0706.0279/ms.tex
1: \documentclass{emulateapj}
2: 
3: \begin{document}
4: 
5: \shortauthors{Luhman et al.}
6: \shorttitle{Edge-on Disk around a Brown Dwarf}
7: 
8: \title{{\it Hubble} and {\it Spitzer} Observations of an Edge-on Circumstellar
9: Disk around a Brown Dwarf\altaffilmark{1}}
10: 
11: \author{
12: K. L. Luhman\altaffilmark{2,3},
13: Luc\'\i a Adame\altaffilmark{4},
14: Paola D'Alessio\altaffilmark{5}, 
15: Nuria Calvet\altaffilmark{6}, 
16: Kim K. McLeod\altaffilmark{7},
17: C. J. Bohac\altaffilmark{8},
18: William J. Forrest\altaffilmark{8},
19: Lee Hartmann\altaffilmark{6}, 
20: B. Sargent\altaffilmark{8},
21: \& Dan M. Watson\altaffilmark{8}
22: }
23: 
24: \altaffiltext{1}{
25: Based on observations made with the NASA/ESA {\it Hubble Space
26: Telescope} and the {\it Spitzer Space Telescope}.
27: The {\it Hubble} observations are associated with proposal ID 10511 and 
28: were obtained at the Space Telescope Science Institute, which is operated by 
29: the Association of Universities for Research in Astronomy, Inc., under NASA 
30: contract NAS 5-26555. {\it Spitzer} is operated by the Jet Propulsion 
31: Laboratory at the California Institute of Technology under NASA contract 1407.}
32: 
33: \altaffiltext{2}{Department of Astronomy and Astrophysics,
34: The Pennsylvania State University, University Park, PA;
35: kluhman@astro.psu.edu.}
36: 
37: \altaffiltext{3}{Visiting Astronomer at the Infrared Telescope Facility, which
38: is operated by the University of Hawaii under Cooperative Agreement no. NCC
39: 5-538 with the National Aeronautics and Space Administration, Office of Space
40: Science, Planetary Astronomy Program.}
41: 
42: \altaffiltext{4}{Instituto de Astronom\'\i a, Universidad Nacional
43:  Aut\'onoma de M\'exico, Apartado Postal 70-264, 
44: Ciudad Universitaria, M\'exico DF, M\'exico.}
45: 
46: \altaffiltext{5}{Centro de Radioastronom\'\i a y Astrof\'\i sica, 
47: Universidad Nacional Aut\'onoma de M\'exico, Apartado Postal 72-3 (Xangari), 
48: Morelia, Michoac\'an, M\'exico.}
49: 
50: \altaffiltext{6}{Department of Astronomy, The University of Michigan, 
51: 500 Church Street, 830 Dennison Building, Ann Arbor, MI.}
52: 
53: \altaffiltext{7}{Whitin Observatory, Wellesley College, Wellesley, MA.}
54: 
55: \altaffiltext{8}{Department of Physics and Astronomy, 
56: The University of Rochester, Rochester, NY.}
57: 
58: \begin{abstract}
59: 
60: 
61: We present observations of a circumstellar disk 
62: that is inclined close to edge-on around a young brown dwarf 
63: in the Taurus star-forming region. 
64: Using data obtained with SpeX at the NASA Infrared Telescope Facility, 
65: we find that the slope of the 0.8-2.5~\micron\ spectrum of the brown dwarf
66: 2MASS~J04381486+2611399 cannot be reproduced with a photosphere reddened by 
67: normal extinction. Instead, the slope is consistent with scattered light, 
68: indicating that circumstellar material is occulting the brown dwarf. 
69: By combining the SpeX data with mid-infrared photometry and 
70: spectroscopy from the {\it Spitzer Space Telescope} and previously published
71: millimeter data from Scholz and coworkers, we construct the spectral energy
72: distribution for 2MASS~J04381486+2611399 and model it in terms of a young 
73: brown dwarf surrounded by an irradiated accretion disk. 
74: The presence of both silicate absorption at 10~\micron\ and silicate emission
75: at 11~\micron\ constrains the inclination of the disk to be $\sim70\arcdeg$,
76: i.e. $\sim20\arcdeg$ from edge-on.
77: Additional evidence of the high inclination of this disk is provided by
78: our detection of asymmetric bipolar extended emission surrounding 
79: 2MASS~J04381486+2611399 in high-resolution optical images obtained with 
80: the {\it Hubble Space Telescope}.
81: According to our modeling for the SED and images of this system, the disk
82: contains a large inner hole that is indicative of a transition disk
83: ($R_{in}\approx58R_\star\approx0.275$~AU) and is somewhat larger than expected 
84: from embryo ejection models ($R_{out}=20$-40~AU vs. $R_{out}<10$-20~AU). 
85: 
86: 
87: 
88: 
89: 
90: 
91: 
92: \end{abstract}
93: 
94: \keywords{accretion disks -- planetary systems: protoplanetary disks -- stars:
95: formation --- stars: low-mass, brown dwarfs --- stars: pre-main sequence}
96: 
97: \section{Introduction}
98: \label{sec:intro}
99: 
100: 
101: 
102: Measurements of the properties of circumstellar accretion disks around young 
103: stars are important because they represent constraints on the initial 
104: conditions of planet formation. 
105: A unique set of measurements can be performed on a disk in the rare case
106: in which it is seen close to edge-on.
107: Because an edge-on disk occults the central star, scattered light
108: from the disk surface dominates the total emergent flux at optical and 
109: near-infrared (IR) wavelengths, making it possible to spatially resolve the disk
110: with high-resolution imaging \citep{bur96}. At mid-IR wavelengths, the outer
111: disk is seen in absorption against the star and inner disk, allowing 
112: spectroscopic measurements of the disk composition \citep{wat04,pon05}.
113: The sample of edge-on disks discovered to date remains small. 
114: Some of the most notable examples are
115: Orion~114-426 \citep{mcc96}, Haro~6-5B \citep{kri98}, HH~30 \citep{bur96}, 
116: IRAS~04302+2247 \citep{luc97,pad99}, DG~Tau~B \citep{pad99},
117: OphE-MM3, CRBR 2422.8-3423 \citep{bra00}, HK~Tau~B \citep{sta98,kor98}, 
118: HV~Tau~C \citep{mon00}, LkH$\alpha$~263~C \citep{jay02,cha02},
119: 2MASS~J1628137-243139 \citep{gro03}, and PDS~144 \citep{per06}.
120: Additional young stars in Taurus exhibit dust lanes that may also
121: trace edge-on disks \citep{har99,pad99}. 
122: 
123: 
124: Because young stars with edge-on disks are seen only in scattered light
125: at optical and near-IR wavelengths,
126: they appear much fainter than unocculted young stars at a given spectral type. 
127: Recent spectroscopic surveys of nearby star-forming regions have identified
128: a number of objects that are subluminous in this manner 
129: \citep{fer01,luh03,sle04}.  
130: One of these sources, 2MASS~J04381486+2611399 in Taurus (henceforth
131: 2M~0438+2611), appears to be a brown dwarf with a mass near 0.05~$M_\odot$ 
132: \citep{luh04tau}.
133: This brown dwarf also exhibits anomalous near-IR colors and strong forbidden 
134: line emission, which are characteristics that are frequently observed in stars
135: with edge-on disks.
136: To definitively establish whether 2M~0438+2611 has an edge-on disk, 
137: we have observed it with near- and mid-IR spectroscopy and high-resolution 
138: optical imaging.
139: In this paper, we present these new data (\S~\ref{sec:obs}) 
140: and fit them with the predictions of accretion disk models to constrain 
141: the inclination and other physical properties of the disk around 2M~0438+2611
142: (\S~\ref{sec:model}).
143: 
144: 
145: 
146: 
147: 
148: \section{Observations and Analysis}
149: \label{sec:obs}
150: 
151: \subsection{Near-infrared Spectroscopy}
152: \label{sec:spex}
153: 
154: To investigate the possibility that 2M~0438+2611 is seen in scattered light,
155: we first obtained a low-resolution near-IR spectrum of it with SpeX 
156: \citep{ray03} at the NASA Infrared Telescope Facility (IRTF). These data
157: were collected on the night of 2004 November 12. They were reduced with the 
158: Spextool package \citep{cus04} and corrected for telluric absorption 
159: \citep{vac03}. The final spectrum extends from 0.8-2.5~\micron\ and 
160: exhibits a resolving power of $R=100$.
161: 
162: 
163: For most of the young brown dwarfs that we have observed in previous studies
164: \citep[e.g.,][]{luh06tau2},
165: the differences in the slopes of their 0.8-2.5~\micron\ spectra are 
166: consistent with differences in extinction. However, this is not the case for
167: 2M~0438+2611, which has an anomalous slope that differs
168: significantly from those of other objects, regardless of any correction for
169: extinction using standard reddening laws \citep{rl85,car89}.
170: This behavior is illustrated in Figure~\ref{fig:spec},
171: where we compare 2M~0438+2611 to a typical
172: unreddened young brown dwarf with the same optical spectral 
173: type\footnote{For the standard spectrum, we use SpeX data for 
174: 2MASS~04442713+2512164 \citep[henceforth 2M~0444+2512,][]{luh04tau} after 
175: dereddening it by $A_V=2.3$, which is the extinction implied by a comparison to 
176: other young brown dwarfs and to dwarf standards. Based on that comparison,
177: the SpeX data for 2M~0444+2512 do not exhibit excess emission from a disk.}
178: The standard was artificially reddened according to 
179: the reddening law of \citet{car89} to the point that it has the same relative
180: fluxes at 0.8 and 2.5~$\micron$ as 2M~0438+2611. However, the shape of this
181: reddened spectrum between 0.8 and 2.5~$\micron$ differs significantly from 
182: that of 2M~0438+2611.
183: An alternative demonstration of this effect is shown in 
184: Figure~\ref{fig:spec} through the quotient these two spectra, which departs
185: from the constant value expected for normal reddening.
186: Thus, reddening cannot explain the observed slope of 2M~0438+2611.
187: Instead, relative to the flux at 0.8~\micron, the spectrum of 2M~0438+2611 
188: becomes redder more slowly with longer wavelengths than expected from standard 
189: extinction laws, which is 
190: consistent with the presence of (blue) scattered light in the spectrum.
191: Meanwhile, the absence of significant residuals in the quotient of 2M~0438+2611 
192: and the standard demonstrates the close match in the depths of the steam 
193: absorption bands, supporting the similarity in spectral types indicated by 
194: previous optical spectroscopy. 
195: 
196: \subsection{Mid-infrared Spectroscopy}
197: 
198: 
199: Photometry at 2~\micron\ from the Two-Micron All-Sky Survey 
200: \citep[2MASS,][]{skr06}, at 3-8~\micron\ from the Infrared Array
201: Camera \citep[IRAC;][]{faz04} aboard the {\it Spitzer Space Telescope}
202: \citep{wer04}, and at 1.3~mm from the 30~m telescope at the Institut de Radio 
203: Astronomie Millim{\'e}trique (IRAM)
204: have previously revealed excess emission in 2M~0438+2611
205: that is indicative of circumstellar dust \citep{luh04tau,luh06tau2,sch06}.
206: To better constrain the properties of this material,
207: we obtained a mid-IR spectrum of 2M~0438+2611 on 2005 March 19
208: with the {\it Spitzer} Infrared Spectrograph \citep[IRS;][]{hou04} as a part 
209: of the Guaranteed Time Observations of the IRS instrument team.  
210: These observations (AOR 12705792) were performed with the short-wavelength, 
211: low-resolution module of IRS, providing data from 5.3-14~\micron\ with a 
212: resolving power of $R=90$. The spectrum was processed with the S14 
213: pipeline at the {\it Spitzer} Science Center. The remaining reduction 
214: was performed with the methods that have been previously 
215: applied to IRS data for other low-mass members of Taurus \citep{fur05}.
216: We have also measured photometry at 24~\micron\ for 2M~0438+2611 from
217: archival images obtained with the Multiband Imaging Photometer for
218: {\it Spitzer} \citep[MIPS;][]{rie04}. Using the methods described
219: by \citet{all07}, we measured a flux of $62.5\pm3.3$~mJy from these MIPS data.
220: 
221: We present the spectra from SpeX and IRS and the photometry from 2MASS, IRAC,
222: MIPS, and IRAM for 2M~0438+2611 in Figure~\ref{fig:sed}. 
223: The IRAC and IRS data agree well with each other, 
224: while the 2MASS photometry and the SpeX data differ significantly in 
225: both color and flux level.  The SpeX data are consistent with a
226: smooth extension of the IRAC and IRS data, while the 2MASS measurements
227: appear to be discontinuous from the latter. 
228: The IRAC and IRS observations were performed only 4 months after the 
229: SpeX observations, while the 2MASS data were obtained 6 years earlier.
230: Thus, one explanation for the discrepancy in the 2MASS photometry relative
231: to the other data is variability, which is plausible for any young star
232: and is particularly likely for an object that is seen in scattered light
233: because of changes in the geometry of the occulting material.
234: Therefore, we exclude the 2MASS photometry for the purposes of modeling 
235: the spectral energy distribution (SED) of 2M~0438+2611 in \S~\ref{sec:model}.
236: 
237: 
238: \subsection{Optical Imaging}
239: \label{sec:wfpc2}
240: 
241: Because of the initial evidence indicating that the SED of 2M~0438+2611 might
242: be dominated by scattered light at optical and near-IR wavelengths 
243: \citep[][\S~\ref{sec:spex}]{luh04tau}, we sought to detect spatially resolved 
244: scattered light through high-resolution broad-band optical imaging with the 
245: {\it Hubble Space Telescope}. In addition, given the presence of 
246: forbidden line emission in optical spectra of this brown dwarf \citep{luh04tau},
247: we performed narrowband imaging centered on [O~I] at 6300~\AA\ in an attempt 
248: to detect resolved line emission from a jet or an outflow. Using the Wide Field 
249: Planetary Camera (WFPC2) aboard {\it Hubble}, we obtained images of
250: 2M~0438+2611 through the F675W, F791W, F850LP, and F631N filters
251: on 2005 October 22. The target was placed near the center of the PC, which
252: has a plate scale of $0.046\arcsec$~pixel$^{-1}$. Two images were taken in each 
253: filter, each with exposure times of 160~sec for F675W, F791W, and F850LP and 
254: 300~sec for F631N. Each pair of images at a given filter was combined 
255: to create a single image.
256: 
257: In the [O~I] image, 2M~0438+2611 is unresolved and no extended emission 
258: is detected. Meanwhile, each of the three broad-band images reveal 
259: both a point source and spatially resolved emission, as shown in 
260: Figure~\ref{fig:image}. The extended emission is elongated and reaches 
261: 0.4-$0.5\arcsec$ from one side of the point source at a position angle of 
262: $245\arcdeg$.  Analysis of these images with the point-spread-function of WFPC2 
263: indicates that a small amount ($\sim0.1\arcsec$) of extended emission is present
264: on the opposite side of the point source as well. 
265: 
266: 
267: 
268: 
269: 
270: \section{Disk Model}
271: \label{sec:model}
272: 
273: \subsection{Model Parameters}
274: \label{sec:para}
275: 
276: 
277: We have modeled the SED of 2M~0438+2611 in Figure~\ref{fig:sed} following the 
278: procedures from \citet{dal98,dal99,dal01,dal06}. 
279: In short, we solve the equations for the disk vertical structure, assuming 
280: it is an $\alpha$-disk heated by viscous dissipation and by stellar 
281: irradiation. The relatively high flux at 1.3~mm of this object 
282: \citep{sch06} suggests the presence of grains that are larger than typical 
283: grains in the interstellar medium. At the same time, the presence of the 
284: 10~\micron\ silicate band and the shape of the extinction of the stellar SED
285: are indicative of small grains. Therefore, we have constructed a disk 
286: model in which a large fraction of the dust has settled in the midplane,
287: growing to a maximum size of $\sim1$~mm, while a small fraction of the
288: dust remains in the upper layers in the form of small interstellar-like 
289: grains \citep{dal06}. 
290: The dust is assumed to be segregated spheres of ``astronomical'' silicates
291: and graphite with abundances and optical constants from \citet{dl84}
292: and \citet{wl01} and a size distribution of $n(a)\sim a^{-3.5}$ where 
293: $a$ is the grain radius \citep{mrn}.
294: For the small grains in the upper layers, the minimum and maximum grain
295: radii are $a_{min}=0.0005$~\micron\ and $a_{max}=0.25$~\micron.
296: For the large grains at the disk midplane, we use 
297: $a_{min}=0.0005$~\micron\ and $a_{max}=1$~mm. The dust-to-gas mass
298: ratio of the small grains is parameterized in terms of $\epsilon$,
299: which is the ratio normalized by the standard interstellar value of $\sim0.01$. 
300: We have calculated models for $0.01<\epsilon<1$. The dust-to-gas mass ratio 
301: of the large grains at the midplane is calculated assuming that the total mass 
302: in grains is conserved at each radius.
303: 
304: 
305: In our model, the dusty disk is truncated at a radius $R_{in}$, where the
306: inner wall of the disk receives radiation from the brown dwarf and accretion
307: shocks at the stellar surface with a normal incidence. A natural
308: explanation for a wall of this kind is that the silicates are sublimated
309: inside $R_{in}$ \citep{dul01,muz04}. In this case, the temperature at the 
310: inner edge of the disk is $T_{wall}=T(R_{in})\sim1400$~K.  
311: However, disks can be truncated at larger radii, corresponding to lower 
312: temperatures (i.e., transitional disks). Therefore, we have modeled the wall
313: emission following the procedures of \citep{dal05}, varying the temperature 
314: of the optically thin dust in the wall from 300 to 1400~K.
315: We have adopted an accretion rate of $3\times10^{-11}$~$\rm M_{\odot}\,yr^{-1}$
316: for our disk model, which is similar to the value derived by \citet{muz05} for
317: 2M~0438+2611 through modeling of the profile of its H$\alpha$ emission line.
318: 
319: 
320: 
321: Although the disk surface density, $\Sigma$, is not an input parameter for our 
322: models, we are able to modify it through the viscosity parameter $\alpha$ since 
323: $\Sigma\sim\frac{\dot{M}}{\alpha}$ for an $\alpha$-disk. 
324: When fitting the millimeter flux, decreasing $\alpha$ (i.e., increasing
325: $\Sigma$) has the effect of decreasing the outer radius of the disk.
326: After exploring models for $10^{-6}\leq\alpha\leq10^{-2}$,
327: we find that the observed millimeter flux constrains $\alpha$ to the low 
328: end of this range if the accretion rate in the outer disk is the same as 
329: the accretion rate onto the brown dwarf \footnote{Disks around T Tauri stars 
330: have been modeled with $\alpha\sim0.01$ \citep{dal98}.}.
331: For a brown dwarf disk, $\alpha\leq10^{-4}$ implies a viscous 
332: timescale of $t_\nu\gtrsim25$~Myr, which is too long to justify our
333: assumption of a steady disk with a constant accretion rate throughout the disk.
334: It is possible that the accretion rate increases with disk radius,
335: as in the disk model from \citet{gam96}. A higher accretion rate in the
336: outer disk would correspond to lower $\alpha$ and shorter $t_\nu$. 
337: In a disk of this kind, material would accumulate in the inner disk, 
338: perhaps in a dead zone. Exploring this possibility in detail would require
339: a disk model that includes a dead zone.
340: 
341: For the photosphere of 2M~0438+2611, we have adopted an effective temperature 
342: of 2838~K \citep{luh04tau}. Because this object is seen in scattered light 
343: at optical and near-IR wavelengths \citep[][\S~\ref{sec:spex}]{luh04tau}, 
344: its extinction cannot be measured from its colors. As a result, 
345: we cannot reliably measure its luminosity with the normal method of 
346: applying an extinction correction to photometry in these bands. 
347: Therefore, we have performed the disk calculations for a range
348: of luminosities that are typical of members of Taurus near the spectral
349: type of 2M~0438+2611, namely $L_{\rm bol}=0.04$, 0.06, 0.08, and 0.1~$L_\odot$. 
350: For 2M~0438+2611, we adopt a mass of 0.05~$M_\odot$, which is the value 
351: implied by its spectral type for a member of Taurus \citep{luh04tau}.
352: As shown in \citet{luh04tau} and \S~\ref{sec:spex}, 2M~0444+2512
353: has the same optical and IR spectral types as 2M~0438+2611. In addition, 
354: although it exhibits mid-IR excess emission that indicates the presence
355: of a disk \citep{luh06tau2}, 2M~0444+2512 does not show excess emission
356: at wavelengths shortward of 2.5~\micron\ in a comparison to SpeX data 
357: for diskless brown dwarfs near the same spectral type \citep{luh06tau1}. 
358: Therefore, we have adopted the extinction-corrected SpeX data for 2M~0444+2512
359: described in \S~\ref{sec:spex} to represent the 0.8-2.5~\micron\ SED of the 
360: photosphere of 2M~0438+2611.
361: We assume that both 2M~0438+2611 and 2M~0444+2512 are at the average 
362: distance of members of Taurus \citep[$d=140$~pc,][]{wic98}. 
363: We measured the average colors between 2MASS $K_s$ and the IRAC bands
364: (3.6, 4.5, 5.8, 8.0~\micron) for diskless late-type members of Taurus 
365: \citep{har05,luh06tau2} and applied them to the SpeX data for 2M~0444+2512
366: to extend the photospheric SED to 8.0~\micron. The SED was extrapolated to 
367: wavelengths beyond 8.0~\micron\ with a Rayleigh-Jeans distribution.
368: 
369: 
370: 
371: 
372: 
373: 
374: 
375: \subsection{Best-fit Model}
376: 
377: A given portion of the observed SED provides constraints on specific 
378: properties of the disk. 
379: We use the $K$-band flux to estimate the optical depth to the brown dwarf, 
380: which in turn constrains the inclination, outer radius, and viscosity parameter
381: of the disk. The silicate feature near 10~\micron\ is sensitive to both 
382: the inclination and the presence of small grains in the upper layers
383: of the disk. The abundance of these small grains is constrained
384: by the 24~\micron\ flux relative to the emission at shorter wavelengths. 
385: The flux at 3-10~\micron\ depends on the inner radius 
386: of the disk and the stellar luminosity, while the millimeter flux is 
387: determined by a combination of the outer radius and the surface density.
388: In this section, we discuss each of these constraints in detail 
389: for 2M~0438+2611 and present the resulting best-fit model for its disk.
390: 
391: 
392: The best fit to the flux at 3-10~\micron\ is provided by a stellar luminosity
393: of 0.06~$L_\odot$ (corresponding to $R_*=1.02$~$R_\odot$) and a 
394: radius of $R_{in}\approx58R_\star\approx0.275$~AU for the wall, which 
395: corresponds to a wall temperature of 400~K.
396: For models with the wall placed at the dust destruction radius, the predicted 
397: fluxes at 3-10~\micron\ are brighter than the observed values. 
398: Thus, our modeling indicates that the disk is inwardly 
399: truncated. The inner disk of 2M~0438+2611 shows the same physical
400: properties as disks that have been identified in the
401: literature as ``transitional disks'', and which have been
402: shown to have truncated optically thick disks at various
403: distances from the central stars, from both SED modeling
404: \citep{cal02,cal05,dal05,muz06} and millimeter interferometry \citep{hug07}.
405: Disks of this kind are thought to harbor forming planets that are 
406: opening gaps in the disk \citep{ric03,qui04}. Photoevaporation is an
407: alternative explanation for these disk gaps, but it is unlikely for
408: 2M~0438+2611 given the low ultraviolet fluxes that are expected from a
409: brown dwarf \citep{muz06}. 
410: 
411: For all reasonable choices of model parameters, the reddened flux from the 
412: stellar photosphere should dominate the total emergent flux at 
413: $\sim2$-3~\micron. Combining the luminosity of 0.06~$L_\odot$ and the 
414: observed $K$-band flux produces a value of $\tau_{2.25\mu m}\approx1.3$ 
415: for the optical depth to the photosphere.
416: 
417: The 10~\micron\ silicate profile of 2M~0438+2611 is distinctive, showing
418: both absorption and emission components. This profile tightly constrains
419: the disk inclination because small angles produce only emission and large
420: ones produce only absorption, as illustrated for the face-on and edge-on 
421: disks around FM~Tau and DG~Tau~B in Figure~\ref{fig:sed2} \citep{wat04,fur06}.
422: For $i\sim70\arcdeg$, the model predicts silicate emission from the disk 
423: and silicate absorption from the highly extinguished wall that combine
424: to form absorption and emission features at 10 and 11~\micron, which closely 
425: matches the data, as shown in Figure \ref{fig:sed2}.
426: While modeling IR and millimeter photometry for 2M~0438+2611, \citet{sch06}
427: also used a high inclination angle of $i\sim80\arcdeg$ for their model of
428: this disk. A high inclination was produced by their model because
429: 2M~0438+2611 is much fainter than the photosphere of typical young brown
430: dwarfs, so a highly inclined, obscuring disk was needed to suppress the
431: near- and mid-IR fluxes of their adopted photospheric template 
432: to the observed levels. Thus, their evidence for a high
433: inclination disk was equivalent to that presented by \citet{luh04tau}, 
434: who showed that 2M~0438+2611 is anomalously faint at near-IR wavelengths for 
435: its spectral type and might be occulted by circumstellar material.  
436: The distinctive silicate features in our IRS spectra 
437: (as well as the extended emission in the WFPC2 images) represent 
438: new evidence of a highly-inclined disk around 2M~0438+2611. 
439: \citet{sch06} suggested that the disk around 2M~0444+2512
440: also might have a high inclination. However, the slope of its 
441: near-IR spectrum is consistent with a small amount of extinction and
442: a normal reddening law, and thus does not indicate the presence of scattered
443: light. In addition, an unpublished IRS spectrum of 2M~0444+2512 does
444: not show the silicate absorption that is seen in 2M~0438+2611  
445: and other highly-inclined disks (Figure~\ref{fig:sed2}).
446: 
447: The flux at 24~\micron\ is sensitive to the degree of dust settling in the disk.
448: We find that small grain abundances of $0.03<\epsilon<0.09$ in the upper layers
449: of the disk are required to explain the observed 24~\micron\ flux.
450: For values of $\epsilon$ that are outside of this range, 
451: the predicted slope of the mid- and far-IR SED is smaller or larger than the 
452: observed one. 
453: 
454: In Table~\ref{tab:models}, we summarize the models that provide the best fits
455: to the SED of 2M~0438+2611 for five values of $\alpha$.
456: The SEDs produced by these models are shown with the observed SED in 
457: Figure~\ref{fig:sed}.
458: These models reproduce the observed SED longward of 2~\micron\ equally
459: well, while the model for $\alpha=10^{-4}$ provides a somewhat better match
460: to the flux at shorter wavelengths. The disk radii of these models range 
461: from 5 to 200~AU. Thus, the SED alone does not provide a
462: useful constraint on the outer radius of the disk. 
463: 
464: To estimate the disk radius for 2M~0438+2611, we compare the images produced by 
465: each of our five models to the WFPC2 images in Figure~\ref{fig:image}.
466: The WFPC2 data show a central peak surrounded by asymmetric bipolar emission.
467: In comparison, the model images for $R_{out}=5$ and 12~AU show a point 
468: source and no detectable extended emission and the model for 
469: $R_{out}=200$~AU produces too much extended emission.
470: Meanwhile, the images from the models for $R_{out}=20$ and 40~AU agree
471: reasonably well with the WFPC2 images of 2M~0438+2611.
472: The relative fluxes of the two lobes of emission are better matched by
473: the model for $R_{out}=20$~AU, while the length of the extended emission is 
474: better matched by $R_{out}=40$~AU. 
475: Thus, the WFPC2 images constrain the disk radius to be $R_{out}=20$-40~AU 
476: and confirm the high inclination that is implied by the SED analysis.
477: In addition, given that the model images are formed by stellar light scattered 
478: at the disk surface and extinguished by the outer disk, 
479: the agreement between the optical SED and the predicted scattered light 
480: spectrum and flux level (Figure~\ref{fig:sed}) is further evidence supporting 
481: our estimate of the depletion factor of small grains in the upper disk layers.
482: 
483: 
484: 
485: 
486: 
487: 
488: 
489: \section{Conclusions}
490: 
491: 
492: The young brown dwarf 2M~0438+2611 exhibits several characteristics that 
493: are often observed in stars with edge-on disks, such as unusually faint 
494: near-IR fluxes for its spectral type, strong emission in forbidden transitions, 
495: and anomalous near-IR colors \citep{luh04tau}. 
496: Through observations presented in this paper, 
497: we have confirmed that 2M~0438+2611 is occulted by a highly inclined disk.
498: This new evidence is summarized as follows:
499: 
500: \begin{enumerate}
501: 
502: \item
503: Based on a comparison to other young brown dwarfs,
504: the slope of the 0.8-2.5~\micron\ spectrum of 2M~0438+2611 
505: cannot be explained as a photosphere reddened by a standard extinction law
506: and instead is consistent with scattered light, which suggests that the 
507: brown dwarf is occulted by circumstellar material.
508: 
509: \item
510: The presence of silicate absorption at $\sim10$~\micron\ and 
511: silicate emission at $\sim11$~\micron\ constrains the disk inclination angle
512: to be near $\sim70\arcdeg$ (i.e., $20\arcdeg$ from edge-on) because higher or
513: lower inclinations would produce only absorption or emission, respectively.
514: 
515: \item
516: We detect asymmetric bipolar emission in WFPC2 images of 2M~0438+2611, which 
517: is consistent with simulated scattered-light images produced our model
518: of a highly-inclined disk around this brown dwarf. 
519: 
520: \end{enumerate}
521: 
522: 
523:  
524: In addition to its inclination, we have been able to constrain several other 
525: properties of the disk around 2M~0438+2611 through modeling of its SED
526: and high-resolution images.
527: These constraints are made possible by the unusual wealth of data available for 
528: this disk, although achieving a self-consistent model that simultaneously
529: reproduces those data has proved to be quite challenging.
530: In our best-fit models, the disk has an inclination of $\sim70\arcdeg$, 
531: an inner radius of $58R_\star\approx0.275$~AU, an outer
532: radius of 20-40~AU, a total mass of 100-200~$M_\oplus$, and an abundance of 
533: small grains in its upper layers of $0.03\lesssim\epsilon\lesssim0.09$, 
534: indicating a large degree of settling to the midplane.
535: Our estimate of the inner radius suggests the presence of a large inner hole,
536: which is a characteristic of transitional disks \citep{cal02,cal05,dal05,muz06}.
537: Meanwhile, the outer radius of the disk around 2M~0438+2611 is somewhat
538: larger than expected from models of embryo ejection 
539: \citep[$R_{out}<10$-20~AU,][]{bat03}.
540: 
541: 
542: 
543: 
544: 
545: 
546: 
547: \acknowledgements
548: 
549: We acknowledge support from 
550: grant GO-10511 from the Space Telescope Science 
551: Institute and grant AST-0544588 from the National Science Foundation (K. L.), 
552: grant 172854 from CONACyT (L. A.), grants from CONACyT and PAPIIT/DGAPA, 
553: M\'exico (P. D.), and NASA grants NAG5-9670 and NAG5-13210 (N. C., L. H.). 
554: 
555: \begin{thebibliography}{}
556: 
557: \bibitem[Allen et al.(2007)]{all07}
558: Allen, P. R., et al. 2007, \apj, 655, 1095
559: 
560: \bibitem[Bate et al.(2003)]{bat03}
561: Bate, M. R., Bonnell, I. A., \& Bromm, V. 2003, \mnras, 339, 577
562: 
563: \bibitem[Brandner et al.(2000)]{bra00}
564: Brandner, W., et al. 2000, \aap, 364, L13
565: 
566: \bibitem[Burrows et al.(1996)]{bur96}
567: Burrows, C. J., et al. 1996, \apj, 473, 437
568: 
569: \bibitem[Calvet et al.(2002)]{cal02}
570: Calvet, N., D'Alessio, P., Hartmann, L., Wilner, D., Walsh, A., \& Sitko, M. 
571: 2002, \apj, 568, 1008 
572: 
573: \bibitem[Calvet et al.(2005)]{cal05}
574: Calvet, N., et al. 2005, \apj, 630, L185
575: 
576: \bibitem[Cardelli et al.(1989)]{car89}
577: Cardelli, J. A., Clayton, G. C., \& Mathis, J. S. 1989, \apj, 345, 245
578: 
579: \bibitem[Chauvin et al.(2002)]{cha02}
580: Chauvin, G., et al. 2002, \aap, 394, 949
581: 
582: \bibitem[Cushing et al.(2004)]{cus04}
583: Cushing, M. C., Vacca, W. D., \& Rayner, J. T. 2004, \pasp, 116, 362
584: 
585: \bibitem[D'Alessio et al.(2001)]{dal01}
586: D'Alessio, P., Calvet, N., \& Hartmann, L. 2001, \apj, 553, 321
587: 
588: \bibitem[D'Alessio et al.(2006)]{dal06}
589: D'Alessio, P., Calvet, N., Hartmann, L., Franco-Hern{\'a}ndez, R., 
590: \& Serv{\'{\i}}n, H. 2006, \apj, 638, 314
591: 
592: \bibitem[D'Alessio et al.(1999)]{dal99}
593: D'Alessio, P., Calvet, N., Hartmann, L., Lizano, S., \& Cant{\'o}, J. 1999, 
594: \apj, 527, 893
595: 
596: \bibitem[D'Alessio et al.(1998)]{dal98}
597: D'Alessio, P., Canto, J., Calvet, N., \& Lizano, S. 1998, \apj, 500, 411
598: 
599: \bibitem[D'Alessio et al.(2005)]{dal05}
600: D'Alessio, P., et al. 2005, \apj, 621, 461
601: 
602: \bibitem[Draine \& Lee(1984)]{dl84} 
603: Draine, B. T. \& Lee, H. M. 1984, \apj, 285, 99
604: 
605: \bibitem[Dullemond et al.(2001)]{dul01}
606: Dullemond, C. P., Dominik, C., \& Natta, A. 2001, \apj, 560, 957
607: 
608: \bibitem[Fazio et al.(2004)]{faz04}
609: Fazio, G. G., et al. 2004, \apjs, 154, 10
610: 
611: \bibitem[Fern\'{a}ndez \& Comer\'{o}n(2001)]{fer01}
612: Fern\'{a}ndez, M., \& Comer\'{o}n, F. 2001, \aap, 380, 264
613: 
614: \bibitem[Furlan et al.(2005)]{fur05}
615: Furlan, E., et al. 2005, \apj, 621, L129
616: 
617: \bibitem[Furlan et al.(2006)]{fur06}
618: Furlan, E., et al. 2006, \apjs, 165, 568
619: 
620: \bibitem[Gammie(1996)]{gam96}
621: Gammie, C. F. 1996, \apj, 457, 355
622: 
623: \bibitem[Grosso et al.(2003)]{gro03}
624: Grosso, N., Alves, J., Wood, K., Neuh\"{a}user, R., Montmerle, T., 
625: Bjorkman, J. E. 2003, \apj, 586, 296
626: 
627: \bibitem[Hartmann et al.(1999)]{har99}
628: Hartmann, L., Calvet, N., Allen, L., Chen, H., \& Jayawardhana, R. 1999, 
629: \aj, 118, 1784
630: 
631: \bibitem[Hartmann et al.(2005)]{har05}
632: Hartmann, L., et al. 2005, \apj, 629, 881
633: 
634: \bibitem[Houck et al.(2004)]{hou04}
635: Houck, J. R., et al. 2004, \apjs, 154, 18
636: 
637: \bibitem[Hughes et al.(2007)]{hug07}
638: Hughes, A. M., et al. 2007, \apj, submitted
639: 
640: \bibitem[Jayawardhana et al.(2002)]{jay02}
641: Jayawardhana, R., Luhman, K. L., D'Alessio, P., \& Stauffer, J. R. 2002, \apj,
642: 571, L51
643: 
644: \bibitem[Koresko(1998)]{kor98}
645: Koresko, C. D. 1998, \apj, 507, L145
646: 
647: \bibitem[Krist et al.(1998)]{kri98}
648: Krist, J. E., et al. 1998, \apj, 501, 841
649: 
650: \bibitem[Lucas \& Roche(1997)]{luc97}
651: Lucas, P. W., \& Roche, P. F. 1997, \mnras, 286, 895
652: 
653: \bibitem[Luhman(2004)]{luh04tau}
654: Luhman, K. L. 2004, \apj, 617, 1216
655: 
656: \bibitem[Luhman(2006)]{luh06tau1}
657: Luhman, K. L. 2006, \apj, 645, 676
658: 
659: \bibitem[Luhman et al.(2003)]{luh03}
660: Luhman, K. L., Stauffer, J. R., Muench, A. A., Rieke, G. H., Lada, E. A.,
661: Bouvier, J., \& Lada, C. J. 2003, \apj, 593, 1093
662: 
663: \bibitem[Luhman et al.(2006)]{luh06tau2}
664: Luhman, K. L., Whitney, B. A., Meade, M. R., Babler, B. L., Indebetouw, R.,
665: Bracker, S., \& Churchwell, E. B. 2006, \apj, 647, 1180
666: 
667: \bibitem[Mathis, Rumpl, \& Nordsieck(1977)]{mrn} 
668: Mathis, J. S., Rumpl,W. \& Nordsieck, K. H. 1977, \apj, 217, 425
669: 
670: \bibitem[McCaughrean \& O'Dell(1996)]{mcc96}
671: McCaughrean, M. J., \& O'Dell, R. C. 1996, \aj, 111, 1977
672: 
673: \bibitem[Monin \& Bouvier(2000)]{mon00}
674: Monin, J.-L., \& Bouvier, J. 2000, \aap, 356, L75
675: 
676: \bibitem[Muzerolle et al.(2004)]{muz04}
677: Muzerolle, J., D'Alessio, P., Calvet, N., \& Hartmann, L. 2004, \apj, 617, 406
678: 
679: \bibitem[Muzerolle et al.(2005)]{muz05}
680: Muzerolle, J., Luhman, K. L., Brice\~{n}o, C., Hartmann, L., \& Calvet, N. 
681: 2005, \apj, 625, 906
682: 
683: \bibitem[Muzerolle et al.(2006)]{muz06}
684: Muzerolle, J., et al. 2006, 643, 1003
685: 
686: \bibitem[Padgett et al.(1999)]{pad99}
687: Padgett, D. L., et al. 1999, \aj, 117, 1490
688: 
689: \bibitem[Perrin et al.(2006)]{per06}
690: Perrin, M. D., Duch\^ene, G., Kalas, P., \& Graham, J. R. 2006, \apj, 645, 1272
691: 
692: \bibitem[Pontoppidan et al.(2005)]{pon05}
693: Pontoppidan, K. M., et al. 2005, \apj, 622, 463
694: 
695: \bibitem[Quillen et al.(2004)]{qui04} 
696: Quillen, A. C., Blackman, E. G., Frank, A., \& Varni{\`e}re, P. 2004, \apj, 
697: 612, L137
698: 
699: \bibitem[Rayner et al.(2003)]{ray03}
700: Rayner, J. T., et al. 2003, \pasp, 115, 362
701: 
702: \bibitem[Rice et al.(2003)]{ric03} 
703: Rice, W. K. M., Wood, K., Armitage, P. J., Whitney, B. A., \& Bjorkman, J. E.
704: 2003, \mnras, 342, 79
705: 
706: \bibitem[Rieke \& Lebofsky(1985)]{rl85}
707: Rieke, G. H., \& Lebofsky, M. J. 1985, \apj, 288, 618
708: 
709: \bibitem[Rieke et al.(2004)]{rie04}
710: Rieke, G.~H.\ et~al.\ 2004, \apjs, 154, 25
711: 
712: \bibitem[Scholz et al.(2006)]{sch06}
713: Scholz, A., Jayawardhana, R., \& Wood, K. 2006, \apj, 645, 1498
714: 
715: \bibitem[Skrutskie et al.(2006)]{skr06}
716: Skrutskie, M., et al. 2006, \aj, 131, 1163
717: 
718: \bibitem[Slesnick et al.(2004)]{sle04}
719: Slesnick, C. L., Hillenbrand, L. A., \& Carpenter, J. M. 2004, \apj, 610, 1045
720: 
721: \bibitem[Stapelfeldt et al.(1998)]{sta98}
722: Stapelfeldt, K. R., et al. 1998, \apj, 502, L65
723: 
724: \bibitem[Vacca et al.(2003)]{vac03}
725: Vacca, W. D., Cushing, M. C., \& Rayner J. T., 2003, \pasp, 115, 389
726: 
727: \bibitem[Watson et al.(2004)]{wat04}
728: Watson, D. M., et al. 2004, \apjs, 154, 391
729: 
730: \bibitem[Weingartner \& Draine(2001)]{wl01}
731: Weingartner, J. C., \& Draine, B. T. 2001, \apj, 548, 296
732: 
733: \bibitem[Werner et al.(2004)]{wer04}
734: Werner, M. W., et al. 2004, \apjs, 154, 1 
735: 
736: \bibitem[Wichmann et al.(1998)]{wic98}
737: Wichmann, R., Bastian, U., Krautter, J., Jankovics, I., \& Ruci\'nski, S. M.
738: 1998, \mnras, 301, L39
739: 
740: \end{thebibliography}
741: 
742: %\clearpage
743: 
744: \begin{deluxetable}{lllll}
745: \tablewidth{0pt}
746: \tablecaption{Disk Models for 2M~0438+2611\label{tab:models}}
747: \tablehead{
748: \colhead{$\alpha$} & 
749: \colhead{$R_{out}$} &
750: \colhead{$M_{disk}$} &
751: \colhead{$i$} &
752: \colhead{} \\
753: \colhead{($\times10^{-4}$)} &
754: \colhead{(AU)} &
755: \colhead{($M_\oplus$)} &
756: \colhead{(deg)} &
757: \colhead{$\epsilon$}}
758: \startdata
759: 0.05 & 5 & 700 & 68 & 0.04 \\
760: 0.5 & 12 & 140 & 70 & 0.03 \\
761: 0.8 & 20 & 130 & 71 & 0.03 \\
762: 1 & 40 & 190 & 67 & 0.09 \\
763: 2 & 200 & 340 & 70 & 0.09 \\
764: \enddata
765: \end{deluxetable}
766: 
767: %\clearpage
768: 
769: \begin{figure}
770: \plotone{f1.eps}
771: \caption{
772: {\it Top}:
773: Near-infrared spectrum of the young brown dwarf 2M~0438+2611 ({\it solid line}) 
774: compared to data for a standard young brown dwarf with the same optical 
775: spectral type ({\it dotted line}).
776: The spectrum of 2M~0438+2611 is normalized at 1.68~\micron.  The standard 
777: spectrum has been reddened by $A_V=5.5$ so that it has the same relative
778: fluxes at 0.8 and 2.5~$\micron$ as 2M~0438+2611. The overall shape of
779: 2M~0438+2611 cannot be reproduced by applying reddening to the standard.
780: {\it Bottom}: The anomalous slope of 2M~0438+2611 is further
781: illustrated by the ratio of the spectra of these two objects
782: ({\it solid line}), which departs significantly from the constant ratio
783: expected for normal reddening of a photosphere. The absence of 
784: significant residuals in this ratio near the steam bands confirms
785: that the two objects have similar spectral types.}
786: \label{fig:spec}
787: \end{figure}
788: 
789: %\clearpage
790: 
791: \begin{figure}
792: %\epsscale{.7}
793: \plotone{f2.eps}
794: \caption{
795: Spectral energy distribution of the young brown dwarf 2M~0438+2611 
796: compared to models for its circumstellar accretion disk. The data for 
797: 2M~0438+2611 consist of photometry from 2MASS, IRAC, MIPS, and IRAM
798: ({\it filled circles}) and spectroscopy from SpeX and IRS ({\it solid lines}).
799: The components of the model for 2M~0438+2611 consist of the stellar 
800: photosphere, light from the stellar photosphere scattered by the disk surface, 
801: emission from the inner disk wall, and emission from the disk 
802: ({\it dashed lines}). 
803: We show the best fit models to the SED for $R_{out}=5$-200~AU
804: ({\it top to bottom}, Table~\ref{tab:models}).
805: For each of these models, the sum of the model 
806: components ({\it dotted line}) agrees with most of the data, except for
807: the 2MASS photometry, which is discrepant from all of the other measurements.
808: The IRAC and IRAM data are from \citet{luh06tau2} and \citet{sch06},
809: respectively. All other measurements are from this work.
810: With the exception of the millimeter measurement, the uncertainties for 
811: the photometric measurements are smaller than the points.
812: }
813: \label{fig:sed}
814: \end{figure}
815: 
816: %\clearpage
817: 
818: \begin{figure}
819: \epsscale{.8}
820: \plotone{f3.eps}
821: \caption{
822: Mid-infrared spectrum of the young brown dwarf 2M~0438+2611 compared
823: to data for a face-on disk \citep[FM~Tau,][]{fur06} and 
824: an edge-on disk \citep[DG~Tau~B,][]{wat04}, which exhibit silicate 
825: emission and absorption, respectively. 
826: To reproduce both the emission and the absorption observed in the spectrum of 
827: 2M~0438+2611, our models for its disk ({\it dotted line}) require an 
828: inclination near $\sim70\arcdeg$ (i.e., $20\arcdeg$ from edge-on).
829: The disk model shown here is for $R_{out}=40$.
830: }
831: \label{fig:sed2}
832: \end{figure}
833: 
834: %\clearpage
835: 
836: \begin{figure}
837: \epsscale{0.6}
838: \plotone{f4.eps}
839: \caption{
840: WFPC2 images of 2M~0438+2611 ({\it top row}) and simulated images produced by
841: the disk models that fit the SED in Figure~\ref{fig:sed} ({\it bottom
842: five rows}).
843: In the model images, the disk is close to edge-on ($i\sim70\arcdeg$) and is 
844: aligned close to north-south. Thus, the extended emission in the horizontal
845: direction is above and below the disk.
846: For each image, the intensity is displayed on a logarithmic scale
847: and the size is $1.5\arcsec\times1.5\arcsec$. 
848: }
849: \label{fig:image}
850: \end{figure}
851: 
852: \end{document}
853: