1: %%
2: %% AASTeX v5.0 LaTeX 2e macros.
3:
4:
5: % \documentclass{aastex}
6:
7: %% preprint produces a one-column, single-spaced document:
8:
9: \documentclass[preprint]{aastex}
10: \usepackage{graphicx}
11:
12: %% preprint2 produces a double-column, single-spaced document:
13:
14: % \documentclass[preprint2]{aastex}
15:
16: %% You can insert a short comment on the title page using the command below.
17:
18: \slugcomment{To appear in the Astrophysical Journal}
19:
20: %% If you wish, you may supply running head information, although
21: %% this information may be modified by the editorial offices.
22: %% The left head contains a list of authors,
23: %% usually a maximum of three (otherwise use et al.). The right
24: %% head is a modified title of up to roughly 44 characters. Running heads
25: %% will not print in the manuscript style.
26:
27: \shorttitle{A procedure to estimate distances}
28: \shortauthors{Allende Prieto}
29:
30:
31: %% This is the end of the preamble. Indicate the beginning of the
32: %% paper itself with \begin{document}.
33:
34: \begin{document}
35:
36: %% LaTeX will automatically break titles if they run longer than
37: %% one line. However, you may use \\ to force a line break if
38: %% you desire.
39:
40: \title{An Empirical Procedure to Estimate Distances to Stellar Clusters}
41:
42:
43: %% Use \author, \affil, and the \and command to format
44: %% author and affiliation information.
45: %% Note that \email has replaced the old \authoremail command
46: %% from AASTeX v4.0. You can use \email to mark an email address
47: %% anywhere in the paper, not just in the front matter.
48: %% As in the title, you can use \\ to force line breaks.
49:
50: \author{Carlos Allende Prieto}
51: \affil{McDonald Observatory and Department of Astronomy, The University of Texas at Austin, Texas 78712-1083}
52: \email{callende@astro.as.utexas.edu}
53:
54: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
55: %% style, abstract will output a Received/Accepted line after the
56: %% title and affiliation information. No date will appear since the author
57: %% does not have this information. The dates will be filled in by the
58: %% editorial office after submission.
59:
60: \begin{abstract}
61:
62: A most desirable feature of a standard candle to estimate astronomical distances is robustness against changes in metallicity
63: and age. It is argued that the radii of main
64: sequence stars with spectral types from solar to A0 show predictable
65: changes with metallicity, and detectable changes with evolution. Such
66: stars populate the solar neighborhood, and therefore benefit from
67: measurements of angular diameters. Also, reliable determinations of their
68: masses and radii are available from observations of eclipsing binaries.
69: Three empirical relationships are defined and suggested for estimating
70: distances to dwarfs from only $BVK$ photometry. Comparison
71: with {\it Hipparcos} trigonometric parallaxes
72: shows that the method provides
73: errors of about 15 \% for a particular star, which can be reduced to
74: roughly 1.5 \% when applied to young clusters (age $\lesssim 1-2$ Gyr)
75: with $\sim 100$ stars of the appropriate spectral types.
76: If reddening is unknown, main sequence stars with effective
77: temperatures close to 8000 K can constrain it, although
78: an estimate of $\mathcal{R} \equiv A(V)/E(B-V)$ is required.
79: \end{abstract}
80:
81: %% Keywords should appear after the \end{abstract} command. The uncommented
82: %% example has been keyed in ApJ style. See the instructions to authors
83: %% for the journal to which you are submitting your paper to determine
84: %% what keyword punctuation is appropriate.
85:
86: \keywords{Galaxy: open clusters and associations: general --- stars: distances --- stars: fundamental parameters}
87:
88:
89: %% From the front matter, we move on to the body of the paper.
90: %% In the first two sections, notice the use of the natbib \citep
91: %% and \citet commands to identify citations. The citations are
92: %% tied to the reference list via symbolic KEYs. The KEY corresponds
93: %% to the KEY in the \bibitem in the reference list below. We have
94: %% chosen the first three characters of the first author's name plus
95: %% the last two numeral of the year of publication as our KEY for
96: %% each reference.
97:
98: \section{Introduction}
99:
100: Reliable distance estimators are scarce, yet required for many
101: purposes. A few examples will suffice here. Spectroscopic parallaxes
102: are typically used to estimate distances to massive stars across the
103: Galaxy (e.g. Reed \& Reed 2000) for the lack of a better tool.
104: Astrometric parallaxes are usually in need of distance estimates for
105: reference stars very close in the sky to the main target
106: (see, e.g., McArthur et al. 1999). Again, spectroscopic
107: classification and {\it standard} brightness expectations are combined
108: to yield a distance. Accurate distances to clusters derived
109: independently from stellar evolution models
110: are key to test both the models and cluster ages inferred for them
111: (Rosenberg et al. 1999). The {\it Hipparcos} mission (ESA 1997) has
112: greatly improved the situation
113: on this front, with very precise mean parallaxes for nearby clusters
114: (Robichon et al. 1999), but a discussion on possible systematic errors
115: for some clusters is still open (see Pinsonneault et al. 1998). Finally,
116: the Hubble Space Telescope Key Project's effort to constrain
117: the value of the Hubble's constant from different standard candles
118: relies on the cepheids period-luminosity relationship, which is itself
119: based on the assumed distance to the Large Magellanic Cloud and mainly
120: limited in accuracy by its uncertainty (Mould et al. 2000). Different
121: methods of deriving that distance have been implemented, and have been
122: recently critically reviewed by Feast (1999): pulsating stars,
123: supernovae, red-clump stars, and eclipsing binaries have been
124: proposed. The situation is worrying, as the different estimates span a
125: range of a 30\%.
126:
127: A desirable virtue of a procedure to estimate distances is robustness
128: against systematic errors. Most distance estimators rely on
129: relationships between observed quantities and fundamental properties of
130: the standard objects employed. Then, one of the main goals is to
131: understand how the properties of the
132: standard object change with possible parameters,
133: typically metal content and age. This is possible
134: for any kind of star, based on an appropriate understanding of
135: stellar structure and evolution, but maximizing simplicity reduces
136: the uncertainties. State-of-the-art stellar interior models provide an
137: extraordinary guide, but the complexity of the physics, and the
138: difficulties in relating observed and fundamental parameters make them
139: of limited accuracy to derive distances.
140:
141:
142: Some basic quantities are within reach from direct measurements for
143: nearby objects: parallaxes and angular diameters. However, nearby
144: means here roughly a hundred parsecs for the parallax (see the {\it
145: Hipparcos} Catalogue) and, highly dependent on the stellar radius, from
146: several tens to a few hundreds of parsecs for the angular diameter
147: (see, e.g., van Belle 1999). More interesting stellar properties,
148: masses and radii, can be directly determined from combined
149: spectroscopic and photometric observations of double-lined eclipsing
150: binaries (see, e.g., Andersen 1991). Some of such systems are indeed
151: close enough to have highly accurate parallaxes, providing additional
152: information (Ribas et al. 1998; Popper 1998). As a result of the
153: distribution of mass and brightness of the stellar population of the
154: solar neighborhood, the most direct
155: measurements of parallaxes, angular diameters, radii, and masses,
156: involve main-sequence stars with masses between one and
157: three times larger than solar. These stars should be
158: preferred to derive empirical relationships between their physical
159: parameters and measurements, and are proposed here for deriving
160: distances to stellar clusters.
161:
162:
163: \section{Description of the proposed method}
164:
165: The suggested recipe is a kind of {\it photometric parallax}. First,
166: it is possible to use a Barnes-Evans like relationship between
167: the surface brightness and a color index to estimate
168: stellar angular diameters from photometric
169: measurements. A second empirical relation between the stellar radius
170: and a color index, or the absolute magnitude, constrains
171: the radius and, when reddening is absent or
172: well known, fixes the distance. At least in theory, a third equation
173: would close the system, allowing for the amount of reddening to
174: be determined.
175:
176: \subsection{Known reddening}
177:
178: The surface brightness, defined as $S_V= 15 + V_0 + 5\log \theta$, where
179: $V_0$ is the Johnson V magnitude and $\theta$ is the stellar angular
180: diameter in arcseconds, has been shown by Di
181: Benedetto (1998) to follow a tight correlation with the $(V-K)_0$ color
182: index. This is understandable because (1) $S_V$ is linearly related with
183: the bolometric correction and the logarithm of the stellar effective
184: temperature ($T_{\rm eff}$), and (2) the $(V-K)_0$ broad-band color index
185: is almost insensitive to any stellar parameter other than $T_{\rm
186: eff}$ (Wesselink 1969, Barnes \& Evans 1976). Di Benedetto fitted a second-order polynomial to the observed
187: $S_V$ as a function of $(V-K)_0$ for nine dwarfs and subgiants, finding
188: a scatter in the fit of 0.03 mag for the range
189: $-0.1 \le (V-K)_0 \le 1.5$ (spectral types $\sim$ A0--G2).
190: This relationship binds the stellar radius and the distance to
191: the star in a very solid manner. The relationship remains
192: almost the same as stars evolve away from the main sequence. Besides,
193: model atmospheres and stellar evolutionary models predict it to hold
194: for any metallicity.
195:
196: Figure \ref{f1} shows the data and the polynomial fitted by Di Benedetto
197: (solid line). The predictions from the isochrones of Bertelli et al.
198: (1994) with ages from 4 Myr up to 20 Gyr and metallicities (Z) from
199: 0.004 to 0.05 have been over-plotted with different line styles and lay
200: on top of the fit for $0.5 \lesssim (V-K)_0 \lesssim 1.5$, but predict a
201: slightly lower surface brightness for bluer colors. Di Benedetto
202: excluded two stars from the fit ($\alpha$ Aql and $\beta$ Car).
203: The star $\alpha$ Aql is clearly off the
204: correlation, but one might argue whether $\beta$ Car
205: or the evolved $\alpha$ Oph should be excluded or not.
206: Including or excluding any of those stars, the coefficients of the fitting
207: polynomial do not change significantly, and for this reason we have
208: adopted those published by Di Benedetto:
209:
210: \begin{equation}
211: \label{dbd}
212: S_V = 2.556 + 1.580 (V-K)_0 - 0.106 (V-K)_0^2 ~{\rm mag},
213: \end{equation}
214:
215: \noindent which is valid in the range $-0.1 \le (V-K)_0 \le 1.5$, and
216: shows a scatter of 0.03 mag. Van Belle (1999) has carried
217: out an equivalent regression between the zero-magnitude angular sizes
218: and the $(V-K)_0$ colors that can be translated to the dashed line in
219: Figure \ref{f1}. His figures are inconsistent with Di Benedetto's, in particular for the Sun.
220:
221: \begin{figure}[ht!]
222: \centering
223: \includegraphics[width=9cm,angle=90]{./f1.ps}
224: \protect\caption[ ]{
225: Relationship between the surface brightness ($S_V$)
226: and the $(V-K)_0$ color index for the ten dwarfs and subgiants (plus a
227: giant, $\alpha$ Oph), and the polynomial fit derived by Di Benedetto
228: (1998; solid line). The dashed line corresponds to the relationship
229: between the $V$ zero-magnitude and the $(V-K)_0$ color derived by van
230: Belle (1999), and all the other curves show the predictions of the
231: isochrones of Bertelli et al. (1994) for different metallicities and
232: ages (see text). The Sun is identified by the usual symbol ($\odot$).
233: \label{f1}}
234: \end{figure}
235:
236:
237: The $S_V$ vs. $(V-K)_0$ relationship allows us to estimate the angular
238: diameter of the star ($\theta$) and, consequently, relate the stellar
239: radius with the distance to, or the parallax ($p$) of, the star:
240:
241: \begin{equation}
242: \label{eq2}
243: \log \frac{R}{R_{\odot}} = 2.031 + \log \theta - \log p,
244: \end{equation}
245:
246: \noindent where $\theta$ and $p$ are in arcseconds, and it has
247: been used that 1 pc is approximately $3.086 \times 10^{16}$ m, and
248: $R_{\odot} \simeq 6.960 \times 10^{8}$ m.
249:
250: A second relationship is required to break the ambiguity between
251: distance and radius. The list of eclipsing binaries compiled by
252: Andersen (1991) can be used to empirically identify the link between
253: radius and absolute visual magnitude for main sequence stars, as shown
254: in Figure \ref{f2}. The list
255: has been supplemented with the systems RT And and CG Cyg (Popper
256: 1994), as they fill an important gap at masses slightly larger than
257: solar. As the
258: stellar radius increases along with the stars' evolution during its main
259: sequence phase, we are interested in narrowing the range of radii as much as
260: possible. The evolution from the Zero Age Main Sequence (ZAMS) can be
261: constrained from the extremely accurate gravities available for these
262: stars. From the stars spectroscopically classified as dwarfs
263: in Andersen's list, I have retained only those whose gravities
264: were not lower than 23 \% (0.1 dex) of the value predicted at
265: the ZAMS by the isochrones. In Figure \ref{f2}, the stars included in
266: the fit are identified with filled circles and error bars, while
267: those not included are represented by crosses. The
268: least-squares fit (solid line) is
269:
270: \begin{equation}
271: \label{eq1}
272: \begin{array}{ll}
273: \log \frac{R}{R_{\odot}} = & 3.820 \times 10^{-1} - 9.801 \times 10^{-2} M_V + 7.636 \times 10^{-3} M_V^2 \\
274: & + 4.895 \times 10^{-4} M_V^3 - 6.133 \times 10^{-5} M_V^4 - 2.831 \times 10^{-5} M_V^5, \\
275: \end{array}
276: \end{equation}
277:
278: \noindent holds for $-4.6 \le M_V \le$ 5.7, and exhibits a standard deviation of 0.02 dex.
279:
280: \begin{figure}[ht!]
281: \centering
282: \includegraphics[width=12cm,angle=90]{./f2.ps}
283: \protect\caption[ ]{
284: Radii vs. absolute visual magnitudes for eclipsing binaries. The crosses have been dismissed as they have already evolved significantly from the ZAMS. The solid line corresponds to a fifth-order polynomial. The dashed, dot-dashed, three-dot-dashed, long-dashed, and dotted lines show isochrones of 4 Myr, and 0.04, 0.4, 4, 20 Gyr, respectively. The arrow marks the range of applicability for the $S_V - (V-K)_0$ relation.
285: \label{f2}}
286: \end{figure}
287:
288:
289: The isochrones with ages younger than $\sim 1$ Gyr (dashed,
290: dot-dashed, and three-dot-dashed lines in Figure \ref{f2}) predict
291: slightly smaller radii for $M_V > 3$. The Sun itself, which has been
292: included in the fitting sample, might be slightly smaller than the
293: trend suggested by the immediately cooler and hotter stars. Pols et
294: al. (1997), in a comparison of evolutionary models against eclipsing
295: binaries, found inconsistencies for the systems whose components were
296: close to or below one solar mass $-$ considering the same sample as
297: here. However, these systems are either active binaries or flare
298: stars. The
299: M1 twins in the system YY Gem ($M_V \simeq 9$), and the cool components
300: of RT And and CG Cyg ($M_V \simeq 6.3$) are far from the range of
301: luminosities of interest for this work (namely, where the relationship
302: between $S_V$ and $(V-K)_0$ has been tuned, identified in Figure
303: \ref{f2} with the arrows) and have been excluded from further
304: consideration, but they fit well in the same picture. The hypothesis
305: that some of these systems are evolving off the main-sequence is not
306: plausible, as those at $M_V \simeq 5.5$ mag are expected to freeze in
307: their ZAMS position for a period of time longer than the age of the
308: Galaxy. The fact that the radii observed at a given $M_V$ is larger
309: than expected at the ZAMS could be induced by the stars undergoing
310: contraction.
311: Alternatively, the disagreement could be related to the stars'
312: membership to binary systems, since models are constructed for
313: isolated (non- interacting) stars. Another symptom of inconsistencies
314: with stellar evolution models has been discussed by Popper (1997) and
315: Clausen et al. (1999): binaries with one component in the mass range
316: $0.7-1.1 M_{\odot}$ do not lie along a unique isochrone.
317: With no definite answers in hand, we adopt the empirical
318: fit, cautioning the reader about possible systematic errors on the
319: cool side of our $T_{\rm eff}$ range.
320:
321: The scatter about the fit is significantly larger than the error bars
322: and, as explained by Andersen (1991), it mostly corresponds to real
323: differences in chemical abundances, age, rotational velocities, etc.,
324: among the stars. While the radii compiled by Andersen (1991)
325: are of strictly observational origin, the absolute magnitudes were
326: derived from effective temperatures and bolometric corrections
327: ($BC$s). The effective temperatures were compiled from a number of
328: heterogeneous sources, including broad-band photometry,
329: intermediate-band photometry, and spectrum synthesis, all dependent on
330: model atmospheres. Bolometric corrections depend on model atmospheres
331: too (see Popper 1980 for the details of the scale used by Andersen
332: 1991). These two factors introduce additional noise in defining the
333: relationship in Eq. \ref{eq1}. More importantly, however, is that the
334: zero point of the $T_{\rm eff}$ and $BC$ scales could induce systematic
335: effects in Eq. \ref{eq1}. From the definition of luminosity,
336: we can write $\sigma^2(\log
337: \frac{L}{L_{\odot}}) = 4 \sigma^2(\log \frac{R}{R_{\odot}}) + 16
338: \sigma^2(\log \frac{T_{\rm eff}}{T_{\odot}})$. As radii and effective
339: temperatures in Andersen's compilation are typically determined within
340: 2 \% and 3-4 \%, respectively, the error in the luminosity is
341: dominated by the uncertainty in the effective temperature. Similarly,
342: from the definition of visual and bolometric magnitudes, we can write
343: $\sigma^2(M_V) = 2.5^2 \sigma^2(\log \frac{L}{L_{\odot}}) +
344: \sigma^2(BC) \simeq 2.5^2 16 \sigma^2 (\log \frac{T_{\rm
345: eff}}{T_{\odot}}) + \sigma^2(BC)$. Given the expected uncertainties in
346: the $BC$s for late-type stars are not expected to be larger than about 0.04
347: mag, we should only worry about the
348: $T_{\rm eff}$ scale, and $\sigma(M_V) \simeq 4.34 \frac{\sigma(T)}{T}$. Very recently, Ribas et al. (2000) have critically revised Andersen's $T_{\rm eff}$s.
349: The updated values, mostly on the scale of modern Kurucz's model atmospheres (Kurucz 1991, 1994), are compared with those published by Andersen in Figure \ref{f3}. The mean difference
350: $\log T_{\rm eff} ({\rm Andersen}) - \log T_{\rm eff} ({\rm Ribas~ et~ al.})$
351: is $-0.0014 \pm 0.0013$,
352: which implies a correction to $M_V$ of roughly 0.014 mag,
353: suggesting that Andersen's values, and therefore Eq. (1),
354: are not significantly biased.
355:
356: \begin{figure}[ht!]
357: \centering
358: \includegraphics[width=10cm,angle=90]{./f3.ps}
359: \protect\caption[ ]{
360: Comparison between the effective temperatures used by Andersen (1991) to estimate absolute $V$ magnitudes, and the recently revised values of Ribas et al. (2000).
361: \label{f3}}
362: \end{figure}
363:
364:
365: Ideally, one would estimate $M_V$ directly from trigonometric parallaxes.
366: {\it Hipparcos} has observed a number of eclipsing binaries with sufficient accuracy (Popper 1998). Unfortunately, most of the components of
367: those systems have evolved significantly from the ZAMS. It is, however, reassuring that the absolute magnitudes derived by Popper (1998)
368: for the few components that appear on,
369: or very close to the ZAMS are in good agreement with the
370: indirect determinations listed by Andersen (1991).
371:
372:
373: The relationship between $\frac{R}{R_{\odot}}$ and $M_V$ for stars
374: near the ZAMS in Eq. \ref{eq1}, can be expressed in terms of the
375: parallax, as $\log p = \frac{1}{5} (M_V - V_0 - 5)$, where $V_0$ is
376: the intrinsic V magnitude of the star. Then, the
377: Eqs. \ref{eq2} and \ref{eq1} form a complete system of
378: equations that provide $\frac{R}{R_{\odot}}$ and $\log p$,
379: once the $V_0$ and $K_0$ intrinsic magnitudes are known.
380:
381: Eclipsing binaries are a special case. The radii of the
382: components can be determined directly and, using the $S_V$ vs.
383: $(V-K)_0$ (or $S_V-T_{\rm eff}$) relationship to infer the angular
384: diameter, Eq. \ref{eq2} provides the distance. This recipe will
385: eliminate the need for spectrophotometry, and the prospective error in
386: the angular diameter might be even smaller,
387: as adopting $\sigma(S_V) \simeq 0.05$,
388: then $\sigma(\theta) \simeq 2$\%. Unfortunately, the only two eclipsing
389: binaries analyzed in the LMC at the time of this writing (Guinan et al.
390: 1998, Ostrov, Lapasset \& Morrell 2000) are way hotter than the range
391: covered by Di Benedetto's $S_V$ vs. $(V-K)_0$ fit.
392:
393: \subsection{Unknown reddening}
394:
395: The pair of relations between $\log \frac{R}{R_{\odot}}$ and $\log p$
396: (Eqs. \ref{eq1} and \ref{eq2}) described in the previous subsection
397: requires a third equation to solve for an
398: additional variable, the interstellar reddening. This implies
399: the adoption of a simplified model for the interstellar absorption, assuming
400: that $\mathcal{R} \equiv A(V)/E(B-V)$ is known ($\mathcal{R}=3.1$ is
401: the mean value for the galactic interstellar medium).
402: It is well-known that $\mathcal{R}$ does not have a universal value, but
403: instead changes spatially, depending on the absorption and scattering
404: properties of the interstellar medium. There are, however, different
405: ways to derive $\mathcal{R}$
406: from empirical determinations of the reddening in
407: different colors (see Fitzpatrick 1999 and references therein).
408:
409: The $(B-V)_0$ color index is available for the components of
410: all binaries previously considered, and can be related to the
411: stellar radius. Figure \ref{f4}, which includes the same stars
412: used to build the $\log \frac{R}{R_{\odot}}$ vs. $M_V$ relationship,
413: shows that the observed
414: radii correlate tightly with the color index, in very good agreement
415: with the ZAMS predicted by the isochrones for $(B-V)_0 \le 0.3$
416: (their code is the same as in Figure \ref{f2}).
417: Again, the match does not hold for
418: cooler stars. The solid line in Figure \ref{f4} corresponds
419: to a fifth-order polynomial, given by:
420:
421: \begin{equation}
422: \label{eq3}
423: \begin{array}{ll}
424: \log \frac{R}{R_{\odot}} = & 2.266 \times 10^{-1} - 5.222 \times 10^{-1} (B-V)_0 + 2.767 (B-V)_0^2 \\
425: & -7.387 (B-V)_0^3 + 7.245 (B-V)_0^4 -2.471 (B-V)_0^5, \\
426: \end{array}
427: \end{equation}
428:
429: \noindent which holds between $-0.3 \le (B-V)_0 \le 0.8$, and shows a standard deviation of 0.03 dex.
430:
431: \begin{figure}[ht!]
432: \centering
433: \includegraphics[width=12cm,angle=90]{./f4.ps}
434: \protect\caption[ ]{
435: Radii vs. $(B-V)_0$ for eclipsing binaries close to the ZAMS. The solid line represents a fifth-order polynomial and the other curves correspond to isochrones of solar composition and different ages, following the convention of Figure \ref{f2}.
436: \label{f4}}
437: \end{figure}
438:
439:
440: This third empirical relationship provides a single value for the
441: stellar radius, once the intrinsic $(B-V)_0$ color is known. The stellar
442: absorption in the V band, $A(V)$, is introduced into the equations
443: through the expressions for the reddening in the two color indices:
444:
445: \begin{eqnarray}
446: \label{eq4}
447: E(B-V) & = & \frac{A(V)}{\mathcal R}\\ \nonumber
448: E(V-K) & = & A(V) \frac{ (\mathcal{R}-0.02)}{1.12 \mathcal{R}}\\ \nonumber
449: \end{eqnarray}
450:
451:
452: \noindent (Fitzpatrick 1999), and adopting (or measuring) a value for $\mathcal{R}$, we have a system of three equations in which $\log \frac{R}{R_{\odot}}$, $\log p$, and $A(V)$ are the unknowns.
453:
454: \subsection{Summary and practical application of the procedure}
455:
456:
457: The procedure to estimate distances can be summarized as follows. When the reddening is known, one can derive the stellar radius and parallax by combining two of the following three equations:
458:
459: \begin{equation}
460: \label{eq5}
461: \left\{
462: \begin{array}{ccc}
463: \log \frac{R}{R_{\odot}} & = & 2.031 + \log \theta - \log p \\
464: \log \frac{R}{R_{\odot}} & = & 3.820 \times 10^{-1} - 9.801 \times 10^{-2} M_V + 7.636 \times 10^{-3} M_V^2 \\
465: & & + 4.895 \times 10^{-4} M_V^3 - 6.133 \times 10^{-5} M_V^4 - 2.831 \times 10^{-5} M_V^5 \\
466: \log \frac{R}{R_{\odot}} & = & 2.266 \times 10^{-1} - 5.222 \times 10^{-1} (B-V)_0 + 2.767 (B-V)_0^2 \\
467: & & -7.387 (B-V)_0^3 + 7.245 (B-V)_0^4 -2.471 (B-V)_0^5, \\
468: \end{array}
469: \right.
470: \end{equation}
471:
472: \noindent where
473:
474: \begin{equation}
475: \left\{
476: \begin{array}{ccc}
477: \log \theta & = & \frac{1}{5} (S_V[(V-K)_0] - V_0 -15)\\
478: M_V & = & V_0 + 5 + 5 \log p. \\
479: \end{array}
480: \right.
481: \end{equation}
482:
483: \noindent $BVK$ and $M_V$ are given in magnitudes, $\theta$ and $p$ in
484: arcseconds, and $S_V[(V-K)_0]$ is in magnitudes, as prescribed in Eq.
485: \ref{dbd}.
486:
487: When reddening is unknown, but $\mathcal{R} \equiv A(V)/E(B-V)$
488: can be estimated, one should solve simultaneously the three equations in the system (\ref{eq5}), where
489:
490: \begin{equation}
491: \left\{
492: \begin{array} {ccc}
493: \log \theta & = & \frac{1}{5} (S_V[(V-K)_0] - V_0 - 15)\\
494: M_V & = & V_0 + 5 + 5 \log p \\
495: V_0 & = & V-A(V) \\
496: (B-V)_0 & = & \displaystyle (B-V)- A(V)/\mathcal{R} \\
497: (V-K)_0 & = & \displaystyle(V-K)-A(V)\left(\frac{\mathcal{R}-0.02}{1.12 \mathcal{R}}\right). \\
498: \end{array}
499: \right.
500: \end{equation}
501:
502: The upper panel of Figure \ref{f5} shows a graphical example
503: corresponding to the analysis of the star HD 224817 (HIP 80), with
504: $B=8.97$, $V=8.40$ and $K=6.92$. The curves labeled as $a)$, $b)$ and
505: $c)$ correspond to the definition of angular diameter (Eq. \ref{eq2}),
506: the $\log \frac{R}{R_{\odot}} - M_{V}$ (Eq. \ref{eq1}), and the $\log
507: \frac{R}{R_{\odot}}- (B-V)_0$ (Eq. \ref{eq3}) relationships,
508: respectively. It has been assumed a null interstellar absorption, a
509: plausible hypothesis for a star at roughly 64 pc from the Sun or,
510: equivalently, with $\log p = -1.81$. The crossing points are: $a-b:
511: -1.77$, $a-c: -1.79$, and $b-c: -1.82$, indicating distances within
512: 10\% of that derived from the parallax measured by {\it Hipparcos}.
513: The power of the suggested procedure will strongly depend on the way
514: each of the three curves responds to changes in the interstellar
515: absorption. The lower panel of Figure \ref{f5} shows such variations
516: and, for this particular case, the three crossing points change nearly
517: in parallel against $A(V)$ -- a remarkably odd feature.
518:
519: \begin{figure}[ht!]
520: \centering
521: \includegraphics[width=8cm,angle=90]{./f5a.ps}
522: \includegraphics[width=8cm,angle=90]{./f5b.ps}
523: \protect\caption[ ]{
524: {\it Upper panel}: Lines defined by the $S_V - (V-K)_0$ (solid line; $a$), $\log \frac{R}{R_{\odot}} - M_V$ (dashed; $b$), and $\log \frac{R}{R_{\odot}} - (B-V)_0$ (dot-dashed; $c$) relationships
525: for the star HD 224817. {\it Lower panel}: Variation of the
526: crossing points between the curves $a-b$ (solid line),
527: $a-c$ (dashed), and $b-c$ (dot-dashed) against $A(V)$.
528: \label{f5}}
529: \end{figure}
530:
531:
532: To determine whether it is possible or not to extract information on
533: the interstellar reddening, I assume a given distance and amount of
534: reddening and then, make use
535: of Bertelli et al's ZAMS to estimate the absolute visual magnitude of
536: stars with different $(V-K)_0$ color. I
537: derive the observed $V$ magnitude and, via Eqs. \ref{eq1},
538: \ref{eq3}, and \ref{eq4}, the observed $B$ and $K$ magnitudes. Then,
539: I determine the variation with reddening of the parallax at which the
540: three curves in the $\log \frac{R}{R_{\odot}} - \log p$ plane cross.
541: Figure \ref{f6} shows the variation of the mean value of
542: $\frac{\partial \log p}{\partial A(V)}$ for the $a-b$ (solid), $a-c$
543: (dashed), and $b-c$ curves. The error bars correspond to the standard
544: deviation of the slopes of the curves, which are close to, but not
545: quite, straight lines in the vicinity of the right $A(V)$. The three
546: lines change in parallel for a star with $(V-K)_0 \simeq 1.5$, with a
547: slope of about $-0.7$, as confirms Figure \ref{f5}b, but the slope of
548: the variation of the crossing points $b-c$ is significantly different
549: from the other two combinations for stars with $(V-K)_0 \simeq 0.4$ or
550: $T_{\rm eff} \simeq 8000$ K. This result suggests that stars with spectral types A3$-$A6 are suitable for extracting the interstellar reddening.
551:
552:
553: \begin{figure}[ht!]
554: \centering
555: \includegraphics[width=9cm,angle=90]{./f6.ps}
556: \protect\caption[ ]{
557: Mean slope of the position of the crossing points between each par of the the $\log \frac{R}{R_{\odot}}$ vs. $\log p$ curves: $a-b$, $a-c$, and $b-c$, for stars of different effective temperatures ($(V-K)_0$ colors).
558: \label{f6}}
559: \end{figure}
560:
561:
562: \section{A check against {\it Hipparcos}. Evolution effects}
563:
564: Di Benedetto (1998) compiled $V$ and $K$ photometry, and derived
565: empirical effective temperatures for a large number of stars selected
566: as flux standards for the Infrared Space Observatory (ISO). A large
567: fraction of them have precise parallaxes measured in the {\it
568: Hipparcos} catalog, and are considered here for testing
569: the suggested procedure to estimate distances. I have restricted the
570: analysis to the Northern Hemisphere stars classified as dwarfs with
571: errors in the parallax smaller than 30 \%. Most of
572: the stars are nearby, and assuming $A(V) =0$ is appropriate in most
573: cases. As stars evolve away from the ZAMS or, more precisely, from the
574: vicinity of the $\log \frac{R}{R_{\odot}} - M_V$ and $\log
575: \frac{R}{R_{\odot}} - (B-V)_0$ lines empirically defined by the
576: components of eclipsing binaries, the use of such relationships will
577: result in underestimated
578: radii or, equivalently, distances. To quantify this effect we can set
579: limits to the luminosity above the ZAMS of the stars analyzed. In
580: practice, when applying the method to stellar clusters, that limit can
581: be set with respect to the lower boundary, in brightness, of the main
582: sequence. An analysis of the HR diagram, mainly the slope of the
583: main sequence band, should reveal if the considered stars
584: have started to depart significantly from the ZAMS.
585:
586: Figure \ref{f7} compares the distances derived from the mean of the
587: crossing points $a-b$, $a-c$ and $b-c$, with those derived from the {\it
588: Hipparcos} parallaxes, restricting the analysis to stars with absolute
589: visual magnitudes: i) 0.5 mag, ii) 1.0 mag, and iii) 1.5 mag brighter
590: than the theoretical ZAMS. The mean and rms differences between the
591: retrieved distances and those derived from the {\it Hipparcos}
592: parallaxes for the case i),
593: $\frac{d({\rm THIS~ WORK})-d({\rm HIP})}{d({\rm HIP})}$, are displayed in
594: Table 1. I have considered the entire sample, as well as several divided
595: subgroups, depending of their effective temperature.
596:
597: \begin{figure}[ht!]
598: \centering
599: \includegraphics[width=12cm]{./f7.ps}
600: \protect\caption[ ]{
601: Comparison between the retrieved distances and those derived from {\it Hipparcos} parallaxes when the limiting luminosity is set to 0.5 (i), 1.0 (ii), and 1.5 (iii) mags above the ZAMS predicted by the evolutionary models of Bertelli et al. (1994).
602: \label{f7}}
603: \end{figure}
604:
605:
606: A first and very important remark is that the limits of 0.5, 1.0 and
607: 1.5 magnitudes above the theoretical ZAMS are not affecting equitably
608: stars of different mass. As shown in Figure \ref{f4}, the observed
609: radii are slightly larger than the isochrones' prediction for stars
610: with $(B-V)_0 > 0.3$ ($T_{\rm eff} \lesssim 7000$ K) by some quantity
611: that can reach up to 0.05 dex. This implies that the chosen criterion
612: is some 0.2 magnitudes more strict for these stars, which dominate the
613: sample, than for their hotter partners, and therefore we should
614: translate the sequence of $0.5-1.0-1.5$ mag into $0.3-0.8-1.3$ mag.
615: Averaging the results for the crossing points $a-b$, $a-c$ and $b-c$,
616: the mean difference between the derived and the {\it Hipparcos}
617: parallaxes (or distances) is $-1 \pm 2$ \%. Using any single crossing
618: point the systematic difference is never larger than 5\%, and
619: restricting the comparison to stars with effective temperatures lower
620: than 7000 K, which constitute more than three quarters of the sample,
621: similar results hold. The twelve stars with $\ge 7000$ K show a
622: systematic departure from the {\it Hipparcos} measurements. This is
623: likely the result of evolutionary effects. After a few Gyr these stars
624: are no longer expected to be near the ZAMS (see the three-dot-dashed
625: line in Fig. \ref{f2}). Finally, it is of interest to mention that if
626: we derive a relationship between $M_V$ and $(B-V)_0$ for the components of
627: eclipsing binaries and we use it to determine
628: photometric parallaxes for the same {\it Hipparcos} sample,
629: the retrieved distances would have an uncertainty that is
630: roughly 50 \% larger.
631:
632: \section{Metallicity effects}
633:
634: We recall that the $S_V - (V-K)_0$ relationship is expected to be
635: extremely independent of the metal abundances, but the proposed
636: procedure involves two other relations that are not. A strong dependence
637: on metallicity would be very negative, for the sample of eclipsing
638: binaries have metal abundances not far from solar, and they constitute the
639: only empirical resource considered. The ZAMS isochrone for solar
640: metallicity departs from the preferred empirical relationships for
641: stars cooler than 7000 K, but at $M_V = 3.16$ ($(B-V)_0 = 0.25$), where
642: the curves have an inflection point, their agreement is perfect. We can
643: make use of that reference to estimate the systematics in case the
644: metal abundance happens to differ from solar.
645:
646: \begin{figure}[ht!]
647: \centering
648: \includegraphics[width=8cm,angle=90]{./f8a.ps}
649: \includegraphics[width=8cm,angle=90]{./f8b.ps}
650: \protect\caption[ ]{
651: Relationships between $\log \frac{R}{R_{\odot}}$ vs. $M_V$ or $(B-V)_0$ for $\log \frac{Z}{Z_{\odot}} = -0.4$, 0, and 0.4, as predicted by the isochrones (dashed lines), or from the polynomial fitted to the eclipsing binaries (of roughly solar composition), scaled with the shape predicted by the isochrones (solid lines). At the inflection points (solid vertical lines), a change of 0.1 dex per dex is observed.
652: \label{f8}}
653: \end{figure}
654:
655:
656: Figure \ref{f8} shows the empirical relations between $\log
657: \frac{R}{R_{\odot}}$ and $M_V$ or $(B-V)_0$ (solid line) and the
658: predictions of the isochrones (dashed line) for solar abundances
659: as well as differences in defect and excess by 0.4 dex. The empirical relations
660: have been scaled as well, adopting the relative shape between models
661: of different metal content, to produce the uppermost (larger radii)
662: and lowermost (smaller radii) solid curves. At the inflection point, for
663: the two relationships, the changes in radius are very close to linear
664: all the way down to $\log \frac{Z}{Z_{\odot}} = -2$, and probably
665: below that, with a positive slope slightly smaller than 0.1 dex per
666: dex. Combining this result with Eq. \ref{eq2}, we can directly
667: translate that to a systematic effect upon the derived distance. In
668: short, the effect is quite small for moderate departures from solar
669: metallicity. For instance, ignoring the fact that a star is about
670: $-0.2$ dex metal deficient compared to the Sun will not introduce a
671: systematic error larger than a 5\%, in the sense that the distance
672: will be overestimated.
673:
674:
675:
676: \section{Summary and discussion}
677:
678: This paper describes a new method to determine distances to unevolved
679: stars. The errors involved in the practical application to individual
680: stars are relatively large ($\sim$ 15 \%), but the procedure is still
681: of great interest when applied to clusters. It is based on empirical
682: relationships, although model atmospheres enter to define a) a
683: limb-darkening law for obtaining true angular diameters from
684: observations, and b) $T_{\rm eff}$ and $BC$, providing $M_V$. The
685: method takes advantage of direct determinations of radii in eclipsing
686: binaries to establish
687: relationships between radius and absolute visual magnitude and radius
688: and $(B-V)_0$ color. It benefits from the solid Barnes-Evans like
689: correlation existing between surface brightness, $S_V$, and the
690: $(V-K)_0$ color index or, equivalently, between the latter and the
691: stellar angular diameter.
692:
693: The procedure, which only makes use of $BVK$ photometry, can be applied
694: without modification to estimate distances to stellar systems with
695: known reddening, and with a chemical composition close to solar. It
696: requires the target stars, dwarfs with effective temperatures between
697: 6000 and 10000 K, not to have evolved significantly from the ZAMS, and
698: therefore, it is restricted to systems with ages up to 2 Gyr,
699: approximately. Unevolved dwarfs with effective temperatures close to
700: 8000 K ($0.3 \lesssim (V-K)_0 \lesssim 0.5$), can be used to constrain
701: the reddening under the assumption that $\mathcal{R} \equiv
702: A(V)/E(B-V)$ is known (maybe adopting the mean value for the galactic
703: interstellar medium, 3.1).
704:
705: Eclipsing binaries including unevolved
706: stars with $6000 \le T_{\rm eff} \le 10000$ K
707: are a particular case where extremely reliable distances can be derived, independently from bolometric corrections and without spectrophotometry.
708:
709: Assuming an LMC distance modulus of 18.7, the stars of interest would
710: have magnitudes of $20.2 \lesssim V \lesssim 23.7$, right at the edge
711: of current capabilities (see, e.g., Brandner et al. 1999). Besides,
712: many galactic clusters constitute appropriate targets for
713: this method, although no complete $BVK$ observations of large numbers
714: of stars in any young open cluster have been published so far. Further improvements can be expected if $K$
715: magnitudes and spectroscopic metallicities become available
716: for the components of eclipsing binaries. The
717: acquisition, careful analysis, and reduction to individual color
718: indices of infrared light curves for eclipsing binaries is very
719: desirable. That would make it possible to extend,
720: and improve, the relationships between
721: fundamental stellar properties and the colors employed here.
722:
723:
724: \acknowledgments
725: I thank David Lambert, Barbara McArthur, Jocelyn Tomkin, and Russel
726: White for comments and discussions. Constructive criticism from the
727: referee, Ignasi Ribas, helped to improve the paper.
728:
729: %% We have used macros to produce journal name abbreviations.
730: %% AASTeX provides a number of these for the more frequently-cited journals.
731: %% See the Author Guide for a list of them.
732:
733: %% Note that the style of the \bibitem labels (in []) is slightly
734: %% different from previous examples. The natbib system solves a host
735: %% of citation expression problems, but it is necessary to clearly
736: %% delimit the year from the author name used in the citation.
737: %% See the natbib documentation for more details and options.
738:
739: \begin{thebibliography}{}
740: \bibitem[]{} Andersen, J. 1991, \araa, 3, 91
741: \bibitem[]{} Barnes, T. G., \& Evans, D. S. 1976, \mnras, 174, 489
742: \bibitem[]{} Bertelli, G., Bressan, A., Chiosi, C., Fagotto, F., \& Nasi, E. 1994, \aaps, 106, 275 %
743: \bibitem[]{} Brandner, W., Grebel, E. K., Zinnecker, H., \& Brandl, B. 1999, in New Views of the Magellanic Clouds, IAU Symp. 190, Y.-H. Chu, N. B. Suntzeff, J. E. Hesser, and D. A. Bohlender, eds., (ASP: San Francisco) p. 366
744: \bibitem[]{} Clausen, J. V., Baraffe, I., Claret, A., \& VandenBerg, D. A. 1999, in Stellar Structure: Theory and Test of Convective Energy Transport, A. Gim\'enez, E. F. Guinan and B. Montesinos, eds.,
745: ASP Conf. Series, Vol. 173, (ASP: San Francisco) p. 265
746: \bibitem[]{} Di Benedetto, G. P., 1998, \aap, 339, 858
747: %\bibitem[]{} Diaz, M. P., \& Hubeny, I. 1999, \apj, 523, 786
748: \bibitem[]{} ESA 1997, The Hipparcos and Tycho Catalogues, ESA SP-1200
749: \bibitem[]{} Feast, M. 1999, in New Views of the Magellanic Clouds, IAU Symp. 190, Y.-H. Chu, N. B. Suntzeff, J. E. Hesser, and D. A. Bohlender, eds., (ASP: San Francisco) p. 542
750: \bibitem[]{} Fitzpatrick, E. L. 1999, \pasp, 111, 63
751: \bibitem[]{} Guinan, E. F., Fitzpatrick, E. L., Dwarf, L. E., Maloney, F. P., Maurone, P. A., Ribas, I., Pritchard, J. D., Bradstreet, D. H., \& Gim\'enez, A. 1998, \apj, 509, L21
752: %\bibitem[]{} Hagen, G. L. 1970, Pub. of the David Dunlab Obs., 4, 1
753: %\bibitem[]{} Johnson, H. L. 1953, \apj, 117, 353
754: %\bibitem[]{} Manteiga, M., Martinez Roger, C., Morales, C., \& Sabau, L. 1991, \aaps, 87, 419
755: \bibitem[]{} McArthur, B. E., Benedict, G. F., Lee, J., Lu, C.-L., van Altena, W. F., Deliyannis, C. P., Girard, T., Fredrick, L. W., Nelan, E., Duncombe, R. L., Hemenway, P. E., Jefferys, W. H., Shelus, P. J., Franz, O. G., \& Wasserman, L. H. 1999, \apj, 520, L59
756: \bibitem[]{} Mould, J. R. et al. 2000, \apj, 529, 786
757: \bibitem[]{} Ostrov, P. G., Lapasset, E., \& Morrell, N. I. 2000, \aap, in press (astro-ph/0003118)
758: \bibitem[]{} Pinsonneault, M. H., Stauffer, J. Soderblom, D. R., King, J. R., \& Hanson, R. B. 1998, \apj, 504, 170
759: \bibitem[]{} Pols, O. R., Tout, C. A., Schr\"oder, K.-P., Eggleton, P. P., \& Manners, J. 1997, \mnras, 289, 869
760: \bibitem[]{} Popper, D. M. 1980, \araa, 18, 115
761: \bibitem[]{} Popper, D. M. 1994, \aj, 103, 1091
762: \bibitem[]{} Popper, D. M. 1997, \aj, 114, 1195
763: \bibitem[]{} Popper, D. M. 1998, \pasp, 110, 929
764: \bibitem[Reed and Reed (2000)]{2000PASP..112..409R} Reed, B.\ C.\, \& Reed, L.\ G.\ 2000, \pasp, 112, 409
765: \bibitem[]{} Ribas, I., Gim\'enez, A., Torra, J., Jordi, C., \& Oblak, E. 1998, \aap, 330, 600
766: \bibitem[]{} Ribas, I., Jordi, C., Torra, J., \& Gim\'enez, A., E. 2000, \mnras, 313, 99
767: \bibitem[]{} Robichon, N., Arenou, F., Mermilliod, J.-C., \& Turon, C. 1999, \aap, 345, 471
768: \bibitem[]{} Rosenberg, A., Saviane, I., Piotto, G., \& Aparicio, A. 1999, \aj, 118, 2306
769: \bibitem[]{} van Belle, G. T. 1999, \pasp, 111, 1515
770: \bibitem[]{} Wesselink, A. J. 1969, \mnras, 144, 297
771: \end{thebibliography}
772:
773:
774: %% Generally speaking, only the figure captions, and not the figures
775: %% themselves, are included in electronic manuscript submissions.
776: %% Use \figcaption to format your figure captions. They should begin on a
777: %% new page.
778:
779: \clearpage
780:
781: %% No more than seven \figcaption commands are allowed per page,
782: %% so if you have more than seven captions, insert a \clearpage
783: %% after every seventh one.
784:
785: %% There must be a \figcaption command for each legend. Key the text of the
786: %% legend and the optional \label in curly braces. If you wish, you may
787: %% include the name of the corresponding figure file in square brackets.
788: %% The label is for identification purposes only. It will not insert the
789: %% figures themselves into the document.
790: %% If you want to include your art in the paper, use \plotone.
791: %% Refer to the on-line documentation for details.
792:
793:
794: %\figcaption[f1.ps]{Relationship between the surface brightness ($S_V$)
795: %and the $(V-K)_0$ color index for the ten dwarfs and subgiants (plus a
796: %giant, $\alpha$ Oph), and the polynomial fit derived by Di Benedetto
797: %(1998; solid line). The dashed line corresponds to the relationship
798: %between the $V$ zero-magnitude and the $(V-K)_0$ color derived by van
799: %Belle (1999), and all the other curves show the predictions of the
800: %isochrones of Bertelli et al. (1994) for different metallicities and
801: %ages (see text). The Sun is identified by the usual symbol ($\odot$).
802: %\label{f1}}
803:
804: %\figcaption[f2.ps]{Radii vs. absolute visual magnitudes for eclipsing
805: %binaries. The crosses have been dismissed as they have already evolved
806: %significantly from the ZAMS. The solid line corresponds to a fifth-order
807: %polynomial. The dashed, dot-dashed, three-dot-dashed, long-dashed, and dotted
808: %lines show isochrones of 4 Myr, and 0.04, 0.4, 4, 20 Gyr, respectively. The
809: %arrow marks the range of applicability for the $S_V - (V-K)_0$
810: %relation.\label{f2}}
811:
812: %\figcaption[f3.ps]{Comparison between the effective temperatures used by
813: %Andersen (1991) to estimate absolute $V$ magnitudes, and the recently revised
814: %values of Ribas et al. (2000). \label{f3}}
815:
816: %\figcaption[f4.ps]{Radii vs. $(B-V)_0$ for eclipsing binaries close to the
817: %ZAMS. The solid line represents a fifth-order polynomial and the other curves
818: %correspond to isochrones of solar composition and different ages, following the
819: %convention of Figure \ref{f2}. \label{f4}}
820:
821: %\figcaption[f5a.ps, f5b.ps]{{\it Upper panel}: Lines defined by the $S_V -
822: %(V-K)_0$ (solid line; $a$), $\log \frac{R}{R_{\odot}} - M_V$ (dashed; $b$),
823: %and $\log \frac{R}{R_{\odot}} - (B-V)_0$ (dot-dashed; $c$) relationships
824: %for the star HD 224817. {\it Lower panel}: Variation of the
825: %crossing points between the curves $a-b$ (solid line),
826: %$a-c$ (dashed), and $b-c$ (dot-dashed) against $A(V)$. \label{f5}}
827:
828: %\figcaption[f6.ps]{Mean slope of the position of the crossing points between
829: %each par of the the $\log \frac{R}{R_{\odot}}$ vs. $\log p$ curves: $a-b$,
830: %$a-c$, and $b-c$, for stars of different effective temperatures ($(V-K)_0$
831: %colors).\label{f6}}
832:
833: %\figcaption[f7.ps]{Comparison between the retrieved distances and those derived %from {\it Hipparcos} parallaxes when the limiting luminosity is set to 0.5 (i), %1.0 (ii), and 1.5 (iii) mags above the ZAMS predicted by the evolutionary
834: %models of Bertelli et al. (1994). \label{f7}}
835:
836: %\figcaption[f8a.ps,f8b.ps]{Relationships between $\log \frac{R}{R_{\odot}}$ vs. %$M_V$ or $(B-V)_0$ for $\log \frac{Z}{Z_{\odot}} = -0.4$, 0, and 0.4, as
837: %predicted by the isochrones (dashed lines), or from the polynomial fitted to
838: %the eclipsing binaries (of roughly solar composition), scaled with the shape
839: %predicted by the isochrones (solid lines). At the inflection points (solid
840: %vertical lines), a change of 0.1 dex per dex is observed. \label{f8}}
841:
842:
843: %% Tables should be submitted one per page, so put a \clearpage before
844: %% each one.
845:
846: %% Two options are available to the author for producing tables: the
847: %% deluxetable environment provided by the AASTeX package or the LaTeX
848: %% table environment. Use of deluxetable is preferred.
849: %%
850:
851: %% Three table samples follow, two marked up in the deluxetable environment,
852: %% one marked up as a LaTeX table.
853:
854:
855: \clearpage
856:
857: \begin{deluxetable}{crrrrr}
858: %\footnotesize
859: \tablecaption{Relative mean (and standard deviation) between the retrieved distances and those derived from {\it Hipparcos} parallaxes. \label{table1}}
860: \tablewidth{0pt}
861: \tablehead{
862: \colhead{Sample} & \colhead{$a-b$} & \colhead{$a-c$} &
863: \colhead{$b-c$} &
864: \colhead{Average} & \colhead{N} }
865: \startdata
866: All & $-0.05$ (0.13) & $-0.02$ (0.15) & 0.04 (0.26) & $-0.01$ (0.17) & 55 \\
867: $T_{\rm eff} < 7000$ & $-0.03$ (0.14) & 0.01 (0.15) & 0.11 (0.25) & 0.03 (0.17) & 43 \\
868: $7000 \leq T_{\rm eff} < 8000$ & $-0.18$ (0.05) & $-0.16$ (0.06) & $-0.11$ (0.09) & $-0.15$ (0.07) & 3 \\
869: $8000 \leq T_{\rm eff} < 9000$ & $-0.13$ (0.02) & $-0.15$ (0.02) & $-0.18$ (0.07) & $-0.15$ (0.03) & 4 \\
870: $T_{\rm eff} \ge 9000$ & $-0.07$ (0.04) & $-0.14$ (0.04) & $-0.26$ (0.04) & $-0.16$ (0.04) & 5 \\
871: \enddata
872: \end{deluxetable}
873:
874:
875:
876: \end{document}
877:
878: %%
879: