1: %\documentclass[12pt,preprint]{emulateapj}
2: \documentclass[12pt,preprint]{aastex}
3: \usepackage{natbib}
4: \citestyle{aa}
5: \begin{document}
6:
7: \title{Spatially Resolving the Inner Disk of TW Hya}
8:
9: \author{J. A. Eisner,\altaffilmark{1,2} E. I. Chiang,\altaffilmark{1,3}
10: \& L.A. Hillenbrand\altaffilmark{4}}
11: \altaffiltext{1}{University of California at Berkeley,
12: Department of Astronomy, 601 Campbell Hall, Berkeley, CA 94720}
13: \altaffiltext{2}{Miller Fellow}
14: \altaffiltext{3}{Alfred P. Sloan Research Fellow}
15: \altaffiltext{4}{California Institute of Technology,
16: Department of Astronomy MC 105-24, Pasadena, CA 91125}
17: \email{jae@astron.berkeley.edu}
18:
19: %\author{J. A. Eisner\altaffilmark{1} \& E. I. Chiang\altaffilmark{2}}
20: %\affil{University of California at Berkeley \\
21: %Department of Astronomy \\
22: %601 Campbell Hall \\
23: %Berkeley, CA 94720}
24: %\email{jae,echiang@astron.berkeley.edu}
25: %\and
26: %\author{L. A. Hillenbrand}
27: %\affil{California Institute of Technology \\
28: %Department of Astronomy MC 105-24 \\
29: %Pasadena, CA 91125}
30: %\email{lah@astro.caltech.edu}
31: %\altaffiltext{1}{Miller Fellow}
32: %\altaffiltext{2}{Alfred P. Sloan Research Fellow}
33:
34: \keywords{stars: planetary systems: protoplanetary
35: disks---stars: pre-main-sequence---stars: individual (TW Hya)}
36:
37: %\slugcomment{Draft of {\bf \today}}
38: %
39: %For submission to {\bf ApJ}}
40:
41:
42: \begin{abstract}
43: We present Keck Interferometer observations of TW Hya that spatially
44: resolve its emission at 2 $\mu$m wavelength. Analyzing
45: these data together with existing $K$-band veiling
46: and near-infrared photometric measurements,
47: we conclude that the inner disk consists of optically thin,
48: sub-micron-sized dust extending from $\sim 4$ AU to within 0.06 AU of the
49: central star. The inner disk edge may be magnetospherically
50: truncated.
51: %We estimate
52: %the mass of inner disk dust to be $3 \times 10^{21}$ g. Furthermore,
53: %Given the inner disk size determined from the interferometric measurement,
54: %the observed near-IR excess can only be reproduced if the dust is composed of
55: %particles with sizes smaller than or about $1$ $\mu$m.
56: Even if we account for the presence of gas in the inner disk, these small
57: dust grains have survival times against radiation blow-out that are orders of
58: magnitude shorter than the age of the system, suggesting
59: continual replenishment through collisions of larger bodies.
60: \end{abstract}
61:
62:
63: \section{Introduction}
64: TW Hya is a nearby \citep[$\sim 51$ pc;][]{MAMAJEK05}, young
65: \citep[$\sim 10$ Myr;][]{WEBB+99}
66: star surrounded by an accretion disk that evinces a large inner hole
67: as judged from the observed spectral energy distribution
68: \citep[SED;][]{CALVET+02}. Unusually low excess emission
69: at wavelengths $\lambda \la 10$ $\mu$m can be modeled with an optically thick
70: disk whose inner edge is located $\sim 4$ AU from the central star.
71: While observations of 10 $\mu$m silicate emission \citep{SLR00,UCHIDA+04}
72: together with non-zero excess at 2 $\mu$m \citep{JV01} suggest the presence of
73: at least some
74: dust grains with sizes less than a few microns at stellocentric distances
75: $R \la 4$ AU, this inner disk material appears optically thin,
76: and has been estimated to constitute less than a lunar mass \citep{CALVET+02}.
77: Detection of warm gas \citep{HERCZEG+04,RETTIG+04} and accretion signatures
78: \citep{MUZEROLLE+00,AB02} confirm that the region inside 4 AU is not
79: devoid of material.
80:
81: At $R \sim 4$ AU, dust temperatures ($\sim 100$ K) are substantially
82: lower than sublimation temperatures for silicate grains
83: \citep[$\ga 1500$ K; e.g.,][]{POLLACK+94}.
84: This suggests the optically thick outer disk is truncated at 4 AU
85: by a mechanism
86: other than dust sublimation. Large holes inferred from SEDs are commonly
87: attributed to planets, which may clear gaps about their orbits.
88: A planet impedes accretion of material outside its orbit, while inner
89: disk material is free to drain onto the central star
90: \citep{GT82,BRYDEN+99,RICE+03}. However, a viscous outer disk causes inward
91: migration of planets and their associated gaps \citep[e.g.,][]{LP86,WARD97}.
92: Thus, inner holes would be filled in on the viscous
93: timescale unless the disk
94: is less massive than the planet, in which case the timescale is lengthened
95: by the mass ratio between the planet and the disk \citep{CHIANG03}.
96: Since the outer disk of TW Hya is massive
97: \citep[$\ga 0.1$ M$_{\odot}$;][]{WEINBERGER+02,WILNER+05},
98: planets are unlikely to
99: preserve inner clearings over the lifetime of the system
100: ($\sim 10$ Myr) unless the outer disk is unusually inviscid
101: \citep[$\alpha \la 10^{-5}$;][]{SS73}.
102:
103: An alternative explanation for SED-inferred holes
104: is that dust grains have grown larger than a few microns, depleting
105: the population of small grains that would produce the near-IR
106: emission. Scattered light and long-wavelength emission from the outer
107: ($R > 10$ AU) disk of TW Hya suggest substantial grain growth,
108: up to cm sizes \citep{WEINBERGER+02,WILNER+05}, supporting the hypothesis
109: of grain coagulation. The small amount of sub-micron-sized dust required to
110: explain emission at $\lambda \la 10$ $\mu$m and the spectral shape of the
111: 10 $\mu$m silicate feature \citep{CALVET+02,UCHIDA+04}
112: may represent the tail at small sizes of a grain size distribution that
113: peaks at sizes much larger than a micron. Furthermore, as we discuss below,
114: small dust grains are short-lived in the TW Hya disk and demand
115: continual replenishment, possibly from collisions of larger parent bodies.
116: The population of sub-micron-sized dust in the inner disk may not be
117: primordial.
118: %and may in fact be a byproduct
119: %of previous dust growth processes that gave rise to large bodies
120: %capable of dust-producing collisions.
121:
122: Here we present observations with the Keck Interferometer that
123: spatially resolve the inner disk around TW Hya for the first time.
124: Previous observations at sub-millimeter to centimeter
125: wavelengths spatially resolved emission at larger radii
126: ($> 10$ AU) and enabled powerful constraints on the
127: outer disk structure and dust properties \citep{QI+04,WILNER+05}.
128: With our near-IR interferometric observations, we extend this analysis to
129: the inner disk. By combining spatially resolved measurements with spectral
130: information, we determine the radial distribution, temperature, and
131: approximate grain sizes of dust. We
132: confirm that the inner disk is populated by small amounts of sub-micron-sized
133: dust, and show that the inner radius\footnote{In the remainder of
134: the paper, the ``inner radius'' refers to the inner edge of the
135: optically thin inner disk, not the inner edge of the optically
136: thick outer disk. We take the latter to be located at
137: $R \sim 4$ AU based on previous
138: modeling by \citet{CALVET+02}.}
139: of this optically thin disk occurs further from
140: the star than previous spatially unresolved observations imply.
141:
142: \section{Observations \label{sec:obs}}
143: We observed TW Hya near $K$-band ($\lambda_K = 2.14$ $\mu$m;
144: $\Delta \lambda =0.3$ $\mu$m) with the Keck Interferometer (KI)
145: on April 21, 2005 (UT 20050421).
146: The KI is a fringe-tracking Michelson interferometer that
147: combines light from the two 10-m Keck apertures
148: and provides an angular resolution of $\la 5$ mas \citep{CW03,COLAVITA+03}.
149: A single 130s observation of TW Hya was obtained between observations of three
150: calibrators (HD 97023, HD 97940, and HD 99934). Figure \ref{fig:rawv2} shows
151: the uncalibrated visibilities of the target and calibrators.
152:
153: Data calibration is described by \citet{EISNER+05}; here
154: we summarize the procedure. We first determine the system
155: visibility (i.e., the point source response of the interferometer) from
156: the weighted mean of observations of unresolved calibrators \citep{BODEN+98}.
157: Source and calibrator data are corrected for
158: detection biases \citep{COLAVITA99} and integrated into 5s
159: blocks. The calibrated $V^2$ for the target source is averaged over all 5s
160: blocks, with uncertainties given by the quadrature addition
161: of the internal scatter and the uncertainty in the system visibility.
162: The calibrated, normalized, squared visibility measured for TW Hya at
163: $r_{uv}=28$ M$\lambda$ ($u=49.6$ m, $v=36.7$ m) is $V^2=0.88 \pm 0.05$.
164:
165:
166: \section{Modeling \label{sec:mod}}
167: We model our measured $V^2$ together with previous photometric
168: measurements at $\lambda=1$--5 $\mu$m \citep{WEBB+99,SLR00},
169: assuming photometric uncertainties of 10\%.
170: The circumstellar-to-stellar flux ratio is critical for modeling
171: the circumstellar component of the visibility and near-IR SED
172: \citep[e.g.,][]{EISNER+04}; we utilize a previous measurement of
173: this ratio at 2 $\mu$m, $r_K = 0.07 \pm 0.04$ \citep{JV01}.
174: Our model consists of the central star and an optically thin disk.
175: We model the central star using a Kurucz stellar atmosphere with
176: radius $R_{\ast} =$ R$_{\odot}$, temperature
177: $T_{\ast} = 4000$ K, mass $M_{\ast}=0.7$ M$_{\odot}$, surface gravity
178: $\log g$ (cm s$^{-2}$) $= 4.5$,\footnote{While the computed
179: $\log g \approx 4.3$ for our assumed values of $R_{\ast}$ and $M_{\ast}$,
180: we adopt $\log g=4.5$ since Kurucz models exist for this value. Our results
181: are insensitive to this small difference.} and distance $d=51$ pc
182: \citep{WEBB+99,JV01,AB02,MAMAJEK05}.
183:
184: The optically thin disk extends from an inner truncation radius,
185: $R_{\rm in}$, to an outer radius, $R_{\rm out}$.
186: Since most of the near-IR emission is generated close to $R_{\rm in}$ where
187: the hottest dust resides, our results are insensitive to $R_{\rm out}$;
188: for simplicity, we assume $R_{\rm out}=4$ AU.
189: %the approximate inner extent
190: %of the optically thick outer disk component inferred from previous
191: %analysis of the broadband SED \citep{CALVET+02}.
192: The mass surface density of the inner disk is parameterized as
193: $\Sigma=\Sigma_0 (R/ {\rm AU})^{-3/2}$, where $R$ is the
194: stellocentric radius and $\Sigma_0$ is the surface density at $R=1$ AU.
195:
196: We assume dust grains are of a single size and
197: adopt a simple prescription for the frequency-dependent dust opacity,
198: $\kappa_{\nu} = \kappa_{K} (\nu/\nu_{K})^{\beta}$ cm$^2$ g$^{-1}$. For
199: sub-micron-sized grains, we set $\beta=1$ and $\kappa_K = 10^3$. For
200: larger grains with sizes $\sim 10 \mu$m, we take $\beta=0$ and
201: $\kappa_K=10^2$. These choices are compatible with previous computations
202: of opacities at $\lambda = 0.1$--5 $\mu$m
203: by \citet{MN93}. Our normalizations are $\sim 10^2$ times
204: higher than theirs since our $\kappa_\nu$ is the dust-mass opacity
205: as opposed to the dust+gas opacity; i.e., the units of $\kappa_\nu$ are
206: cm$^2$ per g of dust.
207: %While we have assumed singly-sized
208: %dust grains, we compare our adopted opacities to those computed for MRN
209: %size distributions \citep{MRN77} that may be more physically plausible.
210: %An MRN distribution with minimum grain sizes larger than a few microns
211: %and maximum grain sizes smaller than a few hundred microns produces
212: %near-IR opacities similar to ours for $\beta=0$, while our $\beta=1$ case
213: %is compatible with MRN opacities as long as $\la \mu$m-sized grains are
214: %present and the maximum dust grain sizes are smaller than $\sim 100$ $\mu$m.
215: %\citep{MN93}.
216:
217: We compute the dust temperature under the assumption that the disk is optically
218: thin:
219: \begin{equation}
220: T_{\rm dust}(R) = T_{\ast} \left(\frac{R_{\ast}}{2R}\right)^{2/(4+\beta)}.
221: \label{eq:tr}
222: \end{equation}
223: The total flux of the inner disk
224: is derived by dividing the disk into annuli,
225: computing the flux for each annulus, and summing the annular fluxes.
226: Similarly, model visibilities are computed for each annulus, and the
227: visibility for the entire inner disk is given by the flux-weighted
228: sum of the annular visibilities. The flux in an annulus of infinitesimal
229: width $dR$ is
230: \begin{equation}
231: dF_{\nu}(R) = \frac{2\pi}{d^2} B_{\nu} \left(T_{\rm dust} \right)
232: \tau_{\nu} \: R \: dR ,
233: \label{eq:fr}
234: \end{equation}
235: where $B_\nu$ is the Planck function and
236: $\tau_{\nu}$ is the vertical optical depth,
237: \begin{equation}
238: \tau_{\nu}(R) = \kappa_{\nu} \Sigma = \kappa_{\nu} \Sigma_0
239: \left(\frac{R}{\rm AU}\right)^{-3/2}.
240: \label{eq:taur}
241: \end{equation}
242: The normalized visibility for an annulus extending from $R_1$
243: to $R_2 = R_1+dR$ is given by the difference
244: of visibilities for uniform disks having radii equal to
245: $R_1$ and $R_2$:
246: \begin{equation}
247: V(R) = \frac{\lambda d}{2\pi r_{uv} (R_2^2 - R_1^2)}
248: \left[R_2J_1 \left(\frac{2\pi r_{uv} R_2}{\lambda d}\right) -
249: R_1J_1 \left(\frac{2\pi r_{uv} R_1}{\lambda d}\right)\right].
250: \end{equation}
251: Here $r_{uv} = 28$ M$\lambda$ is the $uv$ radius, $\lambda=\lambda_K$ is the
252: observing wavelength, and $J_1$ is
253: a first-order Bessel function.
254:
255: The total flux density from the disk and the central star equals
256: \begin{equation}
257: F_{\nu, \rm tot} = F_{\nu, \ast} +
258: \int_{R_{\rm in}}^{R_{\rm out}} d F_{\nu}(R).
259: \end{equation}
260: The veiling at 2 $\mu$m is
261: \begin{equation}
262: r_K = \frac{1}{F_{K, \ast}} \int_{R_{\rm in}}^{R_{\rm out}} d F_{K}(R),
263: \end{equation}
264: where $F_K$ is the flux density at $\lambda = \lambda_K$.
265: The squared visibility of the model at 2 $\mu$m is
266: \begin{equation}
267: V^2 = \left[\frac{F_{K , \ast}V_{\ast}
268: + \int_{R_{\rm in}}^{R_{\rm out}} d F_{K}(R)V(R)}
269: {F_{K, \rm tot}}\right]^2,
270: \end{equation}
271: where $V_{\ast}=1$ is the visibility of the unresolved central star.
272:
273: We solve for the best-fit parameters of the model
274: by computing the $\lambda=1$--5 $\mu$m fluxes, $r_K$, and $V^2$
275: for a grid of values of $R_{\rm in}$ and $\Sigma_0$, and minimizing the
276: $\chi^2$-residuals between model and data.
277: Uncertainties for best-fit parameters are determined
278: from $\chi^2$-error ellipses \citep[e.g.,][]{EISNER+04}.
279: We do not include $\beta$ or $\kappa_K$ as free parameters and
280: instead consider two model cases: $\beta=1$ and $\kappa_K=10^3$ (intended
281: to model sub-micron-sized grains), and $\beta=0$ and $\kappa_K=10^2$
282: (representing grains sized $\sim 10$ $\mu$m).
283:
284: Scattered light from the disk is justifiably ignored in our modeling.
285: %large radii, ignored in our modeling,
286: %could lower measured $V^2$ values and lead to larger inferred inner disk size
287: %scales \citep[e.g.,][]{EISNER+04}.
288: At $\lambda=1.1$ and 1.6 $\mu$m, the disk-scattered flux on
289: angular scales of $0\rlap{.}''4$--$4''$ is estimated to comprise
290: 2.4\% and 2.1\% of the stellar flux, respectively \citep{WEINBERGER+02}.
291: Given the blue color and roughly flat surface brightness profile
292: of the scattered light within $\sim 0\rlap{.}''8$ \citep{WEINBERGER+02},
293: we conclude that the $K$-band scattered flux within the 50 mas field of
294: view of KI is $< 1\%$. Therefore thermal emission, as traced by the
295: $K$-band veiling \citep[$\sim 7\%$ of the stellar flux;][]{JV01},
296: dominates over any scattered emission.
297:
298: While we have assumed TW Hya is a single star, a low-mass stellar companion
299: could contribute to the near-IR visibilities and SED.
300: To the best of our knowledge, no stellar companions have been detected in
301: previous HST imaging or radial velocity monitoring, and thus the presence of
302: a luminous second star appears
303: unlikely. Additional KI observations, or astrometric
304: and/or radial velocity monitoring could test this possibility definitively.
305:
306: \section{Results and Discussion \label{sec:disc}}
307: Figure \ref{fig:mods} shows the best-fit models for $\beta=0$ and $1$,
308: together with the $V^2$ and SED data.
309: Best-fit values for $R_{\rm in}$ and $\Sigma_0$, and their
310: $1\sigma$ uncertainties are listed in Table 1. Our model can
311: reproduce the KI $V^2$ measurement and near-IR SED of TW Hya if $\beta= 1$
312: and $R_{\rm in} \sim 0.06$ AU. While models with $\beta=0$ can fit the
313: SED data well (the larger quantity of SED data relative to the single $V^2$
314: measurement skews the fits accordingly), only $\beta=1$ models
315: can simultaneously reproduce our $V^2$ measurement.
316: From these best-fit parameter values, we compute the temperature at the inner
317: truncation radius $T_{\rm in}$, the dust mass $M_{\rm dust}$,\footnote{The
318: dust mass depends sensitively on $R_{\rm out}$ and the assumed
319: surface density profile. Estimates of $M_{\rm dust}$ are therefore
320: highly uncertain. For comparison, under the assumption of a constant surface
321: density, \citet{CALVET+02} estimate a dust mass approximately four orders
322: of magnitude higher than the values listed in Table \ref{tab:models}.}
323: %the radial
324: %optical depth $\tau_{\rm r}$ (under the assumption of a hydrostatic disk),
325: and the 2 $\mu$m vertical optical depth at the inner edge $\tau_{K,\rm in}$.
326: %Estimated radial optical depths are $< 1$,
327: %justifying our assumption of optically thin dust.
328:
329: The inner radius of the optically thin disk ($0.06$ AU) exceeds
330: that inferred from previous modeling of spatially unresolved data
331: \citep[0.02 AU;][]{CALVET+02}.
332: Our large inner radius leads to an inner disk temperature lower than
333: expected for dust sublimation (Table 1), suggesting that an alternate
334: truncation mechanism is necessary. One possibility is that the inner disk
335: extends inward to the magnetospheric radius $R_{\rm mag}$,
336: where the ram pressure from accretion balances the stellar magnetic pressure.
337: Although hot dust may still exist interior to
338: this radius, its high infall velocity \citep[e.g.,][]{EDWARDS+94} implies an
339: optical depth orders of magnitude lower than that of dust outside
340: $R_{\rm mag}$. Assuming
341: an accretion rate of $5 \times 10^{-10}$ M$_{\odot}$ yr$^{-1}$
342: \citep{MUZEROLLE+00}, stellar magnetic field strength of 2.6 kG
343: \citep{YJV05}, and stellar parameters from \S \ref{sec:mod}, we compute
344: $R_{\rm mag} \sim 0.09$ AU \citep[e.g.,][]{KONIGL91}.
345: This is comparable to our best-fit $R_{\rm in}$ (with $\beta=1$),
346: indicating that the optically thin
347: inner disk may indeed be magnetospherically truncated.
348:
349: The fact that models require $\beta = 1$ to fit the combined $V^2$+SED
350: data indicates that the inner disk contains a population of dust grains
351: having sizes smaller than $\sim 1$ $\mu$m.
352: In fact, we can obtain slightly better fits to our data if we allow
353: $\beta>1$, as one might expect if very small grains ($\la
354: 0.01$ $\mu$m) were present \citep{MN93}.
355:
356: Sub-micron-sized grains are quickly blown out of the inner disk by stellar
357: radiation pressure. Gas friction mediates
358: dust removal; unbound grains achieve a terminal outflow velocity equal to the
359: product of the momentum stopping time and the net outward acceleration due to
360: radiation pressure and gravity
361: \citep[e.g.,][]{WEIDENSCHILLING77}. Estimates of the gas
362: density are necessary to calculate
363: the survival times of small dust grains. Using the measured column
364: density and temperature of H$_2$ in the inner disk \citep{HERCZEG+04},
365: accounting for the possibility that the midplane may be up to $10^3$ times
366: denser than the warm surface \citep{GNI04,NAJITA06}, and assuming
367: a hydrostatic disk, we estimate a midplane gas density at 1 AU
368: of $\la 10^{-15}$
369: g cm$^{-3}$.\footnote{Observations of warm CO imply similar gas densities
370: \citep{RETTIG+04,NAJITA06}.}
371: The stopping time for micron-sized grains at $R \sim 1$ AU
372: is $\ga 10^{-2}$ yr, the terminal velocity is $\ga 1$ km s$^{-1}$, and
373: the removal time is $\la 1$ yr.
374:
375: Small dust grains in the TW Hya inner disk survive for $\la 1$ yr, and
376: are thus ephemeral over the age of the system. Because of the
377: difficulty in transporting sub-micron-sized grains from the outer disk
378: at $R \ga 4$ AU to $R_{\rm in}=0.06$ AU, we argue that inner disk dust
379: is continually re-generated, possibly by collisions of
380: a swarm of larger parent bodies that also reside in the inner disk.
381: %This argument supports the hypothesis that that the dearth of near-IR excess
382: %emission in TW Hya is caused by a lack of small particles.
383:
384: \section{Conclusions}
385: We observed TW Hya with the Keck Interferometer and found the 2 $\mu$m
386: emission to be spatially resolved. We modeled the interferometric
387: data together with previous measurements of the $K$-band
388: veiling and near-IR fluxes, and inferred that the inner disk consists of
389: optically thin dust extending
390: from the edge of the optically thick outer disk
391: \citep[$R \sim 4$ AU;][]{CALVET+02} to $R_{\rm in} = 0.06$
392: AU of the central star.
393: This inner radius is larger than expected from dust sublimation; the
394: truncation may be magnetospheric in origin. The near-IR
395: emitting dust is composed of sub-micron-sized particles which are
396: extremely short-lived; this dust may be
397: replenished by erosive collisions of larger parent bodies in the inner disk.
398:
399: \medskip
400: \noindent{\bf Acknowledgments.}
401: The near-IR interferometry data presented in this paper were obtained
402: with the Keck Interferometer (KI) of the W.M. Keck Observatory, which
403: was made possible by the generous financial support of the W.M. Keck
404: Foundation and is operated as a scientific partnership between the California
405: Institute of Technology, the University of California, and NASA.
406: The authors thank the entire KI team for making these observations possible,
407: and acknowledge the cultural role and reverence that the summit of Mauna Kea
408: has always had within the indigenous Hawaiian community.
409: %We are most fortunate to have the opportunity to conduct
410: %observations from this mountain.
411: The authors are grateful to M. Sitko for
412: providing the 3--5 $\mu$m photometry used in this paper, and to J. Najita and
413: J. Carr for useful comments about the warm CO gas.
414: %J.A.E. acknowledges support from a Miller Research Fellowship.
415:
416:
417: \begin{thebibliography}{36}
418:
419: \bibitem[{{Alencar} \& {Batalha}(2002)}]{AB02}
420: {Alencar}, S.~H.~P. \& {Batalha}, C. 2002, \apj, 571, 378
421:
422: \bibitem[{{Boden} {et~al.}(1998){Boden}, {Colavita}, {van Belle}, \&
423: {Shao}}]{BODEN+98}
424: {Boden}, A.~F., {Colavita}, M.~M., {van Belle}, G.~T., \& {Shao}, M. 1998,
425: in Proc. SPIE Vol. 3350, Robert D. Reasenberg, ed., 872
426:
427: %\bibitem[{{Brittain} {et~al.}(2005){Brittain}, {Rettig}, {Simon}, \&
428: % {Kulesa}}]{BRITTAIN+05}
429: %{Brittain}, S.~D., {Rettig}, T.~W., {Simon}, T., \& {Kulesa}, C. 2005, \apj,
430: % 626, 283
431:
432: \bibitem[{{Bryden} {et~al.}(1999){Bryden}, {Chen}, {Lin}, {Nelson}, \&
433: {Papaloizou}}]{BRYDEN+99}
434: {Bryden}, G., {Chen}, X., {Lin}, D.~N.~C., {Nelson}, R.~P., \& {Papaloizou},
435: J.~C.~B. 1999, \apj, 514, 344
436:
437: \bibitem[{{Calvet} {et~al.}(2002){Calvet}, {D'Alessio}, {Hartmann}, {Wilner},
438: {Walsh}, \& {Sitko}}]{CALVET+02}
439: {Calvet}, N., {D'Alessio}, P., {Hartmann}, L., {Wilner}, D., {Walsh}, A., \&
440: {Sitko}, M. 2002, \apj, 568, 1008
441:
442: \bibitem[{{Chiang}(2003)}]{CHIANG03}
443: {Chiang}, E.~I. 2003, \apj, 584, 465
444:
445: \bibitem[{{Colavita} {et~al.}(2003){Colavita}, {Akeson}, {Wizinowich}, {Shao},
446: {Acton}, {Beletic}, {Bell}, {Berlin}, {Boden}, {Booth}, {Boutell}, {Chaffee},
447: {Chan}, {Chock}, {Cohen}, {Crawford}, {Creech-Eakman}, {Eychaner},
448: {Felizardo}, {Gathright}, {Hardy}, {Henderson}, {Herstein}, {Hess},
449: {Hovland}, {Hrynevych}, {Johnson}, {Kelley}, {Kendrick}, {Koresko}, {Kurpis},
450: {Le Mignant}, {Lewis}, {Ligon}, {Lupton}, {McBride}, {Mennesson},
451: {Millan-Gabet}, {Monnier}, {Moore}, {Nance}, {Neyman}, {Niessner}, {Palmer},
452: {Reder}, {Rudeen}, {Saloga}, {Sargent}, {Serabyn}, {Smythe}, {Stomski},
453: {Summers}, {Swain}, {Swanson}, {Thompson}, {Tsubota}, {Tumminello}, {van
454: Belle}, {Vasisht}, {Vause}, {Walker}, {Wallace}, \& {Wehmeier}}]{COLAVITA+03}
455: {Colavita}, M.~M., et al. 2003, \apjl, 592, L83
456:
457: \bibitem[{{Colavita}(1999)}]{COLAVITA99}
458: {Colavita}, M.~M. 1999, \pasp, 111, 111
459:
460: \bibitem[{{Colavita} \& {Wizinowich}(2003)}]{CW03}
461: {Colavita}, M.~M. \& {Wizinowich}, P.~L. 2003,
462: in Proc. SPIE Vol. 4838, Wesley A. Traub, ed., 79
463:
464: \bibitem[{{Edwards} {et~al.}(1994){Edwards}, {Hartigan}, {Ghandour}, \&
465: {Andrulis}}]{EDWARDS+94}
466: {Edwards}, S., {Hartigan}, P., {Ghandour}, L., \& {Andrulis}, C. 1994, \aj,
467: 108, 1056
468:
469: \bibitem[{{Eisner} {et~al.}(2005){Eisner}, {Hillenbrand}, {White}, {Akeson}, \&
470: {Sargent}}]{EISNER+05}
471: {Eisner}, J.~A., {Hillenbrand}, L.~A., {White}, R.~J., {Akeson}, R.~L., \&
472: {Sargent}, A.~I. 2005, \apj, 623, 952
473:
474: \bibitem[{{Eisner} {et~al.}(2004){Eisner}, {Lane}, {Hillenbrand}, {Akeson}, \&
475: {Sargent}}]{EISNER+04}
476: {Eisner}, J.~A., {Lane}, B.~F., {Hillenbrand}, L., {Akeson}, R., \& {Sargent},
477: A. 2004, \apj, 613, 1049
478:
479: \bibitem[{{Glassgold} {et~al.}(2004){Glassgold}, {Najita}, \& {Igea}}]{GNI04}
480: {Glassgold}, A.~E., {Najita}, J., \& {Igea}, J. 2004, \apj, 615, 972
481:
482: \bibitem[{{Goldreich} \& {Tremaine}(1982)}]{GT82}
483: {Goldreich}, P. \& {Tremaine}, S. 1982, \araa, 20, 249
484:
485: \bibitem[{{Herczeg} {et~al.}(2004){Herczeg}, {Wood}, {Linsky}, {Valenti}, \&
486: {Johns-Krull}}]{HERCZEG+04}
487: {Herczeg}, G.~J., {Wood}, B.~E., {Linsky}, J.~L., {Valenti}, J.~A., \&
488: {Johns-Krull}, C.~M. 2004, \apj, 607, 369
489:
490: \bibitem[{{Johns-Krull} \& {Valenti}(2001)}]{JV01}
491: {Johns-Krull}, C.~M. \& {Valenti}, J.~A. 2001, \apj, 561, 1060
492:
493: \bibitem[{{K\"{o}nigl}(1991)}]{KONIGL91}
494: {K\"{o}nigl}, A. 1991, \apjl, 370, L39
495:
496: \bibitem[{{Lin} \& {Papaloizou}(1986)}]{LP86}
497: {Lin}, D.~N.~C. \& {Papaloizou}, J. 1986, \apj, 309, 846
498:
499: \bibitem[{{Mamajek}(2005)}]{MAMAJEK05}
500: {Mamajek}, E.~E. 2005, ApJ, 634, 1385
501:
502: \bibitem[{{Miyake} \& {Nakagawa}(1993)}]{MN93}
503: {Miyake}, K. \& {Nakagawa}, Y. 1993, Icarus, 106, 20
504:
505: \bibitem[{{Muzerolle} {et~al.}(2000){Muzerolle}, {Calvet}, {Brice{\~n}o},
506: {Hartmann}, \& {Hillenbrand}}]{MUZEROLLE+00}
507: {Muzerolle}, J., {Calvet}, N., {Brice{\~n}o}, C., {Hartmann}, L., \&
508: {Hillenbrand}, L. 2000, \apjl, 535, L47
509:
510: \bibitem[{{Najita}(2006)}]{NAJITA06}
511: {Najita}, J. 2006, in Proceedings IAU Symposium No. 231, D.C. Lis, G.A. Blake,
512: \& E. Herbst, eds., in press
513:
514: \bibitem[{{Pollack} {et~al.}(1994){Pollack}, {Hollenbach}, {Beckwith},
515: {Simonelli}, {Roush}, \& {Fong}}]{POLLACK+94}
516: {Pollack}, J.~B., {Hollenbach}, D., {Beckwith}, S., {Simonelli}, D.~P.,
517: {Roush}, T., \& {Fong}, W. 1994, \apj, 421, 615
518:
519: \bibitem[{{Qi} {et~al.}(2004){Qi}, {Ho}, {Wilner}, {Takakuwa}, {Hirano},
520: {Ohashi}, {Bourke}, {Zhang}, {Blake}, {Hogerheijde}, {Saito}, {Choi}, \&
521: {Yang}}]{QI+04}
522: {Qi}, C., et al. 2004, \apjl, 616, L11
523:
524: \bibitem[{{Rettig} {et~al.}(2004){Rettig}, {Haywood}, {Simon}, {Brittain}, \&
525: {Gibb}}]{RETTIG+04}
526: {Rettig}, T.~W., {Haywood}, J., {Simon}, T., {Brittain}, S.~D., \& {Gibb}, E.
527: 2004, \apjl, 616, L163
528:
529: \bibitem[{{Rice} {et~al.}(2003){Rice}, {Wood}, {Armitage}, {Whitney}, \&
530: {Bjorkman}}]{RICE+03}
531: {Rice}, W.~K.~M., {Wood}, K., {Armitage}, P.~J., {Whitney}, B.~A., \&
532: {Bjorkman}, J.~E. 2003, \mnras, 342, 79
533:
534: %\bibitem[{{Roberge} {et~al.}(2001){Roberge}, {Lecavelier des Etangs}, {Grady},
535: % {Vidal-Madjar}, {Bouret}, {Feldman}, {Deleuil}, {Andre}, {Boggess},
536: % {Bruhweiler}, {Ferlet}, \& {Woodgate}}]{ROBERGE+01}
537: %{Roberge}, A., et al. 2001, \apjl, 551, L97
538:
539: \bibitem[{{Shakura} \& {Sunyaev}(1973)}]{SS73}
540: {Shakura}, N.~I. \& {Sunyaev}, R.~A. 1973, \aap, 24, 337
541:
542: \bibitem[{{Sitko} {et~al.}(2000){Sitko}, {Lynch}, \& {Russell}}]{SLR00}
543: {Sitko}, M.~L., {Lynch}, D.~K., \& {Russell}, R.~W. 2000, \aj, 120, 2609
544:
545: \bibitem[{{Uchida} {et~al.}(2004){Uchida}, {Calvet}, {Hartmann}, {Kemper},
546: {Forrest}, {Watson}, {D'Alessio}, {Chen}, {Furlan}, {Sargent}, {Brandl},
547: {Herter}, {Morris}, {Myers}, {Najita}, {Sloan}, {Barry}, {Green}, {Keller},
548: \& {Hall}}]{UCHIDA+04}
549: {Uchida}, K.~I., et al. 2004, \apjs, 154, 439
550:
551: \bibitem[{{Ward}(1997)}]{WARD97}
552: {Ward}, W.~R. 1997, \apjl, 482, L211
553:
554: \bibitem[{{Webb} {et~al.}(1999){Webb}, {Zuckerman}, {Platais}, {Patience},
555: {White}, {Schwartz}, \& {McCarthy}}]{WEBB+99}
556: {Webb}, R.~A., {Zuckerman}, B., {Platais}, I., {Patience}, J., {White}, R.~J.,
557: {Schwartz}, M.~J., \& {McCarthy}, C. 1999, \apjl, 512, L63
558:
559: \bibitem[{{Weidenschilling}(1977)}]{WEIDENSCHILLING77}
560: {Weidenschilling}, S.~J. 1977, \mnras, 180, 57
561:
562: \bibitem[{{Weinberger} {et~al.}(2002){Weinberger}, {Becklin}, {Schneider},
563: {Chiang}, {Lowrance}, {Silverstone}, {Zuckerman}, {Hines}, \&
564: {Smith}}]{WEINBERGER+02}
565: {Weinberger}, A.~J., et al. 2002, \apj, 566, 409
566:
567: \bibitem[{{Wilner} {et~al.}(2005){Wilner}, {D'Alessio}, {Calvet}, {Claussen},
568: \& {Hartmann}}]{WILNER+05}
569: {Wilner}, D.~J., {D'Alessio}, P., {Calvet}, N., {Claussen}, M.~J., \&
570: {Hartmann}, L. 2005, \apjl, 626, L109
571:
572: \bibitem[{{Yang} {et~al.}(2005){Yang}, {Johns-Krull}, \& {Valenti}}]{YJV05}
573: {Yang}, H., {Johns-Krull}, C.~M., \& {Valenti}, J.~A. 2005, ApJ, in press
574:
575: \end{thebibliography}
576:
577: \epsscale{1.0}
578: \begin{figure}
579: \plotone{f1.eps}
580: \caption{Uncalibrated, squared visibilities ($V^2$) measured by KI for
581: TW Hya and three unresolved calibrator sources. The smaller $V^2$ value
582: of TW Hya relative to the calibrators indicates that this source is angularly
583: resolved. Scatter in the uncalibrated $V^2$ is due to a combination
584: of instrumental and atmospheric effects, including variable performance
585: of the Keck AO systems and phase jitter arising from atmospheric motions
586: and instrumental vibrations.
587: \label{fig:rawv2}}
588: \end{figure}
589:
590: \epsscale{1.0}
591: \begin{figure}
592: \plottwo{f2a.eps}{f2b.eps}
593: \caption{Squared visibilities (left) and near-IR
594: fluxes (right) for a model consisting of a young star surrounded by an
595: optically thin disk,
596: compared to the data \citep[this work;][]{WEBB+99,SLR00}.
597: The flux of the central star is indicated by the solid gray line.
598: Models with large dust grains ($\beta=0$) can not fit the
599: data well, in contrast to models with
600: sub-micron-sized dust grains ($\beta = 1$).
601: \label{fig:mods}}
602: \end{figure}
603:
604:
605:
606: %\clearpage
607: \begin{deluxetable}{lccccccc}
608: %\tabletypesize{\scriptsize}
609: %\tabletypesize{\tiny}
610: \tablewidth{0pt}
611: \tablecaption{Optically thin disk models \label{tab:models}}
612: \tablehead{\colhead{$ $} & \colhead{$\chi_{\rm r}^2$} & \colhead{$R_{\rm in}$}
613: & \colhead{$R_{\rm out}$} & \colhead{$\Sigma_0$} & \colhead{$T_{\rm in}$} &
614: \colhead{$M_{\rm dust}$} & \colhead{$\tau_{K, \rm in}$}
615: \\ & & (AU) & (AU) & (g cm$^{-2}$) & (K) & (g) & }
616: \startdata
617: $\beta=0$ & 1.17 & $0.02 \pm 0.01$ & 4 & $2.0^{+2.0}_{-0.1} \times 10^{-6}$ &
618: 1400 & $1 \times 10^{22}$ & 0.07 \\
619: $\beta=1$ & 0.98 & $0.06 \pm 0.01$ & 4 & $6.3 \pm 1.5 \times 10^{-7}$ &
620: 1120 & $3 \times 10^{21}$ & 0.04 \\
621: %$\beta=2$ & 0.91 & $0.24 \pm 0.05$ &
622: %$1.6 \pm 0.4 \times 10^{-6}$ & 870 & 0.01 & $ 1 \times 10^{21}$
623: \enddata
624: \tablecomments{$R_{\rm in}$ is the best-fit inner radius
625: and $\Sigma_0$ is the dust surface density at $R=1$ AU. The outer
626: disk radius $R_{\rm out}$ is fixed for all models.
627: The inner disk temperature
628: $T_{\rm in}$, dust mass $M_{\rm dust}$,
629: %radial optical depth $\tau_{\rm r}$,
630: and 2 $\mu$m vertical optical depth at the inner edge $\tau_{K,\rm in}$,
631: are computed for the best-fit values of $R_{\rm in}$ and $\Sigma_0$.
632: The value of $M_{\rm dust}$ depends on the assumed surface density profile,
633: $\Sigma \propto R^{-3/2}$; different assumptions regarding this profile
634: yield dust masses that can differ by orders of magnitude.}
635: \end{deluxetable}
636:
637: \end{document}
638: