1: % Revision of 7/25/03, incorporating referee's comments
2: % ApJ manuscript number 58350
3:
4:
5: \documentclass[preprint]{aastex}
6: % \documentclass{aastex}
7:
8: \begin{document}
9:
10: \title{The distances to open clusters
11: from main-sequence fitting. I. New models and a comparison
12: to the properties of the Hyades eclipsing binary vB~22}
13:
14: \author{Marc H. Pinsonneault and Donald M. Terndrup}
15: \affil{Ohio State University, Department of
16: Astronomy, Columbus, OH 43210}
17: \email{pinsono,terndrup@astronomy.ohio-state.edu}
18:
19: \author{Robert B. Hanson}
20: \affil{University of California Observatories/Lick
21: Observatory, Santa Cruz, CA 95064}
22: \email{hanson@ucolick.org}
23:
24: \and
25:
26: \author{John R. Stauffer}
27: \affil{Infrared
28: Processing and Analysis Center, California Institute of
29: Technology, Mail Code 100-22, 770 South Wilson Avenue,
30: Pasadena, CA 91125}
31: \email{stauffer@ipac.caltech.edu}
32:
33: \slugcomment{}
34:
35: \shorttitle{Hyades isochrone}
36: \shortauthors{Pinsonneault et al.}
37:
38: \begin{abstract}
39: In the first of a new series
40: of papers on open cluster distances, we use updated
41: stellar evolution models to construct an isochrone appropriate
42: for the Hyades, and compare it with
43: the Hyades eclipsing binary system
44: vB~22. We find that the absolute and relative luminosities of
45: the two stars are in good agreement with the
46: model, but the radii do
47: not match the values inferred from eclipse data.
48: We present evidence that there is a consistency problem
49: with the flux ratios and the inferred radii, and discuss
50: possible theoretical effects that could be responsible for
51: the mismatch in the radii. We derive a helium abundance
52: for the Hyades of $Y = 0.271 \pm 0.006$, which is
53: equal within the errors to the Sun's initial helium
54: abundance even though the Hyades is considerably more
55: metal-rich.
56: \end{abstract}
57:
58: \keywords{binaries: (eclipsing), stars: distances, stars: abundances}
59:
60: \section{Introduction}
61:
62: The distances to Milky Way star clusters as derived
63: from main-sequence fitting play a critical role in
64: unraveling the history of the Galaxy and, via
65: luminosity calibration of pulsating variables,
66: in finding distances throughout the Local Group.
67: Parallaxes from the Hipparcos
68: satellite have provided precise measurements of the
69: distances to the nearest open clusters, particularly
70: the Hyades \citep{per98}, allowing stringent tests of
71: the predictions of stellar evolutionary models
72: \citep[e.g.,][]{lebreton01}.
73:
74: To go beyond the handful of clusters with trigonometric
75: parallaxes \citep[e.g.,][]{vl99,rob99,makarov02} requires
76: isochrones that are physically accurate and
77: well calibrated over a wide range of temperature --
78: ideally all along the main sequence. As is well
79: known this is a nontrivial
80: exercise, since the luminosity of the main-sequence is
81: a sensitive function of helium abundance and metallicity.
82: Radii from stellar models depend on the treatment of
83: convection, for which only simple phenomenological theories
84: are available. Photometric colors and bolometric corrections are
85: often poorly determined, especially for stars much hotter
86: or cooler than the Sun, and are highly dependent on the
87: details of model atmospheres employed in the computation
88: of the isochrones. Rapidly rotating stars typically
89: have large spots and chromospheric emission, not modelled in the
90: computation of the isochrones, which could
91: affect their colors especially in the blue and ultraviolet
92: \citep{vla87,sta03}, As a result of some or
93: all these effects, it is typically the case
94: that even the best isochrones have do not match the detailed shape
95: of the main sequence as determined from photometry
96: \citep{ter00,cas01,lebreton01}.
97:
98: Despite the complexity of the problem, it is now possible
99: to determine more accurate absolute and relative
100: distances to open clusters. Helioseismology
101: has given us reliable measures of many parameters which
102: directly or indirectly affect the radius, including the
103: helium abundance, the amount of helium diffusion,
104: and opacities in the convective zone \citep[e.g.,][]{bpb01}.
105: Hipparcos parallaxes in the Hyades \citep{per98,debruijne01}
106: and of nearby field stars \citep{jimenez03,per03} can potentially
107: provide the means of empirically correcting the
108: isochrone color-temperature relation.
109:
110: In addition, eclipsing binary stars provide a powerful test of the
111: theory of stellar structure and evolution,
112: particularly the mass/luminosity relation. This is
113: especially true of systems in star clusters, where there
114: are additional constraints on the age and abundances of
115: the stars. As summarized by \citet{lebreton01},
116: the Hyades cluster has five binaries where
117: the components have measured masses. Of these systems
118: only vB~22 has masses with a small enough uncertainty to
119: place powerful constraints on the theoretical models; we
120: will therefore focus on vB~22 (= 818~Tau, HD~27130). In
121: this system,
122: the relative magnitudes of vB~22A and vB~22B have been
123: measured in several colors and the absolute radii have
124: been inferred.
125:
126: We will use vB~22 as a test of both the absolute
127: luminosities and effective temperatures of our models.
128: We contend that the agreement in luminosity that we obtain
129: justifies the construction of an empirical isochrone where
130: the colors as a function of $M_V$ are adjusted to reproduce
131: the morphology of the Hyades color-magnitude diagram.
132: In addition, we will show that the absolute magnitudes of
133: the two components in the $B$, $V$, and $I$ bands provide support
134: for the relative model luminosities and effective temperatures,
135: even though the direct radii inferred from the binaries
136: are not in agreement with the models. This step justifies
137: holding the model effective temperatures fixed and varying
138: the color calibrations when constructing the empirical
139: isochrone, the details of which are discussed
140: in the second paper in this series. Finally, we examine the question
141: of the Hyades helium abundance and the ratio $\Delta Y /
142: \Delta Z $ appropriate for chemical evolution of solar-neighborhood
143: stars.
144:
145: The vB~22 system has been studied extensively since \citet{mcclure82}
146: first used it to determine the distance to the Hyades and to
147: constrain the mass-luminosity relationship. The most recent
148: papers \citep{lastennet99,lebreton01,tr02} have yielded
149: somewhat discordant results. \citet{lebreton01} claim evidence for
150: a low helium abundance, while \citet{lastennet99}
151: and \citet{tr02} find that the luminosities of
152: the models are consistent with the data if the Hipparcos
153: distance is adopted. All authors note the apparent
154: contradiction between the radii of theoretical models and
155: those obtained from the eclipse data. In light of these
156: results we believe that a careful analysis of the data
157: and the theoretical models is warranted, with particular
158: attention to the errors involved.
159:
160: \section{An Isochrone for the Hyades}
161:
162: We begin with a new set of theoretical models described in
163: detail by \citet{spt00}. The essential aspects of these
164: models are repeated here for the reader's convenience.
165:
166: We used the Yale Rotating Evolution Code (YREC) to
167: construct evolutionary tracks over the mass
168: range $0.25 \leq (M / M_\odot) \leq 2.25$. YREC is a Henyey
169: code which solves the equations of stellar structure in
170: one dimension \citep{guenther92} and which follows
171: rotational evolution by treating the star as a
172: set of nested, rotationally deformed shells. For this
173: application, however, we used the code in its
174: non-rotating mode; non-rotating stars of the age and masses
175: considered here are structurally identical to those
176: which are rotating \citep[e.g.,][]{spt00} and solar-like Hyades
177: stars are slow rotators \citep{radick87,psc03}. The chemical
178: composition of each shell is updated separately using
179: the nuclear reaction rates of \citet{gb98}. Composition
180: changes due to microscopic diffusion can be calculated. The initial
181: chemical mixture is the solar mixture of \citet{gn93}, and
182: for the Sun the models have a surface metallicity of $Z =
183: 0.0176$ at the age of the solar system.
184: We use the latest OPAL opacities \citep{ig96}
185: for the interior of the star down to temperatures of
186: $\log T({\rm K}) = 4$. For lower temperatures, we use
187: the molecular opacities of \citet{af94}. For regions of
188: the star with $\log T({\rm K}) \geq 6$, we used the
189: OPAL equation of state \citep{rsi96}. For regions where
190: $\log T(K) \leq 5.5$, we used the equation of state from
191: \citet{saumon95}, which calculates particle densities
192: for hydrogen and helium, including partial dissociation
193: and ionization by both pressure and temperature. In
194: the transition region between these two temperatures,
195: both formulations are weighted with a ramp function and
196: averaged. The equation of state includes both radiation
197: pressure and electron degeneracy pressure. For the surface
198: boundary condition, we experimented with several stellar
199: atmosphere models as described below. For our base
200: case, we adopted $Y = 0.273$ for the Hyades and ignored
201: diffusion (details
202: to follow), and used the standard
203: B\"ohm-Vitense mixing length theory \citep{bv58,cg68}
204: with $\alpha = 1.72$, to match the solar radius
205: ($R_\odot = 6.9598 \times 10^{8}$ m) and luminosity
206: ($L_\odot = 3.8515 \times 10^{26}$ W) at the
207: present age of the Sun (4.57 Gyr).
208:
209: The evolutionary tracks were generated
210: for a Hyades metallicity of [Fe/H] $= +0.13$ and
211: scaled solar abundances \citep{bf90,psc03}. The tracks were
212: interpolated to form an isochrone for an adopted
213: age of 550 Myr, consistent with ages derived from
214: models excluding convective overshoot \citep{per98}.
215: Models with overshoot have ages $\approx 625$ Myr, but since we are
216: dealing with relatively low mass stars,
217: the comparison to vB~22 is completely
218: insensitive to the choice of cluster age.
219: The color-temperature
220: relation in \citet{lej98} was used to generate preliminary
221: colors\footnote{$V - I$ is on the Cousins system,
222: while the $V - K_s$ colors use the (short) $K$ band.}
223: from the model parameters $M_{\rm bol}$ and $T_{\rm eff}$.
224: The resulting isochrone is given as Table 1; the sensitivity
225: of the colors to the choice of color-temperature
226: relation is discussed below. Note that this isochrone is
227: not empirically calibrated to match the photometry of the
228: Hyades main sequence, a necessary
229: procedure discussed at length in our next paper.
230:
231: The principal theoretical uncertainties in the models
232: are the adopted mixture of heavy elements and
233: the physics chosen for the solar calibration,
234: in particular whether microscopic diffusion is included
235: or not. In the Hyades, the average iron abundance
236: [Fe/H] is now determined to high precision ($\pm 0.01$ dex),
237: and the relative abundances are
238: near solar for most of the elements that contribute
239: significantly to the internal opacities \citep{psc03}.
240: The choice of model atmospheres and low-temperature opacities only
241: affects the position in the HR diagram for effective
242: temperatures below 3500 K; similar comments apply to the
243: equation of state. The superb agreement between theory
244: and data for helioseismology provides some real confidence
245: in the accuracy of the ingredients of the models for stars
246: similar to the Sun, such as vB~22.
247:
248: Although the effects of microscopic diffusion do not
249: matter for the Hyades themselves (the cluster is only 12\%
250: the age of the Sun), they make a difference for the choice
251: of a calibrating solar model.
252: The net effect of diffusion is a gradual decrease in the
253: helium and heavy element content in the outer layers with
254: a smaller fractional rise in the central values of these
255: quantities. Compared to models with a uniform composition
256: profile, solar models including diffusion have a higher
257: overall helium abundance and require a higher value of
258: $\alpha$ to match the solar radius and luminosity.
259:
260: The proper thing to do, therefore, would be to calibrate
261: the Hyades models using diffusion models for the Sun. For
262: our base case (the isochrone in Table 1),
263: however, we ignored this and instead used models that
264: do not follow diffusion and which were calibrated using
265: models of the homogeneous Sun even though these are
266: incompatible with seismology. The principal reason for this
267: is that the effect of rotationally induced mixing, which
268: diminishes the effects of diffusion, is difficult to model
269: for stars much hotter than the Sun where convection zones
270: are very thin.
271:
272: For the base case, which we will call the ``no diffusion''
273: models, we chose the helium abundance as illustrated in Figure 1.
274: We assumed that the helium abundance is a function of the
275: heavy element content $Z$ is given by $Y = Y_p
276: + (\Delta Y/ \Delta Z) Z$, where $Y_p$ is the primordial
277: helium abundance. We took $Y_p = 0.245 \pm 0.002$, an
278: intermediate value between estimates from cosmic nucleosynthesis
279: and measures in metal-poor \ion{H}{2} regions
280: \citep[see][and references therein]{bono02,ti02}. The
281: solar helium abundance in models lacking diffusion
282: is $Y_\odot = 0.266 \pm 0.001$; these require a
283: mixing length set by $\alpha = 1.74$. Adopting these
284: values of $Y_p$, $Y_\odot$, and $\alpha$ yields
285: $Y = 0.273$ for the Hyades using the Sun's surface
286: metal abundance of $Z_\odot = 0.0176$.
287:
288: We also computed models using an alternative set of parameters
289: calibrated on solar models that include diffusion \citep{bpb01}.
290: Models compatible with helioseismology that include both
291: rotational mixing and diffusion have surface abundances of
292: $Y_{\odot,{\rm surf}} = 0.249 \pm 0.003$ and
293: $Z_{\odot,{\rm surf}} = 0.0176$, which imply an initial
294: composition of $Y_\odot = 0.274$ and $Z_\odot = 0.019 \pm 0.001$ and
295: $\alpha = 1.85$. Using these values to extrapolate a model for
296: the Hyades at $Z({\rm Hyades}) / Z_\odot = 1.35$, we would
297: derive $Y = 0.280$ for the Hyades. These will be called the
298: ``diffusion'' models; these also follow the effects of
299: microscopic diffusion, even though it does not produce
300: significant effects at the age of the Hyades.
301:
302: \section{Photometry, masses, and radii for vB~22}
303:
304: Table 2 summarizes three high-precision distance estimates
305: to vB~22, from an orbital parallax \citep{ps88}
306: solution\footnote{This value was apparently
307: misquoted by \citet{lebreton01}
308: in their Table 1, but does not affect their analysis.},
309: the Hipparcos trigonometric parallax \citep{per98}, and
310: the kinematic parallax \citep{debruijne01}. These are all
311: in excellent agreement.
312:
313: The basic photometric data for the vB~22 system
314: are summarized in Table 3. The first two rows of
315: that table show the \citet{sm87} photometry for
316: vB~22 (i.e., both stars together) in $B$, $V$,
317: and (Cousins) $I_C$, along with
318: the derived luminosity ratio in each filter,
319: where the errors are taken from
320: that paper. Following this are the
321: apparent magnitudes in each band for the individual
322: components derived from the luminosity ratios. Note that
323: the photometric errors are significant for the secondary and will
324: be accounted for in the discussion below. Finally, we derive the
325: difference in absolute magnitude between the primary and secondary
326: in each filter, and also include the
327: absolute magnitudes that would be obtained from the
328: kinematic parallax \citep{debruijne01} of vB~22,
329: again with errors in the individual components.
330:
331: In Table 4, we compare the model radii and temperatures
332: at the masses derived for each
333: component of vB~22 by \citet{tr02} to their
334: solution (top ten rows
335: of the table); we also compare the models to the earlier estimates from
336: \citet{ps88} (last six rows). The latter values were the ones
337: used by \citet{lebreton01}. In the comparison to
338: the \citet{tr02} solution, we tabulate the properties of the model
339: both with and without diffusion, while we only
340: show the no-diffusion case in comparison to \cite{ps88}. The
341: quantities derived from the model are on the scale where
342: the Sun has $M_{\rm bol}$ is 4.746 and the radii and effective
343: temperatures are obtained from the solar-calibrated
344: helium and mixing length.
345:
346: The first thing to note is that in all cases the model
347: radii at the observed masses are considerably
348: different from the radii derived from the analysis
349: of the eclipses by $2.5 - 4.5$ times the formal errors
350: derived from propagating the mass error.
351: This indicates that there is an inconsistency\footnote{Since
352: $L \propto R^2T^4_{\rm eff}$, one could match the
353: models to the data using two of the quantities
354: $(L, R, T_{\rm eff})$ but not all three simultaneously. Another
355: indication that something is the matter is that
356: both stars formally have the same gravity in the \cite{tr02}
357: analysis, which contradicts the strong prediction from
358: theory that the mean density increases with decreasing mass
359: for main-sequence stars. The models predict
360: $\log g({\rm vB~22A}) - \log g({\rm vB~22B}) = -0.11$.} in the
361: observational determination of mass, radius, and
362: luminosity compared to the model. Since luminosity
363: is determined by a combination of temperature and radius,
364: we need to examine these separately to pinpoint the
365: source of this inconsistency.
366:
367: To quantify the size of the mismatch in temperature,
368: we also show in Table 4 the model temperatures at the
369: observed masses of vB~22A and vB~22B using the
370: mass/radius relation in the models. We also compare
371: this to the case in which we take the observed radius
372: as correct, which would imply a larger temperature for
373: vB~22A where the observed radius is smaller than in
374: the model, and a smaller temperature for vB~22B. Here,
375: the required changes in temperature are considerable,
376: amounting to $150 - 190$ K.
377:
378: The effect of including diffusion
379: is shown in the top part of that Table for the
380: \citet{tr02} solution.
381: For the same starting helium abundance the effects of the
382: precise value of the mixing length are very
383: small for luminosity and
384: modest even for the effective temperature:
385: the diffusion models would be roughly 35 K
386: hotter than the no-diffusion models for the primary
387: and 13 K hotter for the secondary. Thus the inconsistency
388: with the models is not caused by the treatment of diffusion
389: in the solar calibration.
390:
391: Because the Hyades has been well studied spectroscopically,
392: there exist independent estimates of the luminosity/temperature
393: relation. In Figure 2, we show spectroscopically derived
394: temperatures for a subset of stars in the recent study
395: of Hyades abundances by \citet{psc03}. Stars with $T < 6000$ K
396: and with good $BVI_CK_s$ photometry are shown as open
397: points with error bars. The values of $M_V$ are derived
398: from individual kinematic parallaxes \citep{debruijne01};
399: errors in this quantity are dominated by distance errors
400: rather than photometric errors. PSC to not list individual
401: temperature errors, so we took $\pm 50$ K as a representative
402: value, derived from PSC's comparison of their temperatures
403: to those in previous studies. The
404: solid line on that figure is the Hyades isochrone derived in
405: this paper. The filled circles show the temperature of
406: the isochrone at the \cite{tr02} masses. The filled
407: triangles display the temperatures that would be found for
408: the components of vB~22 if the measured radii were correct.
409: The agreement between the isochrone and the spectroscopic
410: temperatures is excellent, which shows that the luminosity/radius
411: relationship in the models is nearly correct, at least
412: under the assumption that all Hyades stars have identical
413: metallicity. If on the other hand the
414: luminosity/temperature relation in the models
415: were adjusted to match the radii in the \citet{tr02}
416: solution (vB~22A hotter by 200 K, vB~22B cooler), then
417: there would be difference of about 0.15 dex
418: between the hottest and coolest stars in the PSC sample
419: (the hotter stars would be come out more metal rich).
420:
421: In Figure 3, we compare the isochrones to the \citet{tr02}
422: solution in $M_V$, $B - V$ and $V - I_C$ as a function of mass.
423: The isochrone is slightly brighter than the data, which indicates
424: that the distance to vB~22 is underestimated or, as we will
425: discuss in $\S$ 4, that the helium abundance we adopted
426: for the Hyades is too high. The color-temperature relation
427: in the isochrone differs from the inferred colors of the
428: binary components, but in this paper we are mainly concerned
429: with comparing the model luminosities to the data.
430:
431: Table 5 summarizes the errors in absolute magnitude,
432: radius, or effective
433: temperature that are contributed by different effects.
434: To compute the result of the uncertainty in mass,
435: we assume the \citet{tr02} error of 0.0062
436: $M_\odot$ for each star.
437: The errors from [Fe/H] assume $\sigma{\rm [Fe/H]} = 0.05$ dex
438: per star,
439: while those listed for the bolometric correction were
440: computed by taking the largest difference between the
441: inferred fluxes for that filter between three different color
442: calibrations (below) and dividing by 2. We have also computed
443: errors that would result if the metal abundance were known
444: with vanishingly small errors; these are shown in the
445: rows labeled ``no Z.'' Because the sign of changes
446: in metallicity is the same for both components the errors
447: in their relative fluxes are smaller; we give these values
448: in the row labeled ``B-A.''
449:
450: We compare the absolute magnitudes of the two
451: components of vB~22 to the models in Table
452: 6, where we employ models lacking diffusion
453: but with two alternative color
454: calibrations, that of \citet{ah95} and \citet{aam96},
455: in addition to the one employed as our base
456: case \citep{lej98}. The comparison is done at
457: fixed mass. In general, the different color
458: calibrations only change the luminosity by a
459: few hundredths of a magnitude. We take the scatter
460: in the luminosities indicating the size of
461: errors in the bolometric corrections; these were
462: shown in Table 5. We show the effects of including
463: diffusion or of forcing the temperature scale to
464: match the observed stellar radii in Table 7. These
465: models are for the \citet{lej98} color calibration
466: only.
467:
468: The agreement between theory and observation is impressive
469: for the no diffusion models and well within the expected
470: errors for both the absolute luminosities and the relative
471: luminosities. Overall the different color calibrations
472: agree best for the $V$ and $B$ bands, while there is more
473: scatter in the predicted $I_C$-band luminosities. The relative
474: fluxes are very close to the predicted level for the $V$
475: and $B$ bands and are mildly inconsistent with the $I_C$ band
476: fluxes, especially for vB~22B where the models are fainter than
477: the data by 0.1 mag or so. This comparison
478: indicates that the problem is most likely
479: in the $I_C$ band bolometric corrections rather than in the
480: $V$ band bolometric corrections.
481:
482: However, the agreement is not preserved if the effective
483: temperatures are altered to the values inferred from the
484: eclipse data. Essentially, choosing a lower effective
485: temperature for the secondary drives down the $V$ and $B$
486: band fluxes while slightly increasing the $I$-band flux.
487: As a result the relative flux differences in the $V$ and
488: $B$ bands become much larger. This result is insensitive
489: to the metallicity of the Hyades because decreases in the
490: metallicity affect the luminosity and effective temperature
491: of both components in the same sense, while the model
492: radii are insensitive to the metallicity. We therefore
493: conclude that the relative fluxes in different bands, the
494: mass-luminosity relationship, and the radii obtained from
495: the eclipse data are not consistent with one another.
496: One of the three must be in error. Because of the
497: insensitivity of the mass-luminosity relationship to errors in the
498: input physics we view it as more likely that there is some
499: unresolved issue in one of the two other ingredients.
500: As in previous analyses of this system, we note that there
501: is no obvious single change in
502: the input physics that can reconcile the models with
503: both components simultaneously. What we have added is evidence that
504: these radii are also inconsistent with the flux ratios.
505:
506: \section{The Helium Abundance of the Hyades}
507:
508: Because the luminosity of stellar models at fixed mass
509: is very sensitive to the helium abundance
510: ($\partial M_{\rm bol}/ \partial Y = -10$),
511: we can formally derive an initial helium abundance for
512: the Hyades: using the \citet{debruijne01} distance to vB~22,
513: we find $Y = 0.271 \pm 0.006$ for the no-diffusion models.
514: The models including diffusion are brighter than those
515: without, indicating that the initial choice of $Y = 0.280$
516: was too high at the assumed distance. Correcting those
517: models brings the estimated helium abundance down to
518: $Y = 0.271$, showing that the derived helium abundance
519: is nearly independent of the details of the solar calibration.
520:
521: If we adopt a primordial helium abundance of
522: 0.245, the Hyades would
523: give a slope $\Delta Y / \Delta Z =
524: 1.11 \pm 0.25$, smaller than the values of 1.5 and 1.2
525: obtained from solar models with and
526: without diffusion respectively. If the
527: more realistic initial solar helium abundance
528: (including diffusion) and the Hyades helium
529: abundance are taken at face value, they
530: suggest a scatter in helium at fixed metal
531: abundance of order 0.009; however,
532: this range is only marginally significant.
533: If we take this
534: as a one sigma error range, it would imply only a
535: small resulting error in a cluster distance modulus at
536: fixed [Fe/H] of order 0.027 magnitudes.
537:
538: The Hyades helium abundance we derive
539: is equal within the errors to the solar value $Y = 0.273$
540: in models excluding diffusion. The
541: solar models with diffusion would predict a higher
542: helium abundance of 0.280, which is
543: not favored by the data; however, it is only of
544: order $2\sigma$ from the measurement.
545: Isochrones are less sensitive to changes in
546: helium than models of a given mass, with a
547: change of 0.01 in helium causing a change in
548: $M_V$ at fixed $T_{\rm eff}$ of 0.03. We can include
549: this in our next paper in the error budget for the absolute distances.
550:
551: This value is higher than that obtained by Lebreton et
552: al.\ (2001), who found $Y = 0.255 \pm 0.009$. There
553: are two reasons for this difference. First, we adopted
554: the revised masses of \citet{tr02}; these yield
555: predicted luminosities that are 0.08 to 0.09 mag
556: fainter than found by \citet{ps88}, resulting in
557: a higher helium abundance. Second, our models employ
558: a number of ingredients not used by Lebreton et al.;
559: as they note, using the OPAL equation of state,
560: Kurucz model atmospheres rather than a gray atmosphere,
561: and a higher mixing length
562: all increase the inferred value of $Y$.
563:
564: The mixture of heavy elements can also have an impact
565: on the properties of the models. There have been two
566: relatively recent revisions of the \citet{gn93}
567: abundance scale for the Sun which was used in
568: this paper. \citet{gs98} have reduced
569: the CNO abundances, while \citet{asplund00} has
570: proposed a downward revision of 9.2\% in the zero-point
571: between the meteoritic and photospheric abundance scales.
572: This would not alter those abundances not tied to the
573: meteoritic scale (C, N, O, Ne, etc.) but would affect
574: important interior opacity sources such as Si and Fe.
575: Helioseismic tests \citep{bpb01}
576: indicate that the \citet{gs98} mixture marginally degrades the
577: agreement with the measured solar convection zone depth,
578: but by a degree that is consistent with other known
579: theoretical uncertainties. The solar helium abundance
580: is insensitive to the CNO abundances. The \citet{asplund00}
581: mixture has a small impact on the solar sound speed, but
582: the lower iron abundance results in a lower solar helium
583: abundance. This is in disagreement with the measured
584: surface abundance of helium at the 2-3 $\sigma $ level;
585: \citet{bpb01} were not able to rule this
586: out because the degree of disagreement is sensitive to
587: systematic errors in the helium abundance determination
588: arising from the equation of state.
589:
590: We have verified that adopting the \citet{gs98} mixture instead
591: of the \citet{gn93} mixture produces only small changes in the
592: isochrones (or even the more sensitive tests possible
593: in vB~22). The \citet{asplund00} mixture would imply a
594: Hyades helium abundance (including diffusion) that is
595: comparable to the \citet{gn93} or
596: \citet{gs98} helium abundance inferred
597: in the absence of diffusion. We have also explored the
598: effects of including small deviations from the solar
599: mix in the Hyades as measured by \citet{psc03}, and found
600: that the impact would be of the same order as the
601: difference between the choice of the solar element mix.
602:
603: \section{Summary and Remarks}
604:
605: In this paper, we have begun a new analysis of open cluster distances
606: by performing stringent tests the of the luminosity/temperature for new
607: isochrones with updated physics. We explored the effect of varying
608: many details of the models, including the abundance mix of
609: helium and metals, and concluded that the models match the
610: relative temperatures and luminosities of the components of the
611: binary vB~22 in the Hyades. Along the way, we derived a helium
612: abundance that is not much different from that in the Sun, even
613: though the Hyades is more metal-rich.
614:
615: The resulting helium abundance is almost insensitive to the details
616: of the solar calibration used to generate the isochrone. Models
617: using a solar calibration consistent with helioseismology predict
618: a high value of $Y$ for the Hyades; the resulting isochrone, however,
619: is too bright at fixed mass, which leads to the conclusion that the
620: Hyades helium abundance is not much different from the Sun's
621: initial value of $Y = 0.272$. This may imply a scatter in the
622: helium abundance of about $\delta Y = 0.01$ at fixed metallicity in
623: the solar neighborhood. While this would produce a real effect
624: on the luminosity of the isochrones at fixed mass, we have argued
625: that the effect on the isochrones at fixed color would be much
626: smaller and would not produce big errors in the distance estimates
627: derived from main-sequence fitting.
628:
629: The principal result of this paper was our demonstration that there is
630: an inconsistency between the solar models and the binary data in
631: mass, luminosity, temperature, and radius. External checks on the
632: abundance and temperature scales from recent high-precision abundances
633: in the Hyades \citet{psc03}, indicates that the problem lies with
634: the observational determination of the radii. We conclude from this
635: that we can adjust the isochrones to match the Hyades photometry by
636: leaving the stellar luminosity and temperatures as they are, and
637: compute corrections to the color/temperature relations; this is the
638: subject of our next paper.
639:
640: The comparison between the model colors and absolute magnitude and
641: the data for vB~22 may be complicated by the presence of spots on
642: the stellar surfaces. In the Pleiades, stars with the luminosities
643: of vB~22B fall up to 0.5 mag below the main sequence in $V$, $B - V$
644: as defined by the Hyades \citet{vla87,sta03}, the difference is negligible
645: when $V$ and $V - I$ are used. \citet{sta03} attribute this to
646: the presence of hot and cool regions on the surfaces of the rapidly-rotating
647: Pleiades stars as compared to their slowly-rotating counterparts
648: in the Hyades. The orbital period of vB~22 is 5.6 days \citep{sm87}.
649: If both members of vB~22 were corotating, this would correspond to equatorial
650: rotation speeds of only 8 km s$^{-1}$ for vB~22A and 6 km s$^{-1}$ for vB~22B
651: (the binaries are well detached).
652: These values are not much different from the average
653: $v \sin i$ for Hyades stars of similar colors in \citet{psc03},
654: and considerably slower than the projected rotation
655: speed ($v \sin i$) for the Pleiades stars with the most
656: anomalous $B - V$ color. The vB~22 system, however, has exhibited
657: flares and color variations outside eclipses, so active regions may indeed be
658: important at some level \citep{sm87}.
659:
660: \acknowledgements
661:
662: The work reported here was supported in part by the National Science
663: Foundation, under grants AST-9731621 and AST-0206008 to the Ohio State
664: University Research Foundation. We wish to thank the anonymous referee
665: for many helpful comments.
666:
667: \begin{thebibliography}
668:
669: % \bibitem[]{}
670:
671: \bibitem[Alexander \& Ferguson(1994)]{af94} Alexander,
672: D. R., \& Ferguson, J. W. 1994, \apj, 437, 879
673:
674: \bibitem[Allard \& Hauschildt(1995)]{ah95} Allard, F., \&
675: Hauschildt, P. H. 1995, \apj, 445, 433
676:
677: \bibitem[Alonso, Arribas, \& Mart\'{\i}nez-Roger(1996)]{aam96}
678: Alonso, A., Arribas, S., \& Mart\'{\i}nez-Roger, C. 1996,
679: \aap, 313, 873
680:
681: \bibitem[Asplund(2000)]{asplund00} Asplund, M. 2000, \aap,
682: 359, 755
683:
684: \bibitem[Bahcall, Pinsonneault, \& Basu(2001)]{bpb01}
685: Bahcall, J. N., Pinsonneault, M. H., \& Basu, S. 2001,
686: \apj, 555, 990
687:
688: \bibitem[Boesgaard \& Friel(1990)]{bf90} Boesgaard, A. M.,
689: \& Friel, E. D. 1990, \apj, 351, 467
690:
691: \bibitem[B\"ohm-Vitense(1958)]{bv58} B\"ohm-Vitense,
692: E. 1958, Z.\ Astrophys., 46, 108
693:
694: \bibitem[Bono et al.(2002)]{bono02} Bono, G., Balbi, A., Cassisi, S.,
695: Vittorio, N., \& Buonanno, R. 2002, \apj, 568, 463
696:
697: \bibitem[de Bruijne, Hoogerwerf, \& de Zeeuw(2001)]{debruijne01}
698: de Bruijne, J. H. J., Hoogerwerf, R., \& de Zeeuw,
699: P. T. 2001, \aap, 367, 111
700:
701: \bibitem[Castellani, Degl'Innocenti, \& Prada Moroni(2001)]{cas01}
702: Castellani, V., Degl'Innocenti, S., \& Prada Moroni, P. G.
703: 2001, \mnras, 320, 66
704:
705: \bibitem[Cox \& Guili(1968)]{cg68} Cox, J. P., \& Guili,
706: R. T. 1968, Principles of Stellar Structure (New York:
707: Gordon and Breach)
708:
709: \bibitem[Grevesse \& Noels(1993)]{gn93} Grevesse, N., \& Noels,
710: A. 1993 in Origin and Evolution of the
711: Elements, ed. M. Prantzos, E. Vangioni-Flam, \& M. Cass\'e
712: (Cambridge: Cambridge Univ. Press), 15
713:
714: \bibitem[Grevesse \& Sauval(1998)]{gs98} Grevesse, N., \& Sauval,
715: A. J. 1993, Space Sci.\ Rev., 85, 161
716:
717: \bibitem[Gruzinov \& Bahcall(1998)]{gb98} Gruzinov, A.,
718: \& Bahcall, J. 1998, \apj, 504, 996
719:
720: \bibitem[Guenther et al.(1992)]{guenther92} Guenther,
721: D. B., Demarque, P., Kim, Y.-C., \& Pinsonneault,
722: M. H. 1992, \apj, 387, 372
723:
724: \bibitem[Iglesias \& Rogers(1996)]{ig96} Iglesias, C. A.,
725: \& Rogers, F. J. 1996, \apj, 464, 943
726:
727: \bibitem[Jimenez et al.(2003)]{jimenez03} Jimenez, R., Flynn, C.,
728: MacDonald, J., \& Gibson, B. K. 2003, preprint (astro-ph/0303179)
729:
730: \bibitem[Lastennet et al.(1999)]{lastennet99} Lastennet, E.,
731: Valls-Gabaud, D., Lejeune, T., \& Oblak, E. 1999, \aap, 349, 485
732:
733: \bibitem[Lebreton, Fernandes, \& Lejeune(2001)]{lebreton01} Lebreton,
734: Y., Fernandes, J., \& Lejeune, T. 2001, \aap, 374, 540
735:
736: \bibitem[van Leeuwen(1999)]{vl99} van Leeuwen, F. 1999, \aap, 341, 71
737:
738: \bibitem[van Leeuwen, Alphenaar, \& Meys(1987)]{vla87}
739: van Leeuwen, F., Alphenaar, P., \& Meys, J. J. M. 1987, \aap, 67, 483
740:
741: \bibitem[Lejeune, Cuisinier, \& Buser(1998)]{lej98} Lejeune, T.,
742: Cuisinier, F., \& Buser, R. 1998, \aap, 313, 873
743:
744: \bibitem[Makarov(2002)]{makarov02} Makarov V. V., 2002, \aj, 124, 3299
745:
746: \bibitem[McClure(1982)]{mcclure82} McClure, R. D. 1982, \apj, 254, 606
747:
748: \bibitem[Paulson, Sneden, \& Cochran(2003)]{psc03} Paulson,
749: D. B., Sneden, C., \& Cochran, W. D. 2003, \aj, 125, 3185
750:
751: \bibitem[Percival, Salaris, \& Kilkenny(2003)]{per03} Percival,
752: S. M., Salaris, M., \& Kilkenny, D. 2003, \aap, 400, 541
753:
754: \bibitem[Perryman et al.(1998)]{per98} Perryman, M. A. C., et al.
755: 1998, \aap, 331, 81
756:
757: \bibitem[Peterson \& Solensky(1988)]{ps88}
758: Peterson, D. M., \& Solensky, R. 1988, \apj, 333, 256
759:
760: \bibitem[Radick et al.(1987)]{radick87} Radick, R. R.,
761: Thompson, D. T., Lockwood, G. W., Duncan, D. K., \&
762: Baggett, W. E. 1987, \apj, 321, 459
763:
764: \bibitem[Robichon et al.(1999)]{rob99} Robichon, N., Arenou, F.,
765: Mermilliod, J.-C., \& Turon, C. 1999, \aap, 345, 471
766:
767: \bibitem[Rogers, Swenson, \& Iglesias(1996)]{rsi96}
768: Rogers, F. J., Swenson, F. J., \& Iglesias, C. A. 1996,
769: \apj, 456, 902
770:
771: \bibitem[Saumon, Chabrier, \& Van Horn(1995)]{saumon95}
772: Saumon, D., Chabrier, G., \& Van Horn, H. M. 1995, \apjs,
773: 99, 713
774:
775: \bibitem[Schiller \& Malone(1987)]{sm87} Schiller,
776: S. J., \& Malone, E. F. 1987, \aj, 93, 1471
777:
778: \bibitem[Sills, Pinsonneault, \& Terndrup(2000)]{spt00}
779: Sills, A., Pinsonneault, M. H., \& Terndrup, D. M. 2000,
780: \apj, 534, 335
781:
782: \bibitem[Stauffer et al.(2003)]{sta03} Stauffer, J. R., Jones,
783: B. F., Backman, D., Hartmann, L. W., Barrado y Nevascu\'es, D.,
784: Pinsonneault, M. H., Terndrup, D. M., \& Muench, A. 2003,
785: preprint
786:
787: \bibitem[Terndrup et al.(2000)]{ter00} Terndrup, D. M., Stauffer,
788: J. R., Sills, A., Yuan, Y., Jones, B. F., Fischer, D., \&
789: Krishnamurthi, A. 2000, \apj, 119, 1303
790:
791: \bibitem[Thuan \& Izotov(2002)]{ti02} Thuan, T. X., \& Izotov,
792: Y. I. 2002, Space Sci. Rev., 100, 263
793:
794: \bibitem[Torres \& Ribas(2002)]{tr02} Torres, G., \& Ribas, I. 2002,
795: \apj, 567, 1140
796:
797: \end{thebibliography}
798:
799: \clearpage
800: \input{tabl1.tex}
801:
802: \input{tabl2.tex}
803:
804: \input{tabl3.tex}
805:
806: \input{tabl4.tex}
807:
808: \input{tabl5.tex}
809:
810: \input{tabl6.tex}
811:
812: \input{tabl7.tex}
813:
814: \clearpage
815:
816: \figcaption[f1.eps]{Derivation of the helium abundance $Y$ for the
817: Hyades isochrone. The lower dashed line shows the extrapolation
818: from the primordial helium abundance through models of the Sun that
819: lack diffusion to the Hyades metal abundance [Fe/H] $= + 0.13 \pm 0.01$
820: ($Z = 0.0237$);
821: this is the base case discussed in the paper and shown in Table 1.
822: The upper line shows the extrapolation through solar models consistent
823: with helioseismology, which would imply a higher helium abundance in the
824: Hyades. Both initial values are shown as open circles. The filled
825: point shows the helium abundance for the Hyades found in this
826: paper.}
827:
828: \figcaption[f2.eps]{Comparison of isochrone and spectroscopic
829: temperature scales for the Hyades. The solid line is the
830: theoretical isochrone in Table 1, while the open points with
831: error bars show spectroscopic temperatures from Paulson, Sneden,
832: \& Cochran (2003) along with absolute visual magnitude computed
833: from Hyades kinematic parallaxes (de Bruijne, Hoogerwerf, \& de Zeeuw
834: (2001). The filled circles are for vB~22A and vB~22B at the
835: isochrone temperature and
836: measured masses for each component, while the triangles indicate the
837: temperatures that would be derived by forcing the models to
838: have the radius indicated by the Torres \& Ribas (2002) solution. }
839:
840: \figcaption[f3.eps]{Comparison of the isochrone to data for vB~22.
841: The isochrone is shown as a solid line, and the values for vB~22A and
842: vB~22B are shown as points with error bars.}
843:
844: \end{document}