ``subdwarfs'' also began
191: being used for a class of underluminous blue stars \citep{Humason1947,
192: Feige1958, Greenstein1966a, Greenstein1966b}. The terminology was
193: based on the understanding that if these were high luminosity blue
194: stars, their distances would be outside the Milky Way, so it was
195: surmised that these stars should be underluminous, and therefore
196: closer. Although their temperatures are similar to O and B dwarfs,
197: their spectral features are, in fact, quite different. Generally,
198: O-type subdwarfs (sdO) and B-type subdwarfs (sdB) both have broad
199: Balmer absorption. sdBs have weak or no He lines, while sdOs have
200: strong He II (4686\AA) or other He II lines \citep{Heber1992}. Since
201: \cite{Feige1958}, such blue objects have been called sdO, sdB, or
202: sdOB-type stars.
203:
204: Thus, we are left with the unfortunate situation that there are two
205: different classes of stars called ``subdwarfs.'' One is located at
206: the cool end of the HR diagram while the other is at the hot end. The
207: two classes of stars are subluminous for completely different
208: astrophysical reasons but share the same ``sd'' spectral
209: classification prefix. Cool subdwarfs usually have low metallicity
210: \citep{Chamberlain1951, Greenstein1966, Mould1976, Allard1995}, so
211: their opacties are different from dwarfs. For example,
212: \cite{Allard1995} discussed the possible opacity sources for a solar
213: type dwarf and a [m/H]=$-$2.5 subdwarfs. TiO dominates the opacity
214: sources in the optical band. However, because of a decreasing
215: metallicity for subdwarfs, TiO opacity decreases dramatically. Hence,
216: this less blanketing from TiO bands causes more continuum flux
217: radiated from deeper and hotter layer of stellar atmosphere and their
218: spectrum falls closer overall to that of a blackbody, so these
219: subdwarfs appear bluer than dwarfs, as shown in
220: Figure~\ref{fig.continuum.model}\footnote{The details of these
221: synthetic spectra, the ``GAIA model grids'', are discussed in section
222: 4.}. To the contrary, hot blue subdwarfs of types O and B are
223: progenitors of white dwarfs, and their subluminous nature is not
224: caused by metallicity at all. Instead, hot subdwarfs represent a
225: stage in the stellar evolution cycle of an evolved star, and they
226: happen to be crossing the main sequence at the moment of observation.
227:
228: Because they {\it are} different kinds of stellar objects, we suggest
229: that the two classes should not share the same spectral classification
230: notation, ``sd.'' \cite{Roman1955} argued that for types later than
231: G0, the spectral notation ``VI'' should be used for stars that are
232: $\sim$1--2 magnitude less luminous than main sequence stars. Although
233: \cite{Jaschek1987} stated that ``this designation (VI) should
234: definitely be abandoned'', no specific reasons were actually given.
235: Here we propose that the luminosity class ``VI'' should be adopted for
236: cool subdwarfs, especially for K and M types. The three primary
237: reasons are:
238:
239: \begin{itemize}
240:
241: \item Cool subdwarfs of types K and M lie clearly below main sequence
242: stars on the HR diagram, forming an additional class of objects.
243: Assigning their types as VI continues the progression outlined by
244: other classes. Like the family of giants, which includes luminous
245: supergiants (I), bright giants (II), normal giants (III), and
246: subgiants (IV), the family of dwarfs includes main sequence dwarfs (V)
247: and their subdwarf (VI) counterparts.
248:
249: \item Cool subdwarfs and OB subdwarfs are completely differently types
250: of objects with different origins. The notation ``sd'' is suitable
251: for OB subdwarfs, which have not yet as a group been established to
252: have any sort of sequence on the HR diagram, at least when using using
253: parallaxes from the Yale Parallax Catalog, Hipparcos, and more recent
254: efforts. On the other hand, there are dozens of cool subdwarfs with
255: parallax measurements, and they do form a coherent group below the
256: main sequence on the HR diagram, making a Roman numeral designation
257: reasonable.
258:
259: \item Historically, the Roman numerals used for luminosity
260: categorization track with different gravities for the stars
261: classified, with higher gravities being assigned higher Roman
262: numerals. We find that assigning ``VI'' for subdwarfs appropriately
263: continues this trend. Figure~\ref{fig.Lopez.Morales} shows the
264: mass-gravity relation using data from Table 1 of \cite{Lopez2007}. In
265: the best represented mass regime from 0.35$M_{\sun}$ to
266: 0.70$M_{\sun}$, a crude trend indicates that lower metallicity stars
267: do indeed have higher gravities. Although much more data are needed
268: to understand clearly how metallicity affects the mass-gravity
269: relation, current evidence supports using the ``VI'' designation for
270: low metallicity subdwarfs, which tend to exhibit higher gravities.
271:
272: \end{itemize}
273:
274: \noindent Thus, in order to separate the OB subdwarfs from the cool
275: subdwarfs, we suggest the `sd'' prefix should not be used for low
276: metallicity (and/or high gravity, as will be shown below) subdwarfs.
277:
278: \section{Observations and Reductions}
279: \subsection{Observations}
280:
281: Our subdwarf targets were selected from several different efforts,
282: including lists of spectroscopically identified subdwarfs
283: \citep{Gizis1997, Reid2005}, subdwarfs with parallax measurements
284: \citep{Jao2005, Costa2005, Costa2006}, candidates from proper motion
285: catalogs \citep{Deacon2005, Subasavage2005a, Subasavage2005b}, and
286: stars with metallicity measurements \citep{Carney1994, Cayrel2001,
287: Nordstrom2004}. We define our subdwarfs of interest to be those with
288: $V-K_{s}\geqslant$ 2.0, [m/H] $\leqslant-$0.5, or having absolute
289: magnitudes at least one magnitude less luminous in $M_{Ks}$ than a fit
290: to main sequence stars of comparable color with trigonometric
291: parallaxes from the RECONS (Research Consortium on Nearby Stars) 10 pc
292: sample \citep{Henry2006}.
293:
294: Spectroscopic observations were made with the 1.5-m and 4.0-m
295: telescopes at CTIO. For the observations on the 1.5-m from 2002 to
296: 2006, the R-C spectrograph with a Loral 1200$\times$800 CCD camera was
297: used with the \#32 grating (in first order) at tilt 15.1$^\circ$. The
298: order-blocking filter OG570 was utilized to provide spectra covering
299: the range of 6000\AA~to 9500\AA~with a resolution of 8.6\AA. The only
300: variation in observing parameters was that during the May and December
301: 2006 observing runs a larger slit width of 6\arcsec~and 4\arcsec,
302: respectively, ~was used instead of the 2\arcsec~slit used in previous
303: runs, in order to minimize the differential color refraction (because
304: the slit orientation was not changed during observations). For
305: observations on the 4.0-m in 2002, the R-C spectrograph with a Loral
306: 3K$\times$1K CCD was used with the \#181 grating (in first order) at
307: tilt 58.8$^\circ$. The order-blocking filter OG515 was utilized to
308: provide spectra covering the range from 5500\AA~to 10000\AA~with a
309: resolution of 6\AA. Fringing at wavelengths longer than
310: $\sim$7000\AA~in the 4.0-m data was removed by customized IDL
311: routines. Bias frames and dome flats (and sky flats at the 1.5-m)
312: were taken at the beginning of each night for calibration. At least
313: two exposures were taken for each object to permit cosmic ray
314: rejection. If stars were faint, additional observations were
315: sometimes made. A 10 second Ne$+$He$+$Ar or Ne only arc lamp spectrum
316: was recorded after each target to permit wavelength calibration.
317: Several spectroscopic flux standard stars found in the $IRAF$
318: spectroscopy reduction packages were observed during each observing
319: run, usually nightly. Reductions were carried out in the standard way
320: using IRAF reduction packages. Wavelength and flux calibrations were
321: done using {\it onedspec.dispcor} and {\it onedspec.calibrate} within
322: IRAF, respectively.
323:
324: Many of the subdwarfs discussed in this paper have new trigonometric
325: parallaxes and $VRI$ photometry acquired during our southern nearby
326: star program, CTIOPI (Cerro Tololo Inter-american Observatory Parallax
327: Investigation, see \citealt{Jao2005}). In a future paper in this
328: subdwarf series, we will present the astrometric (particularly
329: trigonometric parallaxes) and photometric results.
330:
331: \subsection{Identifying Subdwarfs}
332:
333: During our five year spectroscopic campaign, we have acquired spectra
334: for more than 900 objects. To glean subdwarfs from our spectroscopic
335: database, we calculated their spectroscopic indices listed in
336: Table~\ref{tbl.cahn.tio5.index} (targets listed alphabetically),
337: mimicking the methodology outlined by \cite{Gizis1997}.
338:
339: Figure~\ref{fig.cahn.plot} shows TiO5 plotted against CaH1 and
340: CaH2$+$CaH3 for various samples of stars. Our subdwarfs are shown
341: with solid circles. For comparison, small dots indicate main sequence
342: stars from \cite{Hawley1996}, while open triangles and squares
343: represent subdwarfs and ``extreme'' subdwarfs from \cite{Gizis1997}
344: and \cite{Reid2005}. Some stars (solid circles) having indices
345: located near or in the main sequence regions in these plots have been
346: manually checked to confirm that they are subdwarfs. Using the HR
347: diagram in Figure~\ref{fig.hr.plot}, we confirm the low luminosities
348: of our spectroscopically selected subdwarfs ($V-K_{s}>$ 2.7) that have
349: accurate trigonometric parallaxes.
350:
351: Although we focus primarily on the K and M subdwarfs for this study,
352: we include a few G-type subdwarfs among those selected from
353: metallicity measurements in the literature. For reference, we
354: consider early K-type stars to have types K0 to K2, mid K-types to be
355: K3 to K5, and late K-types to be K6 and later. We find that it is
356: difficult, but possible, to separate late G from early K-type stars
357: using our spectral coverage and resolution. Their continuum slopes
358: have only slight differences across our wavelength window coverage,
359: and there are no noticeable absorption differences beyond 7500\AA.
360: Spectra for G1V to K5V types from \cite{Jacoby1984} (resolution
361: $\sim$4\AA) and \cite{Silva1992} (resolution $\sim$11\AA) are plotted
362: in Figure~\ref{fig.jacoby.silva}. The only strong features are the
363: absorption lines of Ba I (6497\AA)\footnote{\cite{Turnshek1985} noted
364: that this feature at 6497\AA~is a blend of different atomic lines,
365: including Fe I, Ba I, Ca I, Mn I, Co I, Ti I and II, and Ni I. Ba I
366: is likely the dominant absorber because it has the largest Einstein
367: coefficient in the NIST atomic spectra database.} and H$\alpha$
368: (6563\AA), with gradually increasing Ba I absorption and decreasing
369: H$\alpha$ absorption as the effective temperature drops. These
370: effects can be seen in both sequences, regardless of the spectral
371: resolution. Our spectral resolution of 6--9\AA~falls between the
372: resolution of the two sequences shown, so we can use the relative
373: absorption strengths of Ba I and H$\alpha$ to separate G and K-type
374: stars. In total, we have identified 88 K and M subdwarfs and five G
375: type subdwarfs, using spectra with coverage from 6000\AA~to 9000\AA.
376:
377: \subsection{Sorting Spectra}
378: \label{sec.sorting.spectra}
379:
380: After reduction, the spectra were sorted into different bins based
381: upon similarity in overall slope and features. This assured that
382: stars in each bin had approximately the same temperature. However,
383: several impediments to clean sorting were encountered:
384:
385: \begin{itemize}
386:
387: \item All mid K-type subdwarfs had spectra virtually indistinguishable
388: from dwarf standard stars, yet they had low metallicity measurements
389: and/or were found below the main sequence on the HR diagram (see top
390: figure in Figure~\ref{fig.problem.spect}).
391:
392: \item Many subdwarf spectra placed into the same bin showed
393: differences only in CaH (see middle figure in
394: Figure~\ref{fig.problem.spect}).
395:
396: \item Many subdwarf spectra matched different dwarf standards at the
397: blue ($\lambda$ $<$ 7570\AA) and red ($\lambda$ $>$ 8200\AA) ends (see
398: bottom figure in Figure~\ref{fig.problem.spect}).
399:
400: \end{itemize}
401:
402: \noindent In order to understand what factors caused these anomalies,
403: we next examine theoretical studies that provide synthetic spectra
404: that can be compared directly to the observed spectra.
405:
406: \section{Grids of Synthetic Spectra}
407:
408: We use grids of synthetic spectra computed with PHOENIX codes
409: (hereafter, GAIA model grids) to understand how subdwarfs' physical
410: parameters (temperature, metallicity and gravity) affect their
411: spectra. The most recently released GAIA model grids
412: \citep{Brott2005} are available at an FTP site in
413: Hamburg\footnote{ftp://ftp.hs.uni-hamburg.de/pub/outgoing/phoenix/GAIA/v2.6.1/}.
414: \cite{Gizis1997}, \cite{Woolf2005}, and \cite{Burgasser2006} have all
415: used these synthetic model grids to characterize subdwarfs, but an
416: older version of the grids was used in all three cases. The version
417: we employ here, 2.6.1, was released in late 2004. A comparison of one
418: pre-2004 spectral model (provided by V. Woolf, private communication)
419: and a new spectral model (from the GAIA model grids) for a cool
420: subdwarf is shown in Figure~\ref{fig.new.old}. Improvements to the
421: new models include (1) an enlarged and enhanced version of the
422: equation of state, (2) more atomic, ionic, and molecular line
423: opacities, (3) inclusion of the formation of dust particles for cool
424: stars, and (4) microturbulence calculations. Additional water and TiO
425: opacities and the inclusion of dust are enhancements particularly
426: applicable to the low mass stars discussed here.
427:
428: The two spectra shown in Figure~\ref{fig.new.old} are virtually
429: identical redward of 7000\AA, but there are significant differences
430: between 6500\AA~and 7000\AA, where the CaH2 and CaH3 bands are found,
431: and these differences will certainly affect evaluations done with the
432: older models. The new model in this region has much shallower
433: absorptions than the old one, which will affect metallicity and
434: gravity estimates. A few narrow absorption features (Li I at 6103\AA,
435: Ca I at 6122\AA~and 6162\AA) are also changed. As outlined in the
436: discussion section $\S$\ref{sec:discussion}, even the latest version
437: of the model grids does not provide ideal matches to real spectra, so
438: further progress can still be made.
439:
440: \subsection{Identifying Mid K-type Subdwarfs}
441: \label{sec.K.subdwarf}
442:
443: We use the latest GAIA model grids\footnote{GAIA model grids also
444: provide various values of $\alpha$-elements (O, Ne, Mg, Si, S, Ar, Ca,
445: and Ti), which yield abundance ratios such as O/Fe, Ne/Fe, etc.
446: \cite{Nissen1997} showed that most field halo stars have
447: [$\alpha$/Fe]$_{*}$$\approx$[$\alpha$/Fe]$_{\sun}$, so we select
448: [$\alpha$/$\alpha_{\sun}$]=0.0.} to calculate predicted CaH band
449: strengths and plot the derived indices against effective temperatures
450: in Figure~\ref{fig.model.index.plot}. We evaluate stars with
451: $T_{eff}$ between 2700K and 4500K and [m/H] between 0.0 and $-$3.0.
452: For the moment, we adopt {\it log g} $=$ 5.0 generically for
453: subdwarfs. This gravity value does not apply to all types of
454: subdwarfs, but we are presently interested in outlining the behavior
455: of the CaH features with metallicity alone.
456:
457: When $T_{eff}$ is less than about 3300K,
458: Figure~\ref{fig.model.index.plot} shows that the CaH1 index decreases
459: (stronger absorption) when metallicity decreases from 0.0 to $-$2.0 at
460: fixed $T_{eff}$. However, the trend reverses for [m/H] $=$ $-$2.5 and
461: $-$3.0. When $T_{eff}$ is between 3300K and 3500K, there is a very
462: weak relation between metallicity and the CaH1 index, but from 0.0 to
463: $-$1.0, the relation (CaH1 index decreases when metallicity decreases)
464: still holds. This relationship is degenerate for lower
465: metallicities. For temperatures hotter than 3500K, the CaH1 index
466: increases (weaker absorption) as metallicity decreases, in contrast to
467: the low temperature region. For the CaH2$+$CaH3 index, the trends are
468: generally the same, except that (1) at temperatures less than 3200K
469: the index decreases (stronger absorption) only for 0.0 to $-$1.0, with
470: a reversal for lower metallicities, and (2) the trend for higher
471: temperature stars (increasing index, weaker absorption with lower
472: metallicity) is the same, but the turnover is near 3200K rather than
473: 3500K.
474:
475: Even more important than these subtleties is that overall, the hotter
476: the subdwarf, the weaker its CaH bands. Note the collapse of any
477: differences between indices for hotter stars of various metallicities
478: in Figure~\ref{fig.model.index.plot}. This collapse makes separating
479: mid K-type subdwarfs from dwarfs based only on spectroscopic indices
480: difficult using our spectral coverage (6000\AA--9000\AA) and
481: resolution (6\AA--8.6\AA). An alternative method, such as the HR
482: diagram shown in Figure~\ref{fig.hr.plot}, sufficiently solves the
483: problem for mid K-type subdwarfs and will be discussed in the next
484: section. Independent metallicity measurements via high resolution
485: spectroscopic observations \citep{Bonfils2005, Bean2006} can also be
486: utilized. Our own spectra are not of sufficiently high resolution to
487: measure metallicities, but all selected K-type subdwarfs have measured
488: [m/H] $\leqslant$ $-$0.5 from other publications.
489:
490: \subsection{Mid K-type Subdwarf Sample from Our Observations}
491:
492: There are 31 subdwarfs having CaH1 indices larger than 0.9 enclosed by
493: the dashed box in Figure~\ref{fig.cahn.plot}. In this region, there
494: are no subdwarfs with previously measured CaH/TiO5 indices to compare
495: to our new sample of mid K-type stars, making it difficult to separate
496: the dwarfs and subdwarfs based on the CaH1 index alone. Fortunately,
497: 21 of these 31 stars have trigonometric parallaxes, and are plotted
498: with concentric circles in Figure~\ref{fig.hr.plot}. At least 15 of
499: these stars are subdwarfs based on their locations one or more
500: magnitudes below the main sequence on the HR diagram. The star with
501: the largest offset is DEN0515-7211\footnote{This star was first
502: reported in \cite{Costa2006} as reference star \#4 in the LHS 1749
503: parallax field. We identify it henceforth as DEN0515-7211 (DENIS$-$P
504: J051545.1$-$721122).}, located at $(V-K)$ = 3.3, $M_{Ks} =$ 9.7. This
505: star is a full 4.5 magnitudes less luminous than the main sequence,
506: but has no CaH or TiO5 features. Two additional stars (G016-009AB and
507: G026-009ACD) above the main sequence are known to be double-line
508: spectroscopic binaries with [m/H] $<-$0.5. These individual targets
509: are discussed in section 6.2. The dwarf/subdwarf status of only four
510: stars of the 21 remain ambiguous --- we suspect that most of them are
511: also subdwarfs, perhaps with as yet undetected companions elevating
512: them into main sequence territory.
513:
514: Generally, the spectroscopic index method fails to distinguish
515: subdwarfs from dwarfs if the derived CaH1 index is greater than 0.9.
516: With the benefit of additional trigonometric parallax information
517: and/or metallicity measurements, however, we can conclude that nearly
518: all of these mid K-type stars are indeed subdwarfs.
519:
520: \section{Late K-type and M-type Subdwarfs from GAIA Model Grids}
521: \label{sec.subdwarf.facts}
522:
523: For late K-type (redder than K5.0) and M-type stars, we use the GAIA
524: model grids to understand how the effective temperatures,
525: metallicities, and gravities affect the shapes and features of
526: subdwarfs' spectra. This analysis allows us to develop a spectral
527: sequence for cool subdwarfs of types K6.0 to M6.0, which is
528: tentatively extended blueward to K3.0 when additional information is
529: incorporated. We plot GAIA synthetic noiseless spectra in
530: Figure~\ref{fig.4800-2800.plot}, at increments of 200K (cooler than
531: 4000K) and 400K (hotter than 4000K) for stars with metallicities of
532: [m/H]$ =$ 0.0, $-$1.0, and $-$2.0. Because gravities have very
533: limited impact on the overall shapes of the spectra (shown in the top
534: panel of Figure~\ref{fig.3500.plot}), we do not show gravity plots
535: with fixed temperatures and metallicities (of course, some features do
536: change markedly with gravity, but not the overall slopes of the
537: spectra). Based on the synthetic spectra from GAIA model grids, cool
538: subdwarf spectra between 6000\AA~and 9000\AA~exhibit the following
539: trends:
540:
541: \begin{itemize}
542:
543: \item The effects of metallicity are minimal in low resolution
544: subdwarf spectra for stars with temperatures of 4400K and hotter.
545: This makes it difficult to separate dwarfs and subdwarfs using low
546: resolution spectra (as discussed in
547: $\S$\ref{sec.K.subdwarf}). However, from these noiseless spectra, we
548: can still identify a few metallic lines showing metallicity
549: trends. The most prominent feature is marked \#1 at 6256\AA~in
550: Figure~\ref{fig.4800-2800.plot}. Unfortunately, we do not see this
551: feature in any of our K-type spectra (nor is it listed in Table 1 of
552: \citealt{Turnshek1985}), so we consider its validity questionable.
553: The next prominent feature, marked \#2 in
554: Figure~\ref{fig.4800-2800.plot}, is Ca I (6162\AA). This line can
555: possibly be used to distinguish subdwarfs from dwarfs (see examples
556: discussed in $\S$\ref{sec.notes}), but in practice it is somewhat
557: difficult to evaluate in real spectra (with noise) at our resolution.
558:
559: \item For stars with temperatures of 2800--4000K, metallicity strongly
560: affects the spectra between 6000\AA~and 8200\AA. This is the region
561: that has been historically used to assign spectral types. In effect,
562: subdwarfs with decreased metallicities have spectra that are
563: ``brightened'' or ``less blanketed'' at the blue end, relative to
564: solar metallicity stars. However, the continuum at wavelengths longer
565: than 8200\AA~for temperatures 3400--4800K is nearly free of
566: metallicity effects. We can therefore use the 8200--9000\AA~region to
567: establish subtypes in the subdwarf spectral sequence because the slope
568: is a function of temperature. This also allows us to mirror the
569: spectral sequence for dwarfs, providing a useful link between the
570: dwarf and subdwarf sequences.
571:
572: \item The TiO5 band strength at 7050--7150\AA~is very sensitive to
573: metallicity for temperatures cooler than 4000K. As shown in the top
574: panel of Figure~\ref{fig.3500.plot}, the TiO5 band strength is
575: effectively independent of gravity. We can therefore use the TiO5
576: feature to separate subdwarfs with different metallicities if their
577: continua (8200--9000\AA) are the same, regardless of their gravities.
578:
579: \item For a star of given temperature, stronger CaH bands could be
580: caused by lower metallicity, as shown in
581: Figure~\ref{fig.4800-2800.plot}, or higher gravity, as shown in the
582: top panel of Figure~\ref{fig.3500.plot}. If two subdwarfs have the
583: same continua from 8200--9000\AA~and the same TiO5 band strength, but
584: their CaH bands are different, we can rank them by their relative
585: gravities.
586:
587: \end{itemize}
588:
589: Consequently, the impediments to sorting cool subdwarf spectra at our
590: resolution discussed in $\S$\ref{sec.sorting.spectra} can be overcome
591: by understanding the trends revealed in GAIA synthetic spectra: (1)
592: the mid K-type dwarfs have the same spectra as dwarfs for our spectral
593: coverage and resolution, (2) CaH features are affected by both
594: metallicity and gravity, while the TiO5 band is affected by
595: metallicity but not gravity, and (3) the continuum from 8200\AA~to
596: 9000\AA~is not strongly affected by either metallicity or gravity, so
597: can therefore be used for spectral sequencing.
598:
599: \section{Additional Evidence Supporting the Metallicity and Gravity Trends
600: Indicated by the GAIA Models}
601: \label{sec.trend}
602:
603: To investigate the metallicity trends seen in GAIA models, we compare
604: our available spectra to metallicity measurements provided
605: independently by others. Measuring M dwarf metallicities is
606: difficult, but several recent attempts have made progress
607: \citep{Valenti1998, Woolf2005, Bonfils2005, Bean2006}. Six red dwarfs
608: that we have observed are included in the study by \cite{Bonfils2005},
609: comprising three pairs of M dwarfs of types M1.0V, M2.5V and M3.0V
610: shown in Figure~\ref{fig:bonfils.recons}. Each pair includes a
611: relatively low metallicity dwarf (gray), and a relatively high
612: metallicity dwarf (black). The red ends (from 8200\AA~to 9000\AA) of
613: each pair match one another, but the blue ends of the lower
614: metallicity members is brighter in each case, as predicted by the GAIA
615: models. The effect is rather subtle, but the derived metallicities
616: for each pair are not wildly different (none of the six stars are
617: subdwarfs), and yet the trend is confirmed in all three cases. One
618: caveat is that the metallicities from these six stars were determined
619: from the polynomial relation in \cite{Bonfils2005}, not measured
620: directly from spectra, but this appears to be the best that can be
621: done given the available data. We conclude that the metallicity trend
622: revealed in the GAIA models (see Figure~\ref{fig.4800-2800.plot}) is
623: sound because it appears to be confirmed in real M dwarf spectra.
624:
625: Contrary to the evidence for the metallicity trend, we have found no
626: direct spectroscopic results to support the gravity trend in dwarfs
627: and subdwarfs. Direct gravity measurements are difficult because (1)
628: stars must have both mass and radius measurements and (2) clean
629: spectra without contamination from companions must be obtained. This
630: limits the available target lists to eclipsing binaries such as those
631: discussed in \cite{Lopez2007} with cleanly deconvolved spectra or
632: visual binaries in which individual radii can be measured via long
633: baseline interferometry. Neither class of objects yet provides a rich
634: dataset for cool dwarfs or subdwarfs.
635:
636: Hence, we rely on other observational or theoretical efforts to
637: investigate the gravity trend. The top panel of
638: Figure~\ref{fig.3500.plot} indicates that other than some sharp
639: metallic lines, CaH bands show the most prominent changes when gravity
640: varies (TiO is unaffected by changing gravity). \cite{Ohman1934}
641: demonstrated that the CaH2 band is found in the spectra of M-type
642: dwarfs, but is not observed in the spectra of M-type giants, thus
643: identifying the CaH2 band as a gravity indicator to separate dwarfs
644: and giants. In addition to the GAIA models, \cite{Mould1976} also
645: showed that for stars with effective temperatures of 3250K, a spectrum
646: from his atmospheric model with {\it log g} $=$ 5.75 has stronger CaH2
647: than a spectrum with {\it log g} $=$ 4.75. If the CaH2 band is a
648: gravity indicator, we may presume that the same gravity effects for
649: CaH1 and CaH3 bands will be seen.
650:
651: As shown in the two panels of Figure~\ref{fig.3500.plot}, the GAIA
652: models imply that the CaH band strengths are indicators of both
653: gravity and metallicity differences. In reality, if two red subdwarfs
654: have spectra with the same overall continua and slopes, as well as
655: matching TiO band strengths (e.g.~middle panel in
656: Figure~\ref{fig.problem.spect}), the only remaining discrepancies will
657: be at the CaH bands. We believe that such differences are caused by
658: different gravities.
659:
660: Obviously, there is not yet a wealth of accurate direct observational
661: results of red dwarfs and subdwarfs that can be used to stress test
662: the metallicity and gravity trends seen in GAIA models. Nonetheless,
663: what little we do have supports the trends, so we use these trends to
664: assist us in establishing the subdwarf spectral sequence discussed
665: below.
666:
667: \section{Subdwarf Spectral Sequence}
668:
669: \subsection{Procedures}
670:
671: In addition to presenting 88 subdwarfs, a goal of this project is to
672: establish a subdwarf spectral sequence that mirrors the sequence for
673: dwarfs and considers the latest available synthetic models. Although
674: synthetic models are not yet capable of fully representing the
675: complicated spectra of these cool stars, the models can be used to
676: investigate the primary factors that affect subdwarfs' spectra.
677: Specifically, we examine the spectra in the framework of what appear
678: to be the three main drivers of the trends observed --- temperature,
679: metallicity, and gravity.
680:
681: We first separate our available subdwarf spectra into several groups
682: that have similar overall slopes. Known subdwarfs such as LHS 12, GJ
683: 161, and LHS 2734A are used as anchor points. Within each group,
684: several stars that have high surface gravities are obvious because
685: their spectra match except in CaH, implying nearly identical
686: temperatures and metallicities.
687:
688: We then compare each subdwarf's spectrum with our sequence of dwarf
689: spectral standards\footnote{Our dwarf standard sequence is a hybrid of
690: cool dwarf standards from \cite{Gray2006} for K stars, and
691: \cite{Boeshaar1976} and \cite{Kirkpatrick1991} for M stars. The dwarf
692: standard sequence is the topic of a future publication
693: \citep{Beaulieu2009}. Our spectral sequencing efforts began before
694: the recent spectral sequence from SDSS \citep{Bochanski2007} was
695: released. We continue to use our standard sequence because (1) stars
696: in our sequence have trigonometric parallaxes so we can understand how
697: metallicities and gravities affect stars' positions on the HR diagram,
698: (2) the SDSS sequence does not include K-type stars, (3) we have
699: acquired spectra for many K-type stars from \cite{Gray2003} and
700: \cite{Gray2006} so we have benchmark K dwarf spectral standards to
701: link to the M dwarf sequence, (4) SDSS standard stars have telluric
702: lines removed (It is difficult for us to make comparisons to SDSS
703: spectra in the 6800--7100\AA~region, which includes O$_{2}$ and
704: H$_{2}$O absorption. This region overlaps with the CaH3 and TiO5
705: bands. Thus, all subdwarfs would appear to have high gravity when
706: compared to the SDSS standards.), and (5) our dwarf spectra have been
707: acquired using the same telescope/instrument/observing protocols as
708: used for the subdwarfs, and the data have been reduced identically, so
709: they are systematically consistent.} that span types from K0.0 to M9.0
710: using an IDL program to find the closest match to the continuum slope
711: in the region 8200--9000\AA. Visual checks of the matches between all
712: subdwarf and standard dwarf spectra are also made to ensure match
713: quality.
714:
715: \subsection{Results and Notes on Objects}
716: \label{sec.notes}
717:
718: Based on the 88 confirmed cool subdwarf spectra we have, we present a
719: sequence of subdwarfs with spectral types spanning K3.0 to M6.0,
720: listed in Table~\ref{tbl.spectral.type}. We also identify five
721: additional G type subdwarfs. After the five G-type subdwarfs, we sort
722: the cool subdwarfs from K3.0[VI] to M6.0VI, using double lines in
723: Table~\ref{tbl.spectral.type} to separate each type.
724:
725: As discussed in $\S$\ref{sec.K.subdwarf}, because mid K subdwarfs are
726: virtually indistinguishable from K dwarfs at our spectral resolution,
727: we use [VI] to indicate their questionable luminosity classes, which
728: are currently based on their metallicities, kinematics, or locations
729: on the HR diagram. We anticipate that higher resolution spectra will
730: reveal these stars to be subdwarfs. A colon after the type indicates
731: that we have had difficulty in assigning a subtype, metallicity, or
732: gravity.
733:
734: Within each type for which sufficient spectra are available, we sort
735: targets by their metallicities (lowest metallicity first). A letter
736: ``m'' indicates a star having the same metallicity as a main sequence
737: dwarf, while more negative signs indicate lower metallicities,
738: e.g. m$--$ is more metal poor than m$-$. We use as many as six
739: negative signs, because in the case of the M1.0VI type, we have 23
740: stars that fall into seven different metallicity categories.
741: Gravities are indicated by ``g'' with additional plus signs for higher
742: gravities, e.g. g$++$ is higher gravity than g$+$. A baseline
743: subdwarf of low metallicity is assigned m$-$ and g. If a subdwarf has
744: a similar metallicity but higher gravity, the designations are m$-$
745: and g$+$. A few stars appear to have solar metallicity and are
746: subluminous only because of high gravities; these stars are assigned m
747: and g$+$. We have removed any stars from this study that might have
748: gravities lower than main sequence stars, i.e.~slightly evolved stars
749: such as subgiants, that would have gravity g$-$.
750:
751: Note that values for metallicity and gravity are not comparable across
752: all spectral types, i.e. M1.0VI with metallicity ``m$-$'' is not
753: equivalent to M2.0VI with metallicity ``m$-$,'' nor is ``g$+$'' for
754: M1.0VI the same as ``g$+$'' for M2.0VI. We can hope to formalize a
755: definitive subdwarf spectral sequence that includes temperature,
756: metallicity, and gravity trends when hundreds of systematically
757: consistent subdwarf spectra and improved model grids are available.
758: Objects with virtually identical spectra are listed in alphabetical
759: and/or numerical order.
760:
761: We discuss each spectral subtype from K3.0[VI] to M6.0VI in the
762: following sections and highlight noteworthy subdwarfs of various
763: types. The order of the highlighted targets is based on their
764: metallicities or gravities. Colored spectra in the Figures in these
765: sections are illustrative only --- they do not represent any numeric
766: metallicities or gravities, so a red-lined spectrum in one type does
767: not necessarily have similar attributes to a red-lined spectrum in
768: another type.
769:
770: \subsubsection{K3.0[VI] to K5.0[VI] types}
771:
772: We begin with K3.0[VI], for which it is still difficult to separate
773: subdwarfs from dwarfs at our spectral resolution.
774: Figure~\ref{fig.4800-2800.plot} shows how similar mid K-type subdwarfs
775: and dwarfs are (temperatures 4400K and 4800K). For these stars, we
776: use other independent published measurements, e.g.~[m/H] values, to
777: confirm their subdwarf natures. We currently call stars with [m/H]
778: $\leq$ $-$0.5 subdwarfs. The K3.0[VI] to K5.0[VI] spectra are shown
779: in Figure~\ref{fig.K3.K5}.
780:
781: {\bf G 016$-$009AB (K3.0[VI])} \cite{Goldberg2002} reported this star
782: to be a double-lined spectroscopic binary with P$=$9.9 days,
783: $T_{eff}=$ 5903K, and having a mass ratio ($M_{1}$/$M_{2}$) of
784: $\sim$1.22. Given that the components are presumably coeval, the
785: combined spectrum should represent the metallicity for each component.
786: Although their estimated temperature is that of a G type star, our
787: spectrum has the same slope as a K3.0V. \cite{Cayrel2001} and
788: \cite{Goldberg2002} reported measurements of [Fe/H]$=-$0.7 and $-$1.0
789: respectively, while \cite{Laird1988} found [m/H]$=-$1.11. Apparently,
790: it is a low metallicity star and its location on the HR diagram is
791: elevated because of multiplicity. This system is also reported to be
792: a photometrically variable system in \cite{Kazarovets2006}. This
793: system demonstrates that low metallicity subdwarfs do not have
794: different spectra from their main sequence counterparts at our
795: wavelength coverage and resolution, so we assign its luminosity class
796: as [VI] as a result its low metallicity measurements.
797:
798: {\bf LHS 2467 (K4.0[VI])} Our spectrum shows that it is a K4.0[VI],
799: rather than a G7V type, as reported in \cite{Bidelman1985}. Its
800: continuum is not as blue as K0.0V and its Ba I/H${\alpha}$ lines are
801: not as weak/strong as shown in Figure~\ref{fig.jacoby.silva} for a G
802: type star. Its weighted mean parallax from \cite{Hipparcos} and
803: \cite{YPC} is 10.63$\pm$1.88 mas and its proper motion is 0\farcs98
804: yr$^{-1}$ \citep{LHS}, indicating $V_{tan}=$ 437 km sec$^{-1}$.
805: \cite{Ryan1991} reported the star to have $V_{rad}=$ 44 km sec$^{-1}$.
806: After removing the solar motion, LHS 2467 has $(U,V,W)=$
807: (281.4,$-$304.8,$-$41.9) km sec$^{-1}$. The extremely high tangential
808: and space velocities are indicative of a subdwarf.
809:
810: {\bf G 026$-$009ACD (K5.0[VI])} The wide (132\arcsec, corresponding to
811: 6520 AU at the system's distance of 49.4 pc) common proper
812: motion companion, B, in this quadruple system is a white dwarf known
813: as G026$-$010 or G026$-$009B, that during its planetary nebula phase
814: should have had limited impact on the metallicity of the close ACD
815: triple. \cite{Peterson1980} reported G026-009AC to be double-lined
816: spectroscopic binary with P$=$3.75 days, with a mass ratio
817: ($M_{1}$/$M_{2}$)$\sim$1.25. \cite{Allen2000} reported a third
818: component, D, 0\farcs7 away from G026-009AC. \cite{Morrison2003} and
819: \cite{Allen2000} both report the system (ACD) to have low metallicity,
820: measuring [Fe/H]$=-$0.91 and $-$1.19, respectively. Both the $V$ and
821: $K_{s}$ magnitudes include three stars, causing the point to be
822: improperly placed on the HR diagram.
823:
824: {\bf G 022-015 (K5.0[VI])} \cite{Woolf2005} reported this system to
825: have [Fe/H]$=-$0.61. G022-015's spectrum is virtually identical to
826: G026-009ACD, as shown by plotting both together in
827: Figure~\ref{fig.K3.K5}.
828:
829: {\bf GJ 223.1 (K5.0[VI])} This star has the same slope as the K5.0V
830: standard and its location on the HR diagram is on the main sequence
831: line. Nonetheless, \cite{Woolf2005} reported it to have
832: [m/H]$=-$0.62, so we tentatively consider it to be a subdwarf until
833: further information indicates otherwise.
834:
835: Although not definitive, the Ca I line at 6162\AA~in these spectra
836: seems to show the metallicity trend seen in the noiseless model
837: spectra plotted in Figure~\ref{fig.4800-2800.plot} and discussed in
838: section~\ref{sec.subdwarf.facts}. LHS 2467 has a much weaker Ca I
839: line, and G016$-$009AB and G026$-$009AB have slightly weaker Ca I
840: lines than seen in the dwarfs' spectra. However, G022$-$015 and
841: GJ223.1 both have approximately the same Ca I line strengths as
842: dwarfs. Contrary to the Ca I line, the Ca II lines at (8542\AA~and
843: 8662\AA) do not show any metallicity trend from our spectra. Because
844: these prominent lines fail to separate subdwarfs from dwarfs at our
845: resolution and S/N, other evidence, i.e.~kinematics, parallaxes or
846: independent metallicity measurements, are required to identify the
847: subdwarfs. An alternative method to separate the subdwarfs from the
848: dwarfs is to obtain spectra that include the MgH bands at 4845\AA~,
849: 5211\AA, and 5621\AA~that \cite{Bessell1982} pointed are sensitive to
850: metal abundance for temperatures hotter than 4500K.
851:
852: Several possibilities may explain the locations of the three
853: subdwarfs, LHS 2467, G 022-015, and GJ 223.1 on the HR diagram just
854: below or on the main sequence line: (1) they are subdwarfs with unseen
855: companions that brighten their $M_{Ks}$ magnitudes, (2) they are
856: slightly evolved and have moved significantly from the subdwarf zero
857: age main sequence line, or (3) their metallicity measurements are not
858: accurate and they are, in fact, main sequence dwarfs.
859:
860: {\bf LHS 125, LHS 232 and LHS 327} The spectra for these three stars,
861: shown in Figure~\ref{fig.K4}, match neither the LHS 2467 K4.0[VI]
862: spectrum nor a K4.0V spectrum, although the spectra have the overall
863: slopes and the stars have $V-K_s$ colors indicative of K4.0 stars.
864: They are tentatively assigned types of K4.0[VI] based on their
865: locations on the HR diagram.
866:
867: \subsubsection{K6.0VI}
868:
869: We identify four stars to have spectral type K6.0VI, shown in
870: Figure~\ref{fig.K6}. Their spectral slopes fall between K5.0V and
871: K7.0V.
872:
873: {\bf LHS 193A} This is a binary with a separation of 12\farcs59
874: comprised of a cool subdwarf (A) and a DC-type white dwarf (B) with a
875: featureless spectrum \citep{Monteiro2006}. At the system's distance
876: of 31.2 pc, the large separation of the pair corresponds to 393 AU,
877: indicating that significant pollution of the subdwarf by the white
878: dwarf during its planetary nebula phase seems unlikely. From
879: Figure~\ref{fig.K6}, it is clear that LHS 193A's spectral slope is
880: between K5.0V and K7.0V. Although the spectrum is a near match to
881: K5.0V between 8200\AA~and 9000\AA, the blue end (6000\AA--7500\AA) is
882: too low to be a K5.0V (as a subdwarf, the blue end of the spectrum
883: would have to be above the dwarf standard). As shown in
884: Figure~\ref{fig.4800-2800.plot}, for a star with 4400K (a K5.0V from
885: \citealt*{Cox2000}), there is effectively no difference in the
886: continua of spectra for dwarfs and subdwarfs. Hence, we assign LHS
887: 193A a type of K6.0VI because its TiO5 band is ``brightened'' (low
888: metallicity), as shown in Figure~\ref{fig.4800-2800.plot}. Given its
889: low metallicity, location on the HR diagram (shown in the inset of
890: Figure~\ref{fig.K6}), its $V_{tan}=$147.3 km sec$^{-1}$
891: \citep{Jao2005}, and its age of 6--9 Gyr based on its white dwarf
892: companion's cooling age \citep{Monteiro2006}, it is most likely a
893: thick disk subdwarf.
894:
895: {\bf LHS 73 and LHS 227} These two objects have virtually identical
896: spectra, so we only show LHS 227 in Figure~\ref{fig.K6} for clarity.
897: The only differences between these spectra and that of LHS 193A are in
898: CaH. Using results from the top panel of Figure~\ref{fig.3500.plot},
899: we find that LHS 73 and LHS 227 have higher gravities than LHS 193A.
900:
901: We have acquired spectra for both LHS 72 and LHS 73, a wide
902: ($\sim$96\arcsec~separation) common proper motion pair.
903: Unfortunately, the spectrum of LHS 72 ($V-K_{s} =$ 3.29, $M_{Ks} =$
904: 6.69) is poor and requires re-observation, although it would fall in
905: the subdwarf region of the HR diagram. LHS 73 has $V-K_{s}=$ 3.42 and
906: $M_{Ks}=$ 7.27 and is plotted in Figure~\ref{fig.K6}, also clearly in
907: the subdwarf region. \cite{Rodgers1974} previously identified both
908: objects as subdwarfs, but no spectral types were given.
909: \cite{Bidelman1985} identified LHS 72 and LHS 73 as K4 and K5 dwarfs,
910: respectively. \cite{Reyle2006} reported LHS 72/73 (they identified
911: LHS 73 as sdK7) to be the nearest subdwarf binary based on their
912: spectroscopic parallax (18.7 pc), although the trigonometric parallax
913: from YPC is 37.6$\pm$8.9 mas (21.5 pc $<$ $d$ $<$ 34.8 pc). The
914: nearest known subdwarf binary system is actually $\mu$ Cas AB, at a
915: distance of 7.5 pc ($\pi$$_{trig}$ $=$ 132.4$\pm$0.6 mas,
916: \citealt{Hipparcos}).
917:
918: {\bf LHS 161} This star was previously reported to be an esdM2.0 extreme
919: subdwarf \citep{Gizis1997}. However, its slope and spectral features,
920: like LHS 227, are too hot for type M2.0VI. As shown in
921: Figure~\ref{fig.K6}, LHS 161 has stronger CaH than LHS 193A and LHS
922: 227, but the rest of the spectrum is the same. We conclude that LHS
923: 161 has higher gravity than the other two subdwarfs. LHS 161's very
924: strong CaH lines and consequent CaH and TiO5 indices place it in the
925: previously called ``extreme'' subdwarf region, which typically implies
926: that the star has very low metallicity. It appears, from our current
927: understanding, that the strong CaH in LHS 161 may not be linked to
928: metallicity, because the TiO5 band is not significantly different from
929: LHS 193A, LHS 73 or LHS 227.
930:
931: In the inset of Figure~\ref{fig.K6} we connect LHS 193A, LHS 73, LHS
932: 227, and LHS 161 on the HR diagram with thick arrows to outline a
933: sequence of increasing gravity effects that shift their locations to
934: the lower right (less luminous and redder). We will soon see more
935: such gravity effects, which are always towards lower luminosities.
936:
937: \subsubsection{K7.0VI}
938: \label{sub.sec.K7.0VI}
939:
940: We identify four stars to have spectral type K7.0VI, shown in
941: Figure~\ref{fig.K7}. This is the earliest spectral type for which we
942: see clear metallicity effects in the blue regions of our spectra.
943:
944: {\bf LHS 2734A} This is the primary in a new common proper motion
945: system discovered during CTIOPI, for which we measure the components
946: to have $\mu$ = 0$\farcs$59 and 0$\farcs$60 yr$^{-1}$ at position
947: angles 228.2$^{\circ}$ and 228.4$^{\circ}$ for the A and B components,
948: respectively. The B component is located 68\farcs8 away at position
949: angle 162.4$^{\circ}$. Unfortunately, both components are beyond our
950: current trigonometric parallax limit (100 pc), but the A component
951: provides a useful subdwarf anchor point even though its location on
952: the HR diagram is unknown.
953:
954: There are three reasons to adopt LHS 2734A as a reliable subdwarf
955: anchor. First, we measure a zero parallax, implying that its
956: $V_{tan}$ is at least 280 km/sec (assuming a distance of 100 pc). A
957: tangential velocity of this size is indicative of a subdwarf. Second,
958: LHS 2734B $(V-K_{s}=$ 3.90), is redder than LHS 2734A ($V-K_{s}=$
959: 3.23), and later type subdwarfs are more easily separated from dwarfs,
960: as shown in Figure~\ref{fig.4800-2800.plot}. As discussed in
961: $\S$\ref{sub.sec.M1}, LHS 2734B is clearly a subdwarf. Assuming this
962: common proper motion pair formed at the same time with similar
963: ``genetics'', both components are subdwarfs. Third,
964: Figure~\ref{fig.K7} shows the comparison between K7.0V and LHS 2734A,
965: in which the blue end of LHS 2734A's spectrum is clearly
966: ``brightened''. This indicates that LHS 2734A has lower metallicity
967: than K7.0V, as expected for a subdwarf.
968:
969: {\bf LHS 164} This star is four full magnitudes below the main
970: sequence line in the inset in Figure~\ref{fig.K7}. We find a good
971: match at the red end between our K7.0V standard and LHS 164's
972: spectrum, while the blue end is elevated. We conclude that LHS 164's
973: metallicity is even lower than LHS 2734A's.
974:
975: Thus, our K7.0V standard, LHS 2734A, and LHS 164 form a sequence of
976: decreasing metallicity. The CaH and TiO5 band depths significantly
977: decrease as the metallicity drops, matching the effects discussed in
978: section~\ref{sec.K.subdwarf} and Figure~\ref{fig.model.index.plot}.
979:
980: {\bf DEN0515$-$7211 and SCR 0708$-$4709} These two spectra are
981: virtually identical, and are assigned type K7.0VI:
982: (Figure~\ref{fig.K7}) because the spectra do not match either the
983: K7.0V standard, LHS 164, or LHS 2734A. DEN0515$-$7211 is nearly five
984: magnitudes less luminous in $M_{Ks}$ than the main sequence line, so
985: it is certainly a subdwarf. SCR 0708$-$4709 ($\mu=$0\farcs402
986: yr$^{-1}$, \citealt{Subasavage2005a}) has almost the same spectrum as
987: DEN0515$-$7211 and falls in the subdwarf region on the reduced proper
988: motion diagram (see \citealt{Subasavage2005a}).
989:
990: \subsubsection{M0.0VI}
991:
992: We show four stars with spectral type M0.0VI in Figure~\ref{fig.M0}.
993: This is the first group in this manuscript for which we see clear
994: effects of both metallicity and gravity.
995:
996: {\bf LHS 244, LHS 418, and LHS 165} These three spectra have the same
997: continua at the red end but are very different at the blue end. The
998: differences are caused by metallicity, as seen in
999: Figure~\ref{fig.4800-2800.plot}, with the main sequence standard--LHS
1000: 244--LHS 418--LHS 165 trending to lower metallicities. The blue ends
1001: of the subdwarf spectra have clearly ``brightened'' along this
1002: sequence, which is consistent with the theoretical models. In
1003: addition, as discussed in $\S$\ref{sec.K.subdwarf}, CaH1 absorption
1004: decreases (increasing index) as metallicity decreases if $T_{eff}$ is
1005: greater than 3500K. LHS 165 is redder in $V-K_{s}$ and has stronger
1006: CaH1 absorption than LHS 418, so we suspect LHS 165 has a higher
1007: gravity (similar to gravity effects shown in Figure~\ref{fig.K6}). We
1008: also find that LHS 418 has weaker CaH1 absorption than LHS 244,
1009: indicating that LHS 244 could also possibly have higher gravity than
1010: LHS 418. LHS 244 is much redder than LHS 418 and LHS 165 in the HR
1011: diagram is because of a combination of gravity and metallicity
1012: effects.
1013:
1014: {\bf LHS 424} This star has the same spectrum and color as LHS 418, so
1015: we assign the same spectral type. We note that LHS 424 is a full
1016: magnitude less luminous in $M_{Ks}$, perhaps hinting that LHS 418 is a
1017: multiple system.
1018:
1019: {\bf LHS 300AB} This is a close binary with a separation of
1020: $\sim$4\farcs3 \citep{Jao2003} comprised of a cool subdwarf (A) and a
1021: DC-type white dwarf (B) with a featureless spectrum
1022: \citep{Monteiro2006}. The spectrum obtained includes both components,
1023: but with $\Delta$VRI $=$ 4.61, 4.85, 4.96 mag \citep{Monteiro2006},
1024: the contamination of the cool subdwarf spectrum from the white dwarf
1025: is negligible. At a distance of 31.0 pc, the projected
1026: separation implies a distance of 133 AU between the two
1027: components, indicating that the subdwarf's composition was unlikely to
1028: be significantly contaminated by the evolved star during its planetary
1029: nebula phase. The spectrum does not match other M0.0VI stars at the
1030: blue end, but it matches the red end of M0.0V and M0.0VI well. We
1031: therefore assign it a type of M0.0VI:.
1032:
1033: \subsubsection{M0.5VI}
1034:
1035: This is the first half spectral type we currently assign for
1036: subdwarfs. We see both metallicity and gravity effects for M0.5VI, as
1037: shown in Figures~\ref{fig.M0.5.1} and \ref{fig.M0.5.2}.
1038:
1039: {\bf LHS 507, SCR 0709$-$4648, and LHS 12} These three stars, along
1040: with our M0.5V standard, illustrate a nice four-step trend in
1041: metallicity. With the lowest metallicity, LHS 507 is significantly
1042: bluer than LHS 12, which is, in turn, bluer than our main sequence
1043: standard. We do not yet have a trigonometric parallax for SCR
1044: 0709$-$4648, so we cannot confirm its location on the HR diagram. LHS
1045: 12 was previously reported to be a type sdM0.0 in \cite{Gizis1997}.
1046:
1047: {\bf LHS 401} This star has stronger CaH bands and therefore higher
1048: gravity than LHS 507, which has extremely weak CaH1. LHS 401 is
1049: redder than LHS 507, as expected, but is slightly more luminous, which
1050: is not consistent with the gravity effects seen for types K6.0VI and
1051: M0.0VI. We believe this inconsistency is a result of the large
1052: absolute magnitude errors for these two objects.
1053:
1054: {\bf LHS 521} This star has a spectrum nearly identical to LHS 507,
1055: except at 6000-6200\AA. Its CaH2 feature is slightly different from
1056: LHS 507, so we assign it a type of M0.5VI:.
1057:
1058: {\bf LHS 367 and LHS 299} Both objects have trigonometric parallaxes.
1059: Their spectra are similar to SCR 0709$-$4648 and LHS 12, respectively
1060: (shown in Figure~\ref{fig.M0.5.2}), so they also have similar offsets
1061: in metallicities from one another. We assign them as M0.5VI: because
1062: of (1) discrepancies at 6000\AA--6200\AA, (2) LHS 299 has deeper CaH1
1063: absorption than LHS 12 and different CaH2 absorption, and (3) their
1064: locations on the HR diagram are redder than their respective
1065: comparison stars. LHS 299 may have a higher gravity than LHS 12, but
1066: a smaller absolute magnitude error is needed to confirm this
1067: speculation.
1068:
1069: {\bf LHS 360} This star appears to have higher gravity than LHS 12
1070: because the CaH features are deeper, but its location on the HR
1071: diagram does not match its stronger gravity because its $V-K_{s}$
1072: color is slightly bluer than LHS 12. \cite{Costa2006} reported that
1073: LHS 360 has $V_{tan}\sim$ 524 km s$^{-1}$. Based on its very high
1074: $V_{tan}$ and location on the HR diagram, we assign it a type of
1075: M0.5VI:.
1076:
1077: \subsubsection{M1.0VI}
1078: \label{sub.sec.M1}
1079:
1080: The M1.0VI type has the largest number (23) of confirmed subdwarfs in
1081: our current sample. The spectra of 16 different M1.0VI stars are
1082: shown in Figures~\ref{fig.M1.1} and~\ref{fig.M1.2}. Many of these
1083: stars have similar colors, as identified in a reduced proper motion
1084: diagram derived using results from our SuperCOSMOS-RECONS (SCR) survey
1085: \citep{Subasavage2005a, Subasavage2005b}. Given the rich dataset for
1086: this type, we first discuss the sample in terms of metallicity and
1087: gravity effects, then discuss individual targets.
1088:
1089: {\it Metallicity effects} Perhaps better than for any other spectral
1090: type available, the M1.0VI stars show a beautiful trend in metallicity
1091: in their spectra, as shown in the top of Figure~\ref{fig.M1.1}. The
1092: red ends of the spectra match the M1.0V standard spectrum, but the
1093: blue ends of the spectra are very different because of metallicity
1094: effects. We assign their metallicities on a scale of m to m$------$,
1095: where m is indistinguishable from the main sequence standard and 6
1096: ``-'' indicate the most severely metal poor subdwarf.
1097:
1098: As metallicity drops, the TiO5 band gradually weakens, as predicted in
1099: the model spectra of Figure~\ref{fig.4800-2800.plot} for stars cooler
1100: than $\sim$4000K. Our M1.0V standard (metallicity m), LHS 109
1101: (metallicity scale m$----$), and LHS 518 (metallicity scale m$------$)
1102: have trigonometric parallaxes, and as predicted from the models, their
1103: positions on the HR diagram shift bluer with decreasing metallicity.
1104: At metallicity m$-$, SCR 2101$-$5437 is slightly metal poor compared
1105: to M1.0V (TiO5), and has slightly higher gravity than M1.0V (CaH).
1106: Unfortunately, none of the SCR objects in Figure~\ref{fig.M1.1} have
1107: trigonometric parallaxes, yet. Thus, they cannot be plotted on the HR
1108: diagram, and the progressive trend in metallicity effects for these
1109: stars cannot yet be shown.
1110:
1111: {\it Gravity effects} The effects of gravity can be seen at three
1112: different metallicities for type M1.0VI. Spectral for metallicity
1113: scale m$------$ subdwarfs are shown in the middle of
1114: Figure~\ref{fig.M1.1}. SCR 0701$-$0655 has higher gravity than LHS
1115: 518. Both spectra are identical, except at CaH. Spectra for
1116: metallicity scale m$----$ subdwarfs are shown in the bottom of
1117: Figure~\ref{fig.M1.1}. A clear trend can be seen. Three of these
1118: five objects have trigonometric parallaxes --- LHS 109, LHS 385 and
1119: LHS 335 --- and their gravity differences shift their locations on the
1120: HR diagram toward redder and less luminous territory, as with
1121: previously discussed types. (LHS 385's high parallax error is the
1122: likely cause of the slight offset between its HR diagram position and
1123: LHS 109's.) Finally, as shown at the top of Figure~\ref{fig.M1.2} for
1124: metallicity scale m$---$, SCR 0654$-$7358 has higher gravity than SCR
1125: 1756$-$5927.
1126:
1127: {\bf LHS 440} was reported by \cite{Bidelman1985} to be type M1.0V.
1128: The second set of spectra in Figure~\ref{fig.M1.2} shows that LHS 440
1129: has deeper absorptions in all three CaH bands than the M1.0V standard,
1130: but is virtually identical at all other wavelengths. This indicates
1131: that LHS 440 probably has higher gravity than M1.0V, making it less
1132: luminous than main sequence stars on the HR diagram. Because it is
1133: just barely one magnitude below the main sequence fit, we currently
1134: assign it a type of M1.0VI:.
1135:
1136: {\bf LHS 1970 and LHS 2734B} both have noisy spectra because they are
1137: faint. As shown in Figure~\ref{fig.M1.2}, their strong CaH bands
1138: indicate that they are high gravity (strong CaH), low metallicity
1139: (weak TiO5) subdwarfs. LHS 1970's location on the HR diagram confirms
1140: that it is a subdwarf and \cite{Gizis1997} reported its spectral type
1141: is esdM2.5. LHS 2734B's spectral similarity to LHS 1970 indicates that
1142: it also has low metallicity and high gravity. Unfortunately, we
1143: measure zero parallax for the LHS 2734AB pair, within the errors, so
1144: we can not plot them on the HR diagram (see
1145: section~\ref{sub.sec.K7.0VI} for discussion of LHS 2734A). We assign
1146: both LHS 1970 and LHS 2734B types of M1.0VI: because of their noisy
1147: spectra.
1148:
1149: {\bf LHS 158} has a spectrum nearly identical to an M1.0V star, but
1150: falls one magnitude below the main sequence line on the HR diagram.
1151: It also has $V_{tan}=$ 191 km sec$^{-1}$ \citep{Jao2005}. The low
1152: luminosity and tangential velocity together imply that LHS 158 may be
1153: a subdwarf, so we assign it a type of M1.0[VI], representing its
1154: uncertain assignment as a subdwarf, as is the case for the mid K-type
1155: subdwarfs.
1156:
1157: \subsubsection{M2.0VI}
1158:
1159: This is the earliest type for which we see that CaH1 absorption
1160: deepens as the metallicity decreases, as shown by the curves in
1161: Figure~\ref{fig.model.index.plot} for $T_{eff}$ less than 3500K.
1162: Figure~\ref{fig.M2} shows the spectra for stars discussed in this
1163: section.
1164:
1165: {\bf LHS 406, GJ 191 (Kapteyn's Star) and WT0233} form a sequence of
1166: decreasing metallicity. These three stars have parallaxes, and their
1167: positions on the HR diagram show a clear trend toward the blue for
1168: stars with decreasing metallicity (shown with open arrows). One of
1169: the prototypes of the subdwarf class, Kapteyn's Star, which has
1170: [Fe/H]$=$$-$0.99 from \cite{Woolf2005}, has previously been reported
1171: to be a type sdM1.0 in \cite{Gizis1997}. We assign a somewhat later
1172: type for this famous star, which at 3.9 pc is the nearest known
1173: subdwarf of any type.
1174:
1175: {\bf GJ 191/LHS 3620 and LHS 406/LHS 127} The relative effects of
1176: gravity can be seen in the spectra of these two pairs. LHS 3620 has
1177: much stronger gravity than GJ 191, while LHS 127 has slightly stronger
1178: gravity than LHS 406. On the HR diagram, LHS 3620 is redder and less
1179: luminous than GJ 191, as expected. LHS 162 has a spectrum (not
1180: plotted) virtually identical to LHS 3620, but is only slightly redder
1181: and less luminous than GJ 191.
1182:
1183: {\bf LHS 318} is almost identical to GJ 191 except for weaker TiO5
1184: absorption. This weaker TiO5 indicates LHS 318 is slightly more metal
1185: poor than GJ 191, so the blue end of the spectrum should be
1186: ``brightened''; however, it is not. LHS 318's location on the HR
1187: diagram in relation to GJ 191 suggests a higher gravity that is not
1188: apparent in its spectrum. We would expect LHS 318 to appear brighter
1189: and bluer than GJ 191 if metallicity is the only difference in these
1190: two stars, so we assign it as M2.0VI:.
1191:
1192: \subsubsection{M3.0VI}
1193:
1194: {\bf LHS 228/LHS 189AB and LHS 326/LHS 398} are pairs of stars with
1195: nearly identical spectra. As shown in Figure~\ref{fig.M3}, LHS 326,
1196: LHS 228 and our M3.0V standard form a metallicity sequence. As in the
1197: case of M2.0VI, CaH1 absorption deepens for M3.0VI stars with
1198: decreasing metallicity. In the case of LHS189AB, \cite{Costa2006}
1199: reported their separation is about 3\arcsec~ and their
1200: $\Delta$$R\sim$0.5mag. At a distance of 22.1 pc, the projected
1201: separation implies a distance of $\sim$66 AU between the two
1202: components. This spectrum has combined both components, so we assign
1203: it type M3.0VI: until completely ``clean'' data can be acquired for
1204: the components, and their individual metallicity scales can be
1205: determined.
1206:
1207: {\bf WT 135} was previously identified by \cite{Henry2002} as type
1208: M2.5V. We assign its metallicity m$-$: because its TiO5 band is the
1209: same as seen in LHS 228, but the blue end of its spectrum is
1210: ``brightened'' (making it somewhat metal poor compared to LHS 228).
1211: It lacks a trigonometric parallax, so we cannot plot it on the HR
1212: diagram.
1213:
1214: {\bf LHS 272} with $V-I=$ 4.29 is similar in color to WT 135 ($V-I=$
1215: 4.32) but appears to have slightly higher gravity (deeper CaH bands).
1216: Overall, the blue end of its spectrum (6000\AA--6300\AA) is comparable
1217: to WT 135, although with portions slightly brighter and portions
1218: slightly fainter. We therefore assign its metallicity to be the same
1219: as WT 135, at m$-$:. It is slightly redder and less luminous than LHS
1220: 228. We use both hollow (metallicity) and solid (gravity) arrows to
1221: indicate its location relative to LHS 228. The competition between
1222: metallicity and gravity effects in this case seems to indicate that
1223: gravity is the dominant factor because LHS 272's position moves to the
1224: red, rather than the blue. LHS 272 was reported as sdM3.0 in
1225: \cite{Gizis1997}.
1226:
1227: {\bf LHS 326, SCR 2204$-$3347/LHS 541 and SCR 1916$-$3638} form a
1228: clear sequence of gravity effects. Other than having a noisier
1229: spectrum (not shown), LHS 541 appears to be identical to SCR
1230: 2204$-$3347. LHS 541's location on the HR diagram relative to LHS 326
1231: reflects the effects of gravity.
1232:
1233: \subsubsection{M3.5VI}
1234:
1235: This half type is assigned because four stars have spectra shown in
1236: Figure~\ref{fig.M3.5} with redder slopes than M3.0VI (compare to LHS
1237: 228, a M3.0VI subdwarf in the second set of spectra), but not as steep
1238: as M4.0VI (see next section). Conveniently, all four stars of this
1239: type have trigonometric parallaxes.
1240:
1241: {\bf LHS 381 and LHS 144} have lower metallicities than observed for
1242: our M3.5V standard. Their spectra redward of 8200\AA~match M3.5V
1243: relatively well and show the same trends predicted by the GAIA models
1244: (Figure~\ref{fig.4800-2800.plot}) for stars with temperatures of about
1245: 3200K. Their CaH1 indices also match the trend discussed in
1246: section~\ref{sec.K.subdwarf}.
1247:
1248: {\bf LHS 375 and LHS 515} have higher gravities than LHS 381 and LHS
1249: 144, respectively. Both objects are redder and less luminous than
1250: their low gravity counterparts on the HR diagram. \cite{Gizis1997}
1251: reported LHS 375 to be an esdM4.0 subdwarf and \cite{Reid2005}
1252: reported LHS 515 to be an esdM5.0 subdwarf. We classify both as
1253: M3.5VI with high gravities.
1254:
1255: \subsubsection{M4.0VI to M6.0VI}
1256:
1257: We identify only four subdwarfs with types M4.0VI to M6.0VI, only one
1258: of which, LHS 2067A, currently has a trigonometric parallax. Because
1259: of the paucity of such objects, we do not yet have a sample
1260: sufficiently large to map out the effects of metallicity and gravity.
1261:
1262: When stars are cooler than $\sim$3200K, their spectra redward of
1263: 7500\AA~change radically between [m/H]=0.0 and $-$1.0 (see
1264: Figure~\ref{fig.4800-2800.plot}), but are rather more stable between
1265: [m/H] = $-$1.0 and $-$2.0. For example, Figure~\ref{fig.M4.6} shows
1266: comparisons between LEHPM 3861 (M4.0VI) and our M3.0V and M4.0V
1267: standards. Clearly, if its spectral type is incorrectly assigned to
1268: be M3.0VI, there is excess red flux. \cite{Lodieu2005} reported LEHPM
1269: 3861 to be a sdM6.0 subdwarf.
1270:
1271: {\bf LEHPM 3861} (M4.0VI), {\bf LHS 1490} (M5.0VI), {\bf LHS 334}
1272: (M6.0VI), and {\bf LHS 2067A} (M6.0VI), whose spectra are shown in
1273: Figure~\ref{fig.M4.6} are assigned subdwarf spectral types later than
1274: M3.5VI. The types are based on fluxes redward of 7500\AA, with
1275: emphasis on three pseudo-continuum peak points at 8130, 8250 and
1276: 8840\AA~(marked in Figure~\ref{fig.M4.6} with dotted lines). As shown
1277: in Figure~\ref{fig.4800-2800.plot} for stars cooler than 3200K, the
1278: flux decreases at these three points as the metallicity drops from
1279: [m/H]=0.0 to $-$1.0 and this trend becomes even more prominent as
1280: $T_{eff}$ decreases. Thus, these three peaks for a low metallicity
1281: subdwarf will not be brighter than a dwarf with a comparable spectral
1282: type.
1283:
1284: {\bf LHS 2067A} is bluer than M6.0V on the HR diagram and is slightly
1285: below the main sequence line. LHS 2067A was previously identified to
1286: be a subdwarf in \cite{Kirkpatrick1995}, but no spectral type was
1287: given. Although its CaH1 and CaH2+CaH3 indices are not in the
1288: subdwarf region (see Figure~\ref{fig.cahn.plot} and
1289: Table~\ref{tbl.cahn.tio5.index}), its spectrum is clearly different
1290: from M6.0V and M7.0V, with stronger CaH1 absorption. It forms a wide
1291: ($\sim$55\arcsec~NE) common proper motion pair with a white dwarf, LHS
1292: 2067B, so it, like LHS 193AB and LHS 300AB, forms an intriguing pair
1293: that can be used for comparing metallicity and white dwarf ages. At
1294: the system's distance of 25.5 pc, the large separation of the
1295: pair corresponds to 472 AU, indicating that significant pollution of
1296: the subdwarf by the white dwarf during its planetary nebula phase
1297: seems unlikely.
1298:
1299: {\bf LHS 334} was reported to be a sdM4.5 subdwarf in \cite{Reid2005}.
1300: We assign it a type of M6.0VI because of the good match of the three
1301: pseudo-continuum peak points to LHS 2067A, which appears to be
1302: slightly more metal rich than LHS 334.
1303:
1304: \section{Application to SDSS Subdwarfs}
1305: \label{sec:SDSS}
1306:
1307: It is useful to apply our spectral typing methodology to the recent
1308: work of \cite{West2004}, who have provided a significant sample of
1309: sixty new subdwarf spectra, all acquired and reduced in a homogeneous
1310: way. The subdwarfs were selected and spectral types were assigned via
1311: their CaH and TiO indices. For comparison, we retrieved
1312: fifty\footnote{The remaining 10 subdwarfs could not be retrieved using
1313: the SDSS DR4 website, even though we used a 1\arcmin~search radius and
1314: coordinates from their table.} of the publicly available spectra and
1315: use our new method to assign types. The brightest star among these
1316: subdwarfs has $r$=17.0, resulting in somewhat noisy spectra for the
1317: sample, so we smoothed the spectra by averaging the flux over five
1318: pixels and normalizing at 7500\AA. Representative spectra are shown
1319: in Figures~\ref{fig.sdss.subdwarf.1} and~\ref{fig.sdss.subdwarf.2}.
1320:
1321: The SDSS spectra have telluric lines removed, so we use M dwarf
1322: standard spectra from \cite{Bochanski2007} that also omit the telluric
1323: lines, rather than our own standards. In addition,
1324: \cite{Bochanski2007} do not present half-type standard spectra, so we
1325: do not present any half-type spectra for the SDSS subdwarfs. Previous
1326: and current types are listed in Table~\ref{tbl.sdss.subdwarf}. The
1327: subdwarfs have types between M1.0 and M3.0 in \cite{Bochanski2007},
1328: while our efforts yield types between M2.0 and M5.0.
1329:
1330: Our point here is not to assign definitive spectral types, but to
1331: illustrate the gravity and metallicity effects in the SDSS dataset,
1332: and provide examples of the application of our proposed spectral
1333: typing method. Stars of a given spectral type, of course, may have a
1334: range of metallicities, as discussed at length in previous sections,
1335: and as is evident in Figure~\ref{fig.sdss.subdwarf.1}. Gravity
1336: effects are also seen, as shown in Figure~\ref{fig.sdss.subdwarf.2}.
1337: Three stars (SDSSJ085843.89$+$511210.1, SDSSJ093141.85$+$453914.5 and
1338: \\SDSSJ145447.32$+$011006.8) previously identified as subdwarfs have
1339: spectra identical to SDSS M dwarfs. Two other stars,
1340: SDSSJ083217.77$+$522408.2 and SDSSJ113501.76$+$033720.3, appear to
1341: have slightly higher gravities (stronger CaH1 line) than dwarfs, but
1342: the rest of these two spectra are almost identical to SDSS M dwarfs.
1343: Therefore, to these two stars we assign types of M2.0VI: and M3.0VI:
1344: with g+:, respectively.
1345:
1346: Unfortunately, the SDSS dwarfs do not have parallaxes that can be used
1347: to confirm their locations on the HR diagram. In addition, we find
1348: many SDSS subdwarfs with identical spectral types that have $g-z$
1349: colors as different as 0.53. It is therefore difficult to verify how
1350: metallicities and gravities affect the SDSS absolute magnitudes and
1351: colors. Nonetheless, the robust sample of \cite{West2004} provides
1352: many additional subdwarfs that can be targeted for further work.
1353:
1354: \section{Discussion}
1355: \label{sec:discussion}
1356:
1357: \subsection{Why Subdwarfs' Physical Parameters Are Not Listed Explicitly}
1358:
1359: Stellar spectra follow trends primarily defined by temperature, yet as
1360: we have seen, metallicities and gravities also have significant
1361: effects on the spectra of cool subdwarfs. Here we have provided a
1362: consistent spectral sequence for subdwarfs that is based on linking
1363: observed subdwarf spectra to spectra acquired for main sequence dwarfs
1364: using the same telescope/instrument/observing protocols. Here we
1365: compare synthetic to observed spectra to evaluate how well we can
1366: assign values for the temperatures, metallicities, and gravities of
1367: cool dwarfs and subdwarfs.
1368:
1369: \subsubsection{Matching Synthetic and Observed Spectra}
1370:
1371: To test the reliability of the model grids and our fitting procedures,
1372: we use our six standard spectral sequence dwarfs with types M0.0V to
1373: M5.0V as test spectra (see Figure~\ref{fig.observed.model}). All are
1374: currently believed to be uncorrupted single red dwarfs because they
1375: show no evidence of multiplicity from combinations of (1) HST/NICMOS
1376: observations, (2) optical speckle observations, (3) optical CCD
1377: imaging, and/or (4) three or more years of astrometric observations
1378: that would reveal perturbations from unseen companions that
1379: contributed significant light to the system.
1380:
1381: We compared the spectral region from 6000\AA~to 9000\AA~after
1382: normalizing both the GAIA grid spectra and ours at 7500\AA, and
1383: applying a Gaussian function to the much higher resolution GAIA model
1384: spectra to match our resolution. If a star had a significant spectral
1385: shift because of radial velocity, we manually offset the spectrum to
1386: match features obvious in the synthetic spectra. We then calculated
1387: the reduced $\chi^{2}$ differences (hereafter, simply $\chi^{2}$)
1388: between the model grid spectra and ours. Because the synthetic
1389: spectra do not have the telluric lines of O$_{2}\alpha$
1390: (6270\AA--6330\AA), O$_{2}$B (6860\AA--6980\AA), O$_{2}$A
1391: (7590\AA--7710\AA) and water (7150\AA--7330\AA~and 8952\AA--9000\AA),
1392: these absorption regions were excluded in the $\chi^{2}$ calculations.
1393:
1394: The selected model grids have effective temperatures of 2400K to
1395: 4500K, [m/H] from $-$2.0 to $+$0.5, and {\it log g} from 4.0 to 5.5 in
1396: steps of 100K, 0.5 dex, and 0.5 dex, respectively. The upper panel of
1397: Figure~\ref{fig.3500.plot} shows that for a metallicity of $-$1.0 and
1398: temperature 3500K, varying the gravity causes changes in only certain
1399: wavelength regions (and some specific lines). This means $\chi^{2}$
1400: is only sensitive to {\it log g} in relatively small spectral regions,
1401: but not for the overall spectrum. On the other hand, as shown in the
1402: lower panel of Figure~\ref{fig.3500.plot}, for the same temperature
1403: star with {\it log g} fixed at 5.0, there are large differences
1404: between spectra when the metallicity is varied.
1405:
1406: Thus, for a star of a given temperature, the gravity changes the
1407: overall spectral shape minimally while the metallicity changes it a
1408: great deal, so we first secure a star's metallicity and then its
1409: gravity. Figure~\ref{fig.GJ701.chi2} is an example for M1.0V standard
1410: star, showing $\chi^{2}$ curves at various metallicities and
1411: gravities. Each curve represents a specific {\it log g} and grid
1412: spectra have temperatures incremented by 100K. The smallest scatter
1413: in the plots is found when [m/H] is $-$0.5, which is adopted as the
1414: star's metallicity. We then examine each point in the [m/H]$=$$-$0.5
1415: panel to find the best fit with the smallest $\chi^{2}$ at this fixed
1416: metallicity. In this case, the best fit has $T_{eff}=$ 3600K, {\it
1417: log g} $=$ 4.5 and $[m/H] =$ $-$0.5.
1418:
1419: \subsubsection{Discrepancies Between the Best Fitting Synthetic and
1420: Observed Spectra}
1421:
1422: In this paper we do not explicitly list temperatures, metallicities,
1423: or {\it log g} values for subdwarfs. A few examples support our
1424: reasoning for not listing these physical parameters.
1425: Figure~\ref{fig.observed.model} shows observed spectra and the best
1426: fit synthetic spectra for six main sequence spectral standard stars
1427: with types M0.0V to M5.0V (in steps of 1.0 subtypes). In general, the
1428: overall slopes of the model spectra fit fairly well, especially for
1429: the earlier types. However, we discuss here six regions labeled at
1430: the top of Figure~\ref{fig.observed.model} that do not match, which
1431: are particularly relevant for the subdwarfs that are the focus of this
1432: paper.
1433:
1434: \begin{enumerate}
1435:
1436: \item Region 1: The observed spectrum is always less luminous than the
1437: model, except for M0.0V.
1438:
1439: \item Region 2: This is the CaH1 absorption region. Observed spectra
1440: have shallower CaH1 features than the models. The depth of CaH1 is
1441: determined by a combination of gravity and metallicity. Decreasing
1442: CaH1 absorption can be caused by either decreasing gravity or
1443: decreasing metallicity. However, fine tuning the gravity or
1444: metallicity affects not only the CaH1 feature, but the TiO5 feature in
1445: region 4.
1446:
1447: \item Region 3: The pseudo-continuum always peaks near 6530\AA~in the
1448: observed spectra. However, the synthetic spectra have this peak
1449: ``red-shifted'' to $\sim$6650\AA.
1450:
1451: \item Region 4: Containing the CaH2, CaH3 and TiO5 features, this is
1452: the most important region for examining the interplay of metallicity
1453: and gravity in cool dwarfs. Unfortunately, the CaH2 feature is
1454: blended with O$_{2}$B so it is not entirely reliable for analysis of
1455: spectra taken through the Earth's atmosphere. The CaH3 and TiO5
1456: features are usually weaker in observed spectra than in the models.
1457: The strength of TiO5 is primarily driven by metallicity, not gravity
1458: (see the bottom panel of Figure~\ref{fig.3500.plot}). Therefore, this
1459: region reveals valuable information about a star's metallicity, in
1460: particular at types later than M2.0.
1461:
1462: \item Region 5: Several Fe I absorption lines (8388\AA, 8440\AA~and
1463: 8718\AA) and a Mg I line (8718\AA) in the model spectra redward of
1464: 8300\AA~are deeper than observed. This indicates poor metallicity
1465: matches and/or poor modeling of those particular lines.
1466:
1467: \item Region 6: Overall, redward of 8000\AA~the M4.0V and M5.0V
1468: matches are not as good as other regions. This likely indicates
1469: fundamental problems with the strengths of some opacity sources
1470: (e.g.~H$_{2}$O and TiO) in the models.
1471:
1472: \end{enumerate}
1473:
1474: The top panel of Figure~\ref{fig.GJ701.LP776} shows the two best
1475: fitting synthetic spectra for our M1.0V standard. The red spectrum
1476: (3600K, [m/H]$ = -$0.5, {\it log g}$ = $4.5) provides the best fit and
1477: the blue spectrum (3400K, [m/H]$ = -$1.0, {\it log g}$ = $4.0) is
1478: second best. There are only slight differences between the two
1479: synthetic spectra --- the CaH1 strengths and continuum fluxes redward
1480: of 8000\AA~are the only notable differences. However, the effective
1481: temperatures for the two models differ by 200K, while the
1482: metallicities and {\it log g} each differ by 0.5 dex. The lower panel
1483: of Figure~\ref{fig.GJ701.LP776} shows the two best fit spectra for our
1484: M3.0V standard. The best fit (red line) yields 3200K, [m/H]$=-$0.5,
1485: {\it log g}$=$4.5 and the second best fit (blue line) yields 3300K,
1486: [m/H]$=$0.5, {\it log g}$=$5.5. However, neither model fit is an
1487: ideal match to the observed spectrum (which is why the $[m/H]$ value
1488: for the two best matches differs by an order of magnitude), and
1489: matches become even poorer for cooler stars.
1490:
1491: As a whole, many of the best fits for the main sequence dwarfs in
1492: Figure~\ref{fig.observed.model} are for metallicities of $-$0.5, which
1493: is somewhat lower than studies that have specifically attempted to
1494: assign metallicities to cool dwarfs in the solar neighborhood. For
1495: example, the mean [Fe/H] from 21 M dwarf secondaries in
1496: \cite{Bonfils2005} is $-$0.09, and the mean [Fe/H] from five M dwarf
1497: secondaries in \cite{Bean2006} is $-$0.17. Also worthy of note is
1498: that the fits were made for metallicities incremented by 0.5 dex, and
1499: stars with metallicities between $-$0.5 and 0.0 may have slightly
1500: better fits for $-$0.5. One might think that interpolation from
1501: existing model grids would allow better fits. There are, however,
1502: degeneracy problems in matching model spectra to observations --- when
1503: fitting the three-dimensional space of temperature, metallicity, and
1504: gravity, two or more synthetic spectra yield low points with similar
1505: $\chi^{2}$ , as is the case for the M1.0V example discussed above.
1506: More worrisome is that the overall discrepancies discussed above
1507: (items 1 to 6) cannot be removed through interpolation in the existing
1508: model grids.
1509:
1510: Figure~\ref{fig.chi2.curve} illustrates how the $\chi^{2}$ values
1511: change for best fit matches of the GAIA models to our spectral
1512: standards with types M0.0V through M5.5V. Spectral type M3.0V is the
1513: latest type for which a reasonably tight plot like the one shown in
1514: Figure~\ref{fig.GJ701.chi2} can be identified. The dotted line in
1515: Figure~\ref{fig.chi2.curve} provides a dividing point between
1516: reasonable matches and poor matches, indicating that once $\chi^{2}$
1517: exceeds 10, the $\chi^{2}$ plots are too scattered to choose a unique
1518: set of model spectra parameters to match observed spectra.
1519:
1520: A specific example of applying the models to one of our subdwarfs
1521: illustrates the large discrepancies that must be overcome to derive
1522: reliable parameter values for metallicity and gravity. When we
1523: applied our fitting algorithm to our observed spectrum for LHS 335
1524: (M1.0VI, shown in Figure~\ref{fig.lhs0335}), we discovered that the
1525: ``best fit'' was quite poor. Although the formal $\chi^{2}$ value was
1526: 5.9 (less than our cutoff of 10, indicating a reliable fit), the
1527: model's CaH3 band is not deep enough, the TiO5 band is too deep, and
1528: the continuum flux redward of 7700\AA~is less than observed.
1529:
1530: For myriad reasons we conclude that we cannot strictly determine
1531: reliable metallicities and {\it log g} values for cool dwarfs using
1532: the fitting method discussed here. Thus, until improved model grids
1533: are available, we defer assignment of numerical values for
1534: temperatures, metallicities and {\it log g} values. However, we
1535: certainly {\it can} use the GAIA model grids to mimic {\it trends}
1536: (discussed in $\S$\ref{sec.subdwarf.facts} and~\ref{sec.trend}) in the
1537: spectra of subdwarfs to compare the stars within a framework of
1538: changing physical parameters.
1539:
1540: \subsection{The Confusion Between {\it sd}, {\it esd} and {\it usd} Prefixes}
1541:
1542: \cite{Gizis1997} proposed that subdwarfs have two subclasses,
1543: ``subdwarfs'' ({\it sd}) and ``extreme subdwarfs'' ({\it esd}) based
1544: on their CaH and TiO5 band strengths. He separated the two classes
1545: using the line shown in the top panel of Figure~\ref{fig.cahn.plot}.
1546: Another term, ``ultra subdwarf'', was proposed by \cite{Caldwell1984}
1547: to describe GJ 59B ($V-K_{s}=$ 3.42, $M_{Ks}=$ 7.89), which is
1548: $\sim$2.5 magnitudes underluminous compared to main sequence stars of
1549: similar color. This star falls in the region including ``extreme
1550: subdwarfs'' in \cite{Gizis1997}. More recently, \cite{Lepine2007}
1551: adopted the ``ultra subdwarf'' ({\it usd}) term for subdwarfs found to
1552: have stronger CaH features than ``extreme subdwarfs''. We believe
1553: these terms confuse the situation and do not address the underlying
1554: astrophysics. We recommend that they not be used for the following
1555: reasons.
1556:
1557: \begin{itemize}
1558:
1559: \item These terms can only be applied to M-type subdwarfs, not K-type
1560: subdwarfs. Both observed spectra and models show that K subdwarfs do
1561: not follow the trends in CaH absorption with metallicity that M
1562: subdwarfs follow. Thus, there is no clear delineation for K-type
1563: subdwarfs.
1564:
1565: \item Empirically, the values of the indices are affected by a
1566: complicated interplay of temperature, metallicity, and gravity
1567: effects. One cannot separate these three factors simply based on the
1568: indices. Typically, when the term ``extreme'' subdwarf is used, it
1569: refers to ``very low metallicity'' alone. But that is only one part
1570: of the portrait that needs to be painted for a given subdwarf. For
1571: example, filled boxes shown in Figure~\ref{fig.hr.plot} indicate stars
1572: previously identified as extreme subdwarfs from their indices. Note
1573: that the spread for ``extreme'' subdwarfs at a given color
1574: ($V-K_{s}\sim$ 3.5) is 3.5 magnitudes, or a factor of $\sim$25 in
1575: luminosity. There is no clear separation in the fundamental HR
1576: diagram between stars termed ``subdwarfs'' and ``extreme subdwarfs''
1577: classified using spectral indices.
1578:
1579: In Figure~\ref{fig.subdwarf.sdm.esdm.index}, we take a detailed look
1580: at astrophysical causes that shift points in the CaH2+CaH3/TiO5
1581: indices plot. Red circles in Figure~\ref{fig.subdwarf.sdm.esdm.index}
1582: represent the lowest metallicity stars of each spectral type we have
1583: presented here. Some of these very low metallicity stars are not
1584: located in the ``extreme'' subdwarf region. Green circles represent
1585: subdwarfs in our sample that have highest gravity at a given type.
1586: Solid lines connect stars having the same metallicity rankings in
1587: Table~\ref{tbl.spectral.type} (components of the common proper motion
1588: binary LHS 2734 AB are connected by a dotted line because they
1589: presumably have the same metallicity). It is clear from these three
1590: pairs that high gravity can push a subdwarf toward or into the extreme
1591: subdwarf region, even if the metallicities of the two objects are
1592: similar. In addition, many other high gravity subdwarfs (green
1593: circles) that do not have particularly low metallicities are also in
1594: the subdwarf region. Finally, the two blue circles represent LHS 440
1595: and SCR 1822-45542, which subdwarfs having the same metallicities but
1596: higher gravities than dwarfs. Their spectroscopic features indicate
1597: that they are subdwarfs, but they do not have low metallicity at all.
1598: Yet, they fall in the subdwarf region of the indices plot, which has
1599: traditionally indicated low metallicity.
1600:
1601: The inset in Figure~\ref{fig.subdwarf.sdm.esdm.index} helps explain
1602: this phenomenon. We have calculated the spectral indices from the
1603: GAIA synthetic spectra with [m/H]$=$0.0, $-$1.0 and $-$2.0 and
1604: temperatures of 2800--4400K. Two different gravities were selected
1605: for each metallicity. The models indicate that both low metallicity
1606: and high gravity push stars toward the extreme subdwarf region. Thus,
1607: if two stars have the same metallicity, the one with higher gravity
1608: will be pushed more toward the extreme subdwarf region, indicating
1609: that this spectral indices plot is not a clear indicator of
1610: metallicity alone.
1611:
1612: \item Theoretically, in order to show how complicated the TiO5 vs CaH1
1613: plot is astrophysically, we use GAIA model grids to do a
1614: demonstration. We have chosen models with 2700K $< T_{eff} <$ 4500K,
1615: $-$3.0 $< [m/H] <$ 0.0 and 4.0 $< {\it log g} <$ 5.5, and have
1616: calculated the output CaH and TiO5 indices, as shown in
1617: Figure~\ref{fig.cah1.tio5.vs.teff}. As is apparent in the plots, for
1618: a given (TiO5, CaH1) indices pair, there are many possible parameter
1619: combinations. Both indices are a function of temperature, gravity and
1620: metallicity, which would require a 6-D plot to describe TiO5 vs CaH.
1621: Unfortunately, one cannot classify stars as subdwarfs, extreme
1622: subdwarfs, or ultra-subdwarfs using 2-D index plots that infer trends
1623: in metallicity alone.
1624:
1625: \end{itemize}
1626:
1627: For these reasons, the indices only indicate a star's location on
1628: these particular spectroscopic indices plots. They do not address low
1629: or very low metallicity only, as implied by past usage, but instead
1630: incorporate effects of both metallicity and gravity for stars of a
1631: given temperature. They do not provide obvious or direct links to
1632: positions on HR diagrams, which provide the astrophysical meaning
1633: underlying stellar classification. Finally, the modifiers ``extreme''
1634: and ``ultra'' themselves have effectively the same meaning and do not
1635: provide meaningful information about their differences.
1636:
1637: \subsection{Why Previous Methods Work for Dwarfs but Not Subdwarfs}
1638:
1639: In the Palomar-MSU spectroscopy survey, \cite{Reid1995} used the TiO5
1640: index and polynomial equations to assign subtypes for M dwarfs.
1641: \cite{Gizis1997} then applied the same methodology to M subdwarfs. As
1642: can be seen in Figure~\ref{fig.cah1.tio5.vs.teff}, TiO5 versus
1643: temperature is effectively linear, especially for {\it log g} $=$ 5.0
1644: and [m/H] $=$ 0.0, so this typing method works successfully for a
1645: fairly homogeneous set of dwarfs (possibly with somewhat different
1646: gravities) --- temperature is the main factor affecting the overall
1647: slope of red dwarf spectra. However, it is not as simple for
1648: subdwarfs because both low metallicity and high gravity stars are
1649: sorted into the same TiO5 values for various combinations. Thus, the
1650: same methodology that works for dwarfs cannot straightforwardly be
1651: applied to subdwarfs.
1652:
1653: \subsection{The Subdwarf Spectral Standards from \cite{Lepine2007}}
1654:
1655: \cite{Lepine2007} recently released a set of spectral standards for
1656: the sdM, esdM and usdM subclasses using wider spectral coverage than
1657: used by \cite{Gizis1997} and revised polynomial equations. Their
1658: project is contemporaneous with ours, but unfortunately there is only
1659: one star in both samples, LHS 228. We assign it a type of M3.0VI
1660: while their type is sdM2.0. As shown in Figure~\ref{fig:lepine.1},
1661: gravity effects come into play in their spectra\footnote{Their
1662: telluric lines have been removed.}. Each pair of spectra shown have
1663: nearly identical continua, with CaH bands being the only significant
1664: spectral difference. Using their method, each of the pairs of
1665: subdwarfs shown is assigned a different sub-type. We believe that by
1666: using the overall shape of the spectrum, with knowledge that both
1667: metallicity and gravity affect certain regions, each pair should have
1668: the same spectral sub-types. Based on the trends from GAIA models,
1669: the pairs of spectra shown have different gravities, as we have also
1670: seen in our spectra and those from SDSS (shown in $\S$\ref{sec:SDSS}).
1671: Figure~\ref{fig:lepine.2} illustrates spectral differences at types
1672: M3.0, M5.0 and M7.5. Although the spectra in each panel have the same
1673: sub-types for sdM, esdM and usdM subdwarfs, they show very different
1674: continua between 6000\AA~and 9000\AA. When examining the overall
1675: spectra, there is no clear morphology trend for each sub-type. We
1676: conclude that using spectral indices alone omits important information
1677: evident in the overall morphology of the spectra that is useful to
1678: spectral classification.
1679:
1680: \section{Conclusions}
1681:
1682: We have discussed 88 cool subdwarfs using spectra covering
1683: 6000--9000\AA. Based on these spectra and the trends from GAIA model
1684: grids, we have redefined the subdwarf spectral sequence, spanning
1685: types K3.0 to M6.0. We find that wide spectral coverage is the key to
1686: defining a subdwarf's spectral type. We consider this to be an
1687: important, but not final, step in defining the subdwarf spectral
1688: sequence.
1689:
1690: Through the understanding of GAIA model grids, we find that the key to
1691: assigning a subdwarf spectral type is to compare the spectrum to dwarf
1692: spectral standards in regions affected minimally by metallicity and
1693: gravity, thereby making a direct link between the dwarf and subdwarf
1694: sequences. Even so, it remains difficult to establish a definitive
1695: sequence for subdwarfs because of the multi-faceted nature of their
1696: spectra. Until we have surveyed a large number of subdwarfs and
1697: covered a multitude of possible temperatures, metallicities, and
1698: gravities, a definitive sequence will remain elusive.
1699:
1700: From an analysis of the history of the term ``subdwarf,'' and the
1701: layout of the fundamental HR diagram, we propose that the suffix
1702: ``VI'' be used, rather than the ``sd'' prefix, as the preferred
1703: spectral classification notation. This reduces the confusion between
1704: cool subdwarfs and hot OB subdwarfs. This is also prudent because
1705: subdwarfs really do form an independent class of stars on the HR
1706: diagram, for which five Roman numerals are in common use, with cool
1707: subdwarfs naturally falling beneath the main sequence V types.
1708:
1709: Overall, we find that trigonometric parallaxes are crucial for
1710: identifying mid K-type subdwarfs, and allow us to understand how the
1711: complex interplay of temperatures, metallicities, and gravities
1712: affects the positions of cool dwarfs on the HR diagram. We found that
1713: mid K-type subdwarfs can not be identified spectroscopically using our
1714: data. There are many G and M-type subdwarfs, so K-type subdwarfs
1715: presumably exist, but additional information is needed to identify
1716: them. Consequently, we use metallicities, kinematics, and parallaxes
1717: to acertain their true natures and use their locations on the HR
1718: diagram to flag them as ``subdwarfs'' (other spectroscopic wavelength
1719: and resolution combinations could also be used). Because they are not
1720: spectroscopically identified as subdwarfs in our spectra, we use a
1721: conservative notation, [VI], for subdwarfs of types K3.0 through K5.0
1722: to indicate their questionable status.
1723:
1724: We have confirmed that spectroscopic indices are useful in separating
1725: late K to late M-type subdwarfs from dwarfs, but that the indices have
1726: limitations when attempting to understand the astrophysical causes
1727: leading to observed subdwarf spectra. When combined with
1728: trigonometric parallax and photometric information, our results show
1729: that for lower metallicities, subdwarfs are generally bluer and
1730: brighter at optical wavelengths, so they slide up and to the left on
1731: the HR diagram (using axes $V-K_{s}$ vs $M_{Ks}$). In contrast,
1732: higher gravities make stars redder and less luminous at optical
1733: wavelengths, so subdwarfs generally slide down and to the right on the
1734: same HR diagram. Because of the complex, and not yet completely
1735: mapped out, interplay of temperatures, metallicities, and gravities,
1736: we conclude that the ``extreme'' and ``ultra'' prefixes {\it only}
1737: outline locations on spectroscopic indices plots, and do not
1738: successfully differentiate the underlying astrophysical causes for
1739: shifts on CaH/TiO5 plots.
1740:
1741: Improvements in the subdwarf spectral sequence can be made by
1742: observing wide common proper motion subdwarf binaries, like LHS 2734AB
1743: discussed here. Assuming identical metallicities, such binaries allow
1744: us to constrain one of the three variables that affects cool dwarf
1745: spectral types. The ultimate subdwarf spectral sequence will be
1746: three-dimensional, with temperature, metallicity, and gravity as
1747: independent variables (see \citealt{Kirkpatrick2005}, Figure 11).
1748:
1749: Previously, gravity effects in cool subdwarfs have been almost
1750: entirely ignored, as metallicity was the factor considered to describe
1751: changing CaH features. Our results clearly show the importance of
1752: gravity effects. An additional factor, contamination from an unseen,
1753: evolved, i.e.~white dwarf, companion could also change the slope of
1754: subdwarf spectra. Thus, comprehensive surveys of subdwarfs for
1755: companions are warranted\footnote{We have an ongoing project to survey
1756: cool subdwarfs for companions. To date, we have found that cool
1757: subdwarfs have a lower binary fraction than cool dwarfs as reported by
1758: \cite{Riaz2008}, so the contamination assumption from possible unseen
1759: companions appears unlikely.}. Finally, this work also shows that
1760: current synthetic spectra provide a useful framework in which to
1761: evaluate cool dwarf spectra, but they do not yet provide perfect
1762: matches, indicating that atmospheric models still require fine-tuning
1763: to make additional advances in the future characterization of cool
1764: dwarfs (P. Hauschildt private communication).
1765:
1766: \begin{acknowledgements}
1767:
1768: We would like to thank the referee, Sandy Leggett, for her very
1769: helpful comments that improved this paper. We also thank Pat Boeshaar
1770: and Doug Gies for their comments and suggestions. Peter Hauschildt
1771: was instrumental in helping us use the GAIA model grids, and Vincent
1772: Woolf contributed the old version of synthetic spectra. We thank
1773: Sebastien Lepine for providing his subdwarf spectra to us before they
1774: were publicly available, and for his comments on section 9.4. We
1775: appreciate the help of many members of the Georgia State University
1776: (GSU) RECONS team in their data acquisition and reduction efforts,
1777: especially Charlie Finch and Jennifer Winters. This work has been
1778: supported at GSU by NASA's Space Interferometry Mission, the National
1779: Science Foundation (NSF, grant AST-0507711), and GSU. Finally, we
1780: wish to thank the other members of the SMARTS Consortium, without whom
1781: the telescopes at CTIO used for this effort might not be available.
1782:
1783: This research has made use of the SIMBAD database, operated at CDS,
1784: Strasbourg, France. This paper has also made use of data from the
1785: SDSS. Funding for the SDSS and SDSS-II has been provided by the
1786: Alfred P. Sloan Foundation, the Participating Institutions, NSF, the
1787: U.S. Department of Energy, NASA, the Japanese Monbukagakusho, the Max
1788: Planck Society, and the Higher Education Funding Council for England.
1789: This work has used data products from the Two Micron All Sky Survey,
1790: which is a joint project of the University of Massachusetts and the
1791: Infrared Processing and Analysis Center at California Institute of
1792: Technology funded by NASA and NSF.
1793:
1794: \end{acknowledgements}
1795:
1796: \begin{thebibliography}
1797:
1798: \bibitem[Adams \& Joy(1922)]{Adams1922} Adams, W.~S., \& Joy, A.~H.\
1799: 1922, \apj, 56, 242
1800:
1801: \bibitem[Adams et al.(1935)]{Adams1935} Adams, W.~S., Joy, A.~H.,
1802: Humason, M.~L., \& Brayton, A.~M.\ 1935, \apj, 81, 187
1803:
1804: \bibitem[Allard \& Hauschildt(1995)]{Allard1995} Allard, F., \&
1805: Hauschildt, P.~H.\ 1995, \apj, 445, 433
1806:
1807: \bibitem[Allen et al.(2000)]{Allen2000} Allen, C., Poveda, A., \&
1808: Herrera, M.~A.\ 2000, \aap, 356, 529
1809:
1810: \bibitem[Bean et al.(2006)]{Bean2006} Bean, J.~L., Sneden, C.,
1811: Hauschildt, P.~H., Johns-Krull, C.~M., \& Benedict, G.~F.\ 2006, \apj,
1812: 652, 1604
1813:
1814: \bibitem[Beaulieu et al.(2009)]{Beaulieu2009} Beaulieu, T.~D., in
1815: preparation, 2009
1816:
1817: \bibitem[Bessell(1982)]{Bessell1982} Bessell, M.~S.\ 1982,
1818: Proceedings of the Astronomical Society of Australia, 4, 417
1819:
1820: \bibitem[Bessel(1990)]{Bessel1990} Bessel, M.~S.\ 1990, \aaps, 83, 357
1821:
1822: \bibitem[Bidelman(1985)]{Bidelman1985} Bidelman, W.~P.\ 1985, \apjs,
1823: 59, 197
1824:
1825: \bibitem[Bochanski et al.(2007)]{Bochanski2007} Bochanski, J.~J.,
1826: West, A.~A., Hawley, S.~L., \& Covey, K.~R.\ 2007, \aj, 133, 531
1827:
1828: \bibitem[Boeshaar(1976)]{Boeshaar1976} Boeshaar, P.~C.\ 1976,
1829: Ph.D.~Thesis
1830:
1831: \bibitem[Bonfils et al.(2005)]{Bonfils2005} Bonfils, X., Delfosse, X.,
1832: Udry, S., Santos, N.~C., Forveille, T., \& S{\'e}gransan, D.\ 2005,
1833: \aap, 442, 635
1834:
1835: \bibitem[Brott \& Hauschildt(2005)]{Brott2005} Brott, I., \&
1836: Hauschildt, P.~H.\ 2005, The Three-Dimensional Universe with Gaia,
1837: 576, 565
1838:
1839: \bibitem[Burgasser et al.(2003)]{Burgasser2003} Burgasser, A.~J., et
1840: al.\ 2003, \apj, 592, 1186
1841:
1842: \bibitem[Burgasser \& Kirkpatrick(2006)]{Burgasser2006} Burgasser,
1843: A.~J., \& Kirkpatrick, J.~D.\ 2006, \apj, 645, 1485
1844:
1845: \bibitem[Caldwell et al.(1984)]{Caldwell1984} Caldwell, J.~A.~R.,
1846: Spencer Jones, J.~H., \& Menzies, J.~W.\ 1984, \mnras, 209, 51
1847:
1848: \bibitem[Carney et al.(1994)]{Carney1994} Carney, B.~W., Latham,
1849: D.~W., Laird, J.~B., \& Aguilar, L.~A.\ 1994, \aj, 107, 2240
1850:
1851: \bibitem[Cayrel de Strobel et al.(2001)]{Cayrel2001} Cayrel de
1852: Strobel, G., Soubiran, C., \& Ralite, N.\ 2001, \aap, 373, 159
1853:
1854: \bibitem[Cayrel de Strobel et al.(1997)]{Cayrel1997} Cayrel de
1855: Strobel, G., Soubiran, C., Friel, E.~D., Ralite, N., \& Francois, P.\
1856: 1997, \aaps, 124, 299
1857:
1858: \bibitem[Chamberlain \& Aller(1951)]{Chamberlain1951} Chamberlain,
1859: J.~W., \& Aller, L.~H.\ 1951, \apj, 114, 52
1860:
1861: \bibitem[Costa \& M{\'e}ndez(2003)]{Costa2003} Costa, E., \&
1862: M{\'e}ndez, R.~A.\ 2003, \aap, 402, 541
1863:
1864: \bibitem[Costa et al.(2005)]{Costa2005} Costa, E., M{\'e}ndez, R.~A.,
1865: Jao, W.-C., Henry, T.~J., Subasavage, J.~P., Brown, M.~A., Ianna,
1866: P.~A., \& Bartlett, J.\ 2005, \aj, 130, 337
1867:
1868: \bibitem[Costa et al.(2006)]{Costa2006} Costa, E., M{\'e}ndez, R.~A.,
1869: Jao, W.-C., Henry, T.~J., Subasavage, J.~P., \& Ianna, P.~A.\ 2006,
1870: \aj, 132, 1234
1871:
1872: \bibitem[Allen's Astrophysical Quantities(2000)]{Cox2000} Cox, N. \
1873: 2000, Allen's astrophysical quantities, New York, AIP Press, Springer
1874:
1875: \bibitem[Deacon et al.(2005)]{Deacon2005} Deacon, N.~R., Hambly,
1876: N.~C., \& Cooke, J.~A.\ 2005, \aap, 435, 363
1877:
1878: \bibitem[Digby et al.(2003)]{Digby2003} Digby, A.~P., Hambly, N.~C.,
1879: Cooke, J.~A., Reid, I.~N., \& Cannon, R.~D.\ 2003, \mnras, 344, 583
1880:
1881: \bibitem[Gizis(1997)]{Gizis1997} Gizis, J.~E.\ 1997, \aj, 113, 806
1882:
1883: \bibitem[Gizis \& Reid(1997)]{Gizis1997b} Gizis, J., \& Reid, I.\
1884: 1997, \pasp, 109, 1233
1885:
1886: \bibitem[Gizis \& Reid(1999)]{Gizis1999} Gizis, J.~E., \& Reid, I.~N.\
1887: 1999, \aj, 117, 508
1888:
1889: \bibitem[Gray et al.(2003)]{Gray2003} Gray, R.~O., Corbally, C.~J.,
1890: Garrison, R.~F., McFadden, M.~T., \& Robinson, P.~E.\ 2003, \aj, 126,
1891: 2048
1892:
1893: \bibitem[Gray et al.(2006)]{Gray2006} Gray, R.~O., Corbally, C.~J.,
1894: Garrison, R.~F., McFadden, M.~T., Bubar, E.~J., McGahee, C.~E.,
1895: O'Donoghue, A.~A., \& Knox, E.~R.\ 2006, \aj, 132, 161
1896:
1897: \bibitem[Greenstein \& Eggen(1966)]{Greenstein1966} Greenstein,
1898: J.~L., \& Eggen, O.~J.\ 1966, Vistas in Astronomy, 8, 63
1899:
1900:
1901: \bibitem[ESA(1997)]{Hipparcos} The {\it Hipparcos} and Tycho
1902: Catalogues, \ 1997, ESA SP-1200 (Noordwijk: ESA)
1903:
1904: \bibitem[Feige(1958)]{Feige1958} Feige, J.\ 1958, \apj, 128, 267
1905:
1906: \bibitem[Giclas et al.(1971)]{Giclas1971} Giclas, H.~L., Burnham,
1907: R., \& Thomas, N.~G.\ 1971, Flagstaff, Arizona: Lowell Observatory, 1971
1908:
1909: \bibitem[Giclas et al.(1978)]{Giclas1978} Giclas, H.~L., Burnham,
1910: R., Jr., \& Thomas, N.~G.\ 1978, Lowell Observatory Bulletin, 8, 89
1911:
1912: \bibitem[Greenstein(1966)]{Greenstein1966a} Greenstein, J.~L.\ 1966,
1913: \apj, 144, 496
1914:
1915: \bibitem[Greenstein \& M{\"u}nch(1966)]{Greenstein1966b} Greenstein,
1916: J.~L., \& M{\"u}nch, G.\ 1966, \apj, 146, 618
1917:
1918: \bibitem[Goldberg et al.(2002)]{Goldberg2002} Goldberg, D., Mazeh, T.,
1919: Latham, D.~W., Stefanik, R.~P., Carney, B.~W., \& Laird, J.~B.\ 2002,
1920: \aj, 124, 1132
1921:
1922: \bibitem[Hartwick et al.(1984)]{Hartwick1984} Hartwick, F.~D.~A.,
1923: Cowley, A.~P., \& Mould, J.~R.\ 1984, \apj, 286, 269
1924:
1925: \bibitem[Hawley et al.(1996)]{Hawley1996} Hawley, S.~L., Gizis, J.~E.,
1926: \& Reid, I.~N.\ 1996, \aj, 112, 2799
1927:
1928: \bibitem[Henry et al.(2002)]{Henry2002} Henry, T.~J., Walkowicz,
1929: L.~M., Barto, T.~C., \& Golimowski, D.~A.\ 2002, \aj, 123, 2002
1930:
1931: \bibitem[Henry et al.(2004)]{Henry2004} Henry, T.~J., Subasavage,
1932: J.~P., Brown, M.~A., Beaulieu, T.~D., Jao, W.-C., \& Hambly, N.~C.\
1933: 2004, \aj, 128, 2460
1934:
1935:
1936: \bibitem[Henry et al.(2006)]{Henry2006} Henry, T.~J., Jao, W.-C.,
1937: Subasavage, J.~P., Beaulieu, T.~D., Ianna, P.~A., Costa, E., \&
1938: M{\'e}ndez, R.~A.\ 2006, \aj, 132, 2360
1939:
1940:
1941: \bibitem[Heber(1992)]{Heber1992} Heber, U.\ 1992, LNP Vol.~401: The
1942: Atmospheres of Early-Type Stars, 401, 233
1943:
1944: \bibitem[Houk \& Cowley(1975)]{Houk1975} Houk, N., \& Cowley, A.~P.\
1945: 1975, Ann Arbor: University of Michigan, Department of Astronomy,
1946: 1975
1947:
1948: \bibitem[Humason \& Zwicky(1947)]{Humason1947} Humason, M.~L., \&
1949: Zwicky, F.\ 1947, \apj, 105, 85
1950:
1951: \bibitem[Jacoby et al.(1984)]{Jacoby1984} Jacoby, G.~H., Hunter,
1952: D.~A., \& Christian, C.~A.\ 1984, \apjs, 56, 257
1953:
1954: \bibitem[Jao et al.(2003)]{Jao2003} Jao, W.-C., Henry, T.~J.,
1955: Subasavage, J.~P., Bean, J.~L., Costa, E., Ianna, P.~A., \&
1956: M{\'e}ndez, R.~A.\ 2003, \aj, 125, 332
1957:
1958: \bibitem[Jao (2004)]{Jao2004} Jao, Wei-Chun, PhD Dissertation, 2004.
1959:
1960: \bibitem[Jao et al.(2005)]{Jao2005} Jao, Wei-Chun, Henry, T.~J.,
1961: Subasavage, J.~P., Brown, M.~A., Ianna, P.~A., Bartlett, J.~L., Costa,
1962: E., \& M{\' e}ndez, R.~A.\ 2005, \aj, 129, 1954
1963:
1964: \bibitem[Jao in preparation (2008)]{RECONS} Jao, Wei-Chun, 2008, in
1965: preparation
1966:
1967: \bibitem[Jaschek \& Jaschek(1987)]{Jaschek1987} Jaschek, C., \&
1968: Jaschek, M.\ 1987, Cambridge: University Press, 1987
1969:
1970: \bibitem[Joy(1947)]{Joy1947} Joy, A.~H.\ 1947, \apj, 105, 96
1971:
1972: \bibitem[Kazarovets et al.(2006)]{Kazarovets2006} Kazarovets, E.~V.,
1973: Samus, N.~N., Durlevich, O.~V., Kireeva, N.~N., \& Pastukhova, E.~N.\
1974: 2006, Informational Bulletin on Variable Stars, 5721, 1
1975:
1976: \bibitem[Kirkpatrick et al.(1991)]{Kirkpatrick1991} Kirkpatrick,
1977: J.~D., Henry, T.~J., \& McCarthy, D.~W., Jr.\ 1991, \apjs, 77, 417
1978:
1979: \bibitem[Kirkpatrick et al.(1995)]{Kirkpatrick1995} Kirkpatrick,
1980: J.~D., Henry, T.~J., \& Simons, D.~A.\ 1995, \aj, 109, 797
1981:
1982: \bibitem[Kirkpatrick(2005)]{Kirkpatrick2005} Kirkpatrick, J.~D.\ 2005,
1983: \araa, 43, 195
1984:
1985: \bibitem[Kleinman et al.(2004)]{Kleinman2004} Kleinman, S.~J., et al.\
1986: 2004, \apj, 607, 426
1987:
1988: \bibitem[Kuiper(1939)]{Kuiper1939} Kuiper, G.~P.\ 1939, \apj, 89, 548
1989:
1990: \bibitem[Kuiper(1940)]{Kuiper1940} Kuiper, G.~P.\ 1940, \apj, 91, 269
1991:
1992: \bibitem[Laird et al.(1988)]{Laird1988} Laird, J.~B., Carney, B.~W.,
1993: \& Latham, D.~W.\ 1988, \aj, 95, 1843
1994:
1995: \bibitem[Leggett(1992)]{Leggett1992} Leggett, S.~K.\ 1992, \apjs, 82,
1996: 351
1997:
1998: \bibitem[Leggett et al.(2000)]{Leggett2000} Leggett, S.~K., Allard,
1999: F., Dahn, C., Hauschildt, P.~H., Kerr, T.~H., \& Rayner, J.\ 2000,
2000: \apj, 535, 965
2001:
2002: \bibitem[L{\'e}pine et al.(2003)]{Lepine2003} L{\'e}pine, S., Rich,
2003: R.~M., \& Shara, M.~M.\ 2003, \aj, 125, 1598
2004:
2005: \bibitem[L{\'e}pine \& Shara(2005a)]{Lepine2005a} L{\' e}pine, S., \&
2006: Shara, M.~M.\ 2005, \aj, 129, 1483
2007:
2008: %\bibitem[L{\'e}pine \& Shara(2005b)]{Lepine2005b} L{\'e}pine, S., \&
2009: %Shara, M.~M.\ 2005, Bulletin of the American Astronomical Society, 37,
2010: %1408
2011:
2012:
2013: \bibitem[L{\'e}pine et al.(2007)]{Lepine2007} L{\'e}pine, S.,
2014: Rich, R.~M., \& Shara, M.~M.\ 2007, \apj, 669, 1235
2015:
2016: \bibitem[Lodieu et al.(2005)]{Lodieu2005} Lodieu, N., Scholz, R.-D.,
2017: McCaughrean, M.~J., Ibata, R., Irwin, M., \& Zinnecker, H.\ 2005,
2018: \aap, 440, 1061
2019:
2020: \bibitem[L{\'o}pez-Morales(2007)]{Lopez2007} L{\'o}pez-Morales,
2021: M.\ 2007, \apj, 660, 732
2022:
2023:
2024: \bibitem[Luyten (1939)]{Luyten1939} Luyten, W.~J., Bruce Proper Motion
2025: Survey III. The Stars of Large Proper Motion and the Luminosity
2026: Function, Minneapolis: University of Minnesota, 1939
2027:
2028: \bibitem[Luyten(1979)]{LHS} Luyten, W.~J.\ 1979, LHS Catalog,
2029: Minneapolis, University of Minnesota, 1979, 2nd ed.
2030:
2031: %\bibitem[Maness et al.(2007)]{Maness2007} Maness, H.~L., Marcy, G.~W.,
2032: %Ford, E.~B., Hauschildt, P.~H., Shreve, A.~T., Basri, G.~B., Butler,
2033: %R.~P., \& Vogt, S.~S.\ 2007, \pasp, 119, 90
2034:
2035: \bibitem[Mould(1976)]{Mould1976} Mould, J.~R.\ 1976, \aap, 48, 443
2036:
2037: \bibitem[Monet et al.(1992)]{Monet1992} Monet, D.~G., Dahn, C.~C.,
2038: Vrba, F.~J., Harris, H.~C., Pier, J.~R., Luginbuhl, C.~B., \& Ables,
2039: H.~D.\ 1992, \aj, 103, 638
2040:
2041: \bibitem[Monteiro et al.(2006)]{Monteiro2006} Monteiro, H., Jao,
2042: W.-C., Henry, T., Subasavage, J., \& Beaulieu, T.\ 2006, \apj, 638,
2043: 446
2044:
2045: \bibitem[Morrison et al.(2003)]{Morrison2003} Morrison, H.~L., et al.\
2046: 2003, \aj, 125, 2502
2047:
2048: \bibitem[Nissen \& Schuster(1997)]{Nissen1997} Nissen, P.~E., \&
2049: Schuster, W.~J.\ 1997, \aap, 326, 751
2050:
2051: \bibitem[Nordstr{\"o}m et al.(2004)]{Nordstrom2004} Nordstr{\"o}m, B.,
2052: et al.\ 2004, \aap, 418, 989
2053:
2054: \bibitem[{\"O}hman(1934)]{Ohman1934} {\"O}hman, Y.\ 1934, \apj,
2055: 80, 171
2056:
2057:
2058: \bibitem[Peterson et al.(1980)]{Peterson1980} Peterson, R.~C.,
2059: Willmarth, D.~W., Carney, B.~W., \& Chaffee, F.~H., Jr.\ 1980, \apj,
2060: 239, 928
2061:
2062: \bibitem[Reid et al.(2001)]{Reid2001} Reid, I.~N., van Wyk, F.,
2063: Marang, F., Roberts, G., Kilkenny, D., \& Mahoney, S.\ 2001, \mnras,
2064: 325, 931
2065:
2066: \bibitem[Reid et al.(1995)]{Reid1995} Reid, I.~N., Hawley,
2067: S.~L., \& Gizis, J.~E.\ 1995, \aj, 110, 1838
2068:
2069: \bibitem[Reid and Gizis(2005)]{Reid2005} Reid, I.~N. and Gizis, J.~E.,
2070: \ 2005, \pasp, 117, 676
2071:
2072: \bibitem[Reyl{\'e} et al.(2006)]{Reyle2006} Reyl{\'e}, C.,
2073: Scholz, R.-D., Schultheis, M., Robin, A.~C., \& Irwin, M.\ 2006,
2074: \mnras, 373, 705
2075:
2076: \bibitem[Riaz et al.(2008)]{Riaz2008} Riaz, B., Gizis, J.~E.,
2077: \& Samaddar, D.\ 2008, \apj, 672, 1153
2078:
2079: \bibitem[Rodgers \& Eggen(1974)]{Rodgers1974} Rodgers, A.~W.,
2080: \& Eggen, O.~J.\ 1974, \pasp, 86, 742
2081:
2082: \bibitem[Roman(1955)]{Roman1955} Roman, N.~G.\ 1955, \apjs, 2, 195
2083:
2084: \bibitem[Ryan and Norris (1991)]{Ryan1991} Ryan, S.~G. and Norris,
2085: J.~E.,\ 1991, \aj, 101, 1835
2086:
2087: \bibitem[Sandage \& Eggen(1959)]{Sandage1959} Sandage, A.~R., \&
2088: Eggen, O.~J.\ 1959, \mnras, 119, 278
2089:
2090: \bibitem[Silva \& Cornell(1992)]{Silva1992} Silva, D.~R. and Cornell,
2091: M.~E.,\ 1992, \apjs, 81, 865
2092:
2093: \bibitem[Smart et al.(2007)]{Smart2007} Smart, R.~L., Lattanzi, M.~G.,
2094: Jahrei{\ss}, H., Bucciarelli, B., \& Massone, G.\ 2007, \aap, 464, 787
2095:
2096: \bibitem[Subasavage et al.(2005a)]{Subasavage2005a} Subasavage, J.~P.,
2097: Henry, T.~J., Hambly, N.~C., Brown, M.~A., \& Jao, W.-C.\ 2005, \aj,
2098: 129, 413
2099:
2100: \bibitem[Subasavage et al.(2005b)]{Subasavage2005b} Subasavage, J.~P.,
2101: Henry, T.~J., Hambly, N.~C., Brown, M.~A., Jao, W.-C., \& Finch,
2102: C.~T.\ 2005, \aj, 130, 1658
2103:
2104: \bibitem[Taylor (2001)]{Taylor2001} Taylor, B.~J., \ 2001, \aap, 398,
2105: 731
2106:
2107: \bibitem[Turnshek et al.(1985)]{Turnshek1985} Turnshek, D.~E.,
2108: Turnshek, D.~A., \& Craine, E.~R.\ 1985, Astronomy and Astrophysics
2109: Series, Tucson: Western Research Company, 1985
2110:
2111: \bibitem[van Altena et al.(1995)]{YPC} van Altena, W.~F., Lee, J.~T.,
2112: \& Hoffleit, D.\ 1995, The General Catalogue of Trigonometric Stellar
2113: Parallaxes (4th ed.; New Haven: Yale Univ. Obs.)
2114:
2115: \bibitem[Valenti et al.(1998)]{Valenti1998} Valenti, J.~A.,
2116: Piskunov, N., \& Johns-Krull, C.~M.\ 1998, \apj, 498, 851
2117:
2118: \bibitem[Veeder (1974)]{Veeder1974} Veeder, G.~J. \ 1974, \aj, 79,
2119: 1056
2120:
2121: \bibitem[Weis(1996)]{Weis1996} Weis, E.~W.\ 1996, \aj, 112, 2300
2122:
2123: \bibitem[West et al.(2004)]{West2004} West, A.~A., et al.\ 2004, \aj,
2124: 128, 426
2125:
2126: \bibitem[Woolf \& Wallerstein(2005)]{Woolf2005} Woolf, V.~M., \&
2127: Wallerstein, G.\ 2005, \mnras, 356, 963
2128:
2129: \end{thebibliography}
2130:
2131: %%%%%%%%%%%%%%%%%%%%%%%%% Beginning of Tables %%%%%%%%%%%%%%%%%%%%%%%%
2132:
2133: %%%%%%%%% table %%%%%%%%%%
2134:
2135: \begin{deluxetable}{lccccc}
2136: \tablewidth{12cm} \tablecaption{Spectroscopic Indices} \tablehead{
2137: \colhead{Object} & \colhead{TiO5} & \colhead{CaH1} & \colhead{CaH2} &
2138: \colhead{CaH3} & \colhead{CaH2$+$CaH3} }
2139: \startdata
2140: DEN0515$-$7211& 1.002 & 1.021 & 1.019 & 1.003 & 2.022 \\
2141: G016$-$009AB & 0.973 & 1.018 & 1.021 & 0.998 & 2.019 \\
2142: G022$-$015 & 0.978 & 0.996 & 0.989 & 0.984 & 1.973 \\
2143: G026$-$009ACD& 0.953 & 0.990 & 0.975 & 0.990 & 1.965 \\
2144: GJ0191 & 0.860 & 0.876 & 0.678 & 0.848 & 1.526 \\
2145: GJ0223.1 & 0.958 & 1.001 & 0.981 & 0.971 & 1.952 \\
2146: LEHPM1628 & 0.954 & 0.676 & 0.628 & 0.754 & 1.382 \\
2147: LEHPM3861 & 0.957 & 0.774 & 0.340 & 0.368 & 0.709 \\
2148: LHS0012 & 0.883 & 0.881 & 0.779 & 0.889 & 1.668 \\
2149: LHS0073 & 0.972 & 0.903 & 0.862 & 0.927 & 1.789 \\
2150: LHS0109 & 0.939 & 0.823 & 0.752 & 0.863 & 1.616 \\
2151: LHS0125 & 0.981 & 1.020 & 1.014 & 0.993 & 2.007 \\
2152: LHS0127 & 0.716 & 0.755 & 0.537 & 0.739 & 1.276 \\
2153: LHS0144 & 0.768 & 0.609 & 0.396 & 0.592 & 0.987 \\
2154: LHS0148 & 0.977 & 0.816 & 0.729 & 0.830 & 1.560 \\
2155: LHS0158 & 0.732 & 0.849 & 0.639 & 0.829 & 1.469 \\
2156: LHS0161 & 0.889 & 0.777 & 0.689 & 0.817 & 1.506 \\
2157: LHS0162 & 0.838 & 0.731 & 0.577 & 0.747 & 1.324 \\
2158: LHS0164 & 0.987 & 1.005 & 1.000 & 1.017 & 2.017 \\
2159: LHS0165 & 0.956 & 0.889 & 0.841 & 0.910 & 1.751 \\
2160: LHS0186 & 0.710 & 0.739 & 0.568 & 0.764 & 1.332 \\
2161: LHS0189AB & 0.630 & 0.733 & 0.492 & 0.719 & 1.211 \\
2162: LHS0193A & 0.944 & 0.959 & 0.920 & 0.966 & 1.886 \\
2163: LHS0227 & 1.000 & 0.895 & 0.883 & 0.949 & 1.832 \\
2164: LHS0228 & 0.644 & 0.744 & 0.480 & 0.701 & 1.181 \\
2165: LHS0232 & 0.998 & 1.000 & 1.020 & 1.005 & 2.025 \\
2166: LHS0244 & 0.902 & 0.853 & 0.731 & 0.869 & 1.601 \\
2167: LHS0272 & 0.862 & 0.715 & 0.527 & 0.736 & 1.264 \\
2168: LHS0299 & 0.903 & 0.904 & 0.756 & 0.884 & 1.641 \\
2169: LHS0300AB & 1.031 & 0.933 & 0.886 & 0.943 & 1.830 \\
2170: LHS0318 & 0.934 & 0.747 & 0.617 & 0.797 & 1.413 \\
2171: LHS0326 & 0.948 & 0.782 & 0.689 & 0.841 & 1.530 \\
2172: LHS0327 & 0.995 & 1.004 & 1.025 & 1.005 & 2.030 \\
2173: LHS0334 & 0.475 & 0.492 & 0.277 & 0.470 & 0.747 \\
2174: LHS0335 & 0.960 & 0.721 & 0.640 & 0.791 & 1.431 \\
2175: LHS0360 & 0.986 & 0.840 & 0.778 & 0.886 & 1.664 \\
2176: LHS0367 & 1.011 & 0.880 & 0.854 & 0.929 & 1.783 \\
2177: LHS0375 & 0.879 & 0.604 & 0.414 & 0.583 & 0.997 \\
2178: LHS0381 & 0.883 & 0.728 & 0.555 & 0.730 & 1.285 \\
2179: LHS0385 & 0.982 & 0.787 & 0.712 & 0.841 & 1.553 \\
2180: LHS0398 & 0.910 & 0.767 & 0.644 & 0.808 & 1.452 \\
2181: LHS0401 & 0.997 & 0.959 & 0.911 & 0.958 & 1.869 \\
2182: LHS0406 & 0.686 & 0.801 & 0.576 & 0.789 & 1.365 \\
2183: LHS0418 & 0.913 & 0.924 & 0.858 & 0.924 & 1.782 \\
2184: LHS0424 & 0.951 & 0.917 & 0.852 & 0.927 & 1.778 \\
2185: LHS0440 & 0.763 & 0.789 & 0.608 & 0.796 & 1.404 \\
2186: LHS0507 & 0.991 & 0.958 & 0.969 & 0.980 & 1.949 \\
2187: LHS0515 & 0.813 & 0.524 & 0.355 & 0.508 & 0.863 \\
2188: LHS0518 & 0.966 & 0.933 & 0.903 & 0.951 & 1.854 \\
2189: LHS0521 & 1.006 & 0.937 & 0.950 & 0.969 & 1.919 \\
2190: LHS0541 & 0.795 & 0.678 & 0.538 & 0.701 & 1.239 \\
2191: LHS1490 & 0.301 & 0.778 & 0.296 & 0.569 & 0.865 \\
2192: LHS1970 & 0.878 & 0.433 & 0.476 & 0.686 & 1.162 \\
2193: LHS2067A & 0.228 & 0.670 & 0.199 & 0.452 & 0.651 \\
2194: LHS2467 & 1.017 & 1.031 & 1.019 & 1.014 & 2.033 \\
2195: LHS2734A & 1.018 & 0.985 & 0.984 & 0.987 & 1.972 \\
2196: LHS2734B & 1.086 & 0.720 & 0.669 & 0.828 & 1.497 \\
2197: LHS3620 & 0.836 & 0.684 & 0.544 & 0.727 & 1.271 \\
2198: SCR0242$-$5935 & 0.910 & 0.906 & 0.838 & 0.891 & 1.728 \\
2199: SCR0406$-$6735 & 1.015 & 0.989 & 0.829 & 0.937 & 1.766 \\
2200: SCR0433$-$7740 & 0.929 & 0.887 & 0.760 & 0.864 & 1.624 \\
2201: SCR0529$-$3950 & 0.817 & 0.886 & 0.637 & 0.815 & 1.452 \\
2202: SCR0629$-$6938 & 0.845 & 0.908 & 0.579 & 0.784 & 1.363 \\
2203: SCR0654$-$7358 & 0.938 & 0.835 & 0.653 & 0.825 & 1.478 \\
2204: SCR0701$-$0655 & 0.976 & 0.886 & 0.780 & 0.881 & 1.662 \\
2205: SCR0708$-$4709 & 0.984 & 0.992 & 0.979 & 0.976 & 1.955 \\
2206: SCR0709$-$4648 & 1.013 & 0.899 & 0.846 & 0.915 & 1.761 \\
2207: SCR1107$-$4135 & 0.994 & 0.839 & 0.783 & 0.892 & 1.675 \\
2208: SCR1433$-$3847 & 1.001 & 0.904 & 0.797 & 0.930 & 1.726 \\
2209: SCR1455$-$3914 & 0.915 & 0.788 & 0.707 & 0.849 & 1.555 \\
2210: SCR1457$-$3904 & 0.905 & 0.824 & 0.707 & 0.838 & 1.545 \\
2211: SCR1613$-$3040 & 0.920 & 0.898 & 0.749 & 0.885 & 1.634 \\
2212: SCR1739$-$8222 & 0.945 & 1.009 & 0.797 & 0.882 & 1.679 \\
2213: SCR1740$-$5646 & 0.915 & 0.696 & 0.496 & 0.637 & 1.133 \\
2214: SCR1756$-$5927 & 0.922 & 0.794 & 0.757 & 0.856 & 1.613 \\
2215: SCR1822$-$4542 & 0.817 & 0.818 & 0.624 & 0.809 & 1.433 \\
2216: SCR1843$-$7849 & 0.869 & 0.810 & 0.715 & 0.829 & 1.545 \\
2217: SCR1913$-$1001 & 0.962 & 0.923 & 0.835 & 0.925 & 1.760 \\
2218: SCR1916$-$3638 & 0.931 & 0.647 & 0.493 & 0.666 & 1.159 \\
2219: SCR1958$-$5609 & 0.915 & 0.865 & 0.805 & 0.926 & 1.731 \\
2220: SCR2018$-$6606 & 0.886 & 0.856 & 0.715 & 0.836 & 1.550 \\
2221: SCR2101$-$5437 & 0.929 & 0.969 & 0.726 & 0.904 & 1.630 \\
2222: SCR2104$-$5229 & 0.940 & 0.907 & 0.772 & 0.865 & 1.638 \\
2223: SCR2109$-$5226 & 0.967 & 0.774 & 0.646 & 0.799 & 1.445 \\
2224: SCR2204$-$3347 & 0.874 & 0.711 & 0.569 & 0.762 & 1.331 \\
2225: SIP1342$-$3534 & 0.684 & 0.664 & 0.462 & 0.676 & 1.138 \\
2226: WT0135 &0.622 & 0.754 & 0.512 & 0.731 & 1.243 \\
2227: WT0233 & 0.939 & 0.933 &0.739 & 0.857 & 1.596 \enddata
2228: \label{tbl.cahn.tio5.index}
2229: \end{deluxetable}
2230:
2231: %%%%%%% TABLE %%%%%
2232:
2233:
2234: \begin{deluxetable}{cclrccrclllccc}
2235: \rotate
2236: \tabletypesize{\scriptsize}
2237: \tablewidth{23cm}
2238: \tablecaption{Spectral Types}
2239: \tablehead{
2240: \colhead{RA} &
2241: \colhead{DEC} &
2242: \colhead{Object} &
2243: \colhead{$K_{s}$} &
2244: \colhead{$V-K_{s}$} &
2245: \colhead{$M_{K_{s}}$} &
2246: \colhead{$V_{tan}$} &
2247: \colhead{Old Type} &
2248: \colhead{New Type} &
2249: \colhead{metallicity} &
2250: \colhead{gravity} &
2251: \multicolumn{3}{c}{Ref} \\
2252: &
2253: &
2254: &
2255: &
2256: &
2257: &
2258: km/sec &
2259: &
2260: &
2261: &
2262: &
2263: $\pi$ &
2264: $V$ &
2265: spect }
2266:
2267: \startdata
2268: 12 06 00.9 & $+$14 38 56.8 & G012$-$016 & 7.931$\pm$0.018 & 2.16 & 4.12 & 98.4 & K2 & GVI &($-$0.52)\tablenotemark{a} & & H& 2& 2\\
2269: 01 04 26.4 & $-$02 21 59.8 & G070$-$035 & 7.116$\pm$0.020 & 2.04 & 4.15 & 55.8 & G5 & GVI &($-$0.67)\tablenotemark{a} & & H& 4& H\\
2270: 02 25 49.8 & $+$05 53 39.5 & G073$-$056 & 10.427$\pm$0.023 & 2.06 & 8.64 & 44.9 & & GVI &($-$1.18)\tablenotemark{b} & & H& 4& \\
2271: 07 54 34.1 & $-$01 24 44.3 & G112$-$054 & 5.425$\pm$0.023 & 2.00 & 4.01 & 23.7 & K1 & GVI &($-$0.94)\tablenotemark{c} & & H& 1& 2\\
2272: 16 13 48.6 & $-$57 34 13.8 & LHS0413 & 5.293$\pm$0.024 & 2.24 & 4.60 & 106.5 & G8/K0V(W) & GVI &($-$1.35)\tablenotemark{c} & & H& 1&10\\
2273: \tableline
2274: \tableline
2275: 15 45 52.4 & $+$05 02 26.6 & G016$-$009AB & 6.880$\pm$0.024 & 2.27 & 3.46 & 59.2 & K2V & K3.0[VI] &m$-$, ($-$0.77)\tablenotemark{d} & g & H& H& 3\\
2276: \tableline
2277: \tableline
2278: 00 50 17.0 & $-$39 30 08.3 & LHS0125 & 11.452$\pm$0.026 & 2.88 & 7.73 & 342.6 & & K4.0[VI] &m$-$ & g & R& R& ~\\
2279: 07 35 46.3 & $+$03 29 36.0 & LHS0232 & 10.841$\pm$0.024 & 2.84 & 6.57 & 346.0 & & K4.0[VI] &m$-$ & g & Y& R& ~\\
2280: 12 25 50.7 & $-$24 33 17.8 & LHS0327 & 10.144$\pm$0.021 & 2.57 & 5.44 & 418.3 & K0V & K4.0[VI] &m$-$ & g & R& R& 2\\
2281: 11 52 32.0 & $+$27 30 51.3 & LHS2467 & 9.806$\pm$0.017 & 2.43 & 4.94 & 437.0 & G7V & K4.0[VI] &m$-$ & g & H& R& 2\\
2282: \tableline
2283: \tableline
2284: 19 07 02.0 & $+$07 36 57.3 & G022$-$015 & 6.469$\pm$0.018 & 2.71 & 4.54 & 97.2 & K5V & K5.0[VI] &m$-$, ($-$0.61)\tablenotemark{e} & g & H& H& 2\\
2285: 21 32 11.9 & $+$00 13 18.0 & G026$-$009ACD & 7.082$\pm$0.029 & 2.64 & 3.62 & 96.9 & K2V & K5.0[VI] &m$-$, ($-$1.05)\tablenotemark{f} & g & H& H& 2\\
2286: 05 54 34.1 & $-$09 23 33.7 & GJ0223.1 & 7.760$\pm$0.020 & 2.96 & 4.84 & 82.1 & K4V & K5.0[VI] &m$-$, ($-$0.62)\tablenotemark{e} & g & H& 1& 9\\
2287: \tableline
2288: \tableline
2289: 04 32 36.6 & $-$39 02 03.4 & LHS0193A & 8.427$\pm$0.023 & 3.23 & 5.80 & 147.3 & & K6.0VI & m$-$ & g & R&16& ~\\
2290: 23 43 16.7 & $-$24 11 16.4 & LHS0073 & 9.393$\pm$0.021 & 3.42 & 7.27 & 322.3 & K5V & K6.0VI & m$-$ & g+ & Y&19& 2\\
2291: 07 13 40.6 & $-$13 27 57.1 & LHS0227 & 11.036$\pm$0.023 & 3.41 & 7.15 & 362.5 & & K6.0VI & m$-$ & g+ & Y& R& ~\\
2292: 02 52 45.7 & $+$01 55 49.4 & LHS0161 & 10.995$\pm$0.019 & 3.65 & 8.01 & 272.6 & esdM2 & K6.0VI & m$-$ & g++ & Y& R& 8\\
2293: \tableline
2294: \tableline
2295: 03 01 40.6 & $-$34 57 56.5 & LHS0164 & 10.641$\pm$0.024 & 2.92 & 8.20 & 192.6 & & K7.0VI & m$--$ & g & R& R& ~\\
2296: 05 15 45.1 & $-$72 11 22.2 & DEN0515$-$7211 & 13.211$\pm$0.040 & 3.29 & 9.72 & 4.7 & & K7.0VI: & m$--$ & g & 7& 7& ~\\
2297: 07 08 32.0 & $-$47 09 30.5 & SCR0708$-$4709 & 10.764$\pm$0.025 & 3.05 & \nodata &\nodata& & K7.0VI: & m$--$ & g & & R& \\
2298: \tableline
2299: 13 25 14.0 & $-$21 27 06.0 & LHS2734A & 12.896$\pm$0.037 & 3.23 & too~far &$>$280 & & K7.0VI & m$-$ & g & R& R& ~\\
2300: \tableline
2301: \tableline
2302: 03 06 28.7 & $-$07 40 41.5 & LHS0165 & 11.006$\pm$0.025 & 3.42 & 7.67 & 333.1 & & M0.0VI & m$---$ & g+: & Y& R& ~\\
2303: \tableline
2304: 11 11 13.7 & $-$41 05 32.7 & LHS0300A & 9.802$\pm$0.023 & 3.38 & 7.23 & 197.5 & & M0.0VI: & m$--$ & g &11&16& ~\\
2305: 16 25 14.0 & $+$15 40 54.2 & LHS0418 & 10.072$\pm$0.018 & 3.37 & 6.44 & 301.5 & K7V & M0.0VI & m$--$ & g & Y& R& 8\\
2306: 16 37 05.6 & $-$01 32 01.6 & LHS0424 & 10.803$\pm$0.022 & 3.37 & 7.39 & 279.6 & & M0.0VI & m$--$ & g & Y& R& ~\\
2307: \tableline
2308: 08 13 27.8 & $-$09 27 56.6 & LHS0244 & 10.729$\pm$0.023 & 3.64 & 6.88 & 412.7 & & M0.0VI & m$-$ & g++: & Y&19& ~\\
2309: \tableline
2310: \tableline
2311: 22 27 59.0 & $-$30 09 30.0 & LHS0521 & 11.463$\pm$0.019 & 3.22 & 8.14 & 222.1 & & M0.5VI: & m$---$ & g &11&11& ~\\
2312: 21 21 34.8 & $-$19 03 38.6 & LHS0507 & 12.053$\pm$0.026 & 3.15 & 7.81 & 355.2 & K/M sd & M0.5VI & m$---$ & g+ & Y& R& Y\\
2313: 15 39 39.0 & $-$55 09 10.0 & LHS0401 & 9.407$\pm$0.019 & 3.31 & 7.33 & 142.2 & & M0.5VI & m$---$ & g++ & Y& R& ~\\
2314: \tableline
2315: 02 42 26.3 & $-$59 35 01.6 & SCR0242$-$5935 & 12.783$\pm$0.031 & 3.40 & \nodata &\nodata& & M0.5VI & m$--$ & g & & R& \\
2316: 07 09 37.2 & $-$46 48 58.8 & SCR0709$-$4648 & 11.491$\pm$0.026 & 3.41 & \nodata &\nodata& & M0.5VI & m$--$ & g & ~& R& ~\\
2317: 11 07 55.8 & $-$41 35 52.7 & SCR1107$-$4135 & 11.474$\pm$0.019 & 3.49 & \nodata &\nodata& & M0.5VI & m$--$ & g & ~& R& ~\\
2318: 17 39 45.4 & $-$82 22 02.2 & SCR1739$-$8222 & 12.190$\pm$0.026 & 3.48 & \nodata &\nodata& & M0.5VI & m$--$ & g & & R& \\
2319: \tableline
2320: 02 02 52.2 & $+$05 42 21.0 & LHS0012 & 8.684$\pm$0.020 & 3.56 & 6.32 & 342.1 & sdM0 & M0.5VI & m$-$ & g & Y&13& 8\\
2321: 11 11 22.6 & $-$06 31 56.4 & LHS0299 & 11.143$\pm$0.024 & 3.64 & 6.54 & 437.3 & & M0.5VI: & m$-$ & g & Y& R& ~\\
2322: 14 18 20.4 & $-$52 24 12.6 & LHS0367 & 9.786$\pm$0.019 & 3.41 & 6.26 & 268.6 & & M0.5VI: & m$-$ & g & 7& 7& ~\\
2323: 04 06 06.7 & $-$67 35 28.8 & SCR0406$-$6735 & 12.804$\pm$0.031 & 3.54 & \nodata &\nodata& & M0.5VI & m$-$ & g & & R& \\
2324: 04 33 26.5 & $-$77 40 09.7 & SCR0433$-$7740 & 13.361$\pm$0.034 & 3.50 & \nodata &\nodata& & M0.5VI & m$-$ & g & & R& \\
2325: 14 33 03.3 & $-$38 46 59.6 & SCR1433$-$3847 & 13.592$\pm$0.046 & 3.62 & \nodata &\nodata& & M0.5VI & m$-$ & g & & R& \\
2326: 18 43 35.7 & $-$78 49 02.5 & SCR1843$-$7849 & 12.591$\pm$0.026 & 3.55 & \nodata &\nodata& & M0.5VI & m$-$ & g & ~& R& ~\\
2327: 19 58 31.2 & $-$56 09 10.6 & SCR1958$-$5609 & 12.525$\pm$0.033 & 3.51 & \nodata &\nodata& & M0.5VI & m$-$ & g & & R& \\
2328: 13 46 55.5 & $+$05 42 56.4 & LHS0360 & 11.662$\pm$0.023 & 3.50 & 6.73 & 524.0 & & M0.5VI: & m$-$ & g+ & 7& 7& ~\\
2329: \tableline
2330: \tableline
2331: 22 20 27.0 & $-$24 21 49.3 & LHS0518 & 10.393$\pm$0.021 & 3.24 & 7.07 & 230.7 & & M1.0VI & m$------$ & g &11&11& ~\\
2332: 07 01 17.7 & $-$06 55 49.3 & SCR0701$-$0655 & 12.996$\pm$0.030 &\nodata & \nodata &\nodata& & M1.0VI & m$------$ & g+ & & & \\
2333: \tableline
2334: 01 53 09.0 & $-$33 25 02.1 & LHS0148 & 12.832$\pm$0.032 & 3.59 & 8.59 & 374.8 & & M1.0VI & m$-----$ & g+: & 6& 6& ~\\
2335: \tableline
2336: 01 53 09.0 & $-$33 25 02.1 & SCR1913$-$1001 & 11.929$\pm$0.028 & 3.69 & \nodata &\nodata& & M1.0VI & m$----$ & g & & R& \\
2337: 00 17 40.0 & $-$10 46 16.9 & LHS0109 & 10.366$\pm$0.021 & 3.51 & 7.63 & 176.1 & K5 & M1.0VI & m$----$ & g+ & Y& R& 2\\
2338: 21 04 00.5 & $-$52 29 43.5 & SCR2104$-$5229 & 12.763$\pm$0.024 & 3.53 & \nodata &\nodata& & M1.0VI & m$----$ & g+ & & R& ~\\
2339: 14 55 35.8 & $-$15 33 44.0 & LHS0385 & 11.062$\pm$0.023 & 3.55 & 7.61 & 403.4 & M0 & M1.0VI & m$----$ & g++ & Y& R& Y\\
2340: 14 55 51.5 & $-$39 14 33.1 & SCR1455$-$3914 & 11.788$\pm$0.024 & 3.65 & \nodata &\nodata& & M1.0VI & m$----$ & g++ & & R& \\
2341: 12 34 53.1 & $+$05 03 54.1 & LHS0335 & 13.000$\pm$0.030 & 3.60 & 9.74 & 247.2 & & M1.0VI & m$----$ & g+++ & Y& R& ~\\
2342: 21 09 02.5 & $-$52 26 17.8 & SCR2109$-$5226 & 13.049$\pm$0.036 & 3.66 & \nodata &\nodata& & M1.0VI & m$----$ & g+++ & ~& R& ~\\
2343: 01 31 04.1 & $-$50 24 54.3 & LEHPM1628 & 13.459$\pm$0.038 & 3.70 & \nodata &\nodata& & M1.0VI & m$----$ & g++++ & ~& R& ~\\
2344: \tableline
2345: 06 54 06.3 & $-$73 58 03.6 & SCR0654$-$7358 & 13.285$\pm$0.048 & \nodata & \nodata &\nodata& & M1.0VI & m$---$ & g & & & \\
2346: 14 57 49.0 & $-$39 04 51.4 & SCR1457$-$3904 & 12.984$\pm$0.030 & 3.69 & \nodata &\nodata& & M1.0VI & m$---$ & g & & R& \\
2347: 16 13 53.5 & $-$30 40 58.4 & SCR1613$-$3040 & 12.383$\pm$0.031 & 3.66 & \nodata &\nodata& & M1.0VI & m$---$ & g & & R& \\
2348: 17 56 27.9 & $-$59 27 18.2 & SCR1756$-$5927 & 12.686$\pm$0.030 & 3.61 & \nodata &\nodata& & M1.0VI & m$---$ & g & ~& R& ~\\
2349: 20 18 28.7 & $-$66 06 44.5 & SCR2018$-$6606 & 12.990$\pm$0.030 & \nodata & \nodata &\nodata& & M1.0VI & m$---$ & g & & & \\
2350: \tableline
2351: 05 29 40.9 & $-$39 50 25.6 & SCR0529$-$3950 & 11.651$\pm$0.021 & \nodata & \nodata &\nodata& & M1.0VI & m$--$ & g & & & \\
2352: \tableline
2353: 21 01 45.6 & $-$54 37 31.9 & SCR2101$-$5437 & 12.078$\pm$0.026 & 3.70 & \nodata &\nodata& & M1.0VI & m$-$ & g & ~& R& ~\\
2354: 08 01 29.0 & $+$10 43 04.2 & LHS1970 & 13.875$\pm$0.032 & 3.84 & 9.43 & 374.8 & esdM2.5 & M1.0VI: & m$-$ & g+: & Y& R& 8\\
2355: 13 25 14.0 & $-$21 27 06.0 & LHS2734B & 14.932$\pm$0.136 & 3.90 & too~far &$>$280 & & M1.0VI: & m$-$ & g+: & R& R& ~\\
2356: \tableline
2357: 02 42 02.9 & $-$44 30 58.7 & LHS0158 & 9.726$\pm$0.021 & 3.91 & 6.89 & 191.4 & & M1.0[VI]& m & g &11&11& ~\\
2358: 17 18 35.0 & $-$43 26 24.0 & LHS0440 & 8.948$\pm$0.023 & 4.03 & 6.78 & 135.6 & M1V & M1.0VI: & m & g+: &11&11& 2\\
2359: 18 22 58.7 & $-$45 42 45.3 & SCR1822$-$4542 & 12.879$\pm$0.027 & 3.92 & \nodata &\nodata& & M1.0VI: & m & g+: & & R& \\
2360: \tableline
2361: \tableline
2362: 07 56 13.4 & $-$67 05 20.6 & WT0233 & 12.627$\pm$0.024 & 3.60 & 7.84 & 326.8 & M0.0VI & M2.0VI & m$---$ & g & 6& 6& 6\\
2363: 11 56 54.8 & $+$26 39 56.3 & LHS0318 & 11.797$\pm$0.018 & 3.65 & 8.77 & 263.8 & & M2.0VI: & m$---$ & g &18& R& ~\\
2364: \tableline
2365: 05 11 40.6 & $-$45 01 06.0 & GJ0191 & 5.049$\pm$0.021 & 3.80 & 7.08 & 161.0 & sdM1.0 & M2.0VI & m$--$, ($-$0.99)\tablenotemark{e} & g & Y&13& 8\\
2366: 02 56 13.2 & $-$35 08 26.9 & LHS0162 & 11.536$\pm$0.021 & 3.82 & 7.13 & 364.1 & M1.0VI & M2.0VI & m$--$ & g+ & 6& 6& 6\\
2367: 21 04 25.5 & $-$27 52 48.5 & LHS3620 & 12.696$\pm$0.027 & 3.93 & 7.94 & 414.6 & & M2.0VI & m$--$ & g+ & R& R& ~\\
2368: \tableline
2369: 04 03 38.4 & $-$05 08 05.4 & LHS0186 & 10.854$\pm$0.024 & 4.02 & 7.21 & 295.6 & & M2.0VI & m$-$ & g & Y& R& ~\\
2370: 15 43 18.3 & $-$20 15 33.0 & LHS0406 & 9.018$\pm$0.021 & 4.04 & 7.40 & 117.4 & M1.0V & M2.0VI & m$-$ & g &11&11& 2\\
2371: 00 55 43.8 & $-$21 13 05.5 & LHS0127 & 11.733$\pm$0.023 & 4.06 & \nodata &\nodata& & M2.0VI & m$-$ & g+ & ~& R& ~\\
2372: 06 29 56.4 & $-$69 38 13.4 & SCR0629$-$6938 & 12.901$\pm$0.034 & \nodata & \nodata &\nodata& & M2.0VI & m$-$ & g+ & & & \\
2373: \tableline
2374: \tableline
2375: 12 24 26.8 & $-$04 43 36.7 & LHS0326 & 11.234$\pm$0.023 & 3.67 & 7.69 & 304.9 & & M3.0VI & m$--$ & g & R& R& ~\\
2376: 15 34 27.7 & $+$02 16 47.5 & LHS0398 & 11.502$\pm$0.025 & 3.74 & 7.52 & 354.3 & & M3.0VI & m$--$ & g & Y&19& ~\\
2377: 23 17 05.0 & $-$13 51 04.1 & LHS0541 & 12.414$\pm$0.026 & 4.05 & 8.22 & 420.0 & sdM3.0 & M3.0VI & m$--$ & g+ & Y& 1& 8\\
2378: 22 04 02.2 & $-$33 47 38.9 & SCR2204$-$3347 & 11.601$\pm$0.027 & 3.84 & \nodata &\nodata& & M3.0VI & m$--$ & g+ & ~& R& ~\\
2379: 17 40 46.9 & $-$56 46 58.0 & SCR1740$-$5646 & 13.195$\pm$0.040 & \nodata & \nodata &\nodata& & M3.0VI & m$--$ & g++ & & & \\
2380: 19 16 46.5 & $-$36 38 05.8 & SCR1916$-$3638 & 12.947$\pm$0.034 & 3.88 & \nodata &\nodata& & M3.0VI & m$--$ & g++ & ~& R& ~\\
2381: \tableline
2382: 04 25 38.4 & $-$06 52 37.0 & LHS0189AB & 10.311$\pm$0.037 & 3.94 & 8.59 & 128.3 & & M3.0VI: & m$-$ & g & 7& 7& ~\\
2383: 07 16 27.7 & $+$23 42 10.4 & LHS0228 & 11.298$\pm$0.018 & 4.20 & 7.54 & 300.6 & sdM2.0 & M3.0VI & m$-$ & g & Y& R&14\\
2384: 13 42 21.2 & $-$35 34 50.7 & SIP1342$-$3534 & 12.935$\pm$0.030 & 4.37 & \nodata &\nodata& & M3.0VI & m$-$: & g & & R& \\
2385: 04 11 27.1 & $-$44 18 09.0 & WT0135 & 9.834$\pm$0.020 & 4.33 & \nodata &\nodata& & M3.0VI & m$-$: & g & ~& 5& ~\\
2386: 09 43 46.2 & $-$17 47 06.2 & LHS0272 & 8.874$\pm$0.021 & 4.29 & 8.34 & 86.3 & sdM3.0 & M3.0VI & m$-$: & g+ & Y& R& 8\\
2387: \tableline
2388: \tableline
2389: 14 50 28.8 & $-$08 38 36.8 & LHS0381 & 11.237$\pm$0.021 & 3.90 & 8.43 & 274.9 & K7.0V & M3.5VI & m$--$ & g & Y&19& Y\\
2390: 14 31 38.4 & $-$25 25 33.9 & LHS0375 & 11.507$\pm$0.022 & 4.12 & 9.61 & 157.5 & esdM4 & M3.5VI & m$--$ & g+ & Y& R& 8\\
2391: \tableline
2392: 01 38 49.0 & $+$11 21 36.7 & LHS0144 & 12.080$\pm$0.026 & 4.24 & 8.57 & 386.6 & & M3.5VI & m$-$ & g & Y& R& ~\\
2393: 21 55 48.0 & $-$11 21 42.1 & LHS0515 & 12.912$\pm$0.032 & 4.50 & 9.36 & 264.5 & esdM5 & M3.5VI & m$-$ & g+ & Y& R&17\\
2394: \tableline
2395: \tableline
2396: 05 00 15.3 & $-$54 06 09.0 & LEHPM3861 & 13.967$\pm$0.059 & 4.47 & \nodata &\nodata& esdM6 & M4.0VI & m$-$ & g & ~& R&15\\
2397: \tableline
2398: \tableline
2399: 03 02 06.3 & $-$39 50 51.8 & LHS1490 & 9.885$\pm$0.023 & 4.37 & \nodata &\nodata& & M5.0VI & m$-$ & g & & R& \\
2400: \tableline
2401: \tableline
2402: 12 34 15.7 & $+$20 37 05.7 & LHS0334 & 13.044$\pm$0.029 & 4.98 & \nodata &\nodata& sdM4.5 & M6.0VI & m$--$ & g & & R&17\\
2403: \tableline
2404: 08 53 57.0 & $-$24 46 54.0 & LHS2067A & 11.571$\pm$0.023 & 6.38 & 9.54 & 76.2 & subdwarf & M6.0VI & m$-$ & g & R& R&12\\
2405:
2406: \enddata
2407:
2408: \tablecomments{The last three columns provide references for
2409: parallaxes, $V$ band photometry, and previous spectroscopy
2410: results. ``:'' indicates a questionable sub-type, metallicity, or
2411: gravity. All $K_{s}$ magnitudes are from the 2MASS all-sky database.
2412: LHS2734A and B have zero parallax so we list them as ``too far'' in
2413: the table.}
2414:
2415: \tablenotetext{a}{Metallicity data are from \cite{Carney1994}.}
2416:
2417: \tablenotetext{b}{Metallicity data are from \cite{Nordstrom2004}.}
2418:
2419: \tablenotetext{c}{Metallicity data are from \cite{Cayrel2001}. There
2420: are multiple metallicities reported for G112-054, so a mean value is
2421: presented.}
2422:
2423: \tablenotetext{d}{\cite{Cayrel2001}, \cite{Goldberg2002} and
2424: \cite{Laird1988} report different metallicities, so a mean value is
2425: given.}
2426:
2427: \tablenotetext{e}{Metallicity data are from \cite{Woolf2005}}
2428:
2429: \tablenotetext{f}{\cite{Morrison2003} and \cite{Allen2000} report
2430: different metallicities, so a mean value is given.}
2431:
2432: \tablerefs{
2433: (1) \citealt{Bessel1990}
2434: ;(2) \citealt{Bidelman1985}
2435: ;(3) \citealt{Cayrel1997}
2436: ;(4) \citealt{Carney1994}
2437: ;(5) \citealt{Costa2003}
2438: ;(6) \citealt{Costa2005}
2439: ;(7) \citealt{Costa2006}
2440: ;(8) \citealt{Gizis1997}
2441: ;(9) \citealt{Hawley1996}
2442: ;(10) \citealt{Houk1975}
2443: ;(11) \citealt{Jao2005}
2444: ;(12) \citealt{Kirkpatrick1995}
2445: ;(13) \citealt{Leggett1992}
2446: ;(14)\citealt{Lepine2007}
2447: ;(15) \citealt{Lodieu2005}
2448: ;(16) \citealt{Monteiro2006}
2449: ;(17) \citealt{Reid2005}
2450: ;(18) \citealt{Smart2007}
2451: ;(19) \citealt{Weis1996}
2452: ;(H) \citealt{Hipparcos}
2453: ;(R) \citealt{RECONS}
2454: ;(Y) \citealt{YPC}}
2455: %;(15) \citealt{Nordstrom2004}%
2456: %;(17) \citealt{Ryan1991}%
2457:
2458: \label{tbl.spectral.type}
2459: \end{deluxetable}
2460:
2461: %%%%%%%%% table %%%%%%%%%%
2462:
2463: \begin{deluxetable}{lcccclll}
2464: \tablewidth{18.5cm}
2465: \tablecaption{SDSS Subdwarfs}
2466: \tablehead{
2467: \colhead{Object} &
2468: \colhead{$r$} &
2469: \colhead{$g-r$} &
2470: \colhead{$g-z$} &
2471: \colhead{Old Type} &
2472: \colhead{New Type} &
2473: \colhead{metallicity} &
2474: \colhead{gravity} }
2475: \startdata
2476: SDSS~J085843.89$+$511210.1 & 20.00 & 1.53 & 3.10 & M2.0VI & M2.0V & m & \nodata \\
2477: SDSS~J093141.85$+$453914.5 & 19.45 & 1.58 & 3.08 & M2.0VI & M2.0V & m & \nodata \\
2478: SDSS~J145447.32$+$011006.8 & 20.35 & 1.44 & 3.45 & M3.0VI & M3.0V & m & \nodata \\
2479: \tableline
2480: \tableline
2481: SDSS~J003755.20$-$002134.2 & 18.70 & & & M1.0VI & M2.0VI & m$-$ & g \\
2482: SDSS~J083217.77$+$522408.2 & 19.61 & 1.49 & 3.14 & M2.0VI & M2.0VI: & m & g$+$: \\
2483: \tableline
2484: \tableline
2485: SDSS~J033408.64$-$072349.2 & 20.21 & 1.89 & 3.35 & M2.0VI & M3.0VI & m$----$ & g$+$: \\
2486: SDSS~J090434.02$+$513153.9 & 19.40 & 1.78 & 3.29 & M1.0VI & M3.0VI & m$----$ & g$+$: \\
2487: SDSS~J161348.84$+$482016.0 & 18.30 & 1.72 & 3.19 & M1.0VI & M3.0VI & m$----$ & g$+$: \\
2488: \tableline
2489: SDSS~J112751.35$-$001246.8 & 20.03 & & & M1.0VI & M3.0VI & m$---$ & g \\
2490: SDSS~J092708.10$+$561648.1 & 19.37 & 1.62 & 3.02 & M1.0VI & M3.0VI & m$---$ & g$+$ \\
2491: SDSS~J100109.54$+$015450.2 & 19.12 & 1.71 & 3.25 & M2.0VI & M3.0VI & m$---$ & g$++$ \\
2492: SDSS~J101031.13$+$651327.6 & 19.39 & 1.68 & 3.30 & M1.0VI & M3.0VI & m$---$ & g$++$ \\
2493: \tableline
2494: SDSS~J002228.00$-$091444.8 & 18.92 &-0.89 & 3.03 & M1.0VI & M3.0VI & m$--$ & g \\
2495: SDSS~J024501.77$+$003315.8 & 19.38 & 1.54 & 3.15 & M1.0VI & M3.0VI & m$--$ & g \\
2496: SDSS~J081329.95$+$443945.6 & 19.39 & 1.68 & 3.35 & M2.0VI & M3.0VI & m$--$ & g \\
2497: SDSS~J090238.75$+$471813.6 & 19.97 & 1.73 & 3.21 & M2.0VI & M3.0VI & m$--$ & g \\
2498: SDSS~J092534.16$+$524442.4 & 19.79 & 1.71 & 3.31 & M3.0VI & M3.0VI & m$--$ & g \\
2499: SDSS~J092745.78$+$582122.7 & 20.47 & 1.78 & 3.56 & M3.0VI & M3.0VI & m$--$ & g \\
2500: SDSS~J095147.77$+$003612.0 & 18.27 & 1.58 & 3.15 & M2.0VI & M3.0VI & m$--$ & g \\
2501: SDSS~J115900.70$+$665214.3 & 19.37 & 1.61 & 3.35 & M3.0VI & M3.0VI & m$--$ & g \\
2502: SDSS~J125919.29$-$025402.3 & & & & M2.0VI & M3.0VI & m$--$ & g \\
2503: SDSS~J173452.52$+$603603.1 & 18.78 & 1.71 & 3.36 & M2.0VI & M3.0VI & m$--$ & g \\
2504: SDSS~J215937.69$+$005536.2 & 18.96 & 1.61 & 3.22 & M2.0VI & M3.0VI & m$--$ & g \\
2505: SDSS~J221500.88$+$005217.2 & 19.08 & 1.60 & 3.33 & M2.0VI & M3.0VI & m$--$ & g \\
2506: SDSS~J145547.00$+$602837.3 & 19.18 & 1.70 & 3.39 & M2.0VI & M3.0VI & m$--$ & g$+$ \\
2507: SDSS~J104320.47$+$010439.4 & 19.16 & 1.66 & 3.37 & M2.0VI & M3.0VI & m$--$ & g$++$ \\
2508: \tableline
2509: SDSS~J084105.39$+$032109.6 & 20.05 & 1.45 & 3.18 & M3.0VI & M3.0VI & m$-$ & g \\
2510: SDSS~J091451.98$+$453152.8 & 19.02 & 1.64 & 3.49 & M3.0VI & M3.0VI & m$-$ & g \\
2511: SDSS~J093024.66$+$554447.7 & 19.14 & 1.56 & 2.96 & M3.0VI & M3.0VI & m$-$ & g \\
2512: SDSS~J094306.37$+$465701.4 & 19.74 & 1.59 & 3.29 & M3.0VI & M3.0VI & m$-$ & g \\
2513: SDSS~J224854.83$-$091723.2 & 19.85 & 1.57 & 3.39 & M3.0VI & M3.0VI & m$-$ & g \\
2514: SDSS~J235830.60$-$011413.2 & 19.99 & 1.54 & 3.33 & M2.0VI & M3.0VI & m$-$ & g \\
2515: SDSS~J113501.76$+$033720.3 & & & & M3.0VI & M3.0VI: & m & g$+$: \\
2516: \tableline
2517: \tableline
2518: SDSS~J105122.43$+$603844.8 & 17.15 & 1.65 & 3.21 & M2.0VI & M4.0VI & m$---$ & g \\
2519: SDSS~J092429.76$+$523410.7 & 18.70 & 1.67 & 3.21 & M2.0VI & M4.0VI & m$---$ & g \\
2520: \tableline
2521: SDSS~J031314.28$-$000619.8 & 20.33 & 1.48 & 3.39 & M2.0VI & M4.0VI & m$--$ & g \\
2522: SDSS~J082230.00$+$471645.8 & 19.39 & 1.67 & 3.38 & M2.0VI & M4.0VI & m$--$ & g \\
2523: SDSS~J143930.77$+$033317.3 & 19.38 & 1.80 & 3.58 & M3.0VI & M4.0VI & m$--$ & g$+$ \\
2524: \tableline
2525: SDSS~J003541.84$+$003210.1 & 20.15 & 1.64 & 3.37 & M3.0VI & M4.0VI & m$-$ & g \\
2526: SDSS~J003701.37$-$003248.3 & 20.15 & 1.60 & 3.50 & M3.0VI & M4.0VI & m$-$ & g \\
2527: SDSS~J010811.89$+$003042.4 & 17.34 & 1.61 & 3.43 & M3.0VI & M4.0VI & m$-$ & g \\
2528: SDSS~J083002.73$+$483251.6 & 19.92 & 1.90 & 3.69 & M3.0VI & M4.0VI & m$-$ & g \\
2529: SDSS~J171745.22$+$625337.0 & 18.66 & 1.65 & 3.57 & M3.0VI & M4.0VI & m$-$ & g \\
2530: SDSS~J221625.03$-$003122.5 & 19.28 & 1.58 & 3.37 & M3.0VI & M4.0VI & m$-$ & g \\
2531: SDSS~J223802.82$-$082532.4 & 19.71 & 1.78 & 3.59 & M3.0VI & M4.0VI & m$-$ & g \\
2532: SDSS~J224605.41$+$141640.6 & 17.00 & 1.57 & 3.48 & M3.0VI & M4.0VI & m$-$ & g \\
2533: SDSS~J230303.49$-$010656.7 & 18.95 & 1.62 & 3.42 & M3.0VI & M4.0VI & m$-$ & g \\
2534: SDSS~J235116.25$-$003104.8 & 19.48 & 1.49 & 3.31 & M3.0VI & M4.0VI & m$-$ & g \\
2535: \tableline
2536: \tableline
2537: SDSS~J150511.33$+$620926.3 & 18.61 & 1.81 & 3.45 & M3.0VI & M5.0VI & m$--$ & g \\
2538: \tableline
2539: SDSS~J230805.24$+$001812.7 & 19.36 & 1.70 & 3.72 & M3.0VI & M5.0VI & m$-$ & g \\
2540: \enddata
2541:
2542: \tablecomments{Column definitions are the same as in
2543: Table~\ref{tbl.spectral.type}. Double lines separate each type.}
2544:
2545: \label{tbl.sdss.subdwarf}
2546: \end{deluxetable}
2547:
2548: %%%%%%%%% table 3 %%%%%%%%%%
2549:
2550: %%%%%%%%%%%%%%%%%%%%%%%%%%%%% End of Tables %%%%%%%%%%%%%%%%%%%%%%%%%%
2551:
2552: %%%%%%%%%%%%%%%%%%%%%%%% Beginning of Figures %%%%%%%%%%%%%%%%%%%%%%%
2553:
2554: \clearpage
2555:
2556: %%%%%%%%% Figure 1 %%%%%%%%%%
2557:
2558: \begin{figure}[h]
2559: \centering
2560: \includegraphics[angle=90, scale=0.7]{f1.ps}
2561:
2562: \caption{Synthetic spectra for stars having $T_{eff}$ $=$ 3500K and
2563: {\it log g} $=$ 5.0. Red, green, and blue lines represent different
2564: metallicities, 0.0, $-$2.0 and $-$4.0. Note that the relative amounts
2565: of blue and red fluxes trends toward bluer objects at lower
2566: metallicities.}
2567:
2568: \label{fig.continuum.model}
2569: \end{figure}
2570: \clearpage
2571:
2572: %%%%%%%%% Figure 2 %%%%%%%%%%
2573:
2574: \begin{figure}
2575: \centering
2576: \includegraphics[angle=90, scale=0.7]{f2.ps}
2577:
2578: \caption{The mass-gravity relation is shown using data from Table 1 of
2579: \cite{Lopez2007}. A solid line represents a polynomial fit to all of
2580: the points, simply to be used as a guide. Different shades for the
2581: points represent different metallicity measurements. Stars with
2582: unknown metallicities are plotted as open circles.}
2583: \label{fig.Lopez.Morales}
2584: \end{figure}
2585:
2586: %%%%%%%%% Figure 2 %%%%%%%%%%
2587:
2588: \begin{figure}
2589: \centering
2590: \includegraphics[angle=90, scale=0.5]{f3a.ps}
2591: \includegraphics[angle=90, scale=0.5]{f3b.ps}
2592:
2593: \caption{The CaH1 and CaH2+CaH3 versus TiO5 indices are plotted for
2594: our identified subdwarfs (solid circles). For comparison, known cool
2595: dwarfs (dots), subdwarfs (open triangles), and ``extreme'' subdwarfs
2596: (open boxes) from \cite{Hawley1996}, \cite{Gizis1997}, and
2597: \cite{Reid2005} are also shown. The dashed box indicates subdwarfs
2598: that do not exhibit strong spectroscopic indices, but which are either
2599: below the main-sequence on the HR diagram or have published
2600: metallicities [m/H] less than or equal to $-$0.5. A solid line
2601: indicates the separation between regular subdwarfs and extreme
2602: subdwarfs adopted by \cite{Gizis1997}.}
2603:
2604: \label{fig.cahn.plot}
2605: \end{figure}
2606: \clearpage
2607:
2608: %%%%%%%%% Figure 3 %%%%%%%%%%
2609:
2610: \begin{figure}
2611: \centering
2612: \includegraphics[scale=0.6, angle=90]{f4.ps}
2613:
2614: \caption{HR diagram for subdwarfs listed in
2615: Table~\ref{tbl.cahn.tio5.index} and shown in
2616: Figure~\ref{fig.cahn.plot} that have trigonometric parallaxes. Open
2617: circles indicate subdwarfs and filled boxes indicate ``extreme''
2618: subdwarfs based on spectroscopic indices. Concentric circles indicate
2619: stars with CaH1 index greater than 0.9 that are difficult to
2620: distinguish from main sequence stars at our spectral resolution. Open
2621: boxes indicate confirmed spectroscopic binaries. A solid line
2622: indicates a fit to main-sequence dwarfs, primarily from
2623: \cite{Henry2004} with extra dwarf standard stars from \cite{Gray2003}.
2624: The dashed line is one magnitude fainter than this solid line. Note
2625: that the K-type subdwarf sequence merges with the K dwarf sequence at
2626: the blue end of this ($M_{K_{s}}$ vs $V-K_{s}$) plot. The single
2627: point at $V-K_{s}$ = 6.4 is LHS 2067A. The spectral types for dwarfs
2628: are given at the bottom of the figure as references.}
2629:
2630: \label{fig.hr.plot}
2631: \end{figure}
2632: \clearpage
2633:
2634: %%%%%%%%% Figure 4 %%%%%%%%%%
2635:
2636: \begin{figure}
2637: \hspace{-2cm}
2638: \includegraphics[scale=0.8, angle=90]{f5.ps}
2639:
2640: \caption{Spectra for G and K-type stars from \cite{Jacoby1984} (left
2641: panel, 4\AA~resolution) and \cite{Silva1992} (right panel,
2642: 11\AA~resolution) are shown. Because the red cutoff is 7400\AA~in
2643: \cite{Jacoby1984}, the results from \cite{Silva1992} are also plotted
2644: to 7400\AA, and both sets of spectra are normalized at 7400\AA. The
2645: two tick marks indicate the Ba I (left) and H$\alpha$ (right)
2646: absorption features. \cite{Silva1992} have removed telluric
2647: absorption features.}
2648:
2649: \label{fig.jacoby.silva}
2650: \end{figure}
2651: \clearpage
2652:
2653: %%%%%%%%% Figure 5 %%%%%%%%%%
2654:
2655: \begin{figure}
2656: \centering
2657: \includegraphics[scale=0.7, angle=0]{f6.ps}
2658:
2659: \caption{Top: An mid K subdwarf spectrum (black) is virtually
2660: identical to an mid K dwarf (red). Middle: The spectra of two early
2661: M subdwarfs differ only at CaH. Bottom: One subdwarf's spectrum
2662: (black) matches an M0.5V spectral standard (red) at the blue end, but
2663: matches an M2.0V spectral standard (red) at the red end. The deep
2664: telluric band, O$_{2}$A (7570\AA--7700\AA) has been removed.}
2665: \label{fig.problem.spect}
2666:
2667: \end{figure}
2668: \clearpage
2669:
2670: %%%%%%%%% Figure 6 %%%%%%%%%%
2671:
2672: \begin{figure}
2673: \centering
2674: \includegraphics[scale=0.5, angle=90]{f7.ps}
2675:
2676: \caption{Two synthetic spectra from old (red) and new (black) model
2677: grids are shown. The physical parameters for both models are $T_{eff}
2678: =$ 3600K, {\it log g} = 5.0 and [m/H]$ = -$2.0. Major differences are
2679: seen between 6500\AA~and 7000\AA, and in the absorption lines of Li I
2680: (6103\AA) and Ca I (6122\AA~and 6162\AA).}
2681:
2682: \label{fig.new.old}
2683: \end{figure}
2684: \clearpage
2685:
2686: %%%%%%%%% Figure 7 %%%%%%%%%%
2687:
2688: \begin{figure}
2689: \centering
2690: \includegraphics[scale=0.5, angle=90]{f8a.ps}
2691: \includegraphics[scale=0.5, angle=90]{f8b.ps}
2692:
2693: \caption{CaH1 (top) and CaH2$+$CaH3 (bottom) indices from GAIA model
2694: grids are plotted against $T_{eff}$. The different line styles
2695: represent different [m/H]. The spectral types for dwarfs are given at
2696: the top of the figure as references.}
2697:
2698: \label{fig.model.index.plot}
2699: \end{figure}
2700: \clearpage
2701:
2702: %%%%%%%%% Figure 8 %%%%%%%%%%
2703:
2704: \begin{figure}
2705: \centering
2706: \includegraphics[scale=0.7]{f9.ps}
2707:
2708: \caption{GAIA synthetic spectra from 4800K to 2800K are shown. All
2709: spectra are noiseless and have {\it log g} $=$ 5.0. Red, green and
2710: blue lines represent [m/H] $=$ 0.0, $-$1.0 and $-$2.0, respectively.
2711: Effective temperatures for each set of spectra are given above each
2712: group of lines. The feature marked \#1 is not seen in any of our
2713: spectra. The Ca I (6162\AA) feature marked \#2 is seen in our
2714: spectra.}
2715:
2716: \label{fig.4800-2800.plot}
2717: \end{figure}
2718: \clearpage
2719:
2720: %%%%%%%%% Figure 9 %%%%%%%%%%
2721:
2722: \begin{figure}
2723: \centering
2724: \includegraphics[scale=0.5, angle=90]{f10a.ps}
2725: \includegraphics[scale=0.5, angle=90]{f10b.ps}
2726:
2727: \caption{The top plot shows GAIA model grids at fixed metallicity
2728: ([m/H]$=-$1.0) and effective temperature (3500K). Black, red, green
2729: and yellow lines represent various {\it log g} = 4.0, 4.5, 5.0 and
2730: 5.5, respectively. It is clear that CaH bands will be affected by
2731: changing gravity but TiO5 is not. The bottom plot shows model grids
2732: at fixed {\it log g} $=$ 5.0 and $T =$ 3500K. Black, red and green
2733: lines represent various [m/H] $=$ 0.0, $-$1.0 and $-$2.0,
2734: respectively. Note that model grids do not have telluric lines. }
2735:
2736: \label{fig.3500.plot}
2737: \end{figure}
2738: \clearpage
2739:
2740: \begin{figure}
2741: \centering
2742: \includegraphics[scale=0.65]{f11.ps}
2743: \caption{Using our spectra and metallicities measured independently by
2744: others, three different types of M dwarfs show the metallicity trend
2745: predicted by GAIA models. The gray lines represent lower metallicity
2746: stars in each pair. Metallicities from \cite{Bonfils2005} are given
2747: in each panel for the stars.}
2748:
2749: \label{fig:bonfils.recons}
2750: \end{figure}
2751:
2752: %%%%%%%%% Figure 10 %%%%%%%%%%
2753:
2754: \begin{figure}
2755: \centering
2756: \includegraphics[scale=0.6]{f12.ps}
2757:
2758: \caption{K3.0[VI] to K5.0[VI] spectra are shown compared to K dwarf
2759: standard spectra (red). The two telluric bands (O$_{2}$ A and B) have
2760: been removed for clarity. The three thick tick-marks represent the
2761: locations of Ca I (6162\AA) and Ca II (8542\AA~and 8662\AA) absorption
2762: features. The inset plot shows the locations of the stars on the HR
2763: diagram, where filled circles represent subdwarfs and triangles
2764: represent main sequence standards of types K3.0V, K4.0V, and K5.0V
2765: (left to right). For clarity, the axis labels for the inset plot are
2766: not shown but are always $V-K_{s}$ vs. $M_{Ks}$. The errors in
2767: absolute magnitudes are shown. However, because the errors for
2768: $V-K_{s}$ are equal to or smaller than the filled circles, they are
2769: not shown. The solid line represents a fitted main sequence line and
2770: the dashed line is one magnitude fainter than the solid line.}
2771:
2772: \label{fig.K3.K5}
2773: \end{figure}
2774: \clearpage
2775:
2776: %%%%%%%%% Figure 11 %%%%%%%%%%
2777:
2778: \begin{figure}
2779: \centering
2780: \includegraphics[scale=0.7]{f13.ps}
2781:
2782: \caption{Spectra of LHS 125, LHS 232 and LHS 327 (all K4.0[VI]), are
2783: shown. Note that the blue end of these spectra are at 6500\AA~because
2784: of problems with these particular spectra between 6000\AA~and 6500\AA.
2785: The two telluric O$_{2}$ A and B bands have been removed. The
2786: triangle in the inset plot represents a K4.0V standard star. Symbols
2787: and lines have the same meanings as in Figure~\ref{fig.K3.K5}.}
2788:
2789: \label{fig.K4}
2790: \end{figure}
2791: \clearpage
2792:
2793: %%%%%%%%% Figure 12 %%%%%%%%%%
2794:
2795: \begin{figure}
2796: \centering
2797: \includegraphics[scale=0.7]{f14.ps}
2798:
2799:
2800: \caption{Spectra of LHS 193A, LHS 227 and LHS 161 (all K6.0VI), are
2801: shown with our K5.0V and K7.0V standard spectra. The telluric
2802: O$_{2}$ A band has been removed. The thick arrows in the HR
2803: diagram point toward sequentially higher gravity subdwarfs.
2804: Symbols and lines have the same meanings as in
2805: Figure~\ref{fig.K3.K5}.}
2806:
2807:
2808: \label{fig.K6}
2809: \end{figure}
2810: \clearpage
2811:
2812: %%%%%%%%% Figure 13 %%%%%%%%%%
2813:
2814: \begin{figure}
2815: \centering
2816: \includegraphics[scale=0.7]{f15.ps}
2817:
2818: \caption{Spectra of LHS 164, LHS 2734A (both K7.0VI), DEN 0515$-$7211,
2819: and SCR0708-4709 (both K7.0VI:) are shown, with our K7.0V standard
2820: spectrum (represented by a triangle in the inset plot). The telluric
2821: O$_{2}$ A band has been removed. Symbols and lines have the same
2822: meanings as in Figure~\ref{fig.K3.K5}.}
2823:
2824: \label{fig.K7}
2825: \end{figure}
2826: \clearpage
2827:
2828: %%%%%%%%% Figure 14 %%%%%%%%%%
2829:
2830: \begin{figure}
2831: \centering
2832: \includegraphics[scale=0.7]{f16.ps}
2833:
2834: \caption{Spectra of LHS 165, LHS 418, LHS 244 (all M0.0VI), and LHS
2835: 300AB (M0.0VI:) are shown with our M0.0V standard spectrum
2836: (represented by a triangle in the inset plot). The telluric O$_{2}$ A
2837: band has been removed. The hollow arrow indicates the shift on the HR
2838: diagram caused by decreasing metallicity. Symbols and lines have the
2839: same meanings as in Figure~\ref{fig.K3.K5}.}
2840:
2841: \label{fig.M0}
2842: \end{figure}
2843: \clearpage
2844:
2845: %%%%%%%%% Figure 15 %%%%%%%%%%
2846:
2847: \begin{figure}
2848: \centering
2849: \includegraphics[scale=0.7]{f17.ps}
2850:
2851: \caption{Spectra of LHS 507, SCR 0709-4648, LHS 12, LHS 401 (all
2852: M0.5VI), and LHS 521 (M0.5VI:) are shown with our M0.5V standard
2853: spectrum (represented by a triangle in the inset plot). The telluric
2854: O$_{2}$ A band has been removed. The hollow arrow indicates the shift
2855: on the HR diagram caused by decreasing metallicity. The solid arrow
2856: indicates the shift caused by higher gravity. Symbols and lines have
2857: the same meanings as in Figure~\ref{fig.K3.K5}.}
2858:
2859: \label{fig.M0.5.1}
2860: \end{figure}
2861: \clearpage
2862:
2863: %%%%%%%%% Figure 16 %%%%%%%%%%
2864:
2865: \begin{figure}
2866: \centering
2867: \includegraphics[scale=0.7]{f18.ps}
2868:
2869: \caption{Spectra of LHS 367, LHS 299, and LHS 360, are shown (all
2870: M0.5VI:). The telluric O$_{2}$ A band has been removed.}
2871:
2872: \label{fig.M0.5.2}
2873: \end{figure}
2874: \clearpage
2875:
2876: %%%%%%%%% Figure 17 %%%%%%%%%%
2877:
2878: \begin{figure}
2879: \centering
2880: \includegraphics[scale=0.7]{f19.ps}
2881:
2882: \caption{Progressions in metallicity and gravity effects in M1.0VI
2883: subdwarf spectra are shown. The top set of spectra illustrates
2884: spectra with metallicities increasing from black to red (top to
2885: bottom, metallicity scale from m$------$ to m). The red line is our
2886: M0.5V standard spectrum (represented by a triangle in the inset
2887: plot). The bottom two sets of spectra illustrate gravity effects for
2888: metallicity scales m$------$ and m$----$. The telluric O$_{2}$ A band
2889: has been removed. Symbols and lines have the same meanings as in
2890: Figure~\ref{fig.M0.5.1}.}
2891:
2892: \label{fig.M1.1}
2893: \end{figure}
2894: \clearpage
2895:
2896: %%%%%%%%% Figure 18 %%%%%%%%%%
2897:
2898: \begin{figure}
2899: \centering
2900: \includegraphics[scale=0.7]{f20.ps}
2901:
2902: \caption{The top set of spectra illustrates gravity effects for
2903: metallicity scale m$---$. Spectra of LHS 440, LHS 1970, LHS LHS 2734B
2904: (all M1.0VI), and LHS 158 (M1.0[VI]), are also shown compared
2905: individually to our M1.0V standard. The telluric O$_{2}$ A band has
2906: been removed.}
2907:
2908: \label{fig.M1.2}
2909: \end{figure}
2910: \clearpage
2911:
2912: %%%%%%%%% Figure 19 %%%%%%%%%%
2913:
2914: \begin{figure}
2915: \centering
2916: \includegraphics[scale=0.7]{f21.ps}
2917:
2918: \caption{Spectra for WT 233, GJ 191, LHS 406, LHS 3620, LHS 127 (all
2919: M2.0VI) and LHS 318 (M2.0VI:) are shown with our M2.0V standard
2920: spectrum (represented by a triangle in the inset plot). The telluric
2921: O$_{2}$ A band has been removed. Symbols and lines have the same
2922: meanings as in Figure~\ref{fig.M0.5.1}.}
2923:
2924: \label{fig.M2}
2925: \end{figure}
2926: \clearpage
2927:
2928: %%%%%%%%% Figure 20 %%%%%%%%%%
2929:
2930: \begin{figure}
2931: \centering
2932: \includegraphics[scale=0.7]{f22.ps}
2933:
2934: \caption{Spectra for six M3.0VI are shown, with our M3.0V standard
2935: spectrum in the top set of spectra (represented by a triangle in the
2936: inset plot). LHS 326 and LHS 228 show metallicity changes, while LHS
2937: 326, SCR2204$-$3347 and SCR1916$-$3638 show effects of gravity. WT
2938: 135 and LHS 272 are M3.0VI:. Note that the K I lines at 7665\AA~and
2939: 7699\AA~have appeared and blended in the O$_{2}$ A band. Symbols and
2940: lines have the same meanings as in Figure~\ref{fig.M0.5.1}.}
2941:
2942: \label{fig.M3}
2943: \end{figure}
2944: \clearpage
2945:
2946: %%%%%%%%% Figure 21 %%%%%%%%%%
2947:
2948: \begin{figure}
2949: \centering
2950: \includegraphics[scale=0.7]{f23.ps}
2951:
2952: \caption{Spectra for LHS 381 and LHS 144 (both M3.5VI) are shown with
2953: our M3.5V standard spectrum (represented by a triangle in the inset
2954: plot). These three stars show a dramatic metallicity sequence for
2955: M3.5. The spectrum of LHS 228 (M3.0VI) is compared to that of LHS 381
2956: to illustrate the slope difference between M3.0VI and M3.5VI
2957: (8200\AA--9000\AA). LHS 375 and LHS 515 have higher gravities than
2958: LHS 381 and LHS 144, respectively. Note that K I lines (7665\AA~and
2959: 7699\AA) have appeared and blended in the O$_{2}$ A band. Symbols and
2960: lines have the same meanings as in Figure~\ref{fig.M0.5.1}.}
2961:
2962: \label{fig.M3.5}
2963: \end{figure}
2964: \clearpage
2965:
2966: %%%%%%%%% Figure 22 %%%%%%%%%%
2967:
2968: \begin{figure}
2969: \centering
2970: \includegraphics[scale=0.7]{f24.ps}
2971:
2972: \caption{LEHPM 3861 (M4.0VI), LHS 1490 (M5.0VI), LHS 334 (M6.0VI), and
2973: LHS2067A (M6.0VI) subdwarf spectra are shown compared to our standard
2974: M3.0-7.0V standards. Only LHS2067A and the M6.0V standard (filled
2975: triangle) are shown in the HR diagram. Vertical dotted lines
2976: represent the three peaks we use to assist us to assign spectral
2977: types. The K I lines at 7665\AA~and 7699\AA~are clearly seen blended
2978: in the O$_{2}$ A band. Symbols and lines have the same meanings as in
2979: Figure~\ref{fig.M0.5.1}.}
2980:
2981: \label{fig.M4.6}
2982: \end{figure}
2983: \clearpage
2984:
2985: %%%%%%%%% Figure 23 %%%%%%%%%%
2986:
2987: \begin{figure}[h]
2988: \hspace{-2cm}
2989: \includegraphics[scale=0.8, angle=90]{f25.ps}
2990:
2991: \caption{Spectra are shown for subdwarfs from \cite{West2004}
2992: separated into types assigned by us by matching standard spectra (red)
2993: from \cite{Bochanski2007} in the region 8200\AA--9000\AA. Other
2994: colors (black, blue, green, and yellow) represent different
2995: metallicities at each type. Metallicity increases from black to red.}
2996:
2997: \label{fig.sdss.subdwarf.1}
2998: \end{figure}
2999: \clearpage
3000:
3001: %%%%%%%%% Figure 24 %%%%%%%%%%
3002:
3003: \begin{figure}[h]
3004: \centering
3005: \includegraphics[scale=0.7]{f26.ps}
3006:
3007: \caption{These plots indicate the gravity effects for SDSS subdwarfs
3008: of type M3VI and M4VI. The black, blue and green lines represent
3009: increasing gravities: g, g$+$ and g$++$.}
3010:
3011:
3012: \label{fig.sdss.subdwarf.2}
3013: \end{figure}
3014: \clearpage
3015:
3016: %%%%%%%%% Figure 25 %%%%%%%%%%
3017:
3018: \begin{figure}
3019: \centering
3020: \includegraphics[scale=0.7]{f27.ps}
3021:
3022: \caption{Observed spectra of dwarf standard stars (red lines)
3023: compared with the best fitting synthetic spectra (black lines).
3024: Telluric lines are not present in the synthetic spectra. Six regions
3025: labeled from 1 to 6 are discussed in the text.}
3026:
3027: \label{fig.observed.model}
3028: \end{figure}
3029: \clearpage
3030:
3031: %%%%%%%%% Figure 26 %%%%%%%%%%
3032:
3033: \begin{figure}
3034: \centering
3035: \includegraphics[scale=0.7, angle=90]{f28.ps}
3036:
3037: \caption{Synthetic models were compared to the observed spectrum of
3038: our M1.0V standard. The resulting $\chi^{2}$ values are plotted
3039: against temperature for models with five different metallicities, at
3040: four different gravities each. The [m/H]$=-$0.5 plot has the tightest
3041: curves at different {\it log g}.}
3042:
3043: \label{fig.GJ701.chi2}
3044: \end{figure}
3045: \clearpage
3046:
3047: %%%%%%%%% Figure 27 %%%%%%%%%%
3048:
3049: \begin{figure}
3050: \centering
3051: \includegraphics[scale=0.5, angle=90]{f29a.ps}
3052: \includegraphics[scale=0.5, angle=90]{f29b.ps}
3053:
3054: \caption{Comparisons of observed spectra for M1.0V (top) and M3.0V
3055: (bottom) are shown against the two best fitting synthetic spectra
3056: for each. The red line is the best fit, where $T_{eff}$/[m/H]/{\it
3057: log g} $=$ 3600/$-$0.5/4.5 for M1.0, and 3200/$-$0.5/4.5 for M3.0.
3058: The blue line is the second best fit, where the values are
3059: 3400/$-$1.0/4.0 for M1.0 and 3300/0.5/5.5 for M3.0. Telluric bands
3060: are marked as gray boxes.}
3061:
3062: \label{fig.GJ701.LP776}
3063: \end{figure}
3064: \clearpage
3065:
3066: %%%%%%%%% Figure 28 %%%%%%%%%%
3067:
3068: \begin{figure}
3069: \centering
3070: \includegraphics[scale=0.5, angle=90]{f30.ps}
3071:
3072: \caption{The best $\chi^{2}$ values are shown for synthetic spectra
3073: fits to observed spectra for stars of types M0.0V to M5.5V (labeled
3074: relative to each type). The dashed line indicates the selected limit
3075: for reliability of fitting model spectra to observed spectra --- above
3076: this line fits are deemed unreliable.}
3077:
3078: \label{fig.chi2.curve}
3079: \end{figure}
3080: \clearpage
3081:
3082: %%%%%%%%% Figure 29 %%%%%%%%%%
3083:
3084: \begin{figure}
3085: \centering
3086: \includegraphics[scale=0.5, angle=90]{f31.ps}
3087:
3088: \caption{The spectrum of the subdwarf LHS 335 (black) is shown with
3089: the corresponding best fitting synthetic spectrum (red).}
3090:
3091: \label{fig.lhs0335}
3092: \end{figure}
3093:
3094: %%%%%%%%% Figure 30 %%%%%%%%%%
3095:
3096: \begin{figure}[h]
3097: \centering
3098: \includegraphics[scale=0.7, angle=90]{f32.ps}
3099: \vspace{0.5cm}
3100:
3101: \caption{The CaH2+CaH3 versus TiO5 indices are plotted for subdwarfs
3102: discussed in this paper (filled circles). For comparison, known cool
3103: dwarfs (dots), subdwarfs (open triangles), and ``extreme'' subdwarfs
3104: (open boxes) from \cite{Hawley1996}, \cite{Gizis1997}, and
3105: \cite{Reid2005} are also shown. Green circles are subdwarfs with the
3106: highest gravities at a given type (g with most $+$) in
3107: Table~\ref{tbl.spectral.type}. The two blue circles represent LHS 400
3108: and SCR 1822$-$4542, which have higher gravities than M dwarfs.
3109: Red circles represent the lowest metallicity stars in
3110: Table~\ref{tbl.spectral.type} at a given type (m with most $-$).
3111: Solids lines indicate pairs of stars having the same number of $-$ in
3112: metallicity, but different gravities. A dotted line connects LHS
3113: 2734A and LHS 2734B, which is the only wide CPM binary in our sample.
3114: For comparison, the inset plot illustrates the indices (TiO5
3115: vs. CaH2$+$CaH3) calculated from GAIA models for temperatures
3116: 2800--4400K. Axis labels are omitted, but both axes have the same
3117: ranges as the main figure. Three different metallicities, 0.0, $-1.0$
3118: and $-$2.0, and two gravities, $\log g$=4.5 (solid line) and 5.5 (dash
3119: line), are shown. Connections are drawn for stars with the same
3120: temperatures but different gravities.}
3121:
3122: \label{fig.subdwarf.sdm.esdm.index}
3123: \end{figure}
3124: \clearpage
3125:
3126: %%%%%%%%% Figure 31 %%%%%%%%%%
3127:
3128: \begin{figure}
3129: \centering
3130: \includegraphics[scale=0.6, angle=90]{f33a.ps}
3131: \\
3132: \includegraphics[scale=0.6, angle=90]{f33b.ps}
3133:
3134: \caption{CaH1 and TiO5 indices from GAIA model grids are plotted
3135: against effective temperature. The black, blue, red and green lines
3136: indicate {\it log g}$=$4.0, 4.5, 5.0, 5.5, respectively. Note that a
3137: given CaH1 and TiO5 value may correspond to many combinations of
3138: metallicities and gravities.}
3139:
3140: \label{fig.cah1.tio5.vs.teff}
3141: \end{figure}
3142:
3143: %%%%%%%%% Figure 32 %%%%%%%%%%
3144:
3145: \begin{figure}
3146: \centering
3147: \includegraphics[scale=0.7, angle=90]{f34.ps}
3148:
3149: \caption{Spectra from \cite{Lepine2007} are shown, with their
3150: published types. The colors (black and blue) represent different
3151: spectra, and their colors match with labels. Each pair has the almost
3152: the same overall spectral shape, but each star is assigned a different
3153: sub-type and identification as extreme or ultra. Spectra are
3154: normalized at 7500\AA.}
3155:
3156: \label{fig:lepine.1}
3157: \end{figure}
3158:
3159: %%%%%%%%% Figure 33 %%%%%%%%%%
3160:
3161: \begin{figure}
3162: \centering
3163: \includegraphics[scale=0.8]{f35.ps}
3164: \vspace{1cm}
3165:
3166: \caption{Spectra from \cite{Lepine2007} are shown for spectral types
3167: for M3.0, M5.0 and M7.5. The black, blue and red spectra represent
3168: sdM, esdM and usdM, respectively. Spectra are normalized at 7500\AA.}
3169:
3170: \label{fig:lepine.2}
3171: \end{figure}
3172:
3173: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3174:
3175: \end{document}
3176: