1: % TWHYA.TEX -- AASTEX macro package paper.
2: % John E. Gizis
3: % 39012
4:
5: \documentclass[12pt,preprint]{aastex}
6:
7:
8: %\received{}
9: %\revised{}
10: %\accepted{}
11: %\ccc{}{}
12: %\cpright{}{}
13:
14: % Authors may supply running head information, if they wish to do so
15:
16: \shortauthors{Gizis, Shipman and Harvin}
17: \shorttitle{Brown Dwarf}
18:
19: % Beginning
20:
21: \begin{document}
22:
23: \title{First Ultraviolet Spectrum of a Brown Dwarf: Evidence for H$_2$ Fluorescence and Accretion\footnote{Based on observations made with the NASA/ESA Hubble Space Telescope, obtained at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. These observations are associated with program \# 9841.}
24: }
25:
26: \author{John E. Gizis, Harry L. Shipman, James A. Harvin}
27: \affil{Department of Physics and Astronomy, University of Delaware,
28: Newark, DE 19716}
29:
30: \begin{abstract}
31: We analyze an HST/STIS ultraviolet spectrum of the young brown dwarf
32: 2MASSW J1207334-393254, a member of the ten million-year old
33: TW Hya Association that has a planetary-mass companion.
34: We detect and identify numerous emission lines. CIV and other
35: ions are seen that arise in hot gas.
36: We identify a series of lines with Lyman-pumped H$_2$ molecular lines,
37: indicating that cool gas is also present.
38: Overall, this substellar object shows many of the same characteristics as
39: classical T Tauri stars. We interpret our results as direct evidence of
40: accretion from a circumstellar gas disk, consistent with previous claims.
41: The lack of SiIV emission from the accreting gas indicates that silicon has
42: been depleted into grains.
43: \end{abstract}
44:
45: \keywords{stars: formation --- stars: low-mass, brown dwarfs ---
46: circumstellar matter --- planetary systems: protoplanetary disks
47: --- stars: individual (2MASSW J1207334-393254)}
48:
49: \section{Introduction}
50:
51: It has long been speculated that the star formation process results in both stars
52: above and brown dwarfs below the hydrogen-burning limit. With the discovery of
53: many brown dwarfs in star-forming regions, their formation and early evolution
54: can now be observed. Infrared excesses indicate that many brown dwarfs
55: are surrounded by dusty disks \citep{jaya,liu,mohanty_ir}. In at least one case, a brown dwarf
56: of $0.015 M_\odot$, just above the deuterium-burning limit, has a disk \citep{luhman}.
57: Twenty-seven out of seventy-four young ($<3$ Myr) very-low-mass stars and
58: brown dwarfs also have optical emission lines including broad H$\alpha$, which are
59: interpreted as due to accretion \citep{mohanty_bd}. The accreting brown dwarfs are thus considered analogs of substellar classical T Tauri stars. The star forming region brown dwarfs are too
60: distant to resolve close binaries, but Hubble Space Telescope studies of nearby older field
61: brown dwarfs indicate that 20\% of them are binaries with separations
62: $\lesssim 10$ A.U. \citep{hstl2}. In summary, there is ample evidence that
63: in many respects the formation of brown dwarfs is similar to that of stars.
64:
65: The closest known brown dwarf with a disk is 2MASSW J1207334-393254\footnote{This source appears as 2MASS J12073346-3932539 in the final 2MASS release.} . Gizis (2002) discovered
66: this $\sim 0.03 M_\odot$ brown dwarf in a search for substellar members of the
67: $\sim 10$ Myr old TW Hya Association (TWA) and noted its strong H$\alpha$ emssion.
68: Mohanty et al. (2003) found that the emission is variable and very broad, and argued that it
69: is due to accretion. \citet{twhyax} set an upper limit to the presence of X-rays, and
70: argued that this showed that no more that 10\% of the H$\alpha$ emission
71: is chromospheric, thus supporting the accretion origin of the H$\alpha$ emission.
72: \citet{gemini} detected an infrared excess at 8.7 and 10.4 microns
73: and argued for the presence of a disk. \citet{jaya} earlier showed there is no L' (3.8 micron) excess,
74: indicating the disk may have a growing inner hole as in the TWA members TW Hya and
75: Hen 3-600A \citep{jaya99a,jaya99b}.
76: It is worth noting that the system's youth is confirmed by its low
77: surface gravity (Gizis 2002) and lithium (Mohanty et al. 2003), and
78: TWA Membership is confirmed by both radial velocity (Mohanty et al. 2003) and proper motion
79: (Gizis 2002, Scholz et al. 2005). TW Hya itself is at a distance of 55 parsecs \citep{hipparcos}, but it has
80: been argued that 2M1207 is further away, perhaps at 70 parsecs (Sterzik et al. 2004).
81: As if this system were not interesting enough, \citet{planet1} detected a candidate
82: giant planet companion at a separation of 0.78 arcseconds or 55 A.U..
83: The L dwarf nature of the companion has been confirmed by \citep{planet2}.
84: Proper motion confirmation is underway.
85:
86: In this paper, we present an ultraviolet spectrum of 2M1207. Unlike objects in
87: dusty star-forming regions, the TWA is relatively free of extinction, making 2M1207
88: an especially favorable target for ultraviolet spectroscopy.
89:
90: \section{Observations and Data Reduction}
91:
92: We observed 2M1207 on 24-25 July 2004 using Hubble Space Telescope and the
93: Space Telescope Imaging Spectrograph (STIS) in spectroscopic mode.
94: The grating was G140L with the FUV MAMA detectors using the 0.2 arcsecond slit.
95: This setup gives coverage from 1100 \AA~ to 1700 \AA.
96: Three exposures of 2150 seconds (starting at UT 23:24:43), 2900 seconds (UT 00:46:34) and 2900 seconds (UT 02:22:31) were obtained. We used
97: the flux-calibrated two-dimensional images as created by the standard pipeline.
98: No continuum is detected in the spectra, but numerous spectral lines are
99: strongly detected. We traced these using standard IRAF tasks and extracted the
100: spectra. No significant differences were seen in the spectra extracted from
101: the three individual images (specifically, the CIV fluxes agree to within 8\% in the three exposures); for the remainder of this paper, we use the spectrum
102: extracted from an image created by averaging the three individual exposures.
103: This spectrum is shown in Figure 1 and the measured line fluxes are listed in Table 1.
104: To check on our procedures, we used the same procedures on the three field
105: late-M dwarfs (VB8, VB10 and LHS 2065) observed with the same setup by \citet{hawleyuv}, and found the resulting spectra were consistent with the published ones.
106:
107: \section{Analysis}
108: Superficially, the ultraviolet spectrum strongly resembles IUE spectra of classical T Tauri stars
109: \citep{valenti} while differing in many respects from the field late-M dwarfs. Most
110: strikingly, numerous H$_2$ fluorescence lines are present.
111: \citet{twhyauv1} present line identifications for the H$_2$ lines observed
112: in high-resolution STIS observations of TW Hya. All of our detected
113: features match lines pumped by Ly $\alpha$ 1215\AA and 1216\AA photons.
114: We do indeed detect strong Ly$\alpha$ directly in our spectrum, but its strength must be affected by interstellar absorption and geocoronal contamination. the lack of spatial
115: extent of the emission lines indicates that the emission is from the inner
116: $\sim 0.2$ arcseconds (14 A.U.).
117:
118: In addition to the molecular H$_2$ emission, we detect ions that must arise from
119: very hot gas: CIV, CII, CIII, HeII, NV, and OI.
120: All are present in both classical T Tauri stars and field M dwarfs. However, the
121: emission lines differ significantly from the field dwarfs.
122: In 2M1207, the two components of the CIV double are nearly
123: equal strength (note that the apparently stronger blue component is blended with
124: an H$_2$ line), whereas in the field dwarfs the blue component is twice as strong.
125: This indicates 2M1207's CIV emission is optically thick, while the field dwarf's
126: transition region emission is optically thin.
127: We fail to detect SiIV, which would be easily detectable if in the same ratio
128: as the field dwarfs, and which should arise in similar temperatures as the CIV.
129: In classical T Tauri stars, this Si feature is also sometimes much weaker than expected \citep{valenti}.
130:
131: We interpret these observations as showing that 2M1207 is actively accreting from
132: a circumstellar gas disk, following the scenarios already worked out for
133: classical T Tauri stars \citep{twhyauv2}[and references therein] and as
134: advocated by \citet{mohanty} for this source.
135: This circumstellar gas is, for the first time in a brown dwarf, detected
136: directly in the form of the H$_2$ emission lines. The accretion columns have
137: already been described by the observations of broad H$\alpha$ emission \citet{mohanty}.
138: Near the surface, this gas is (shock) heated to high temperatures ($T \approx 10^5$K).
139: These hotspots are responsible for the CIV emission. The lack of Si emission in
140: the hotspots is due to its depletion into dust grains in the circumstellar disk.
141: Indeed, the dusty disk with silicate emission has been detected in the mid-infrared \citep{gemini}.
142: The lack of detectable X-ray coronal emission confirms that
143: chromospheric and transition region emission is relatively weak
144: in 2M1207 compared to the emission from mechanisms described above
145: \citep{twhyax}. Accretion is evidently variable within a night as well as on long timescales given
146: the observed variability of H$\alpha$ \citep{mohanty,twhyax}, complicating any
147: effort to compare non-simultaneous datasets. Overall, 2M1207 seems to be
148: very similar to TW Hya itself. Three other TWA stars, Hen 3-600A, TWA 14 and TWA 5A \citep{muze01,mohanty}, also appeared to accreting.
149:
150: These observations are particularly interesting given that 2M1207 has a giant planetary mass
151: companion candidate \citep{planet1,planet2}. \citet{twhyair} argue that TW Hya originally had an unusually large circumstellar disk of $\sim 25$\% of the primary mass. The
152: long accretion lifetime and observed similarity of 2M1207 raises the possibility that
153: perhaps 2M1207 also had an unusually large disk. The candidate companion
154: would represent 10-20\% of the primary's mass; it is an suggestive coincidence that
155: 2M1207 might have an unusually massive disk and has an unusual planetary-mass
156: companion. However, since the estimated mass of the planet would then still be
157: comparable to the disk it is difficult to imagine that the companion formed like
158: the planets in our own Solar System. The most likely explanation is that the system formed as
159: double brown dwarfs.
160: The system is more widely separately than any field double brown dwarfs, although existing
161: searches were not sensitive to such extreme mass ratios \citep{hstl2}. Planet formation
162: may also be occurring in the inner disk where we observe depleted gas.
163:
164: \section{Summary}
165:
166: We have obtained the first ultraviolet spectrum of a {\it{bona fide}} brown dwarf. We observe
167: CIV and other emission lines which must arise in hot gas, which we identify as
168: due to accretion, and H$_2$ lines which must arise in much colder gas, which
169: we identify as circumstellar gas. This brown dwarf is very similar to a classical
170: T Tauri star. The emission lines differ significantly from those of late-M field dwarfs,
171: and we therefore rule out origin in a transition region between the chromosphere and corona.
172:
173: Further studies of this fascinating system are in order. We have presented our observations in
174: this paper, but further modeling is required in order to estimate the accretion mass rate and other
175: properties of the system. We are obtaining Spitzer mid-infrared photometry of 2M1207 in order to characterize the disk. Sub-mm observations are needed to estimate the disk mass.
176:
177: \acknowledgments
178:
179: We thank Jeff Valenti, Dermott Mullan, and Stan Owocki for useful conversations.
180: Support for program \#9841 was provided by NASA through a grant from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555.
181:
182:
183: %Tables look like this
184: \begin{deluxetable}{llr}
\tablewidth{0pc}
\tablenum{1}
\tablecaption{Emission Line Strengths}
\tablehead{
\colhead{Ion} &
\colhead{$\lambda$ (\AA)} &
\colhead{Flux ($10^{-16}$~erg/sec/cm$^2$)}
}
\startdata
CIII & 1176 & 1\\
NV & 1238 & 5\\
SI & 1296 & $<1$\\
OI & 1304 & 2\\
CII & 1335 & 5\\
OI & 1356 & 1 \\
SiIV & 1394 & $<1.5$ \\
SiIV & 1403& $<1.5$\\
CIV & 1549 & 29\\
HeII & 1640 & 6\\
\enddata
\end{deluxetable}
185: %----------------------BIBLIOGRAPHY---------------------------
186: \begin{thebibliography}{}
187:
188: \bibitem[Chauvin et al.(2004)]{planet1} Chauvin, G., et al. 2004,
189: \aap, 425, L29
190:
191: \bibitem[Gizis(2002)]{gizis2002} Gizis, J.E., 2002, \apj, 575, 484
192:
193: \bibitem[Gizis et al.(2003)]{hstl2} Gizis, J.E., Reid, I.N., Knapp, G.R.,
194: Liebert, J., Kirkpatrick, J.D., Koerner, D.W., \& Burgasser, A. J. 2003, \aj, 125, 3302
195:
196: \bibitem[Gizis \& Bharat(2004)]{twhyax} Gizis, J.E., \& Bharat, R. 2004, \apj, 608, L113
197:
198: \bibitem[Hawley et al.(2003)]{hawleyuv} Hawley, S.L., Gizis, J.E., \&
199: Reid, I.N. 1996, \aj, 112, 2799
200:
201: \bibitem[Herczeg et al.(2002)]{twhyauv1} Herczeg, G.J., Linsky, J.L., Valenti, J.A.,
202: Johns-Krull, C.M., \& Wood, B.E. 2002, \apj, 572, 310
203:
204: \bibitem[Herczeg et al.(2004)]{twhyauv2} Herczeg, G.J., Wood, B.E., Linsky, J.L., Valenti, J.A.,
205: \& Johns-Krull, C.M. 2004, \apj, 607, 369
206:
207: \bibitem[Jayawardhana et al.(1999a)]{jaya99a} Jayawardhana, R., Hartmann, L., Fazio, G.,
208: Fisher, R.S., Telesco, C.M., Pina, R.K. 1999a, \apj. 520, L41
209:
210: \bibitem[Jayawardhana et al.(1999b)]{jaya99b} Jayawardhana, R., Hartmann, L., Fazio, G.,
211: Fisher, R.S., Telesco, C.M., Pina, R.K. 1999b, \apj. 521, L129
212:
213: \bibitem[Jayawardhana et al.(2003)]{jaya} Jayawardhana, R., Ardila, D. R., Stelzer, B., \& Haisch, K.E. 2003, \aj, 126, 1515
214:
215: \bibitem[Kirkpatrick et al.(1999)]{k99} Kirkpatrick, J.D., Reid, I.N.,
216: Liebert, J., Cutri, R.M., Nelson, B., Beichman, C.A., Dahn, C.C.,
217: Monet, D.G., Gizis, J.E., \& Skrutskie, M.F. 1999, \apj, 519, 802
218:
219: \bibitem[Liu et al.(2003)]{liu} Liu, M.C., Najita, J., \& Tokunaga, A.T. 2003, \apj, 585, 372
220:
221: \bibitem[Luhman et al.(2005)]{luhman} Luhman, K.L., et al. 2005, \apj, 620, L51
222:
223: \bibitem[Mohanty et al.(2003)]{mohanty} Mohanty, S., Jayawardhana, R.,
224: \& Barrado y Navascues, D. 2003, \apj, 593, L109
225:
226: \bibitem[Mohanty et al.(2004)]{mohanty_ir} Mohanty, S., et al. 2004, \apj, 609, L33
227:
228: \bibitem[Mohanty et al.(2005)]{mohanty_bd} Mohanty, S., Jayawardhana, R.,
229: \& Basri, G. 2005, \apj, in press (astro-ph/0502155)
230:
231: \bibitem[Muzerolle et al.(2001)]{muze01} Muzerolle, J., Hillenbrand, L., Calvet, N., Hartmann, L., \& Briceno, C. in
232: Young Stars Near Earth: Progress and Prospects, ASP Conference Series Vol. 244. Edited by Ray Jayawardhana and Thomas Greene. San Francisco: Astronomical Society of the Pacific, ISBN: 1-58381-082-X, 2001., p.245
233:
234: \bibitem[Perryman et al.(1997)]{hipparcos} Perryman, M.A.C., et al.
235: 1997, \aap, 323, L49
236:
237: \bibitem[Schneider et al.(2005)]{planet2} Schneider, G., et al. 2005, AAS Meeting 205, \#11.14
238:
239: \bibitem[Scholz et al. (2005)]{sss} Scholz, R.-D., McCaughrean, M.J., Zinnecker, H., \&
240: Lodieu, N. 2005, \aap, 430, L49
241:
242: \bibitem[Sterzik et al.(2004)]{gemini} Sterzik, M.F., Pascucci, I., Apai, D., van der Bliek, N,
243: \& Dullemond, C.P. 2004, \aap, 427, 245
244:
245: \bibitem[Valenti et al.(2000)]{valenti} Valenti, J.A., Johns-Krull, C.M., \& Linsky, J.L. 2000,
246: \apjs, 129, 399
247:
248: \bibitem[Weinberger et al.(2002)]{twhyair} Weinberger, A.J., et al. 2002, \apj, 566, 409
249:
250: %\bibitem[Weinberger et al.(2004)]{twair} Weinberger, A.J., Becklin, E.E., Zuckerman, B., \& Song, I. 2004, \aj, 127, 2246
251:
252:
253: \end{thebibliography}
254:
255: \begin{figure}
256: \epsscale{0.7}
257: \plotone{f1.eps}
258: \caption{HST/STIS spectrum of 2M1207. We detect numerous ions. The positions of
259: SiIV and SiII are marked, but they are not detected. The diamonds mark the positions of
260: the H$_2$ lines pumped by 1215\AA~ and 1216\AA~ identified in TW Hya by \citet{twhyauv1}.
261: Most are clearly detected, and those absent are significantly weaker in the TW Hya spectrum
262: and thus are expected to be too faint to detect.
263: \label{fig1}}
264: \end{figure}
265:
266: \end{document}
267: