1: \documentclass[12pt,preprint]{aastex}
2:
3: %\def\today{\ifcase\month\or January\or February\or March\or April\or
4: %May\or June\or July\or August\or September\or October\or November\or
5: %December\fi \space\number\day, \number\year}
6: %\newcount\nextyear
7: %\nextyear=\number\year \advance\nextyear by 1
8:
9: \def\kms{\ifmmode{\rm km\th s^{-1}}\else km\th s$^{-1}$\fi}
10: \def\th{\thinspace}
11: \def\msol{$M_{\odot}$} \def\bvo{(B$-$V)$_{0}$}
12:
13: \shortauthors{Boden et al.}
14: \shorttitle{V773~Tau~A}
15:
16: \begin{document}
17:
18: \title{Dynamical Masses for Pre-Main Sequence Stars: A
19: Preliminary Physical Orbit for V773 Tau A}
20:
21: \email{**Accepted Form -- 15 June 2007**}
22:
23: \author{Andrew F.\ Boden\altaffilmark{1,2},
24: Guillermo Torres\altaffilmark{4},
25: Anneila I.\ Sargent\altaffilmark{3},
26: Rachel L.\ Akeson\altaffilmark{1},
27: John M.\ Carpenter\altaffilmark{3},
28: David A.\ Boboltz\altaffilmark{5},
29: Maria Massi\altaffilmark{6},
30: Andrea M.~Ghez\altaffilmark{7},
31: David W.\ Latham\altaffilmark{4},
32: Kenneth J.~Johnston\altaffilmark{5},
33: Karl M.~Menten\altaffilmark{6},
34: Eduardo Ros\altaffilmark{6}
35: }
36:
37: \email{bode@ipac.caltech.edu}
38:
39: \altaffiltext{1}{Michelson Science Center, California
40: Institute of Technology, 770 South Wilson Ave., Pasadena CA 91125}
41:
42: \altaffiltext{2}{Department of Physics and Astronomy, Georgia State
43: University, 29 Peachtree Center Ave., Science Annex, Suite 400,
44: Atlanta GA 30303}
45:
46: \altaffiltext{3}{Division of Physics, Math, and Astronomy, California
47: Institute of Technology, MS 105-24, Pasadena, CA 91125}
48:
49: \altaffiltext{4}{Harvard-Smithsonian Center for Astrophysics, 60
50: Garden St., Cambridge MA 02138}
51:
52: \altaffiltext{5}{U.S.~Naval Observatory, 3450 Massachusetts Ave, Washington DC 20392}
53:
54: \altaffiltext{6}{Max-Planck-Institut f\"ur Radioastronomie, Auf dem H\"ugel 69, D-53121 Bonn, Germany}
55:
56:
57: \altaffiltext{7}{Division of Astronomy and Astrophysics, UCLA, Los Angeles, CA 90095-1562}
58:
59: %%\altaffiltext{4}{Present address: Palomar Observatory, California
60: %%Institute of Technology, 770 South Wilson Ave., Pasadena CA 91125}
61:
62:
63: \begin{abstract}
64:
65: We report on interferometric and radial-velocity observations of the
66: double-lined 51-d period binary (A) component of the quadruple
67: pre-main sequence (PMS) system V773 Tau. With these observations we
68: have estimated preliminary visual and physical orbits of the
69: V773~Tau~A subsystem. Among other parameters, our orbit model
70: includes an inclination of 66.0 $\pm$ 2.4 deg, and allows us to infer
71: the component dynamical masses and system distance. In particular we
72: find component masses of 1.54 $\pm$ 0.14 and 1.332 $\pm$ 0.097
73: M$_{\sun}$ for the Aa (primary) and Ab (secondary) components
74: respectively.
75:
76: Our modeling of the subsystem component spectral energy distributions
77: finds temperatures and luminosities consistent with previous studies,
78: and coupled with the component mass estimates allows for comparison
79: with PMS stellar models in the intermediate-mass range. We compare
80: V773~Tau~A component properties with several popular solar-composition
81: models for intermediate-mass PMS stars. All models predict
82: masses consistent to within 2-sigma of the dynamically determined
83: values, though some models predict values that are more consistent than
84: others.
85:
86:
87: \end{abstract}
88:
89: \keywords{binaries: spectroscopic --- stars: fundamental parameters
90: --- stars: pre-main sequence --- stars: individual (V773~Tau)}
91:
92:
93:
94: \section{Introduction}
95: \label{sec:introduction}
96:
97: Accurate determinations of the physical properties of stars
98: (e.g.~mass, radius, temperature, luminosity, elemental abundances,
99: rotation, etc.) provide fundamental tests of stellar structure and
100: evolution models. Among the areas where our understanding of stellar
101: structure is most uncertain is in pre-main sequence (PMS) stars,
102: particularly for low-mass systems \citep[see][and references for
103: summaries]{Palla2001,Hillenbrand2004,Mathieu2007}. Previously we have
104: reported results from our program of studying PMS binary systems with
105: the intent of establishing dynamical mass and luminosity constraints
106: \citep{Boden2005b}. Here we report on observations of the double-lined
107: spectroscopic binary V773~Tau~A, part of the PMS quadruple system
108: V773~Tau.
109:
110:
111:
112: \objectname[V773 Tau]{V773~Tau} (HDE~283447, HBC~367) is among the
113: most interesting and well-studied systems in the Taurus-Auriga star
114: forming region. V773~Tau was first identified as a T~Tauri star by
115: \citet{Rydgren1976} based on H$\alpha$ and Ca II H and K emission,
116: high lithium abundance, photometric variability, and K2 spectral type.
117: The object presented an enigmatic mixture of classical (CTTS) and
118: weak-lined T Tauri (WTTS) characteristics until it became apparent
119: there are multiple components; V773~Tau was resolved as a visual
120: binary independently by \citet{Ghez1993} and \citet{Leinert1993} (with
121: components designated here as A and B). \citet{Martin1994} suggested,
122: and \citet[herein W1995]{Welty1995} established A as a short-period
123: (51 d) double-lined spectroscopic binary. In high-angular resolution
124: studies \citet[herein D2003]{Duchene2003} and \citet{Woitas2003}
125: independently identified an additional, ``infrared'' component in the
126: system (herein designated C -- note D2003 use an alternate component
127: notation), making V773~Tau an apparent compact quadruple system with
128: four stars within roughly 100 AU \footnote{There are at least three
129: different component notations used in describing the V773~Tau system.
130: Herein we have described the 51-d SB2 subsystem as ``A'', with stellar
131: components Aa and Ab; this is consistent with nomenclature used in
132: \citet{Massi2006} and \citet{Woitas2003}. Conversely, D2003 refer to
133: the 51-d SB2 subsystem as ``AB''. Finally, based on arguments made by
134: \citet{Hartigan1994} on the ``likely'' association of V773~Tau and
135: \objectname[FM Tau]{FM~Tau}, the Washington Double Star Catalog
136: \citep{Mason2001} describes the four-star complex as ``A'' (with
137: FM~Tau as ``B''), and the 51-d SB2 as ``Aa'' with components ``Aa1''
138: and ``Aa2''.}. Using spatially-resolved photometry, D2003 find the A,
139: B, and C components to have distinctly different spectral energy
140: distributions. D2003 conclude that V773~Tau contains stars at very
141: different PMS evolutionary states: the optically brighter A stars are
142: WTTS in character, while the circumstellar material inferred from the
143: significant IR excess is associated with the B and C components.
144:
145:
146: Further, the V773~Tau~A subsystem is a bright, polarized, and
147: highly-variable radio source. \citet{Kutner1986} discovered radio
148: continuum emission from the system, appearing as the most luminous
149: Taurus-Auriga T Tauri star in the survey by \citet{O'Neal1990}.
150: \citet{Feigelson1994} published a comprehensive multiwavelength study
151: of the V773~Tau system, observing both polarized and flaring radio
152: emission. W1995 argued that both A components are chromospherically
153: active based on variations in their optical spectrum and estimates of
154: relatively rapid rotation ($v \sin i$ $\sim$ 40 km s$^{-1}$); this
155: point was confirmed by \citet[herein P1996]{Phillips1996}, who
156: resolved the radio emission into a clear binary morphology apparently
157: corresponding to the two A components. \citet{Massi2002} identified
158: an increase in flare activity around A periastron, and
159: \citet{Massi2006} studied one such flare in detail; more frequent and
160: powerful radio flaring at periastron is apparently due to interaction
161: of component magnetic fields while in close proximity. Finally,
162: \citet[herein L1999]{Lestrade99} used Very Long Baseline
163: Interferometry (VLBI) astrometry to measure the trigonometric parallax
164: for V773~Tau~A as 6.74 $\pm$ 0.25 mas, corresponding to a system
165: distance of 148.4 $\pm$ 5.5 pc.
166:
167:
168: Here we report on observations of the V773~Tau~A subsystem made with
169: the Keck Interferometer \citep[KI, see][]{Colavita2003}, the Very Long
170: Baseline Array \citep[VLBA, see][]{Napier1994,Zensus1995}, and
171: spectroscopic radial velocities from the Harvard-Smithsonian Center
172: for Astrophysics (CfA) ``Digital Speedometers'' \citep{Latham1992}.
173: These observations resolve V773~Tau~A and allow us to estimate the
174: visual and physical (i.e.~three-dimensional) orbits, and determine the
175: component dynamical masses and system distance. We describe these
176: orbit models as {\em preliminary} because they weakly depend on
177: ongoing observations and modeling of the larger V773~Tau system.
178: Combining derived dynamical masses and system distance with component
179: radiometric modeling, we estimate component temperatures and
180: luminosities, allowing us to make direct comparisons with a number of
181: popular solar-composition PMS stellar models.
182:
183:
184:
185:
186: \section{Observations and Orbital Solution}
187: \label{sec:observations}
188:
189: \paragraph{KI Observations}
190: The KI interferometric observable used for these measurements is the
191: fringe contrast or {\em visibility} (squared, V$^2$) of an observed
192: brightness distribution on the sky. Orbital analysis methods for such
193: wide-band V$^2$ observations are discussed in \citet{Boden2000} and
194: not repeated here. V773~Tau~A was observed by KI in $K$-band
195: ($\lambda \sim 2.2 \mu$m) on five nights between 2002 Oct 24 and 2006
196: Dec 8, a data set spanning roughly four years and 30 orbital periods.
197: V773~Tau~A and calibration objects were typically observed multiple
198: times during each of these nights, and each observation (scan) was
199: approximately 130 sec long. For each interferometric scan we computed
200: a mean \citep[incoherent wide-band;][]{Colavita1999,Colavita2003}
201: V$^2$ value from the scan data, and the error in the V$^2$ estimate is
202: inferred from the rms internal scatter. For this analysis we have
203: used \objectname[HD 27741]{HD~27741} (G0 V) and \objectname[HD
204: 28483]{HD~28483} (F6 V) as calibration objects. Calibrating our
205: interferometric data with respect to these objects results in 27
206: calibrated visibility scans on V773~Tau~A over the five epochs.
207: Because the AO-corrected Keck Telescopes separately resolve the
208: V773~Tau A-B-C complex (D2003), and the KI beam combiner is fed by
209: single-mode fiber, no significant light from the B or C components
210: enters the fringe camera while observing A \citep[the $K$-band
211: projected FWHM size of the KI single-mode fiber is 50 mas, so the B
212: component is approximately five e-folding lengths away from A at the
213: epoch of our observations;][D2003]{Colavita2003}. Consequently no
214: special provisions for extra light are necessary in processing KI
215: observations of V773~Tau~A.
216:
217:
218: %% KI V2 table (table1.tex)
219: \begin{table}
220: \dummytable\label{tab:V2Table}
221: \end{table}
222:
223: Our KI V$^2$ observations are summarized in Table \ref{tab:V2Table}.
224: The V$^2$ data set is depicted in Figure~\ref{fig:v2Data}, along with
225: V$^2$ predictions from our ``Hybrid'' orbit model (discussed below).
226:
227: \begin{figure}[ph]
228: \epsscale{0.9}
229: %%\plotone{figures/fit.v2.strip.ps}
230: \plotone{f1.eps}
231: %%\includegraphics[angle=-90,width=16cm]{figures/fit.v2.strip.ps}
232: %%\includegraphics[angle=-90,width=16cm]{f2.eps}
233: \caption{KI Data/Model Comparisons for V773~Tau~A V$^2$ Observations.
234: Here we give comparisons for our five epochs of KI V$^2$ observations
235: and V$^2$ predictions from our ``Hybrid'' orbital model
236: (Table~\ref{tab:orbit}). In each case the data and model are shown.
237: \label{fig:v2Data}}
238: \end{figure}
239:
240:
241: \paragraph{VLBI Observations}
242: P1996 observed V773~Tau~A with the VLBA, and with similar spatial
243: frequencies as our KI observations ($\sim$ 50 M$\lambda$) resolved the
244: typical 2 mas separation between the A components. P1996 Figure 6
245: depicts a VLBA contour map of V773~Tau~A at a particular epoch (1992
246: Sept 11), and clearly shows a binary source structure. P1996
247: interpreted this binary morphology as radio flare emission from the
248: two subsystem components. P1996 argued that the radio emission is
249: coincident with the stars (to within a ``few'' stellar diameters), so
250: a component separation interpreted as at-epoch relative astrometry
251: was estimated by P1996.
252:
253: A team led by one of us (M.M.) has studied the physics of V773~Tau~A
254: radio flares. In addition to previous results
255: \citep[e.g.][]{Massi2002,Massi2006}, V773~Tau~A was observed using the
256: VLBA in coordination with the Effelsberg 100\,m radio telescope
257: (herein VLBA+EB). These 8.4-GHz (X-band) observations were conducted
258: around apastron on seven consecutive days (2004 March 11-17
259: inclusive). As in the case of P1996, analysis of these observations
260: revealed what appears to be both A components simultaneously
261: exhibiting radio flares in most of these observations. However some
262: of these ``double-flare'' observations show complex emission
263: morphologies, complicating estimation of the binary relative
264: separation. Reviewing the maps from these seven epochs, we decided to
265: use astrometry from observations where there was significant
266: (i.e.~that the peak emission was greater than six sigma above the rms
267: noise level) emission from both components, and the component
268: structures were largely unresolved compared to the VLBA synthetic
269: beam. These criteria resulted in our retaining three of the seven
270: 2004 epochs for orbital analysis purposes; the astrometry resulting
271: from these observations is summarized in
272: Table~\ref{tab:VLBIastrometry}. A sample image constructed from one
273: of the three astrometric epochs (2004 March 15) is shown in
274: Figure~\ref{fig:VLBAmap}. The astrometric results of both the P1996
275: and these VLBI observations are depicted in
276: Figure~\ref{fig:orbitPlot}; these observations agree well with our
277: orbit model for V773~Tau~A described below. The radio morphologies
278: seen in the full VLBI+EB data set will analyzed in a forthcoming
279: publication (Massi et al 2007, in prep).
280:
281: %% VLBI astrometry table (table2.tex)
282: \begin{table}
283: \dummytable\label{tab:VLBIastrometry}
284: \end{table}
285:
286: \begin{figure}[ph]
287: \epsscale{0.6}
288: %%\plotone{figures/V773_Tau_E_nolabel_stars2.ps}
289: \plotone{f2.eps}
290: \caption{Sample VLBA Map from 2004 March 15. Here we give a contour
291: map of the total intensity continuum emission toward V773~Tau~A from
292: VLBA+EB observations of 2004 March 15. Contours are at -3, 3, 4.5, 6,
293: 7.5, 9, 12, and 15 $\sigma$ of the rms noise in the image (0.1
294: mJy/beam). The FWHM beam size (lower left) is 1.36$\times $0.50~mas
295: at a position angle of $-$11.6$^{\circ}$. The peak flux density in
296: the image is 0.55~mJy/beam. V773~Tau~A is offset in these data and
297: image as a calibrator (J0403+2600, not shown) served as the phase
298: center. Heavy crosses are rendered at the predicted stellar
299: separation based on our ``V$^2$ \& RV'' orbit from
300: Table~\ref{tab:orbit} (i.e. no radio data included in the orbit
301: modeling), registered to the center of the apparent primary position.
302: The sizes of the crosses indicate the uncertainty in the separation
303: estimate ($\sim$ 0.1 mas per axis).
304: \label{fig:VLBAmap}}
305: \end{figure}
306:
307:
308:
309: \begin{figure}[p]
310: \epsscale{0.8}
311: %%\plotone{figures/v773TauA.orbit.ps}
312: \plotone{f3.eps}
313: \caption{Orbit of V773~Tau~A as derived in our ``Hybrid'' solution
314: (Table~\ref{tab:orbit}). Upper Panel: the relative visual orbit model
315: of V773~Tau~A is shown, with the primary and secondary objects
316: rendered at T$_0$ (periastron). The specific epochs where we have KI
317: V$^2$ phase coverage are indicated on the relative orbit (they are not
318: separation vector estimates), and the VLBI separations from P1996 (in
319: blue) and this work (in red) are also shown. Component diameter
320: values are estimated and rendered to scale. Lower Panel: the
321: double-lined radial velocity orbit model and data described here.
322: \label{fig:orbitPlot}}
323: \end{figure}
324:
325:
326: \paragraph{RV Observations}
327: W1995 established V773~Tau~A as a double-lined spectroscopic binary,
328: and published an orbital solution. RV monitoring of V773~Tau with the
329: CfA ``Digital Speedometers'' \citep{Latham1992} was carried out at the
330: Oak Ridge Observatory (Harvard, Massachusetts), the F.L.~Whipple
331: Observatory (Mount Hopkins, Arizona), and the MMT Observatory (also on
332: Mount Hopkins, Arizona). The nearly identical instruments at these
333: facilities record a single echelle order and provide a wavelength
334: coverage of 45~\AA\ centered at 5187~\AA, with a resolving power of
335: $\lambda/\Delta\lambda \approx 35,\!000$. Wavelength calibration was
336: achieved in the usual manner through exposures of a Th-Ar lamp taken
337: before and after each science exposure. The velocity zero point was
338: monitored by means of exposures of the dusk and dawn sky, and small
339: run-to-run corrections were applied to the velocities derived below as
340: described by \cite{Stefanik:99}.
341:
342: Our spectroscopic observations of V773~Tau began roughly at the same
343: time as the W1995 data set and have continued to the present. In total
344: we have collected 52 spectra spanning the period 1991 Sep 25 to 2007
345: Jan 7, with signal-to-noise ratios of 16--35 per resolution element of
346: 8.5~\kms. The Radial Velocities (RVs) derived from these observations
347: are summarized in Table~\ref{tab:RVtable}. RVs were derived using the
348: two-dimensional cross-correlation technique TODCOR \citep{Zucker:94},
349: with templates adopted from a large library of calculated spectra
350: based on Kurucz model atmospheres \citep[see also][]{Nordstrom:94,
351: Latham:02}. The best-matching templates were chosen by running large
352: grids of cross-correlations as described by \citet[][]{Torres2003},
353: which also yielded estimates of some of the stellar parameters. For an
354: assumed Taurus solar metallicity \citep[see][]{Padgett1996} and
355: surface gravities appropriate for these stars ($\log g \approx 4.0$;
356: see Table~5), we derived effective temperatures of $4900 \pm 150$~K
357: and $4740 \pm 200$~K for Aa and Ab, respectively. The adopted errors
358: are conservative estimates to account for the abundance
359: uncertainty. Formal errors are 50~K smaller, and are based on the
360: scatter from the different spectra. Using the same methods we
361: estimated the projected rotational velocity for the primary star to be
362: $v \sin i = 38 \pm 4$~\kms, in good agreement with the determination
363: by W1995 (41.4~\kms; no uncertainty reported). We were not able to
364: determine the $v \sin i$ for the fainter secondary, possibly due to
365: the relatively low signal-to-noise ratios of our spectra. The
366: templates for our radial-velocity determinations were selected to have
367: temperatures of 5000~K and 4750~K (the nearest values in our template
368: library) for Aa and Ab, and rotational velocities of 40~\kms\ for both
369: stars. The light ratio between the secondary and the primary was
370: determined with TODCOR by leaving it as an additional free parameter
371: during the velocity determinations, as described by \cite{Zucker:94}.
372: We obtained $\ell_{Ab}$/$\ell_{Aa}$ = 0.37 $\pm$ 0.03 at a mean
373: wavelength of 5187~\AA.
374:
375: The presence of two additional spatially unresolved components in the
376: V773~Tau system may in principle affect both our determination of the
377: stellar parameters of Aa and Ab and their radial velocities. While the
378: infrared companion V773~Tau~C \citep[D2003,][]{Woitas2003} is much too
379: faint to cause any contamination, V773~Tau~B is estimated to be
380: $\sim$1.7 magnitudes fainter than the spectroscopic binary in the
381: optical (D2003), and could possibly have an effect. We attempted to
382: detect the lines of this component in our spectra using an extension
383: of TODCOR to three dimensions \citep[see][]{Zucker:95}, but were
384: unsuccessful. On this basis we conclude that it does not significantly
385: affect our results above. However, its presence does complicate the
386: analysis of the RV data for the spectroscopic binary in that the
387: motion in the A-B orbit needs to be taken into account. We discuss
388: this further below, and in a forthcoming publication (Torres et al
389: 2007, in prep).
390:
391:
392:
393:
394: %% RV table (table3.tex)
395: \begin{table}
396: \dummytable\label{tab:RVtable}
397: \end{table}
398:
399:
400: \paragraph{Orbital Solution}
401: We have combined the astrometric (KI, P1996, VLBA+EB) and radial
402: velocity data described above to estimate the visual and physical
403: (i.e.~three-dimensional) orbits for V773~Tau~A. In particular,
404: methods for modeling the wide-band visibility data used in this
405: analysis are described in \citet{Boden2000}, and not repeated here.
406: Our ``Hybrid'' orbital solution (Table~\ref{tab:orbit}) is depicted in
407: Figure~\ref{fig:orbitPlot}. The upper panel depicts the relative
408: visual orbit model, with the primary (Aa) component rendered at the
409: origin, and the secondary (Ab) component rendered at periastron. The
410: KI V$^2$ data phase coverage is indicated on the visual orbit with
411: points (they are {\em not} separation vectors); the phase coverage of
412: the V$^2$ data is sparse relative to other similar analyses
413: \citep[e.g.][]{Boden2000}, but phase coverage provided by VLBI
414: observations from P1996 and this work complements the V$^2$ coverage.
415: Further, the incorporation of the VLBI separations also breaks the
416: 180$^\circ$-$\Omega$ degeneracy inherent in $V^2$-only analyses
417: \citep[e.g.][]{Boden2000,Boden2005a}, so the A-orbit rendered in
418: Figure~\ref{fig:orbitPlot} is indeed as it appears on the sky. The
419: apparent size of the V773~Tau~A components are estimated
420: (\S~\ref{sec:SED}, Table~\ref{tab:physics}) and rendered to scale;
421: these same diameter values are used in the orbital modeling. The
422: lower panel depicts radial velocity curves from the ``Hybrid'' orbit
423: model and radial velocity data described above.
424:
425:
426: V773~Tau~A presents two significant challenges for RV orbit
427: estimation. The first is that both components are relatively rapid
428: rotators: W1995 found the $v \sin i$ for both stars to be around 40 km
429: s$^{-1}$; this rotational broadening is seen in our spectra as well,
430: and it significantly degrades the precision of the velocities.
431: Secondly, and most notably, is the A-barycenter motion that results
432: from the A-B orbit. D2003 Figure 4 depicts the significant A-B
433: orbital motion over the time since the detection of V773 Tau B, and
434: our RV data set (starting in 1991) not only spans essentially all of
435: this period, but includes the A-B periastron passage expected in 1996
436: (D2003). It is therefore important to account for the A-B orbital
437: motion in the modeling of the A component RVs. Modeling of the A-B
438: orbit would not have been possible at the time of W1995, but becomes
439: feasible with the present data set.
440:
441: A separate RV-only orbital solution was carried out in which, in
442: addition to solving for preliminary elements of the A orbit, we
443: modeled the Keplerian motion of A around the A-B barycenter. In this
444: analysis we held fixed some of the elements of the outer orbit to the
445: values given by D2003 (specifically, the period, eccentricity,
446: longitude of periastron, and time of periastron passage) and we solved
447: for the velocity semi-amplitude of the barycenter of A ($K_A = 6.11
448: \pm 0.34$ km s$^{-1}$). This motion was then removed from our
449: original velocities prior to combining them with the KI and VLBI data.
450: A more complete treatment would incorporate all the astrometric
451: measurements for the A-B orbit (e.g., D2003) along with the velocities
452: and astrometry for the inner subsystem in a full solution solving for
453: both orbits simultaneously. Such a study is beyond the scope of the
454: present work, but will be described in a forthcoming publication
455: (Torres et al 2007, in prep).
456:
457:
458: There are significant differences between several of our orbital
459: parameters for V773~Tau~A and the corresponding values from W1995,
460: most notably in the period and K$_{Aa}$ (see Table~\ref{tab:orbit}).
461: We attribute these differences to a combination of at least three
462: effects: {\it i}) limited phase coverage of the A orbit by W1995,
463: particularly near A apastron; {\it ii}) possible biases in the W1995
464: RVs due to line blending, exacerbated by the rapid rotation of Aa and
465: Ab along with the lower spectral resolution of that study compared to
466: ours; and {\it iii}) residual effects from the un-modeled motion in the
467: outer orbit in W1995's solution.
468:
469:
470:
471: P1996 argue that their resolved VLBI image of V773~Tau~A measures the
472: relative component positions to within a ``few'' stellar diameters
473: (the primary apparent diameter is approximately 0.15 mas; see
474: \S~\ref{sec:SED}). However, it is important to recognize that VLBI
475: and KI V$^2$ data measure different emission mechanisms, and further,
476: one would not expect individual radio flares to be centered on the
477: stellar photospheres (measured by the KI V$^2$ data). This potential
478: photosphere/radio offset motivated us to approach integrating the
479: radio data into the orbit model cautiously. In our initial screening
480: of the VLBI relative astrometry we compared the radio observations
481: with position predictions from a V$^2$ and RV orbit model
482: (Table~\ref{tab:orbit}). We found the model reasonably reproduced the
483: observed radio separations with a typical residual of 0.2 -- 0.3 mas
484: ($\approx$ 4 R$_*$), slightly larger than the formal errors expected
485: from the VLBI synthsized beam and the snr in the image, but in good
486: agreement with the expected IR/radio offsets. The agreement we see
487: between the radio separations and the V$^2$ \& RV orbit model makes it
488: clear that the VLBI data do contain useful information on the
489: component separations. We therefore felt confident in integrating the
490: KI V$^2$ and four VLBI observations (including the P1996 epoch) into
491: our orbit model estimation. An analysis of the radio mophologies seen
492: in the VLBA+EB observations will be presented in a future publication
493: (Massi et al 2007, in prep).
494:
495:
496: Figure \ref{fig:v2Data} depicts direct comparisons between our KI
497: V$^2$ observations and predictions from our V773~Tau~A ``Hybrid''
498: orbit model from Table \ref{tab:orbit} (the five nights of data are
499: each rendered in separate sub-panels). The model is seen to be in
500: good agreement with the KI data. Further, Figure~\ref{fig:orbitPlot}
501: shows a direct comparison between the VLBA separations and the
502: ``Hybrid'' visual orbit model. Again, the body of the VLBI data are
503: in good agreement with the orbit model, and the VLBI phase coverage
504: complements the phase coverage provided by the KI V$^2$ data.
505:
506:
507: Orbital parameter estimates for the V773~Tau~A subsystem are
508: summarized in Table~\ref{tab:orbit}. Included for relative comparison
509: are the orbital parameters determined by W1995, our RV double-lined
510: orbit (``RV''), the KI V$^2$ \& RV solution (``V$^2$ \& RV''), and the
511: full integrated solution including the RV, V$^2$, and VLBA data
512: (``Hybrid''). The ``Hybrid'' designation for the composite orbit
513: model recognizes its unique inclusion of both near-IR and radio
514: interferometry data; this model is our favored orbit solution and is
515: used in the remaining analysis. Again note that the A
516: barycenter velocity is artificially near zero in our orbital solutions
517: as the RV measurements have been corrected for our model of the A-B
518: orbital motion.
519:
520:
521: \begin{deluxetable}{lcccc}
522: %%%\tabletypesize{\footnotesize}
523: \tablecolumns{5}
524: \tablewidth{0pc}
525: %%\rotate
526:
527: \tablecaption{V773~Tau~A Orbital Parameters.
528: \label{tab:orbit}
529: }
530:
531: \tablecomments{Summarized here are
532: orbital parameters for the V773~Tau~A subsystem as determined by W1995
533: and present results. We give three separate fits to our data sets: RV
534: only, KI V$^2$ integrated with RV (``V$^2$ \& RV''), and KI V$^2$,
535: VLBI, and RV (``Hybrid'') . Barycenter velocity ($\gamma$) estimates
536: (shown in italics) are artificially near zero because we have
537: estimated and removed a model of the A-B orbital motion (see
538: discussion in text). $\Omega$ is quoted in a position angle
539: convention.}
540:
541:
542: \tablehead{
543: \colhead{Orbital} & \colhead{W1995} & \multicolumn{3}{c}{This Work} \\
544: \colhead{Parameter} & & \colhead{RV} & \colhead{V$^2$ \& RV} & \colhead{Hybrid}
545: }
546: \startdata
547: Period (d) & 51.075 $\pm$ 0.018 & 51.1055 $\pm$ 0.0023 & 51.1033 $\pm$ 0.0022 & 51.1039 $\pm$ 0.0021 \\
548: T$_{0}$ (MJD) & 49330.44 $\pm$ 0.56 & 53059.87 $\pm$ 0.36 & 53059.50 $\pm$ 0.35 & 53059.73 $\pm$ 0.33 \\
549: $e$ & 0.267 $\pm$ 0.016 & 0.275 $\pm$ 0.012 & 0.2708 $\pm$ 0.0094 & 0.2717 $\pm$ 0.0085 \\
550: K$_{Aa}$ (km s$^{-1}$) & 32.6 $\pm$ 0.7 & 35.95 $\pm$ 0.53 & 35.96 $\pm$ 0.53 & 35.90 $\pm$ 0.53 \\
551: K$_{Ab}$ (km s$^{-1}$) & 43.1 $\pm$ 2.1 & 41.6 $\pm$ 1.4 & 41.5 $\pm$ 1.4 & 41.5 $\pm$ 1.4 \\
552: $\gamma$ (km s$^{-1}$) & 13.8 $\pm$ 0.9 & {\em 0.00 $\pm$ 0.31} & {\em 0.05 $\pm$ 0.31} & {\em 0.02 $\pm$ 0.32} \\
553: $\omega_{Aa}$ (deg) & 10.5 $\pm$ 5.1 & 5.1 $\pm$ 2.6 & 3.4 $\pm$ 2.5 & 4.6 $\pm$ 2.4 \\
554: $\Omega$ (deg) & & & 61.5 $\pm$ 2.5 & 63.5 $\pm$ 1.7 \\
555: $i$ (deg) & & & 65.9 $\pm$ 2.8 & 66.0 $\pm$ 2.4 \\
556: $a$ (mas) & & & 2.777 $\pm$ 0.056 & 2.811 $\pm$ 0.047 \\
557: $\Delta$ $K$ (mag) & & & 0.559 $\pm$ 0.082 & 0.551 $\pm$ 0.078 \\
558: %%%$\Delta $ (mag) & {\em 1.11 $\pm$ 0.15} & & & \\
559: %%$\chi^2$/DOF & & {\em 1.0} & {\em 1.0} & 1.06 \\
560: %%$\overline{|R_{V^2}|}$/$\sigma_{V^2}$ & & 0.059/0.11 & & 0.059/0.11 \\
561: %%$\overline{|R_{RV}|}$/$\sigma_{RV}$ (km s$^{-1}$) & & & 0.80/1.16 & 0.80/1.16 \\
562: \hline
563: \enddata
564:
565: \end{deluxetable}
566:
567:
568:
569: \section{V773~Tau~A physical properties}
570: \label{sec:physics}
571:
572: The orbital parameters from Table~\ref{tab:orbit} allow us to directly
573: compute many of the physical properties of the V773~Tau~A subsystem
574: and its components. Physical parameters derived from our ``Hybrid''
575: integrated visual/spectroscopic orbit are summarized in Table
576: \ref{tab:physics}. The dynamical masses resulting from our orbit
577: model for the two A components are 1.54 $\pm$ 0.14 and 1.332 $\pm$
578: 0.097 M$_{\sun}$ for Aa and Ab respectively. This primary mass is in
579: acceptable agreement with inferences by \citet[herein
580: WG2001]{White2001} and \citet{Palla2001} based on radiometric
581: properties (i.e.~luminosity and effective temperaure). Because it
582: bears on the A-B orbit analysis (Torres et al 2007, in prep) we
583: estimate the A subsystem mass as 2.87 $\pm$ 0.18 M$_{\sun}$.
584:
585:
586: L1999 estimated a V773~Tau system distance of 148.4 $\pm$ 5.5 pc. The
587: distance determination to V773~Tau~A derived from our visual and
588: physical orbital solution is 136.2 $\pm$ 3.7 pc, corresponding to an
589: orbital parallax of 7.34 $\pm$ 0.20 mas, and consistent with the L1999
590: result at 8.1\% and 1.9-sigma. L1999 describe their distance estimate
591: error as ``formal'', and it apparently did not account for the orbital
592: motion of A in their data reduction. With these caveats we believe
593: the L1999 distance estimate and ours are in acceptable agreement.
594:
595:
596: At the distance of V773~Tau, neither of the A subsystem components are
597: significantly resolved by the KI K-band fringe spacing, and we must
598: resort to model diameters for the components. We have estimated the
599: V773~Tau~A component apparent diameters through spectral energy
600: distribution modeling; details of the spectral energy distribution
601: modeling are given in \S\ref{sec:SED}. We find apparent diameters of
602: 0.152 $\pm$ 0.014 and 0.119 $\pm$ 0.011 mas for the Aa and Ab
603: components respectively. With our system distance estimate these
604: estimated diameters correspond to physical radii of 2.22 $\pm$ 0.20
605: and 1.74 $\pm$ 0.19 R$_\sun$, and (combined with the mass estimates)
606: log surface gravities of 3.930 $\pm$ 0.094 and 4.081 $\pm$ 0.088 for
607: the Aa and Ab components respectively. Similarly we have estimated
608: absolute magnitudes in $V$ and $K$, and a $V-K$ color index for each
609: of the components (Table~\ref{tab:physics}), but these are of limited
610: precision as they are dominated by the uncertainty in the extinction
611: to the system (\S~\ref{sec:SED}).
612:
613:
614:
615: \begin{deluxetable}{ccc}
616: \tabletypesize{\small} \tablecolumns{3} \tablewidth{0pc}
617:
618: \tablecaption{V773~Tau~A Physical Parameters.
619: \label{tab:physics}
620: }
621:
622: \tablecomments{Summarized here are
623: physical (and related) parameters for the V773~Tau~A subsystem as
624: derived primarily from the ``Hybrid'' orbit solution
625: (Table~\ref{tab:orbit}) and radiometric modeling (\S~\ref{sec:SED}).}
626:
627:
628: \tablehead{
629: \colhead{Physical} & \colhead{Primary} & \colhead{Secondary} \\
630: \colhead{Parameter} & \colhead{(Aa)} & \colhead{(Ab)} \\
631: }
632:
633: \startdata
634: a (AU) & 0.1777 $\pm$ 0.0042 & 0.2053 $\pm$ 0.0070 \\
635: Mass (M$_{\sun}$) & 1.54 $\pm$ 0.14 & 1.332 $\pm$ 0.097 \\
636: \cline{2-3}
637: %%Sp Type () & \multicolumn{2}{c}{K0 IV} \\
638: System Mass (M$_{\sun}$) & \multicolumn{2}{c}{2.87 $\pm$ 0.18} \\
639: System Distance (pc) & \multicolumn{2}{c}{136.2 $\pm$ 3.7} \\
640: $\pi_{orb}$ (mas) & \multicolumn{2}{c}{7.34 $\pm$ 0.20} \\
641: \cline{2-3}
642: $T_{\rm eff}$ (K) & 4900 $\pm$ 150 & 4740 $\pm$ 200 \\
643: Model Diameter (mas) & 0.152 $\pm$ 0.014 & 0.119 $\pm$ 0.011 \\
644: Bolometric Flux (10$^{-9}$erg cm$^{-2}$ s$^{-1}$)
645: & 4.44 $\pm$ 0.40 & 2.38 $\pm$ 0.15 \\
646: Luminosity (L$_{\sun}$)
647: & 2.56 $\pm$ 0.35 & 1.37 $\pm$ 0.15 \\
648: Radius (R$_{\sun}$) & 2.22 $\pm$ 0.20 & 1.74 $\pm$ 0.19 \\
649: $\log g$ & 3.930 $\pm$ 0.094 & 4.081 $\pm$ 0.088 \\
650:
651: M$_K$ (mag) & 1.75 $\pm$ 0.12 & 2.30 $\pm$ 0.13 \\
652: M$_V$ (mag) & 3.90 $\pm$ 0.16 & 4.98 $\pm$ 0.16 \\
653: $V$-$K$ (mag) & 2.15 $\pm$ 0.14 & 2.68 $\pm$ 0.14 \\
654: \enddata
655:
656: \end{deluxetable}
657:
658:
659: \subsection{Spectral Energy Distribution Modeling}
660: \label{sec:SED}
661: Because interferometric observations potentially resolve the stellar
662: components in a binary system, we always construct spectral energy
663: distribution (SED) models for binary systems to estimate a priori
664: apparent diameters. Sample results of our SED modeling of V773~Tau~A
665: are shown in Figure~\ref{fig:V773TauA_SED}. Our modeling is based on
666: visible and near-IR photometry on A from WG2001 and D2003 respectively
667: (summarized in Table~\ref{tab:Fluxtable}), and component flux ratios
668: from W1995 and our visible spectroscopy and near-IR KI V$^2$ data
669: (discussed above). Using a custom two-component SED modeling code, we
670: have modeled the A subsystem flux and ratios using solar-abundance,
671: intermediate surface gravity SED templates from \cite{KuruczModels}
672: and \citet{Lejeune1997,Lejeune1998}. We model the flux and ratio data
673: with a large grid of SED templates for each component ranging in
674: effective temperature (spectral type) from 4250 -- 5500 K, surface
675: gravity of log g = 4.0, and including extinction as a free variable in
676: the parameter estimation. Constrained by the ratio data, this SED
677: modeling preferred component temperature {\em differences} over a
678: reasonably narrow range ($\Delta$ $T_{\rm eff} \sim$ 150--250 K)
679: consistent with our spectroscopic study. But because the component
680: colors are somewhat degenerate with reddening from dust extinction we
681: found acceptable solutions over a relatively broad range ($\sim$
682: 500~K) of component temperatures. We have therefore adopted templates
683: corresponding to the spectroscopic temperatures (4900 and 4740 K for
684: Aa and Ab respectively). These temperatures acceptably agree with the
685: composite temperature estimated by WG2001, but the hotter primary in
686: our modeling prefers a slightly higher extinction value (A$_V$ = 1.80
687: $\pm$ 0.15 vs 1.39 $\pm$ 0.17 from WG2001). The resulting component
688: bolometric fluxes and luminosities (the second based on our distance
689: determination) are summarized in Table~\ref{tab:physics}. In
690: particular we find luminosities of 2.56 $\pm$ 0.35 and 1.37 $\pm$ 0.15
691: L$_\sun$ for Aa and Ab respectively; these values are in good
692: agreement with the range of previous component luminosity estimates
693: from W1995, \citet{Ghez1997}, and D2003.
694:
695:
696:
697: %% Photometry table (table6.tex)
698: \begin{table}
699: \dummytable\label{tab:Fluxtable}
700: \end{table}
701:
702:
703: \begin{figure}[ph]
704: %%\includegraphics[angle=-90,width=14.5cm]{figures/v773TauA.binary.sed.pg1.ps}\\
705: \includegraphics[angle=-90,width=14.5cm]{f4a.eps}\\
706: %%\includegraphics[angle=-90,width=15.5cm]{figures/v773TauA.binary.sed.pg7.ps}\\
707: \includegraphics[angle=-90,width=15.5cm]{f4b.eps}\\
708: \caption{Sample Spectral Energy Distribution Model for V773~Tau~A.
709: Here SED templates from \citet{Lejeune1997,Lejeune1998} have been used
710: to simultaneously model published flux measurements (from WG2001 \&
711: D2003) and flux ratio estimates (Table \ref{tab:orbit}; bottom) from
712: our KI V$^2$ and spectroscopic measurements.
713: \label{fig:V773TauA_SED}}
714: \end{figure}
715:
716:
717: \section{Discussion}
718: \label{sec:discussion}
719:
720: \subsection{Comparison with stellar evolution models}
721: \label{sec:modelcomp}
722:
723: \citet{Hillenbrand2004} and \citet{Mathieu2007} have recently
724: summarized comparisons between PMS stellar models and PMS stars with
725: dynamical mass determinations. In particular \citet{Hillenbrand2004}
726: has shown that PMS models are in good agreement with
727: dynamically-determined masses above 1.2 M$_\sun$. However, below 1.2
728: M$_\sun$ the existing models do a poorer job of matching observed
729: component properties, tending to predict hotter and more luminous
730: stars for a given mass.
731:
732:
733: Figure~\ref{fig:modelCompare} presents theoretical HR-diagram
734: (luminosity/$T_{\rm eff}$) comparisons between our determinations of
735: V773~Tau~A component radiometric properties and solar-composition PMS
736: models from \citet{Siess2000}, Yonsei-Yale \citep[herein
737: Y$^2$]{Demarque2004}, \citet{DAntona1997}, and \citet{Palla1999}.
738: \citep[Popular models by][were excluded only because they do not cover
739: the mass range for these stars.]{Baraffe1998} These models span a wide
740: range of physical ingredients and assumptions, some more realistic
741: than others for the particular mass range of interest. In these
742: figures evolutionary tracks at fixed mass between 1.2 and 1.8 M$_\sun$
743: (in steps of 0.1 M$_\sun$ as available from the relevant models) are
744: shown in solid lines, with heavy lines indicating tracks at masses
745: close to dynamical masses from Table~\ref{tab:physics}. Superimposed
746: in dotted lines are isochrones at a relevant range of ages (1, 3, 5,
747: 7, and 10 Myrs as available from the relevant models). Finally, the
748: Y$^2$ models support computing tracks at user-defined masses; tracks
749: at the estimated component masses from Table~\ref{tab:physics} are
750: shown for Aa (1.54 M$_\sun$, blue) and Ab (1.332 M$_\sun$, red).
751:
752:
753: As expected from \citet{Hillenbrand2004}, all the surveyed PMS models
754: are in reasonable (e.g. two-sigma) agreement for the V773~Tau~A
755: component mass range. The models indicate that the V773~Tau~A
756: components are still in their quasi-isothermal (``Hayashi'')
757: contraction phase. More quantitatively some model predictions seem to
758: match the A component properties better than others. For instance,
759: the solar-abundance \citet{Siess2000} (upper left) and
760: \citet{Palla1999} (lower-right) tracks near the dynamical masses tend
761: to under-predict our estimated component temperatures and
762: luminosities. In particular for the \citet{Siess2000} models, this is
763: similar to our previous results for HD~98800~B \citep{Boden2005b}.
764: This apparent temperature discrepancy seems to reinforce a suggestion
765: by \citet{Montalban2004} that the solar-abundance \citet{Siess2000}
766: models appear to be systematically cool compared with other models.
767: Conversely, the \citet{DAntona1997} tracks (lower-left) at our mass
768: values appear to slightly over-predict our component temperatures and
769: luminosity estimates. The Y$^2$ model predictions (upper-right;
770: including tracks for the specific masses of the components) are seen
771: to match our parameter determinations very well.
772:
773: \begin{figure}[tp]
774: \epsscale{1.1}
775: %%\plottwo{figures/v773TauA.Siesscomparisons.solarZ.eps}{figures/v773TauA.Y2comparisons.solarZ.eps}\\
776: \plottwo{f5a.eps}{f5b.eps}\\
777: %%\plottwo{figures/v773TauA.DM97comparisons.solarZ.eps}{figures/v773TauA.PS99comparisons.solarZ.eps}
778: \plottwo{f5c.eps}{f5d.eps}
779: \caption{V773~Tau~A Components Compared With PMS Models. Here we show
780: the V773~Tau~A components in theoretical HR diagrams
781: (luminosity/$T_{\rm eff}$), and solar-composition PMS evolutionary
782: tracks and isochrones by \citet{Siess2000} (top left), Y$^2$
783: \citep{Demarque2004} (top right), \cite{DAntona1997} (bottom left),
784: and \citet{Palla1999} (bottom right). All panels show available
785: tracks in the 1.2 -- 1.8 M$_{\sun}$ range, and isochrones at 1, 3, 5,
786: 7, and 10 Myrs. In the each panel we highlight the available tracks
787: that best corresponding to our dynamical mass values; the Y$^2$ models
788: uniquely allow us to compute tracks at these values. There is
789: excellent agreement between our component parameter estimates and the
790: predictions of the Y$^2$ tracks; other models included here are less
791: consistent with the inferred component parameters.
792: \label{fig:modelCompare}}
793: \end{figure}
794:
795:
796: Figure~\ref{fig:modelCompare2} gives a final model comparison.
797: Recently \citet{Montalban2006} have updated \citet{DAntona1997} models
798: with a revised ``2D radiative-hydrodynamic'' (2D RHD) convection
799: treatment. Figure~\ref{fig:modelCompare2} shows the result of this
800: updated modeling with tracks from \citet[][in black]{DAntona1997} and
801: \citet[][in blue]{Montalban2006}; the presentation is similar to that
802: given in Figure~\ref{fig:modelCompare}, except we have displayed only
803: mass tracks near the component dynamical mass estimates (1.3 and 1.5
804: M$_\sun$). The 2D RHD convective treatment increases the convective
805: ``efficiency'' and energy transport in the atmosphere, moving the
806: tracks toward lower temperatures during the ``Hayashi'' phase, and
807: brings the model predictions into excellent agreement with the
808: component parameters presented here.
809:
810: Isochrones from the models we considered would predict a range of
811: component ages from 2 -- 7 Myrs, but restricting attention to the
812: \citet{Montalban2006} and Y$^2$ models, a more plausible age estimate
813: is 3 $\pm$ 1 Myrs, with both components appearing coeval to within
814: present uncertainties. Such a system age is broadly consistent with
815: age estimates for the Taurus population using a variety of
816: evolutionary models \citep[e.g.~WG2001;][]{Palla2002,Luhman2004}. More
817: specifically this age estimate would put V773~Tau~A in the putative
818: intermediate-age population for Taurus as defined by
819: \citet{Palla2002}.
820:
821:
822:
823: \begin{figure}[p]
824: \epsscale{0.7}
825: %%\plotone{figures/v773TauA.DM97comparisons3.solarZ.eps}
826: \plotone{f6.eps}
827: \caption{V773~Tau~A Components Compared With PMS Models From
828: \citet{DAntona1997} and \citet{Montalban2006}. The presentation is
829: similar to Fig~\ref{fig:modelCompare}, but here we focus on tracks
830: near the A-component dynamical masses (1.3 and 1.5 M$_\sun$). The
831: \citet[][tracks in blue, isochrones in heavy dotted
832: lines]{Montalban2006} models updated the older \citet[][tracks in
833: black, isochrones in light dotted lines]{DAntona1997} models with a
834: revised convection model, bringing the updated models into excellent
835: agreement with the mass and radiometric parameters estimated here.
836: \label{fig:modelCompare2}}
837: \end{figure}
838:
839:
840: \subsection{Summary and Conclusions}
841: \label{sec:conclusions}
842:
843: We have presented new interferometric (Tables~\ref{tab:V2Table} and
844: \ref{tab:VLBIastrometry}) and RV data (Table~\ref{tab:RVtable}) on the
845: PMS binary subsystem V773~Tau~A. Modeling these data in conjunction
846: with older VLBI astrometry from P1996 allows us to estimate the visual
847: and physical orbits for V773~Tau~A, physical parameters of the
848: components, and place significant constraints on PMS stellar evolution
849: models. We characterize the V773~Tau~A orbit model presented here as
850: {\em preliminary}; we plan additional RV, VLBA, and KI observations
851: that will further refine the physical orbit model for the A subsystem
852: and constrain the component parameters. Further, ongoing RV and
853: astrometric observations will continue to in order to probe the A-B
854: orbit, and this in turn will refine the A subsystem RVs and orbit.
855: Details of A-B orbit modeling will be documented in a future
856: publication (Torres et al 2007).
857:
858:
859:
860: The V773~Tau physical parameters inferred from our orbit model are in
861: good agreement with previous results. The most notable difference in
862: our results with previous work is in the system distance; our distance
863: estimate is at modest variance with the VLBA parallax from L1999.
864: However, the L1999 data reduction did not account for the binarity of
865: A which is clearly resolved in multiple epochs (P1996, this work).
866: With that caveat we believe the L1999 distance and ours are in
867: acceptable agreement. Because our distance estimate depends only on
868: the apparent and physical orbits, we believe it is likely accurate at
869: its stated uncertainty. We estimate our relative component mass
870: errors at 9.1\% and 7.3\% for Aa and Ab respectively, making them
871: comparable with other PMS stellar dynamical mass determinations
872: \citep{Hillenbrand2004,Mathieu2007}.
873:
874:
875: It is noteworthy that the P1996 and present VLBI data agree well with
876: the A-orbit model presented here (including a model that excludes the
877: VLBI data themselves, e.g.~see Figure~\ref{fig:VLBAmap}). Narrowly
878: this would seem to support the P1996 interpretation that both A
879: components are active radio emitters. More broadly, the ability to
880: compare the radio morphology with at-epoch estimates of the component
881: positions would seem to present an opportunity to constrain the radio
882: emission and magnetic field topology in these stars. A more complete
883: analysis of the VLBI data in this context will be made in a
884: forthcoming publication (Massi et al 2007, in prep).
885:
886:
887: V773~Tau~A radiometric estimates must necessarily account for (and are
888: complicated by) extinction. Our estimate of the extinction is in
889: reasonable agreement with earlier work
890: \citep[e.g. W1995;][]{Ghez1997,White2001}, but prefers a slightly
891: higher extinction when constrained by the spectroscopic temperatures
892: (which are quasi-independent of the extinction). With this
893: uncertainty in mind, we have made relatively conservative estimates of
894: the errors in the radiometric properties.
895:
896:
897: Even with the relatively conservative error estimates, the combination
898: of our component dynamical mass and radiometric parameter estimates
899: appear to differentiate between the range of solar-abundance PMS
900: models considered here. In particular, we find better agreement
901: between the estimated component parameters and predictions of
902: solar-abundance models from Y$^2$ and \citet{Montalban2006} than other
903: models considered here, and these two models agree on an V773~Tau age
904: estimate of approximately 3 $\pm$ 1 Myr; this is in good agreement
905: with the multiple epochs of Taurus star formation \citep{Palla2002}.
906: This age would place V773~Tau in the second of three age groups
907: defined by \citet{Palla2002}, i.e., stars with age $2 < t < 4$ Myr.
908: Such and younger stars are typically concentrated along the ``central
909: filaments'' of the Taurus molecular clouds systems. V773~Tau appears
910: in fact located at the periphery of a $\sim 0.7\times0.5$ pc-sized
911: ``core'' imaged in the high column density tracing C$^{18}$O molecule
912: by \citet[their core No.~5]{Onishi1996, Onishi1998}. From the observed
913: intensity of the C$^{18}$O $J=1-0$ line one can derive an H$_2$ column
914: density that translates into an A$_V > 5$. Extinction maps from
915: Digitized Sky Survey source counts around V773 Tau's position also
916: indicate an A$_V$ around 5, but have a ``resolution'' of several arc
917: minutes \citep{Dobashi2005}. These extinction values are several
918: times larger than our and other extinction determinations for
919: V773~Tau~A (\S~\ref{sec:SED}), and seem to indicate that the system is
920: located just at the near border of a denser cloud core.
921:
922:
923: Relative to the component parameters estimated here, the
924: \citet{Siess2000} and \citet{Palla1999} models predict cooler and less
925: luminous components, while the \citet{DAntona1997} models predict
926: hotter and more luminous components. However these conclusions would
927: change if the component mass estimates were to systematically change
928: (i.e. both components either more or less massive than present
929: estimates) by only about two sigma. Qualitatively this seems
930: relatively unlikely, but given the uncertainties in this study and the
931: added complexity of multiple orbits in the V773~Tau system it would be
932: prudent to withhold final judgment until the modeling of the A-B orbit
933: improves with time and an expanded astrometric and radial velocity
934: data set.
935:
936: %%% old
937: %%Finally, it is notable that we have restricted our comparisons of
938: %%V773~Tau~A component properties with solar-composition models.
939: %%Fundamentally this is a practical consideration; some of the model
940: %%sets considered here only support solar abundances. Nominally
941: %%abundances in Taurus-Auriga (and by consequence V773~Tau) are thought
942: %%to be near solar values \citep[see][]{Padgett1996}. However, as in
943: %%other sectors of the HR diagram, more secure abundance determinations
944: %%in star forming regions will be critical for definitive tests of PMS
945: %%models.
946:
947: %%% new
948: Finally, it is important to note that placing constraints on stellar
949: evolutionary models in general, and PMS evolutionary models in
950: particular requires precise measurement of more than just component
951: dynamical mass, but also equally important parameters such as
952: luminosity, $T_{\rm eff}$, and abundance. Even at the present level
953: of the work here (i.e.~dynamical masses better than 10\% 1-sigma),
954: uncertainties in $L$ and $T_{\rm eff}$ are significant factors in
955: limiting conclusions we can draw about model predictions. As for
956: abundance, we have restricted our comparisons of V773~Tau~A component
957: properties with solar-composition models. Fundamentally this is a
958: practical consideration; some of the model sets considered here only
959: support solar abundances. Nominally abundances in Taurus-Auriga (and
960: by consequence V773~Tau) are thought to be near solar values
961: \citep[see][]{Padgett1996}. However, as in other sectors of the HR
962: diagram, more secure abundance determinations in star forming regions,
963: along with improved measurements of luminosity and temperature will be
964: critical for definitive tests of PMS models.
965:
966:
967:
968: \acknowledgements
969:
970:
971: Part of this work was performed at the Michelson Science Center (MSC),
972: California Institute of Technology under contract with the National
973: Aeronautics and Space Administration (NASA).
974:
975: Some of the data presented herein were obtained at the W.M.~Keck
976: Observatory, which is operated as a scientific partnership among the
977: California Institute of Technology, the University of California and
978: the NASA. The Observatory was made possible by the generous financial
979: support of the W.M.~Keck Foundation. We gratefully acknowledge the
980: support of personnel at the Jet Propulsion Laboratory, W.M.~Keck
981: Observatory, and the MSC in obtaining KI observations of V773~Tau~A.
982: The authors wish to recognize and acknowledge the very significant
983: cultural role and reverence that the summit of Mauna Kea has always
984: had within the indigenous Hawaiian community. We are most fortunate
985: to have the opportunity to conduct observations from this mountain.
986:
987: The CfA RV observations presented here were originally advocated by
988: Robert Mathieu; we thank him for his gracious contribution of these
989: data in their use here. We thank P.~Berlind, M.~Calkins, J.~Caruso,
990: R.J.~Davis, G.~Esquerdo, J.~Peters, A.~Milone, and R.P.~Stefanik for
991: obtaining many of the spectroscopic observations. GT acknowledges
992: partial support from NASA's MASSIF SIM Key Project (BLF57-04) and NSF
993: grant AST-0406183.
994:
995: We thank F.~D'Antona, J.~Montalb\'an, S.~Stahler, and F.~Palla for
996: sharing their pre-main sequence evolutionary models, and for fruitful
997: discussions on applying them. Thanks also to G.~Duchene for sharing
998: IR photometry for V773~Tau~A (including his kind permission to report
999: these data here in Table~\ref{tab:Fluxtable}), and the anonymous
1000: referee whose many thoughtful comments helped greatly improve this
1001: manuscript.
1002:
1003: This research has made use of services of the MSC at the California
1004: Institute of Technology; the SIMBAD database, operated at CDS,
1005: Strasbourg, France; of NASA's Astrophysics Data System Abstract
1006: Service; and of data products from the Two Micron All Sky Survey,
1007: which is a joint project of the University of Massachusetts and the
1008: Infrared Processing and Analysis Center, funded by NASA and the
1009: National Science Foundation.
1010:
1011:
1012:
1013: \begin{thebibliography}{}
1014:
1015: %%\bibitem[Andersen(1991)]{Andersen1991}
1016: %% Andersen, J. 1991, \aapr~3, 91
1017:
1018: %%\bibitem[Artymowicz \& Lubow(1994)]{Artymowicz1994}
1019: %%Artymowicz, P.~and Lubow, S.~1994 \apj~421, 651.
1020:
1021: \bibitem[Baraffe et al.(1998)]{Baraffe1998}
1022: Baraffe, I.~et al~1998, \aap~337, 403.
1023: %%Baraffe, I., Chabrier, G., Allard, F., and Hauschildt, P.~1998, \aap~337, 403.
1024:
1025: %%\bibitem[Benedict et al.(2001)]{Benedict2001}
1026: %%Benedict, G. et al.~2001, \aj~121, 1613.
1027:
1028: \bibitem[Boden et al.(2000)]{Boden2000}
1029: Boden, A., et al~2000, \apj~536, 880.
1030:
1031: \bibitem[Boden et al.(2005a)]{Boden2005a}
1032: Boden, A., Torres, G., \& Hummel, C.~2005a, \apj~627, 464.
1033:
1034: \bibitem[Boden et al.(2005b)]{Boden2005b}
1035: Boden, A.~et al~2005b, \apj~635, 442.
1036:
1037: \bibitem[Colavita(1999)]{Colavita1999}
1038: Colavita, M.~1999, \pasp~111, 111.
1039:
1040: \bibitem[Colavita et al(2003)]{Colavita2003}
1041: Colavita, M.~et al~2003, \apj~592, L83.
1042:
1043: \bibitem[D'Antona \& Mazzitelli(1997)]{DAntona1997}
1044: D'Antona, F.~\& Mazzitelli, I.~1997, Mem.~S.A.It., 68, 807
1045:
1046: \bibitem[Demarque et al(2004)]{Demarque2004}
1047: Demarque, P.~et al~2004, \apjs~155, 667.
1048:
1049: \bibitem[Dobashi et al.(2005)]{Dobashi2005}
1050: Dobashi, K.~et al~2005, \pasj, 57, 417.
1051: %%Dobashi, K., Uehara, H., Kandori, R., Sakurai, T., Kaiden, M., Umemoto, T., \& Sato, F.\ 2005,
1052:
1053:
1054: \bibitem[Duchene et al(2003)]{Duchene2003}
1055: Duchene, G.~et al.~2003, \apj~592, 288 (D2003).
1056:
1057: %%\bibitem[ESA(1997)]{HIP1997}
1058: %% ESA 1997, The Hipparcos and Tycho Catalogues, ESA SP-1200.
1059:
1060: \bibitem[Feigelson et al(1994)]{Feigelson1994}
1061: Feigelson, E.~et al~1994, \apj~432, 373.
1062:
1063: \bibitem[Ghez et al(1993)]{Ghez1993}
1064: Ghez, A., Neugebauer, G., \& Matthews, K.~1993, \aj~106, 2005.
1065:
1066: \bibitem[Ghez et al(1997)]{Ghez1997}
1067: Ghez, A., White, R., \& Simon, M.~1997, \apj~490, 353.
1068:
1069: \bibitem[Hartigan et al(1994)]{Hartigan1994}
1070: Hartigan, P., Strom, K., \& Strom, S.~1994, \apj~427, 961.
1071:
1072: \bibitem[Hillenbrand \& White(2004)]{Hillenbrand2004}
1073: Hillenbrand, L.~\& White, R.~2004, \apj~604, 741.
1074:
1075: %%\bibitem[Hummel et al.(2001)]{Hummel2001}
1076: %% Hummel, C.~et al.~2001, \aj~121, 1623.
1077:
1078: %%\bibitem[Kastner et al(2002)]{Kastner2002}
1079: %% Kastner, J.~et al~2002, \apj~567, 434.
1080:
1081: \bibitem[Kurucz(2001)]{KuruczModels} Kurucz, R.~2001, Kurucz models
1082: used here are available at http://cfaku5.cfa.harvard.edu.
1083:
1084: \bibitem[Kutner et al(1986)]{Kutner1986}
1085: Kutner, M., Rydgren, A., \& Vrba, F.~1986, \aj~92, 575.
1086:
1087:
1088: \bibitem[Latham(1992)]{Latham1992}
1089: Latham, D.\ W. 1992, in IAU Coll.\ 135, Complementary Approaches to
1090: Double and Multiple Star Research, ASP Conf.\ Ser.\ 32, eds.\ H.\ A.\
1091: McAlister \& W.\ I.\ Hartkopf (San Francisco: ASP), 110
1092:
1093:
1094: \bibitem[Latham et al.(2002)]{Latham:02}
1095: Latham, D.~W.~et al~2002~\aj~124, 1144.
1096:
1097: %%%, Stefanik, R.~P., Torres, G., Davis, R.~J., Mazeh, T.,
1098: %%%Carney, B.~W., Laird, J.~B., \& Morse, J.~A.\ 2002, \aj, 124, 1144
1099:
1100:
1101: \bibitem[Leinert et al.(1993)]{Leinert1993}
1102: Leinert, C.~et al~1993, \aap~278, 129.
1103:
1104: \bibitem[Lejeune et al(1997)]{Lejeune1997}
1105: Lejeune, T.~et al~1997, \aaps~125, 229.
1106:
1107: \bibitem[Lejeune et al(1998)]{Lejeune1998}
1108: Lejeune, T.~et al~1998, \aaps~130, 65.
1109:
1110: \bibitem[Lestrade et al(1999)]{Lestrade99}
1111: Lestrade, J-F, et al~1999, \aap~344, 1014 (L1999).
1112:
1113: \bibitem[Luhman(2004)]{Luhman2004}
1114: Luhman, K.~2004, \apj~617, 1216.
1115:
1116: \bibitem[Mason et al(2001)]{Mason2001}
1117: Mason, B.~et al~2001, \aj~122, 3466.
1118:
1119: \bibitem[Montalb\'an et al(2004)]{Montalban2004}
1120: Montalb\'an, J.~et al~2004, \aap~416, 1081.
1121:
1122: \bibitem[Montalb\'an \& D'Antona(2006)]{Montalban2006}
1123: Montalb\'an, J.~\& D'Antona, F.~2004, \mnras~370, 1823.
1124:
1125: \bibitem[Mart\'in et al(1994)]{Martin1994}
1126: Mart\'in, E.~et al~1994, \aap~282, 503.
1127:
1128: \bibitem[Massi et al(2002)]{Massi2002}
1129: Massi, M., Menten, M., \& Neidhofer, J.~2002, \aap~382, 152.
1130:
1131: \bibitem[Massi et al(2006)]{Massi2006}
1132: Massi, M.~ et al~2006, \aap~453, 959.
1133:
1134: \bibitem[Mathieu et al(2007)]{Mathieu2007}
1135: Mathieu, R.~et al~2007, Proc PPV, U.~Ariz Press.
1136:
1137: \bibitem[Napier et al(1994)]{Napier1994}
1138: Napier, P.~et al~1994,
1139: %%, Bagri, D., Clark, B., Rogers, A.,
1140: %%Romney, J., Thompson, A., \& Walker, R.~1994,
1141: Proc IEEE, 82, 658.
1142:
1143:
1144: \bibitem[Nordstr\"om et al.(1994)]{Nordstrom:94}
1145: Nordstr\"om, B.~et al~1994, \aap~287, 338.
1146:
1147: %%, Latham, D.\ W., Morse, J.\ A., Milone, A.\ A.\ E.,
1148: %%Kurucz, R.\ L., Andersen, J., \& Stefanik, R.\ P. 1994, \aap, 287, 338
1149:
1150:
1151: \bibitem[O'Neal et al(1990)]{O'Neal1990}
1152: O'Neal, D.~et al~1990, \aj~100, 1610.
1153:
1154: \bibitem[Onishi et al.(1996)]{Onishi1996}
1155: Onishi, T.~et al~1996, \apj~465, 815.
1156: %%Onishi, T., Mizuno, A.,
1157: %%Kawamura, A., Ogawa, H., \& Fukui, Y.\ 1996, \apj, 465, 815
1158:
1159: \bibitem[Onishi et al.(1998)]{Onishi1998}
1160: Onishi, T.~et al~1998, \apj~502, 296.
1161: %%Onishi, T., Mizuno, A.,
1162: %%Kawamura, A., Ogawa, H., \& Fukui, Y.\ 1998, \apj, 502, 296
1163:
1164: \bibitem[Padgett(1996)]{Padgett1996}
1165: Padgett, D.~1996, \apj~471, 847.
1166:
1167: \bibitem[Palla \& Stahler(1999)]{Palla1999}
1168: Palla, F.~\& Stahler, S.~1999, \apj~525, 772.
1169:
1170: \bibitem[Palla \& Stahler(2001)]{Palla2001}
1171: Palla, F.~\& Stahler, S.~2001, \apj~553, 299.
1172:
1173: \bibitem[Palla \& Stahler(2002)]{Palla2002}
1174: Palla, F.~\& Stahler, S.~2002, \apj~581, 1194.
1175:
1176: \bibitem[Pickles(1998)]{Pickles1998}
1177: Pickles, A.~1998, \pasp~110, 863.
1178:
1179: \bibitem[Phillips et al(1996)]{Phillips1996}
1180: Phillips, R.~et al~1996, \aj~111, 918 (P1996).
1181:
1182: %%\bibitem[Press et al(1992)]{Press1992}
1183: %%Press, W.~et al~1992, Numerical Recipes (New York:Cambridge University Press)
1184:
1185: \bibitem[Rydgren et al(1976)]{Rydgren1976}
1186: Rydgren, A.~Strom, S., \& Strom, K.~1976, \apjs~30, 307.
1187:
1188: \bibitem[Siess et al(2000)]{Siess2000}
1189: Siess L., Dufour E., \& Forestini M. 2000, \aap~358, 593.
1190:
1191: \bibitem[Stefanik et al.(1999)]{Stefanik:99}
1192: Stefanik, R.\ P., Latham, D.\ W., \& Torres, G. 1999, in Precise
1193: Stellar Radial Velocities, IAU Coll.\ 170, ASP Conf.\ Ser., 185, eds.\
1194: J.\ B.\ Hearnshaw \& C.\ D.\ Scarfe (San Francisco: ASP), 354.
1195:
1196: %%\bibitem[Simon et al(2000)]{Simon2000}
1197: %% Simon, M, Dutrey, A.~\& Guilloteau, S.~2000, \apj~545, 1034.
1198:
1199:
1200: %%\bibitem[Stelzer \& Schmitt(2004)]{Stelzer2004}
1201: %% Stelzer, B.~and Schnitt, J.~2004, \aap~418, 687.
1202:
1203: \bibitem[Tamazian et al(2002)]{Tamazian2002}
1204: Tamazian, V.~et al.~2002, \apj~578, 925.
1205:
1206: \bibitem[Torres et al.(2003)]{Torres2003}
1207: Torres, G. et al.~2003, \aj~125, 825.
1208:
1209: \bibitem[Walker \& Walstencroft(1988)]{Walker1988}
1210: Walker, H.~\& Walstencroft, R.~1988, \pasp~100, 1509.
1211:
1212: \bibitem[Webb et al(1999)]{Webb1999}
1213: Webb, R.~et al~1999, \apj~512, L63.
1214:
1215:
1216: \bibitem[Welty(1995)]{Welty1995}
1217: Welty, A.~1995, \aj~110, 776 (W1995).
1218:
1219: \bibitem[White \& Ghez(2001)]{White2001}
1220: White, R.~\& Ghez, A.~2001, \apj~556, 265 (WG2001).
1221:
1222: \bibitem[Woitas(2003)]{Woitas2003}
1223: Woitas, J.~2003, \aap~406, 685.
1224:
1225: \bibitem[Zensus et al.(1995)]{Zensus1995}
1226: Zensus, J., Diamond, P., \& Napier, P.~1995, ASP Conf.~Ser.~ 82: Very
1227: Long Baseline Interferometry and the VLBA
1228:
1229: \bibitem[Zucker \& Mazeh(1994)]{Zucker:94}
1230: Zucker, S.~\& Mazeh, T.~1994, \apj~420, 806.
1231:
1232: \bibitem[Zucker et al.(1995)]{Zucker:95}
1233: Zucker, S., Torres, G.,~\& Mazeh, T.~1995, \apj~452, 863.
1234:
1235: \end{thebibliography}
1236:
1237: \end{document}
1238: