0801.0402/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[twocolumn]{emulateapj}
3: 
4: %% preprint produces a one-column, single-spaced document
5: %% manuscript produces a one-column, double-spaced document
6: %% preprint2 produces a double-column, single-spaced document
7: 
8: %% If you wish, you may supply running head information, although
9: %% this information may be modified by the editorial offices.
10: %% The left head contains a list of authors,
11: %% usually a maximum of three (otherwise use et al.).  The right
12: %% head is a modified title of up to roughly 44 characters.
13: %% Running heads will not print in the manuscript style.
14: 
15: \newcommand{\Lsun}{$L_{\odot}$}
16: \newcommand{\Msun}{$M_{\odot}$}
17: 
18: \shorttitle{A sub-AU outwardly truncated accretion disk}
19: \shortauthors{McClure et al.}
20: 
21: 
22: \begin{document}
23: 
24: \title{A sub-AU outwardly truncated accretion disk around a classical T Tauri star}
25: 
26: %% Use \author, \affil, and the \and command to format
27: %% author and affiliation information.
28: 
29: \author{M. K. McClure\altaffilmark{1}, W. J. Forrest\altaffilmark{1},
30: B. A. Sargent\altaffilmark{1}, Dan M. Watson\altaffilmark{1},
31: E. Furlan\altaffilmark{2}, P. Manoj\altaffilmark{1}, 
32: K. L. Luhman\altaffilmark{3}, N. Calvet\altaffilmark{4}, C. Espaillat\altaffilmark{4}, P. D'Alessio\altaffilmark{5}, L. W. Hartmann\altaffilmark{4}, 
33: C. Tayrien\altaffilmark{1}, S. T. Harrold\altaffilmark{1}}
34: 
35: \altaffiltext{1}{Department of Physics and Astronomy, University of Rochester, 
36: Rochester, NY 14627; melisma@astro.pas.rochester.edu, forrest@pas.rochester.edu, 
37: bsargent@pas.rochester.edu, dmw@pas.rochester.edu, manoj@pas.rochester.edu, ctayrien@mail.rochester.edu, 
38: sharrold@mail.rochester.edu}
39: \altaffiltext{2}{NASA Astrobiology Institute, and Department of Physics and 
40: Astronomy, UCLA, 430 Portola Plaza, Los Angeles, CA 90095, NASA Postdoctoral Program Fellow; furlan@astro.ucla.edu}
41: \altaffiltext{3}{Department of Astronomy and Astrophysics, The Pennsylvania State University, University Park, PA 16802; kluhman@astro.psu.edu}
42: \altaffiltext{4}{Department of Astronomy, The University of Michigan, 
43: 500 Church St., 830 Dennison Bldg., Ann Arbor, MI 48109; ncalvet@umich.edu, ccespa@umich.edu, lhartm@umich.edu}
44: \altaffiltext{5}{Centro de Radioastronom{\'i}a y Astrof{\'i}sica, Universidad Nacional Aut{\'o}noma de M{\'e}xico, 58089 Morelia, 
45: Michoac{\'a}n, M{\'e}xico; p.dalessio@astrosmo.unam.mx}
46: 
47: \begin{abstract}
48: We present the Spitzer Infrared Spectrograph (IRS) spectrum of SR20, a 5--10 AU binary T Tauri system in the 
49: $\rho$ Ophiuchi star forming region.  The spectrum has features consistent with the presence of a disk; 
50: however, the continuum slope is steeper than the $\lambda^{-4/3}$ slope of an infinite geometrically thin, optically thick disk, 
51: indicating that the disk is outwardly truncated.  Comparison with photometry from the literature shows a large increase in the mid-infrared flux from 
52: 1993 to 1996.  We model the spectral energy distribution and IRS spectrum with a wall $+$ optically thick irradiated disk, 
53: yielding an outer radius of 0.39$_{+0.03}^{-0.01}$ AU, much smaller than predicted by models of binary orbits.  Using a two temperature 
54: $\chi^2$ minimization model to fit the dust composition of the IRS spectrum, we find the disk has experienced significant grain growth: its
55: spectrum is well-fit using opacities of grains larger than 1 $\mu$m.  We conclude that the system experienced a significant gravitational perturbation in 
56: the 1990s.  
57: 
58: 
59: \end{abstract}
60: 
61: \keywords{circumstellar matter --- binaries: close --- stars: individual (Em* SR20) ---
62: stars: pre-main sequence --- infrared: stars}
63: 
64: \section{Introduction}
65: 
66: Classical T Tauri stars are typically surrounded by optically thick accretion disks, indicated by strong and broad 
67: H$\alpha$ emission lines and by a characteristic infrared excess.  Over the last twenty years, observations 
68: have revealed that many of these young stars are components in binary, and sometimes higher order, systems \citep[e.g.][]{simon92, ghez93}.  
69: A binary system may produce up to three distinct disks: a circumprimary, circumsecondary, and circumbinary disk 
70: \citep[][(AL94; JM97)]{artymowicz94, jensen97}.  Gravitational interactions between binaries affect the structure of their 
71: disks; in particular, the maximum outer radius for a circumstellar disk in a binary system is approximately 0.18 to 0.4 times the 
72: semi-major axis (AL94), depending on the mass ratio and orbital eccentricity.
73: 
74: SR20 is one such close binary in the $\rho$ Ophiuchi star forming region.  Its components have an angular separation 
75: ranging from 0.038 to 0.071\arcsec \citep{ghez93, ghez95} which, at a distance to Ophiuchus of 140 pc \citep{dz99} 
76: corresponds to a projected linear separation of 5.3--9.9 AU.  The secondary is 2.2 magnitudes fainter than the primary at 
77: 2.2 $\mu$m \citep{ghez93} and therefore is less massive than the primary, diskless, or both.  H$\alpha$ emission has been detected from SR20 with 
78: equivalent widths ranging from 15 to 21 \AA \citep[][(W05)]{r80, ba92, wilking05}.  Further observations by \citet{tbc03} 
79: did not find a spectro-astrometric signature in the H$\alpha$ line, indicating that there is little to no accretion in the secondary relative to the primary.  
80: Based on the lack of evidence for a circumsecondary disk, we assume that the infrared excess of SR20 is dominated by a circumprimary accretion disk.
81: 
82: Here we present Spitzer Space Telescope Infrared Spectrograph (IRS) \citep{houck04} observations of the SR20 system.  
83: We model the structure of the system and the dust it contains and analyze the nature of the SR20 system.  
84: While instances of sculpture of proto-planetary accretion disks by planetary 
85: or stellar companions \emph{within} the disk are becoming widely known \citep[e.g. JM97,][]{furlan07}, SR20 is an apparently 
86: rarer example of sculpture from \emph{without}.
87: 
88: \section{Observations and Data Reduction}
89: 
90: We observed SR20 on 2006 April 16 (Spitzer AOR 12698368), with the short wavelength (SL; 5.2--14 $\mu$m) and long wavelength (LL; 
91: 14.0--36.1 $\mu$m) low spectral resolution ($\lambda$/$\Delta\lambda$=60--120) IRS modules.  At each order of each module, the object 
92: was observed twice, once in each of the nominal nod positions, $1/3$ of the way from the ends of the slit, with exposure times for each 
93: observation of 6 seconds in SL and 14 seconds in LL.  The spectrum was extracted from the Spitzer Science Center (SSC) S14.0 pipeline basic calibrated data 
94: using the SMART software tool \citep{higdon04} following the same procedure as \citet{furlan06} with the exception of background 
95: subtraction, which was done using the opposite nod position.
96: 
97: Assuming nod-to-nod flux differences were caused by slight mispointing, for each module we scaled the nod with the lower flux to 
98: match the flux level of the nod with the higher flux.  Scale factors were typically within a few \% of unity.  
99: The resulting nods were averaged to obtain the final spectrum, and the spectral uncertainties are estimated to be half 
100: the difference between the two independent spectra from each nod position.  We estimate the spectrophotometric accuracy of 
101: the final result to be around 5\%.
102: 
103: \section{Analysis}
104: \subsection{Spectral Energy Distribution}
105: \label{SED_section}
106: 
107: The 5--36 $\mu$m IRS spectrum of SR20 (Figure \ref{SR20_dust}) has a clear infrared excess and weak silicate dust emission 
108: at 9.8 and 18.0 $\mu$m, consistent with the presence of a circumstellar disk.  This excess cannot be attributed 
109: to an optically thin shell, as in mass-losing giant stars, since the 5--8 $\mu$m H$_2$O absorption features indicate that 
110: the underlying continuum is optically thick. 
111: 
112: Here we adopt a spectral type of G7 and $A_V$ = 5.7 mag, determined from optical data by W05.  
113: To correct the photometry and IRS spectrum for continuum slope fitting, we applied a ``standard'' extinction correction model with 
114: $A_V/\tau_{9.7}$ = 12.5 and $R_V$ = 3.1 \citep[][D03]{draine03}.  For the dust composition fitting, we corrected more carefully over the silicate features 
115: by scaling our previous correction model to $A_V/\tau_{9.7}$ = 25, which may be more appropriate for silicate features in dark clouds \citep{chiar07}.  
116: For consistency, we compared results for both dust composition and disk structure and they are independent of the correction used. 
117: 
118: After exinction correction, we constructed two SEDs using the $A_V/\tau_{9.7}$ = 12.5 corrected spectrum. 
119: The first contains optical data, the near infrared data closest in time to our spectrum, the IRS spectrum and 
120: a 5600 K, solar metallicity NextGen stellar photosphere model \citep{nextgen} (Figure \ref{SR20_SED}).  We chose to normalize 
121: the photosphere model to the 2MASS J-band flux density rather than that at I-band due to the greater variability at I (1.0 magnitude vs. 0.4 magnitudes at J).  
122: It is unlikely that SR20 suffers from significant veiling at J, since the line strengths in the \citet{lr99} K-band spectrum are comparable with a late G photosphere.
123: The second SED shows the near infrared data, sorted into four epoch ``bins'': 1974, 1993, 
124: 1996-1997, and 1999-2004 (Figure \ref{SR20_3epoch}).  Significantly, in 1974 and 1993 the excess from 2 to 10 $\mu$m 
125: was small but increased substantially by 1996.  Changes in the 2--10 $\mu$m excess since then have been small.
126: 
127: It can be seen from Figure 2 that SR20 exhibits excess emission above the photosphere over $\lambda$ = 1.25--36 $\mu$m and that the slope of the 
128: continuum emission is much steeper than $\lambda^{-4/3}$, the slope of an infinite geometrically-thin, optically-thick irradiated disk 
129: \citep{hartmann99}.  There is a correlation between the continuum slope and dust settling in the disk; for the IRS Taurus sample, 
130: the bluest and presumably most settled T Tauri disks had continuum slopes approaching, but not equal to, $\lambda^{-4/3}$ 
131: \citep{furlan06}, as demonstrated by the IRS spectrum of IS Tau, also in Figure \ref{SR20_SED}.  We interpret a slope steeper 
132: than $\lambda^{-4/3}$ to indicate that the circumstellar disk is outwardly truncated.
133: 
134: 
135: \subsection{Wall + Disk Model}
136: \label{Sys_Model_section}
137: 
138: To investigate whether outward truncation is plausible and determine the degree to which it occurs, we modeled the disk over a range of radii.  We began by modeling 
139: the inner edge of the disk with a vertical ``wall'' located at the dust destruction radius \citep{natta01}, calculated following the methods of \citet{dalessio05}, 
140: which include the wall atmosphere.  The wall emission depends on its radius, $R_{in}$, height $Z$, and the inclination to the 
141: line of sight, $i$.  In turn, $R_{in}$ depends on the dust composition and size distribution in the wall and on the sum of the stellar, $L_*$, and accretion, 
142: $L_{acc}$, luminosities.  We use a size distribution  $\propto a^{-3.5}$ between $a_{\rm min} = 0.005 \mu$m and $a_{\rm max}$.  For $L_*$, we adjusted the value from 
143: W05 to 140 pc, yielding 6.76 L$_{\odot}$.  For a list of stellar parameters, see Table \ref{Stellar parameters}.\citet{natta06} estimated an upper limit to 
144: $L_{acc}$ of 0.03 L$_{\odot}$, using the luminosity of the Pa$\beta$ line. They adopted an extinction of $A_J$ = 1.3, which corresponds to $A_V = 4.6$, lower than the 
145: one used here. Scaling up the J band flux according to the difference of extinction values, we get an upper limit to the accretion luminosity of 0.05 L$_{\odot}$, which is 
146: much lower than $L_*$.  Therefore, we neglect the accretion luminosity in the calculation of the dust destruction radius and use as input parameters the maximum grain size, 
147: $a_{max}$, the inclination, and the height of the wall.  Varying $a_{\rm max}$  changed $R_{in}$ and the peak-to-continuum ratio of the silicate features, which formed 
148: in the wall atmosphere, with large $a_{\rm max}$ producing a wall radius closer to the star and a lower peak-to-continuum ratio.  We were able to fit most of the SED with 
149: just the emission from a wall with spherical solid amorphous olivine grains that have grown to 1.0 $\mu$m.  With these grains, the dust destruction radius is located at 
150: 0.34 AU. In the vertical wall approximation, the maximum emission occurs for $\sim 50 \degr$ \citep{dull01}. At this inclination, best fit is for a wall height $Z$ = 
151: 2.11 R$_*$, or 2.6 scale heights, which is lower than the expected range of 4-5 scale heights \citep{dull01}. Lower inclinations would require higher walls to fit 
152: the near-infrared excess.  The wall emission alone is insufficient to explain the IRS spectrum past 10.0 $\mu$m, so we added a very small flat, optically thick 
153: irradiated disk component \citep{hartmann99}.  The temperature distribution of the disk component is given by
154: 
155: %\begin{equation}
156: $T(r) = \left[\frac{I_0}{\sigma}\left[{\rm sin}^{-1}\frac{R_{*}}{r} - 
157: \frac{R_{*}}{r}\left(1 - \frac{{R_{*}}^2}{r^2}\right)^{\frac{1}{2}}\right]\right]^{\frac{1}{4}}$
158: %\end{equation}
159: 
160: \noindent where $r$ is the cylindrical radius, and $I_0=\frac{\sigma{T_{e}}^4}{\pi}$.  From 1.25--971 $\mu$m, we numerically integrated the corresponding flux density 
161: over the range of radii from $R_{\rm in}$, the wall radius, to $R_{\rm out}$, the outer radius.  By adding a disk component with $R_{\rm in}$ = 0.34 AU, $R_{\rm out}$ 
162: = 0.39$_{+0.03}^{-0.01}$ AU, and $i$ = 50$\degr$ to the wall model at the same inclination with 1.0 $\mu$m solid olivine grains, the SED of SR20 is well fit (Figure 
163: \ref{Radius_grid}).  The only discrepancy is on the longer wavelength side of the silicate features, which could be fit better with slightly larger grains.  
164: As noted, we could get similar fits in the near-IR with lower inclinations and taller walls. However, the emission of the optically thick disk scales as 
165: $ {\rm cos} i$, increasing as the inclination decreases. As a result, to obtain the same fit to the mid-infrared, the disk radius would have to be smaller.  
166: A similar conclusion could be expected if the wall were round and its maximum emission occurred at low inclinations \citep{IN05}. Therefore, 0.39 AU
167: is an upper limit for the radius of the disk. 
168: 
169: To understand the relationship between the binary companion and circumprimary disk, we need information about the binary orbit.  
170: We were unable to determine a unique solution (there are only three separation measurements in the literature), but we placed rough 
171: constraints on the semi-major axis and eccentricity.  In the limit of the eccentricity approaching unity, the semi-major axis is half the maximum separation. 
172: Therefore, regardless of the eccentricity, the semi-major axis of SR20 must be greater than 5 AU.  For our preliminary orbit estimate, values between 8 and 
173: 12 AU fit best along with lower eccentricities of 0.2--0.3.
174: 
175: \subsection{Dust Composition Model}
176: \label{Dust Comp}
177: 
178: Although our wall $+$ irradiated disk model produced a good fit to the IRS spectrum, it uses only one silicate component at a time with a fixed set of $a_{\rm max}$.  
179: To perform a more detailed analysis, we calculate a two temperature model of the silicate dust composition by $\chi^2$ minimization with respect to weights proportional 
180: to dust mass, black body solid angle, and temperature for black bodies and dust grains \citep{sargent06, kastner06, chen06} 
181: with optical constants from \citet{day79} and \citet{koike03}.  Grains larger than 8 $\mu$m in radius, which do not contribute 
182: to the silicate features, could still contribute to the black body components.  The best fit ($\chi_\nu$ = 4.0) occurs at 
183: 410 and 1221 K with the components listed in Table \ref{Stellar parameters}, which are shown in Figure \ref{SR20_dust}.  
184: Interestingly, the only significant components besides the black bodies are large (5 $\mu$m) porous (60\% vacuum), amorphous olivine 
185: (Mg$_2$SiO$_4$) grains, which is consistent with the 10 $\mu$m peak to continuum ratio of 1.35:1.  Submicron amporphous olivine and pyroxene, 
186: crystalline silica, and large amorphous pyroxene, all of which are typical of T Tauri disks, were included in the modeling but found not to be present.  
187: Forsterite was present in the best-fit model, but at less than 1$\sigma$ levels, so we do not consider it to be significant.  For $d$ = 140 pc, our fit yields 9.55 $\times$ 
188: $10^{-5}$ $M_{\rm lunar}$ of large amorphous olivine grains, combining the contributions at both temperatures, with an upper limit of 1.36 $\times$ $10^{-5}$ $M_{\rm lunar}$ 
189: on sub-micron grains.
190: 
191: \section{Discussion and Conclusions}
192: \label{discussion}
193: 
194: The SED, system geometry, and dust composition shed light on the nature of the SR20 system, while simultaneously 
195: raising further questions.
196: 
197: The IRS spectrum is well-fit by our composite wall $+$ circumprimary disk truncated beyond 0.39 AU.  Outward truncation could be caused by 
198: the orbit of a companion or lack of replenishment from a circumbinary disk.  Using the tidal truncation models of AL94, the range of likely eccentricities (0.2-0.3) 
199: and semi-major axes (8-12 AU) of the A-B orbit are not extreme enough to cause truncation.  Even with the lower limit on the semi-major axis, 5 AU, and the highest 
200: order resonances and eccentricities tested in AL94, the minimum stable outer truncation radius caused by the known binary is 0.9 AU, considerably larger than the 0.39 AU 
201: we derive (AL94).  However, the AL94 models may not be appropriate for a disk with such a small ratio between the outer and inner radii, so we cannot {\it absolutely} 
202: rule out truncation by the known binary, at least on this basis.  Alternatively, it may be that SR20 is a hierarchical triple.  A companion object with an eccentricity 
203: $<$ 0.8 could orbit in the region between 0.98 AU and 2.17 AU from the primary, causing a disk truncation at 0.39 AU.  Such a close, low mass companion would be difficult 
204: to detect.
205: 
206: Based on the 0.39 AU truncation radius, we speculate that replenishment of the circumprimary disk by a circumbinary disk occurred in the past.  
207: A disk between 0.1 and 10 AU would accrete on a timescale of 1,000--100,000 years \citep{quillen04} and the median age of the Ophiuchus cluster is 2.1 Myr (W05).  
208: SR20 has been observed at 800, 850, and 1300 $\mu$m with only 3 $\sigma$ upper limit results, implying less than $5.4$ $\times$ $10^{4}$ $M_{lunar}$ (540 $M_{lunar}$ 
209: of dust) in disk material \citep{aw07}.  Low circumbinary disk masses are typical of binaries with semi-major axes between 1 and 50 AU \citep{jensen94}, 
210: so submillimeter non-detection does not indicate a depleted circumbinary disk; however, it is unlikely that lack of replenishment is responsible for the 0.39 AU 
211: truncation radius.
212: 
213: A nearby companion truncating the circumprimary disk is perhaps also the best explanation for the photometric variability.  Before 1993, the infrared excess in 
214: SR20 started in the N band, indicating that there were few small dust grains close to the central star.  
215: However, the H$\alpha$ emission indicates that a gaseous accretion disk, perhaps mixed with much larger planetesimals, was present.  
216: Between 1993 and 1996, the infrared excess increased sharply, indicating an increase in the amount of dust in the disk, which could 
217: be explained by a gravitational interaction between the known binary and a third, unseen, body, causing the unseen companion to orbit 
218: closer to the primary, perturbing the circumprimary disk.  Coupling between the gas and dust could confine the newly created dust to the accretion disk.  
219: The appeareance of the observed excess over 3 years, preceded by 20 years without an excess and followed by 10 years of little variation in the observed excess, seems 
220: consistent with truncation by a nearby source with a shorter period than is plausible for the known companion. 
221: 
222: The silicate dust composition of the system is not typical of a Class II object; our fit indicates that the optically thin 
223: portion of the disk is comprised primarily of large grains ($>$ 1 $\mu$m).  Although having a large fraction of large grains is consistent 
224: with debris disks or advanced dust processing in the inner regions of a protoplanetary disk \citep{chen06}, this is the first time we have seen a Class II disk 
225: with negligible submicron grains.  Taken with the aforementioned explanations of the 1993--1996 variability, the large grains 
226: could support a collision scenario:  gravitional interactions between the known secondary and the unseen companion cause collisions 
227: between planetesimals, producing dust grains that are detected as the current disk. If there were a population of planetesimals -- invisible in our spectra -- comprising 
228: an extension of the disk beyond the outer truncation radius of the dust, truncation by the known companion would be more probable.
229: 
230: Near infrared interferometry and orbital monitoring would be necessary to determine the nature of this complex system more precisely.  
231:  
232: 
233: \acknowledgements
234: This work is based on observations made with the {\it Spitzer Space Telescope}, 
235: which is operated by the Jet Propulsion Laboratory, California Institute of Technology, 
236: under NASA contract 1407 and made use of the ADS, SIMBAD, and Vizier utilities. Support was provided by NASA through 
237: contract number 1257184 issued by JPL/Caltech, JPL contract 960803 to Cornell University, 
238: and Cornell subcontracts 31419-5714 to the University of Rochester.  
239: 
240: \begin{deluxetable}{lccc}
241: \tabletypesize{\scriptsize}   %\small (11pt), \footnotesize (10 pt), \scriptsize (8pt)
242: \tablewidth{\linewidth}
243: \tablecaption{Stellar and model properties \label{Stellar parameters}}
244: \tablehead{
245: \colhead{Component} & \colhead{Value} & \colhead{Reference}
246: }
247: \startdata
248: \multicolumn{4}{c}{Stellar parameters}\\
249: \hline
250: $SpT$ & G7 & W05 \\
251: $T_{e}$ (K) & 5584  & W05 \\
252: $L_{*}$ (L$_{\odot}$) & 6.76 & W05, adjusted$^a$ \\
253: $M_{*}$ (M$_{\odot}$) & 1.86  & \citet{siess2000}\\
254: $R_{*}$ (R$_{\odot}$) & 2.57 & \citet{siess2000} \\
255: $L_{acc}$ (L$_{\odot}$ & $<$ 0.05 & \citet{natta06}, adjusted$^a$ \\
256: $A_V$ (mag) &  5.7 & W05 \\
257: \hline
258: \multicolumn{4}{c}{Dust Components $>$ 1$\sigma$}\\
259: \hline
260: Parameter & Value$^b$ & $1\sigma$ uncertainty$^c$ \\
261: \hline
262: (410 K) \\
263: black body & $7.12 \times 10^{-17}$ &  $3.84 \times 10^{-18}$ \\
264: large Mg$_2$SiO$_4$$^d$ & $3.41 \times 10^{-20}$ & $5.18 \times 10^{-21}$ \\
265: \hline
266: (1221 K) \\
267: \hline
268: black body &  $2.56 \times 10^{-17}$ &  $4.70 \times 10^{-19}$ \\
269: large Mg$_2$SiO$_4$$^d$ & $3.58 \times 10^{-21}$ & $5.97 \times 10^{-22}$ \\
270: \tableline
271: \enddata
272: \tablecomments{
273: $^a$ See section 3.2 for details.
274: $^b$ Blackbody components are solid angles in units of sterradians, while dust components are column density times solid angle in g/cm$^2$.  
275: $^c$ Uncertainties are the increments away from the best fit values enough to increase $\chi_\nu$ by 1.0.  $^d$ Amorphous olivine}
276: \end{deluxetable}
277: 
278: \begin{figure}
279: %\epsscale{0.7}
280: \plotone{f5.eps}
281: \caption{Dust composition model overplotted on the extinction corrected IRS spectrum.  Light grey and dark grey represent 410 K and 
282: 1221 K components respectively.  Components are black bodies (dash/dot), large amorphous olivine (solid), and forsterite (dotted).  
283: Forsterite is present at less than $1\sigma$ strength and not considered significant. 
284: \label{SR20_dust}}
285: \end{figure}
286: 
287: \begin{figure}
288: %\epsscale{0.7}
289: \plotone{f2.eps}
290: \caption{``Current'' SED of SR20, with 1983-84 UBVRI \citep{ba92}, 1999 JHK (2MASS), 2004 IRAC (C2D), and MIPS 24 $\mu$m photometry 
291: with our 2006 IRS spectrum.  Data were extinction corrected using $A_V$=5.7 and the D03 law.  UBVRI data are plotted with bars to indicate the extent of 
292: variability.  Overplotted are the IRS spectrum of IS Tau ($\times$ 15) and a line representing $\lambda^{-4/3}$. \label{SR20_SED}}
293: \end{figure}
294: 
295: \begin{figure}
296: %\epsscale{0.7}
297: \plotone{f3.eps}
298: \caption{SR20 photometry from four epochs: 1974: JHKL \citep{rss76}; 1993: JHKL'M'N (JM97); 1996-7: JK (DENIS), 3.6 $\mu$m 
299: (ISO-PHT, sky+source), 6.7 and 14.3 $\mu$m \citep{bontemps01}; 1999-2004: JHK (2MASS), IRAC (C2D), 24 $\mu$m 
300: (Extracted from MIPS AOR 4321280 with an aperture of radius 14.94\arcsec\, sky annulus from 29.88 to 42.33\arcsec\, and aperture 
301: correction factor of 1.143 \citep{su06} using the SSC's APEX software package \citep{mm05}).  Photometry is plotted with 
302: error bars, most of which are smaller than the plot symbols. \label{SR20_3epoch}}
303: \end{figure}
304: 
305: \begin{figure}
306: %\epsscale{0.7}
307: \plotone{f4.eps}
308: \caption{Best fitting model (grey solid line):  d = 140 pc, $i$ = 50\degr\, $R_{\rm in}$ = 0.34 AU, and $R_{\rm out}$ = 0.39 AU.
309: \label{Radius_grid}}
310: \end{figure}
311: 
312: 
313: %\clearpage
314: % Figures and tables here
315: \begin{thebibliography}{References}
316: 
317: \bibitem[Andrews \& Williams(2007)]{aw07} Andrews, S., \& Williams, J.  2007, ApJ, in press
318: \bibitem[Artymowicz \& Lubow(1994)]{artymowicz94} Artymowicz, P., \& Lubow, S. H.  1994, \apj, 421, 651
319: \bibitem[Bontemps et al.(2001)]{bontemps01} Bontemps, S., et al.  2001, \aap, 372, 173
320: \bibitem[Bouvier \& Appenzeller(1992)]{ba92} Bouvier, J., \& Appenzeller, I.  1992, \aaps, 92, 481
321: \bibitem[Chen et al.(2006)]{chen06} Chen, C. H., et al.  2006, \apjs, 166, 351
322: \bibitem[Chiar et al.(2007)]{chiar07} Chiar, J. E., et al.  2007, \apj, 666, 73
323: \bibitem[D'Alessio et al.(2005)]{dalessio05} D'Alessio, P., et al.\ 2005, \apj, 621, 461 
324: \bibitem[de Zeeuw et al.(1999)]{dz99} de Zeeuw, P. T., et al.  1999, \aap, 117, 354
325: \bibitem[Day(1979)]{day79} Day, K. L.  1979, \apj, 234, 158 
326: \bibitem[Draine(2003)]{draine03} Draine, B. T.  2003, Ann. Rev. Astr. Ap., 41, 241
327: \bibitem[Dullemond et al.(2001)]{dull01} Dullemond, C.~P., Dominik, C., \& Natta, A.\ 2001, \apj, 560, 957 
328: \bibitem[Furlan et al.(2006)]{furlan06} Furlan, E., et al.  2006, \apj, 165, 568
329: \bibitem[Furlan et al.(2007)]{furlan07} Furlan, E., et al.  2007, \apj, 664, 1176
330: \bibitem[Ghez et al.(1993)]{ghez93} Ghez, A. M., Neugebauer, G., \& Matthews, K.  1993, \aj, 106, 2005
331: \bibitem[Ghez et al.(1995)]{ghez95} Ghez, A. M., et al.  1995, \aj, 110, 753
332: \bibitem[Hartmann(1998)]{hartmann99} Hartmann, L.  1998, apsf, book
333: \bibitem[Hauschildt et al.(1999)]{nextgen} Hauschildt, P. H., et al.  1999, \apj, 312, 377
334: \bibitem[Higdon et al.(2004)]{higdon04} Higdon, S. J. U., et al.  2004, \pasp,116, 975
335: \bibitem[Houck et al.(2004)]{houck04} Houck, J. R., et al.  2004, \apjs, 154, 18
336: \bibitem[Isella \& Natta(2005)]{IN05} Isella, A., \& Natta, A.\ 2005, \aap, 438, 899
337: \bibitem[Jensen et al.(1994)]{jensen94} Jensen, E. L. N., Mathieu, R. D., \& Fuller, G. A.  1994, \apjl, 429, L29 
338: \bibitem[Jensen \& Mathieu(1997)]{jensen97} Jensen, E. L. N., \& Mathieu, R. D.  1997, \aj, 114, 301
339: \bibitem[Kastner et al.(2006)]{kastner06} Kastner, J. H., et al.  2006, \apj, 638, 29
340: \bibitem[Luhman \& Rieke(1999)]{lr99} Luhman, K. L., \& Rieke, G. H.  1999, \apj, 525, 4
341: \bibitem[Makovoz \& Marleau(2005)]{mm05} Markovoz, D., \& Marleau, F. R.  2005, \pasp, 117, 1113
342: \bibitem[Muzerolle et al.(2004)]{muzerolle04} Muzerolle, J., et al.  2004, \apj, 617, 406 
343: \bibitem[Natta et al.(2001)]{natta01} Natta, A., et al.  2001, \aap, 371, 186
344: \bibitem[Natta et al.(2006)]{natta06} Natta, A., Testi, L., \& Randich, S.  2006, \aap, 452, 245 
345: \bibitem[Quillen et al.(2004)]{quillen04} Quillen, A. C., et al.  2004, \apj, 612, 137
346: \bibitem[Rydgren, Strom \& Strom(1976)]{rss76} Rydgren, A. E., Strom, S. E., \& Strom, K. M.  1976, \apjs, 30, 307
347: \bibitem[Rydgren(1980)]{r80} Rydgren, A. E.  1980, \aj, 85, 438
348: \bibitem[Sargent et al.(2006)]{sargent06} Sargent, B., et al.  2006, \apj, 645, 395
349: \bibitem[Siess et al.(2000)]{siess2000} Siess L., Dufour E., \& Forestini M.  2000, \aap, 358, 593
350: \bibitem[Simon et al.(1992)]{simon92} Simon, M., et al.  1992, \apj, 384, 212 
351: \bibitem[Koike et al.(2003)]{koike03} Koike, C., et al.  2003, \aap, 399, 1101 
352: \bibitem[Su et al.(2006)]{su06} Su, K. Y. L., et al.  2006, \apj, 653, 675
353: \bibitem[Takami, Bailey \& Chrysostomou(2003)]{tbc03} Takami, M., Bailey, J., \& Chrysostomou, A.  2003, \aap, 397, 675
354: \bibitem[Wilking et al.(2005)]{wilking05} Wilking, B. A., et al.  (2005), \aj, 130, 1733
355: 
356: \end{thebibliography}
357: 
358: 
359: \end{document}