0707.1754/ms.tex
1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2005 December 5
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8: 
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12: 
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: \documentclass[12pt,preprint]{aastex}
19: 
20: %% manuscript produces a one-column, double-spaced document:
21: 
22: %%\documentclass[manuscript]{aastex}
23: 
24: %% preprint2 produces a double-column, single-spaced document:
25: 
26: %\documentclass[preprint2]{aastex}
27: 
28: %% Sometimes a paper's abstract is too long to fit on the
29: %% title page in preprint2 mode. When that is the case,
30: %% use the longabstract style option.
31: 
32: %% \documentclass[preprint2,longabstract]{aastex}
33: 
34: %% If you want to create your own macros, you can do so
35: %% using \newcommand. Your macros should appear before
36: %% the \begin{document} command.
37: %%
38: %% If you are submitting to a journal that translates manuscripts
39: %% into SGML, you need to follow certain guidelines when preparing
40: %% your macros. See the AASTeX v5.x Author Guide
41: %% for information.
42: 
43: \newcommand{\vdag}{(v)^\dagger}
44: \newcommand{\myemail}{stelzer@astropa.unipa.it}
45: 
46: %% You can insert a short comment on the title page using the command below.
47: 
48: \slugcomment{Accepted for publication in ApJ}
49: 
50: %% If you wish, you may supply running head information, although
51: %% this information may be modified by the editorial offices.
52: %% The left head contains a list of authors,
53: %% usually a maximum of three (otherwise use et al.).  The right
54: %% head is a modified title of up to roughly 44 characters.
55: %% Running heads will not print in the manuscript style.
56: 
57: \shorttitle{Variability of the Accreting Brown Dwarf 2M1207}
58: \shortauthors{Stelzer et al.}
59: 
60: %% This is the end of the preamble.  Indicate the beginning of the
61: %% paper itself with \begin{document}.
62: 
63: \begin{document}
64: 
65: %% LaTeX will automatically break titles if they run longer than
66: %% one line. However, you may use \\ to force a line break if
67: %% you desire.
68: 
69: \title{Emission Line Variability of the Accreting Young Brown Dwarf 2MASSW J1207334-393254: From Hours to Years}
70: 
71: %% Use \author, \affil, and the \and command to format
72: %% author and affiliation information.
73: %% Note that \email has replaced the old \authoremail command
74: %% from AASTeX v4.0. You can use \email to mark an email address
75: %% anywhere in the paper, not just in the front matter.
76: %% As in the title, use \\ to force line breaks.
77: 
78: \author{Beate Stelzer}
79: \affil{INAF -- Osservatorio Astronomico di Palermo, Piazza del Parlamento 1, 90134 Palermo, Italy}
80: \email{stelzer@astropa.unipa.it}
81: 
82: \and 
83: 
84: \author{Alexander Scholz}
85: \affil{SUPA, School of Physics and Astronomy, University of St.Andrews, North Haugh, St. Andrews, Fife KY\,16~9SS, United Kingdom}
86: 
87: \and 
88: 
89: \author{Ray Jayawardhana}
90: \affil{Department of Astronomy \& Astrophysics, University of Toronto, 50 St. George Street, Toronto, ON\,M5S\,3H4, Canada}
91: 
92: 
93: 
94: %% Notice that each of these authors has alternate affiliations, which
95: %% are identified by the \altaffilmark after each name.  Specify alternate
96: %% affiliation information with \altaffiltext, with one command per each
97: %% affiliation.
98: 
99: 
100: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
101: %% style, abstract will output a Received/Accepted line after the
102: %% title and affiliation information. No date will appear since the author
103: %% does not have this information. The dates will be filled in by the
104: %% editorial office after submission.
105: 
106: \begin{abstract}
107: We have obtained a series of high-resolution optical spectra for 2MASSW J1207334-393254 
108: (2M\,1207).
109: %, the closest and oldest known accreting brown dwarf and the best-studied 
110: %sub-stellar member of the TW\,Hydrae association. 
111: Two consecutive observing nights at 
112: the ESO Very Large Telescope with the UVES spectrograph yielded a timeseries with a 
113: resolution of $\sim 12$\,min.
114: % over a wide spectral range ($\sim 3750-5000$\,\AA~ and $\sim 5700-9500$\,\AA). 
115: Additional high-resolution optical spectra were obtained months 
116: apart at the Magellan Clay telescope using the MIKE instrument. Combined with previously 
117: published results, these data allow us to investigate changes in the emission line 
118: spectrum of 2M\,1207 on timescales of hours 
119: %, i.e., during its rotational period ($\sim 24$\,h), 
120: to years. Most of the emission line profiles of 2M\,1207 are broad, 
121: in particular that of H$\alpha$, indicating that the dominant fraction of the emission 
122: must be attributed to disk accretion rather than to magnetic activity. From the 
123: H$\alpha$ $10$\,\% width we deduce a relatively stable accretion rate between 
124: $10^{-10.1...-9.8}\,{\rm M_\odot/yr}$ for two nights of consecutive observations.
125: Therefore, either the accretion stream is nearly homogeneous over (sub-)stellar longitude  
126: or the system is seen face-on. 
127: %The former interpretation seems more likely, as high 
128: %inclination has been inferred previously from the shape of the H$\alpha$ profile and the 
129: %timescale of its variability. 
130: Small but significant variations are evident throughout our 
131: near-continuous observation, and they reach a maximum after $\sim 8$\,h, roughly the 
132: timescale on which maximum variability is expected across the rotation cycle. Together 
133: with past measurements, we confirm that the accretion rate of 2M\,1207 varies by more 
134: than one order of magnitude on timescales of months to years. Such variable mass accretion 
135: yields a plausible explanation for the observed spread in the $\dot{M} - M$ diagram. 
136: The magnetic field required to drive the funnel flow is on the order of a few 
137: hundred G. 
138: Despite the obvious presence of a magnetic field, no radio nor X-ray emission 
139: has been reported for 2M\,1207. Possibly strong accretion suppresses magnetic 
140: activity in brown dwarfs, similar to the findings for higher mass T Tauri stars. 
141: \end{abstract}
142: 
143: \keywords{Accretion, Accretion Disks -- Stars: Low-Mass, Brown Dwarfs -- 
144: Stars: Circumstellar Matter -- Line: Formation, Profiles -- Stars: Planetary Systems} 
145: 
146: \section{Introduction}
147: 
148: In recent years, there is increasing evidence that young brown dwarfs evolve similarly to 
149: their higher-mass counterparts, the T Tauri stars (TTS). 
150: In particular, their formation process includes
151: a phase of mass accretion from a circumstellar disk, 
152: whose presence is probed by infrared emission in excess of the photosphere 
153: \citep[e.g., ][]{Natta02.1, Jayawardhana03.1, Luhman05.3, Allers06.1}. 
154: Spectroscopic features, such as the strength and shape of H$\alpha$ emission, directly
155: trace the accreting material. 
156: %H$\alpha$ emission in late-type stars may have different origins. 
157: In the stellar regime, `weak TTS' have no optically thick circumstellar matter, 
158: and their H$\alpha$ emission is purely chromospheric. 
159: In `classical TTS', the line emission is dominated by contributions from accretion. Next to
160: displaying much stronger lines, their line profiles are broader due to the high velocities in
161: the accreting gas and Stark broadening \citep{Muzerolle01.1}. Winds may play a role as well,
162: and are recognized by blueshifted absorption components and forbidden emission lines
163: \citep{Mundt84.1, Edwards87.1, Hamann94.1, Hartigan95.1}.
164: Analogous observations have been made for brown dwarfs, where classical TTS-like H$\alpha$ emission
165: persists down to the deuterium burning limit \citep{Mohanty05.1}. Forbidden emission lines, 
166: %such as [O\,I] and [S\,II], 
167: tracing outflows have been reported from a small number of brown dwarfs 
168: \citep[e.g.,][]{Fernandez01.1,Barrado04.4}. 
169: 
170: Brown dwarfs are intrinsically faint objects, therefore a detailed study of magnetic 
171: activity and accretion in young brown dwarfs is restricted to the most nearby ones. 
172: A particularly well-suited playground is the TW Hya association (TWA), with
173: four confirmed brown dwarfs at an age of $\sim 8$\,Myr and a distance of $\sim 50$\,pc. 
174: %One of them, TWA\,5B, is a companion separated by $2^{\prime\prime}$ from 
175: %its primary, an early M-type TTS \citep{Webb99.1}.
176: %2M\,1207 and 2M\,1139 were identified as TWA members on basis of the 2\,MASS survey 
177: %\citep{Gizis02.2}; recently a proper motion survey of the Supercosmos Sky Survey data
178: %uncovered a further brown dwarf, SSSPM\,1102 \citep{Scholz05.5}. All these objects are 
179: %proper motion members of the TWA, and are confirmed spectroscopically to be substellar
180: %(spectral types M8 and later). 
181: %
182: 2M\,1207 is the best-studied among the brown dwarfs in the TWA. 
183: Its youth was confirmed by low gravity %\citep{Gizis02.2} 
184: and Lithium absorption, %\citep{Mohanty03.2}, 
185: and its TWA membership by radial velocity %\citep{Mohanty03.2} 
186: and proper motion %\citep{Scholz05.5} 
187: studies \citep{Gizis02.2, Mohanty03.2, Scholz05.5}. 
188: \cite{Mamajek05.1} has revised the previous distance estimate of 2M\,1207 using 
189: the moving cluster method, resulting in $53 \pm 6$\,pc. 
190: 2M\,1207 is the oldest brown dwarf known to 
191: actively accrete from a disk, thus representing a benchmark in constraining 
192: the disk lifetimes in the substellar regime. 
193: The object has recently aroused strong interest because of the discovery 
194: of a planetary mass companion \citep{Chauvin04.1, Chauvin05.1}. 
195: This companion is at a separation of $40$\,AU and not 
196: expected to influence the magnetic and accretion activity of the brown dwarf.
197: 
198: Stellar astronomers are intrigued by 2M\,1207 being the closest known brown dwarf with a disk. 
199: No substantial $L^{\prime}$ excess was detected by \cite{Jayawardhana03.1}, 
200: but mid-IR photometry established the presence of circumstellar material \citep{Sterzik04.1}.
201: This is also corroborated by {\em Spitzer} photometry, where 2M\,1207 is detected 
202: in all IRAC bands and at $24\,\mu$m above photospheric levels \citep{Riaz06.1}.
203: The spectral energy distribution of 2M\,1207 shows evidence for dust evolution,
204: similar to the higher-mass disk-bearing TTS in the TWA. 
205: 2M\,1207 is therefore a lower-mass analog of the classical TTS. 
206: 
207: Strong H$\alpha$ emission has been reported right with the identification of 2M\,1207
208: as a TWA candidate \citep{Gizis02.2}. 
209: Subsequent
210: high-resolution spectra have shown that the H$\alpha$ profile is broadened and
211: asymmetric, thus displaying characteristic signatures of accretion \citep{Mohanty03.2}. 
212: The H$\alpha$ equivalent width measured by \cite{Mohanty03.2} was only
213: about one-tenth of the earlier value obtained by \cite{Gizis02.2},  
214: and this variability was interpreted as a sign for variable mass accretion rate. 
215: In a subsequent study of emission line variability, aiming at constraining these variations, 
216: 2M\,1207 presented dramatic changes in the shape and intensity of its H$\alpha$ line in 
217: data obtained two months apart 
218: %suggesting strong variations in the accretion rate on time scales of several weeks
219: \citep{Scholz05.4}.
220: The observed profile variability was consistent with a scenario of asymmetric accretion
221: %(by a magnetic funnel)
222: with a direct view into the (magnetic) funnel flow
223: %accretion column onto a star seen at low inclination
224: \citep{Scholz06.1}.
225: In this picture, variability is also expected on the time scale of the rotation
226: period ($\sim 1$\,d),
227: as the hot accretion spots are corotating with the star. Existing data have remained inconclusive
228: in this respect due to poor sampling. Therefore, we have collected new high spectral resolution
229: data with the aim to monitor the H$\alpha$ variability on shorter time scales. 
230: 
231: Here we present the results obtained during two consecutive nights of observation, and
232: from five additional spectra taken weeks to months apart. 
233: The observations are described in Sect.~\ref{sect:obs}. 
234: In Sect.~\ref{sect:profiles} we present a detailed analysis of the characteristics of 
235: the H$\alpha$ emission and other prominent emission lines in the optical spectrum of 2M\,1207. 
236: The results are discussed in Sect.~\ref{sect:discussion}, where the spectral variability
237: is examined on various timescales from hours to years. %Furthermore, %in Sect.~\ref{sect:discussion} 
238: Estimates for the accretion rate and the magnetic field strength of 2M\,1207 are presented, 
239: and put in context with other (sub)stellar objects. Finally, we compare the accretion and
240: activity characteristics of all known brown dwarfs in TWA. 
241: Concluding remarks are found in Sect.~\ref{sect:conclusions}.  
242: 
243: 
244: 
245: \section{Observations}\label{sect:obs}
246: 
247: 2M\,1207 was observed with the {\em Ultraviolet and Visual Echelle Spectrograph} (UVES) 
248: at the ESO VLT\,2 during two nights May 8-9, 2006 in the dichroic mode that covers 
249: the full spectral range of UVES by simultaneous use of both the blue and the red arm. 
250: We used the standard setting DICHR\,\#2 centered at $437$\,nm for the
251: blue arm and at $760$\,nm for the red arm. 
252: %(the former allows simultaneous observations of all CaII NIR triplet lines and Halpha)
253: This setup yields simultaneous spectral coverage from  $3730-4990$\,\AA~ on the blue chip 
254: and from $5650-9460$\,\AA~ on the red mosaic. 
255: A slit width of $0.8^{\prime\prime}$ was chosen 
256: resulting in a resolving power $R \approx 50 000$. 
257: The CCDs were rebinned to $2 \times 2$ pixels.
258: 
259: The integration time was $600$\,s. During the two observing nights, we obtained a 
260: total of $59$ exposures of 2M\,1207 at blue and red wavelengths. Except for 
261: some constraints due to calibrations and weather the frames were taken consecutively, 
262: such that the time resolution for the investigation of spectral changes is $\sim 12$\,min 
263: over a large part of the observations.
264: 
265: %equipped with a mosaic of two CCDs. The mosaic is made of a 2K $\times$ 4K EEV chip 
266: %(pixel size 15 m) for the blue part of the red arm and an MIT-LL CCD for the red part of the red arm.
267: 
268: The data were processed with the UVES pipeline implemented in the MIDAS environment 
269: \citep{Ballester00.1}. 
270: Reduction steps included bias subtraction, flatfield correction, and wavelength calibration. 
271: The final one-dimensional spectra have been corrected for contaminating sky light
272: and cosmic rays have been removed. % with an algorithm that identifies spikes in the extracted spectrum. 
273: Each spectrum was corrected for the radial velocity of 2M\,1207 
274: \citep[$11.2$\,km/s; ][]{Mohanty03.2} and for the barycenter motion of the Earth. 
275: %The wavelength regions used for the extraction of line and continuum fluxes are summarized
276: %in Table~\ref{tab:lines}. 
277: 
278: We extracted a spectrum from each of the individual science frames. In addition, 
279: to increase the S/N for the detection and measurements of weak spectral features, we built
280: averages of each four consecutive frames. Thus, the effective exposure times of these merged spectra 
281: are $40$\,min. % ($4$ frames added) or $3000$\,s ($5$ frames added). 
282: %This way we obtain $14$ average spectra. 
283: Three individual frames with poor S/N were not considered
284: in the merging process, such that we obtain $14$ average spectra from the two nights.
285: The spectra were rebinned to $R=50000$ around individual emission lines 
286: before the scientific analysis was carried out. 
287: 
288: In February and April 2006, we obtained five additional exposures of 2M\,1207 using 
289: the Magellan Inamori Kyocera Echelle (MIKE) spectrograph at the Magellan Clay telescope 
290: on Las Campanas, Chile. These data were reduced using standard routines within IRAF, 
291: %\footnote{IRAF...}, 
292: including bias subtraction, flatfielding, wavelength and relative flux calibration,
293: and removal of cosmic rays. For more details of the reduction procedure, see
294: \cite{Scholz06.1}. Exposure times for the MIKE spectra were
295: typically $10$\,min, and the nominal resolution is $R \approx 25000$.
296: 
297: 
298: \section{Emission line spectrum}\label{sect:profiles}
299: 
300: \subsection{H$\alpha$ emission}\label{subsect:halpha}
301: 
302: The main aim of this study is to examine the variability of the H$\alpha$ emission of 2M\,1207
303: with good time resolution and on time-scales of $\sim 1$\,d, corresponding to its 
304: suspected rotation period. An estimate for the period is given by \citet{Scholz05.4}. In brief, 
305: the rotational velocity of 2M\,1207 is $v \sin{i} = 13 \pm 2$\,km/s \citep{Mohanty03.2}. 
306: Together with a radius of $0.27\,R_\odot$ extracted from evolutionary models \citep{Chabrier00.2},
307: the period is expected to be $P / \sin{i} \approx 25.2$\,h. Assuming a random position for
308: the inclination, the most probable value for $\sin{i}$ is $0.81$, and for this value the rotation period
309: is $20.4$\,h. Magnetospheric accretion models \citep{Hartmann94.1} have shown that 
310: the presence of a red absorption feature in the Balmer lines (frequently seen in 2M\,1207; see below)
311: indicates high inclination. For $i \geq 60^\circ$ the expected period is $\geq 22$\,h. 
312: 
313: The H$\alpha$ emission of 2M\,1207 is thought to be dominated by accretion: the broad wings with velocities
314: of up to $\pm 200$\,km/s can not be explained by magnetic activity, the rotational velocity is comparatively
315: small and can not be responsible for the broad wings, and 
316: signatures for winds are very weak \citep{Mohanty05.1,Whelan07.1}. 
317: 
318: The most widely used indicator for identifying accreting 
319: stars is the equivalent width $W_{\rm EQ}$ of the H$\alpha$ line, where weak emission is
320: due to chromospheric magnetic activity and strong emission, above 
321: a certain threshold of $W_{\rm EQ}$, is attributed to accretion. 
322: However, its dependence on the
323: continuum level precludes $W_{\rm EQ}$ as a universal measure. To take account of
324: the declining continuum for later spectral types, the threshold for $W_{\rm EQ}$ that
325: separates accretors from non-accretors among TTS is usually adapted in discrete
326: steps from $\sim 5$\,\AA~ for spectral types earlier than M0 to something like 
327: $20$\,\AA~ for late M-type stars \citep{Martin98.4}. 
328: % \citep[see e.g. ][]{Barrado03.1}. 
329: %Even more problematic is the fact that 
330: Furthermore, 
331: equivalent width measurements of faint continuum sources such as brown dwarfs are intrinsically associated with 
332: considerable uncertainties. 
333: This is due to the low S/N and the presence of overlapping absorption features
334: that form a `pseudo-continuum' which depends on the spectral resolution. 
335: To avoid these problems, 
336: following a suggestion by \cite{White03.1}, the full width
337: at $10$\,\% of the peak height ($W_{\rm 10\,\%}$) of H$\alpha$ 
338: has been established as more reliable accretion indicator: 
339: Objects with $W_{\rm 10\,\%} > 200$\,\AA~ can be considered accretors \citep{Jayawardhana03.2}. 
340: 
341: 
342: \subsubsection{Profiles and line widths}\label{subsubsect:Ha_profile_width}
343: 
344: We use $W_{\rm 10\,\%}$ as the primary accretion indicator, but examine also other parameters
345: that characterize the line intensity and shape. Fig.~\ref{fig:ha_timeseries} shows 
346: the time-series of these parameters obtained during our run with UVES on May 8-9, 2006. 
347: The $10$\,\% width is clearly above the accretion threshold 
348: throughout our observation ($W_{\rm 10\,\%} \approx 280...320$\,km/s); 
349: on average it is higher during the second night, 
350: in the course of which it slightly declines. 
351: The equivalent width has ups and downs during the first night, and 
352: has its maximum at the beginning of the second night, after which it 
353: declines dramatically and in a continuous way until the end of the observation. 
354: %As detailed in Sect.~\ref{subsect:acc_lines}, the errors in $W_{\rm EQ}$ are significant, 
355: The errors in $W_{\rm EQ}$ are dominated by the (faint) continuum level.  
356: We estimated the uncertainty of the continuum flux by varying the sky extraction area, 
357: and found that it is of the order of $15$\,\%. Then we computed error bars for $W_{\rm EQ}$
358: assuming for the continuum flux $0.85 \cdot I_{\rm cont}$ and $1.15 \cdot I_{\rm cont}$, respectively.
359: Fig.~\ref{fig:ha_timeseries} shows that when taking into account these uncertainties, 
360: the observed variations of $W_{\rm EQ}$ remain significant. 
361: 
362: The H$\alpha$ emission is double-peaked throughout all our observations from February to May 2006.   
363: Fig.~\ref{fig:Ha_profiles} displays the time-series of all $14$ normalized average line profiles
364: $\bar{I_{\rm n}}(\lambda)$ from the UVES run (May 2006), and Fig.~\ref{fig:Ha_profiles_mike} shows the
365: $5$ normalized profiles obtained with MIKE (February and April 2006).  
366: For clarity, in Figs.~\ref{fig:Ha_profiles} and~\ref{fig:Ha_profiles_mike} the individual spectra have been 
367: shifted with respect to each other in the vertical direction. 
368: 
369: \cite{Scholz05.4} have shown that the double-peaked H$\alpha$ profile of 2M\,1207 
370: can be interpreted as a broad emission line onto which a red-shifted absorption component is superposed. 
371: For the minimum of the absorption feature we measure a position that varies 
372: in the range $\Delta v_{\rm min} \approx -10...+70$\,km/s with respect to the expected wavelength
373: ($\Delta v_{\rm min} \approx +30...+70$\,km/s in February 2006, 
374: $\Delta v_{\rm min} \approx +10$\,km/s in April 2006, and 
375: $\Delta v_{\rm min} \approx -10...+40$\,km/s in May 2006). 
376: From Fig.~\ref{fig:ha_timeseries} it appears that twice during the extensive monitoring in May 2006 
377: $v_{\rm min}$ undergoes a discrete jump from high redshift towards the expected line center. 
378: These times correspond to line profiles in which the absorption trough is wide and asymmetric,
379: such that the position of the minimum can not be determined precisely 
380: (Fig.~\ref{fig:Ha_profiles}).  
381: %and therefore the minimum position is not well-defined. 
382: As a measure for the asymmetry of the emission component, Fig.~\ref{fig:ha_timeseries} shows the difference between
383: the $10$\,\% width on the blue (left) and on the red (right) side of $\Delta v=0$.
384: %The red asymmetry 
385: %of the H$\alpha$ emission is documented by the negative values of $W_{\rm 10\,\%}(l-r)$. 
386: %The most symmetric and best centered line profile is seen at the beginning of the second night,
387: %and the redshift increases in the course of the second night 
388: In the course of the UVES observations, $W_{\rm 10\,\%}(l-r)$ varied from $-40...+40$\,km/s.
389: In particular, during the second night the line moved systematically from the blue
390: to the red. This is in contrast to the absorption feature shown above to be redwards of $\Delta v = 0$ 
391: throughout the observation  
392: (cf. also Fig.~\ref{fig:Ha_profiles}). 
393: 
394: We examined the shape of the H$\alpha$ line and its change in time by 
395: modeling each of the $14$ average profiles %$\bar{I_{\rm n}}(\lambda)$ 
396: with two quasi-Gaussians, 
397: one with positive normalization representing the emission component and 
398: one with negative normalization for the absorption component. 
399: Each of these has three free parameters (position, width, and normalization). %,
400: %i.e. the fit functions deviate from a `canonical' Gaussian in that the normalization 
401: %is independent of the width. 
402: %Before the fitting process
403: %all profiles have been normalized to the continuum near the line. 
404: The equivalent widths of the emission $W_{\rm EQ; em}$ and the absorption $W_{\rm EQ; abs}$ 
405: component derived from the fits turn out to be strongly correlated (Fig.~\ref{fig:Ha_flux}). 
406: Since $W_{\rm EQ; em}$ is by definition `normalized' to the continuum this relation is
407: not trivial.
408: In Fig.~\ref{fig:Ha_flux} we also plot the {\em observed} flux, i.e. the difference between 
409: the flux in the emission profile and the absorption profile, versus the $10$\,\% width. 
410: 
411: The H$\alpha$ $10$\,\% width of 2M\,1207 has been measured at eight epochs so far 
412: (see Table~\ref{tab:w10}). 
413: In most of the high-resolution spectra collected so far 2M\,1207 
414: showed $W_{\rm 10\,\%}$ clearly above the accretion threshold, 
415: although in some occasions there were only marginal signs of accretion 
416: ($W_{\rm 10\,\%} \sim 200$\,km/s). 
417: 
418: 
419: 
420: \subsubsection{Quantifying variability}\label{subsect:var_quant}
421: 
422: To quantify the H$\alpha$ variability as a function of wavelength we examined the 
423: normalized variance profile
424: \begin{equation}
425: \sigma^2_{\rm n}(\lambda) = \sigma^2(\lambda) / \bar{I}_{\rm n}(\lambda),  
426: \label{eq:varprof}
427: \end{equation}
428: where the variance profile 
429: $\sigma^2(\lambda) = \frac{1}{N-1} \sum_{i=1}^N{[I_{\rm n,i}(\lambda)  - \bar{I}_{\rm n}(\lambda)]^2}$
430: measures flux variations across the line profile \citep{Johns95.1}.  
431: In the bottom panels of Fig.\ref{fig:Ha_profiles} we plot $\sqrt{\sigma^2_{\rm n}}$ 
432: for three distinct parts of the observation in May 2006. From left to right, are shown 
433: the variances obtained from 
434: all $19$ spectra from the first night, the first $20$ and the last $20$ spectra of the
435: second night, respectively. 
436: This way, the individual $\sqrt{\sigma^2_{\rm n}}$ give information about the variations as a function
437: of wavelength on short time-scales (within the $\sim 4$\,h covered by $19$ respectively $20$ 
438: consecutive frames), 
439: and a comparison of the different $\sqrt{\sigma^2_{\rm n}}$'s probes changes up to $\sim 1$\,d. 
440: 
441: The normalized variance in the continuum expected from Poisson noise is
442: $\sigma^2_{\rm n, exp}(\lambda) = \sigma^2_{\rm n}(\lambda) (\bar{\sqrt{I}}/\bar{I})^2$. 
443: The zero-variability level $\sqrt{\sigma^2_{\rm n, exp}}$ 
444: is shown as dashed line in the bottom panels of Fig.~\ref{fig:Ha_profiles}.
445: Its comparison with the $\sqrt{\sigma^2_{\rm n}}$'s shows clear signs for variability in the
446: line profile. 
447: Throughout the time series, the variance profile (i.e. variability as a function of wavelength)
448: features three distinct peaks, roughly at positions of $-100...-150$, $-20...+20$, and $+100...+150$\,km/s.
449: We note that the same structure was seen by \cite{Scholz06.1} in their H$\alpha$ 
450: time series of 2M\,1207.
451: The position of these peaks (indicating the velocities of maximum changes in the H$\alpha$ flux)
452: is roughly coinciding with the position of the two maxima and the minimum in the H$\alpha$
453: profile. 
454: 
455: The variance profile yields only the time average of the changes as a function of wavelength. 
456: To study the profile variations in more detail we examine 
457: the change of the flux between consecutive spectra across the line. 
458: The relative flux change is given by 
459: \begin{equation}
460: D_{\rm ji}(\lambda) = \frac{1}{2} \times \frac{I_{\rm n,j}(\lambda) - I_{\rm n,i}(\lambda)}{I_{\rm n,j}(\lambda) + I_{\rm n,i}(\lambda)},
461: \label{eq:ampl_change_1}
462: \end{equation}
463: where $I_{\rm n}(\lambda)$ is the normalized profile, $i$ is a running number denoting the
464: exposures and $j = i + 1$. 
465: These relative flux changes $D(\lambda)$ are used to investigate variations 
466: as a function of the time lag between two spectra \citep{Johns95.1}. 
467: To this end, the $D(\lambda)$'s are computed not only for two subsequent spectra but 
468: for all pairs of the $59$ spectra from May 8-9, 2006, 
469: i. e. for each time lag $\delta t = t2 - t1 > 0$ sampled by our UVES data. 
470: Then, we group the $D(\lambda)$'s in time lag bins of $1$\,h (${\rm \Delta t}$) 
471: and calculate the standard deviation $\sigma_{\rm \Delta t}(\lambda)$ 
472: of the $D(\lambda)$'s in each of these bins. Finally, we calculate for each time bin 
473: the average of these standard deviations over the whole line profile ($\bar{\sigma}_{\rm \Delta t}$). 
474: The result is shown in Fig.~\ref{fig:ds}, where 
475: the numbers on top of the panel indicate the number of data points in each of the time lag bins. 
476: The gap is produced by the absence of observations with time lags between
477: $\approx 9...16$\,h, i.e. the data to the left of this gap corresponds to spectra from 
478: the same night, and the data to the right corresponds to spectra from two different nights. 
479: The level of profile variability doubles  
480: from $\sim 2$\,\% for the shortest time lag ($1$\,h) to   
481: a maximum of $\approx 4$\,\%, that is reached for a timelag of $\sim 8$\,h.
482: 
483: To test whether the variations are correlated across the line profile we computed the correlation
484: coefficient for each pair of spectral bins using the profiles $\bar{I}_{\rm i}(\lambda)$,  
485: with $i=1,...,N$. 
486: This results in a correlation matrix $r_{\rm i,j}$ 
487: of dimension $N_{\rm b} \times N_{\rm b}$, where $N_{\rm b}$
488: is the number of spectral bins. The confidence level of each coefficient in the matrix is computed
489: using the error function 
490: \begin{equation}
491: E_{\rm i,j} = \frac{\mid r_{\rm i,j} \mid \sqrt{N_{\rm b}}}{\sqrt{2}},
492: \label{eqw:erf}
493: \end{equation}
494: such that $1-E_{\rm i,j}$ is the confidence level for the correlation to be true. 
495: 
496: %In Fig.~\ref{fig:corr_matrix_Ha}, the correlation matrix 
497: %$r_{\rm i,j}$ is displayed as contour plot. Only values with
498: %confidence level $>99.9$\,\% are considered in this representation.
499: %
500: The correlation matrix has been computed for different subgroups of H$\alpha$ profiles from 
501: the UVES run. Fig.~\ref{fig:corr_matrix_Ha} displays the most interesting correlation matrices
502: $r_{\rm i,j}$ as contour plots. Only values with
503: confidence level $>99.9$\,\% are considered in this representation. The first two
504: diagrams from the left represent the matrices of each $19$ consecutive spectra, obtained in
505: the first and the second observing night, respectively. The cross-shaped pattern seen in the 
506: leftmost diagram
507: is peculiar. The center of the cross is offset by $\approx +25$\,km/s to the red side, and it
508: represents the spectral region near the minimum of the line profile (see Fig.~\ref{fig:ha_timeseries}). 
509: Closer examination shows that this pattern stems from a short time interval. 
510: %during which the profile changes near the absorption minimum in an uncorrelated manner. 
511: This is demonstrated in the rightmost diagram of Fig.~\ref{fig:corr_matrix_Ha}, 
512: that displays the correlation matrix for exposures $i = 7,...,10$, 
513: roughly corresponding to the second average profile in the leftmost panel of Fig.~\ref{fig:Ha_profiles}.
514: Clearly, during this time interval, the profile changes near the 
515: absorption minimum are uncorrelated with all velocities except for the velocities in close vicinity. 
516: The double-square shaped pattern observed during the second night 
517: (middle panel of Fig.~\ref{fig:corr_matrix_Ha}) is more typical \citep[cf.][]{Johns95.1}. 
518: The absence of contours in the upper left and in the lower right
519: of this graph means that red and blue wings vary in an uncorrelated fashion. The range near
520: $v \approx +75$\,km/s, corresponding to the position of the right emission peak, 
521: is particularly well correlated with the whole line. 
522: 
523: 
524: 
525: \subsection{Other signatures of accretion}\label{subsect:acc_lines}
526: 
527: Our high S/N UVES spectra allow us to present the first time series for other optical emission
528: lines of 2M\,1207. 
529: In particular, we detect the Balmer lines up to H$\epsilon$. This latter one 
530: is partly blended with Ca\,II\,H\,$\lambda3968$.
531: 
532: The shape of the H$\beta$ profile
533: follows closely that of H$\alpha$ discussed in detail above. All higher Balmer lines are predominantly 
534: characterized by a double-peaked profile, but at times turn into a single-peaked shape. 
535: This is demonstrated in Fig.~\ref{fig:balmer}, that shows the contemporaneous Balmer series
536: for two epochs. 
537: Each panel in Fig.~\ref{fig:balmer} is obtained from an $8$-frame average. 
538: When present, the absorption minimum is found at roughly the same position for all Balmer lines, 
539: but the ratio of the emission fluxes left and right of the minimum decreases
540: systematically for the higher Balmer lines, i.e. the lower Balmer lines are red-dominated and 
541: the higher Balmer lines are blue-dominated. 
542: 
543: In contrast to the Balmer lines, the Ca\,II H+K emission is narrow with FWHM of only $\sim 20$\,km/s 
544: (see Table~\ref{tab:eqw}). 
545: Furthermore, we detect several lines of He\,I ($\lambda4471$, $\lambda5876$, $\lambda6678$, $\lambda7066$) 
546: and the Ca\,II infrared triplet (hereafter Ca\,IRT) in emission. All these lines
547: are predominantly detected in accretors and, therefore, thought to be related to the accretion process. 
548: The Ca\,II\,$\lambda8662$ flux was shown to be directly related to the mass 
549: accretion rate \citep{Mohanty05.1}. 
550: The three lines of the Ca IRT appear with roughly similar strength,
551: inconsistent with formation in an outflow \citep{Reipurth86.1, Fernandez01.1}. 
552: No Na\,D emission is seen, consistent with the prediction of magnetospheric accretion models 
553: for low accretion rates \citep{Muzerolle01.1}. 
554: 
555: We measured $W_{\rm EQ}$ and the FWHM of the stronger ones of the emission lines discussed above. 
556: The continuum was estimated from two $5$\,\AA~ wide line-free 
557: regions left and right of the respective line. 
558: The observed range of $W_{\rm EQ}$ is reported for each line in Table~\ref{tab:eqw} together with
559: its approximate uncertainty. Given the often loose use of equivalent widths in the literature, 
560: it seems appropriate to stress that $W_{\rm EQ}$ is associated with considerable errors in 
561: objects with faint continuum. This is clear from the numbers given in Table~\ref{tab:eqw}. 
562: The errors cited in that table do not include systematic uncertainties, again affecting 
563: mainly the (faint) continuum, that we discovered in the process of defining the sky extraction 
564: area. Since we are interested in relative changes within the UVES dataset, systematic 
565: uncertainties are not a big concern. However, they should be kept in mind when comparing data 
566: obtained with different instruments, and analysed in different ways by different astronomers. 
567: 
568: All emission lines are stronger in the second night of the UVES observations, and they are particularly
569: weak towards the end of the first night (around MJD\,53863.25  = UT 05\,h on May 8, 2006). 
570: At the same moment the shape of some of the strongest emission lines
571: underwent a change: 
572: (i) He\,I\,$\lambda5876$ temporarily becomes flat-topped and symmetrical, while
573: it is markedly asymmetric with an extended red wing throughout most of the two nights 
574: (see Fig.~\ref{fig:heI}); 
575: (ii) in H$\alpha$ the $10$\,\% width is at its minimum, and 
576: the deficiency in the red vs. the blue peak is particularly strong; 
577: (iii) in the higher Balmer lines the red portion of the line disappears almost completely, 
578: and (iv) some of the weaker lines nearly disappear at the same time. 
579: 
580: 
581: 
582: \section{Discussion}\label{sect:discussion}
583: 
584: 
585: \subsection{Spectral variability}\label{subsect:specvar}
586: 
587: We have examined the optical emission line variability of 2M\,1207 at high spectral
588: resolution 
589: in data obtained at various epochs through February, April, and May 2006.
590: In particular, we present for the first time a near continuous spectral sequence
591: from two consecutive nights with a time-resolution of $\approx 12$\,minutes. 
592: The study presented here complements previous work that was characterized by much poorer
593: temporal sampling. 
594: 
595: 
596: 
597: \subsubsection{Short timescales (days to hours)}\label{subsubsect:short}
598: 
599: \citet{Scholz05.4} have attributed the drastic changes of the H$\alpha$ line profile seen in 
600: earlier high-resolution spectra of 2M\,1207 to varying viewing angle of the
601: accretion column(s) over the course of the brown dwarf's rotation cycle. The absorption component
602: seemed to appear and reappear on a time-scale of $\sim 1$\,day, 
603: an interval approximately coincident with the expected rotation period, 
604: therefore suggesting that the accretion spots moved with respect to the observer. 
605: However, with a maximum of $3$ spectra per night, 
606: the available data did not provide adequate sampling across the presumed rotation cycle. 
607: 
608: Within the two nights of near-continuous observation in May 2006 
609: there is clear variability present in the H$\alpha$ emission and other emission lines, 
610: albeit much less pronounced and systematic than expected from previous data. 
611: Here, we summarize its characteristics: 
612: (i) The line profile is double-peaked throughout the two nights, displaying the central
613: absorption reversal characteristic for accretion in star-disk systems seen near edge-on; 
614: (ii) the absorption feature moves between $-10...+40$\,km/s; 
615: (iii) the absorption minimum widens twice during the observations with a time interval of
616: $\sim 23$\,h;
617: (iv) variability across the line is inferred from the variance profile and the correlation matrix; and 
618: (v) the line flux varies by a factor of $\approx 6$, 
619: correlated with the (small) variations of the $10$\,\% width.
620: 
621: %According to \cite{Hartmann94.1} near-central absorption 
622: %features in Balmer lines from TTS are due to the lower temperature
623: %in the outer regions of the accretion stream
624: %where low-velocity photons emitted near the accretion shock are more 
625: %likely to be absorbed than faster ones. 
626: %% that forms the central portion of the line. 
627: The variable redshift and width of the absorption minimum implies that 
628: the viewing angle of the absorbing layers of the accretion column changed throughout the observation. 
629: \cite{Bouvier03.1} explained the radial velocities of the redshifted absorption components of the
630: TTS AA\,Tau with changes in the structure of the magnetosphere. % combined with the viewing conditions: 
631: In this scenario, during times of high observed
632: absorption redshift an observer at high inclination looks flat onto the magnetic field lines
633: that carry the absorbing material,
634: while low absorption redshift implies a large angle between the absorbing material and the line-of-sight.
635: The different viewing angles can be realized by inflating or deflating the star-disk field. 
636: In the case of 2M\,1207, the observed changes that occur on time-scales shorter than the rotation period
637: may suggest that the magnetosphere is somewhat inhomogeneous across stellar longitude. 
638: However, the changes are not as dramatic as seen before. 
639: The range of observed redshifts in the absorption minimum of 2M\,1207 is comparable to that of AA\,Tau
640: but there is no blueshifted absorption. Therefore, for 2M\,1207 outflows do not play a significant
641: role in the formation of the H$\alpha$ line. 
642: 
643: Although there is no obvious periodicity, the rotation of the star is the most
644: likely cause of the short-term changes of the H$\alpha$ profile summarized above.
645: We have shown that the fractional profile changes reach a maximum after a time-lag of $\sim 8$\,h,
646: and they remain high on a plateau for larger timelags. (The exact position of the maximum is
647: unclear due to the absence of data with timelags between $9-15$\,h.) This timescale is roughly 
648: consistent with one half the expected rotation period, and therefore with the time interval in which
649: maximum changes due to rotational effects are expected. 
650: 
651: If possible contributions from chromospheric activity and winds are neglected, 
652: the emission line flux is expected to be a measure for the accretion rate, and this is supported
653: by its correlation with the $10$\,\% width. 
654: %2M\,1207 has shown little or no signs for magnetic activity 
655: %\citep[no X-ray and no radio emission; ][]{Gizis04.1,Osten06.3}, and the absence of forbidden
656: %line emission in our high-resolution spectra suggests that winds do not play a significant role. 
657: This correlation is not trivial because the fit function's
658: normalization is independent of its width, unlike in a `canonical' Gaussian. 
659: We conclude, from the correlation observed between the equivalent widths of the emission and
660: the absorption components  
661: (Fig.~\ref{fig:Ha_flux}) 
662: that the amount of absorbing material is proportional to the total amount of material crossing 
663: the line-of-sight. 
664: 
665: With the exception of the Ca\,II\,H + K lines, the strong emission lines 
666: (higher Balmer lines, He\,I and Ca\,IRT) follow the H$\alpha$ emission closely 
667: in their shape and variability, indicating a common origin, 
668: i.e. accretion, and chromospheric contributions are not significant. 
669: In particular, throughout our observations the lines of the Balmer series 
670: are mostly double-peaked. Only the higher lines H$\delta$ and H$\epsilon$ at times turn into a single-peaked shape. 
671: 
672: 
673: 
674: 
675: \subsubsection{Intermediate timescales (weeks to months)}\label{subsubsect:intermediate}
676: 
677: The H$\alpha$ line is characterized by a double-peaked profile in all spectra obtained
678: from February to May 2006. In the spectra from February and April 
679: the redshift of the absorption feature is systematically larger with respect to
680: the observations from May of the same year, and consequently  
681: the line profiles observed in February and April are also more asymmetric
682: than the ones from May. However, in the data from 2006 
683: the variations are smoother and less dramatic than the ones from 2005, 
684: for which \cite{Scholz05.4} reported alternating single- and double-peaked profiles 
685: within individual observing nights.  
686: 
687: %The assumption of high inclination is, therefore, supported by the presence of the 
688: %absorption reversal. 
689: If the previous assumption of a high-inclination system rotating at a rate of 
690: $\sim 1$\,d is correct, the 
691: data from 2006 are compatible with a rather homogeneous distribution of accretion elements,
692: i.e. magnetosphere, across the (sub)stellar longitudes, with smaller inhomogeneities as discussed
693: in Sect.~\ref{subsubsect:short}. In contrast, during 2005 the absorption reversal appeared
694: and disappeared within one rotation cycle, implying strong variations in the accretion structure
695: across the stellar surface. 
696: We conclude that the distribution of the accretion elements or the structure of the 
697: magnetosphere has changed within $\sim 1$\,yr. 
698: In both years, 2005 and 2006, the mass accretion rate, 
699: as inferred from the $10$\,\% width of the H$\alpha$ emission 
700: \citep[see e.g.][ and Sect.~\ref{subsect:macc}]{Natta04.2},
701: has shown fluctuations by a factor $\sim 2$ 
702: on time-scales of weeks to months (cf. Table~\ref{tab:w10}).
703: 
704: 
705: 
706: \subsubsection{Long timescales (years)}\label{subsubsect:long}
707: 
708: Three years have passed since the first high-resolution
709: spectrum of 2M\,1207 was obtained, and in this period there seems to be a tendency  
710: towards enhanced mass accretion rate. 
711: Note, however, that the large H$\alpha$ equivalent width measured by 
712: \cite{Gizis02.2} in a low-resolution spectrum counteracts this trend, and 
713: fluctuations on shorter timescales (discussed above) may be superposed on the long-term evolution. 
714: Further observations are
715: required to confirm our speculation about the secular evolution of the H$\alpha$ emission.  
716: 
717: 
718: 
719: 
720: \subsection{Accretion rate and magnetic field}\label{subsect:macc}
721: 
722: \cite{Natta04.2} have parametrized the accretion rate as a function of $W_{\rm 10\,\%}$ for TTS 
723: and brown dwarfs. 
724: From their Eq.~1, we find an accretion rate of $10^{-10.1...-9.8}\,{\rm M_\odot/yr}$ 
725: for the range of H$\alpha$ $10$\,\% widths measured for 2M\,1207 during the UVES run in May 2006. 
726: In previous measurements, $\dot{M}_{\rm acc}$ was as low as $\sim 10^{-11.2}\,{\rm M_\odot/yr}$,  
727: i.e. 2M\,1207 changes its accretion rate by a factor of $2$ on the timescale of
728: days, and at least by one order of magnitude over the years. 
729: 
730: In Fig.~\ref{fig:mdot} we put 2M\,1207 in the context with other accreting (sub)stellar objects.
731: A correlation between accretion rate and mass 
732: ($\dot{M}_{\rm acc} \sim M^\alpha$ with $\alpha \approx 2$) has been reported in the literature.
733: In Fig.~\ref{fig:mdot} we display this relation using data from 
734: the references given in \cite{Muzerolle05.1}.
735: The range of $\dot{M}_{\rm acc}$ inferred from all available data of 2M\,1207 is 
736: indicated by large circles connected by a vertical line. 
737: 
738: Various explanations have been proposed for the correlation between $\dot{M}_{\rm acc}$ and $M$,  
739: and its large ($>2$\,dex) scatter. 
740: \cite{Padoan05.1} see in the $\dot{M}_{\rm acc} - M$ relation
741: a consequence of large-scale Bondi-Hoyle accretion. In their scenario, 
742: the scatter is produced by the different physical 
743: conditions in different star forming environments. \cite{Alexander06.1} assumed that the $\dot{M}_{\rm acc} - M$ 
744: relation reflects the initial parameters of the disk-star system, and they showed that the scatter 
745: can be reproduced by a combination of varying disk initial parameters and 
746: the viscous evolution of the disk in the course of which the mass accretion rate decreases. 
747: According to \cite{Gregory06.1} the $\dot{M}_{\rm acc} - M$ 
748: relation arises naturally from models of magnetospheric accretion assuming a realistic field
749: geometry. 
750: Finally, \cite{Mohanty05.1} suggested declining disk ionization with decreasing 
751: stellar mass as the origin of the $\dot{M}_{\rm acc} - M$ relation,
752: implying that most of the scatter is due to intrinsic variability in accretion rates. 
753: This view was disputed by \cite{Natta06.1} who did not find strong variations of the 
754: near-IR emission lines indicative of accretion 
755: in a sample of $14$ objects from the $\rho$\,Oph star forming region.
756: On the other hand, it has been demonstrated
757: for a few examples of individual objects, that indeed accretion rates do vary substantially
758: \citep{Gullbring96.2,Alencar05.1,Scholz05.4,Scholz06.1}. 
759: We point out that the range of mass accretion rates derived for 2M\,1207 
760: is comparable to the spread in the $\dot{M}_{\rm acc} - M$ relation. 
761: Therefore, variability might indeed explain this spread. 
762: Repeated measurements of $\dot{M}_{\rm acc}$ for known accretors will be helpful to verify our suggestion. 
763: 
764: An alternative method to derive accretion rates has been proposed by \cite{Mohanty05.1}.
765: They established that both classical TTS and brown dwarfs follow 
766: a linear correlation between the Ca\,II\,$\lambda8662$ emission flux and
767: $\dot{M}_{\rm acc}$ determined independently from H$\alpha$ modeling or veiling measurements. 
768: Following their recipe, we estimate the continuum flux $F_{\rm cont}$ 
769: of 2M\,1207 at $8662$\,\AA~ from the synthetic 
770: DUSTY spectra of \cite{Allard00.1}. We adopt $T_{\rm eff} = 2800$\,K, $\log{g} = 4.0$, 
771: and negligible veiling \citep[see discussion in][ for justification of these choices]{Mohanty05.1}. 
772: The line flux is then given by $F_{\rm CaII} = F_{\rm cont} \cdot W_{\rm EQ}$.
773: For the highest equivalent width measured in May 2006 this results in 
774: $\log{F_{\rm CaII}}\,[{\rm erg/cm^2/s}] \approx 5.1$, 
775: and the fit relation for the low-mass sample of \cite{Mohanty05.1} yields 
776: an accretion rate of $\dot{M}_{\rm acc} \approx 10^{-10.3}\,{\rm M_\odot/yr}$. This is in rough
777: agreement with the value we derived above from the H$\alpha$ $10$\,\% width. We caution, that using 
778: the Ca\,II\,$\lambda8662$ line for estimating the mass accretion rate assumes a negligible contribution 
779: from activity. This seems to be true for brown dwarfs on a statistical basis 
780: \citep[`the $\lambda8662$ component occurs almost exclusively in accreting objects'; ][]{Mohanty05.1},
781: and for 2M\,1207, in particular, by the lack of measureable effects from chromospheric emission
782: in other lines.
783: 
784: According to \cite{Koenigl91.1} the mass accretion rate of TTS 
785: is related to their magnetic field strength,
786: $B \sim \dot{M}^{1/2} \cdot M_*^{1/4} \cdot R_*^{-3} \cdot R_{\rm t}^{7/4}$. 
787: We assume $R_{\rm t} \sim 2 R_*$ for the disk truncation radius \citep{Muzerolle00.2}. The mass and
788: radius of 2M\,1207 are $M_* = 0.03\,M_\odot$ and $R_* = 0.27 R_\odot$ \citep{Chabrier00.2}. 
789: Taking account of the scaling factors, 
790: the average accretion rate of 2M\,1207 observed in May 2006 ($\dot{M} \sim 10^{-10}\,{\rm M_\odot/yr}$)   
791: yields an approximate value for the surface field, $B \approx 200$\,G. 
792: The same equation yields for a TTS with $0.8\,{\rm M_\odot}$, $1.5\,{\rm R_\odot}$, and
793: $\dot{M} \sim 10^{-8}\,{\rm M_\odot/yr}$ a surface field of $\approx 600$\,G. 
794: This validates the remark made by \cite{Scholz06.1} 
795: %had argued that, based on the same relation, 
796: who, without citing numbers for the field strength, had argued that 
797: the magnetic field for a typical brown dwarf is expected to be roughly half that of a typical TTS. 
798: %They did not cite any value for the field strength. 
799: %Their estimate is compatible
800: %with our result, despite the uncertainties from the assumptions made for the ... in the scaling factor. 
801: The magnetic field of TTS obtained this way is somewhat
802: lower than measured values, which are typically in the kilogauss range 
803: \citep[e.g.][ and references herein]{Yang05.1}. 
804: There are, however, considerable uncertainties connected with the use of the Koenigl-relation, 
805: such as the simplified assumption of a dipolar field and the location of the inner disk 
806: radius with respect to the corotation radius that determines the scaling factor. 
807: Given these ambiguities, % in the model definitions, 
808: the order of magnitude agreement between the estimated and observed fields is plausible. 
809: % should be considered no more than an at face value. 
810: 
811: 
812: 
813: 
814: 
815: 
816: \subsection{Accretion versus activity in TWA brown dwarfs}\label{subsect:acc_act}
817: 
818: Given the presence of a magnetic field, inferred from the accretion rate, the absence of 
819: strong magnetic activity on 2M\,1207 is remarkable. A very stringent upper limit to its 
820: X-ray emission was placed by a $50$\,ksec {\em Chandra} observation \citep{Gizis04.1}. The 
821: radio emission was constrained to the lowest value among the brown dwarfs in the 
822: TWA \citep{Osten06.3}. In Table~\ref{tab:bds} we summarize the accretion and activity measures 
823: for all four TWA brown dwarfs. 
824: 
825: The data in Table~\ref{tab:bds} has been collected from the literature except for the X-ray
826: data for SSSPM\,1102. This object was identified as a probable substellar TWA member in a 
827: recent proper motion study of the Supercosmos Sky Survey data \citep{Scholz05.5}. SSSPM\,1102 is in
828: the field-of-view of an archived {\em XMM-Newton} observation (Obs-ID\,0112880201). We ran the
829: standard Science Analysis System source detection process, and find no X-ray source at the 
830: optical position of SSSPM\,1102. To estimate an upper limit to its X-ray luminosity, 
831: we measured the counts at the expected position of SSSPM\,1102
832: and in an annulus around this position (representing the background). After applying the appropriate area 
833: scaling factor and extracting the on-source time from the exposure map, 
834: an upper limit to the net source count rate was estimated using the algorithm of \cite{Kraft91.1}. 
835: The distance to SSSPM\,1102 has been given by \cite{Mamajek05.1} ($d \sim 43$\,pc). This yields
836: $L_{\rm x} < 5.3 \cdot 10^{26}$\,erg/s in the $0.5-10$\,keV band. 
837: 
838: %The comparison with TWA-5\,B, the
839: It is constructive to compare 2M\,1207 to TWA-5\,B, the 
840: best-studied substellar object in the TWA after 2M\,1207. 
841: %shows the 
842: Matching their properties shows the dichotomy of one brown dwarf that accretes but is not 
843: magnetically active, and another brown dwarf that is magnetically active but does not accrete. 
844: Presumably, some mechanism suppresses activity in accreting substellar objects, as seems to be
845: the case for classical TTS \citep{Preibisch05.1,Telleschi06.1}. Possible explanations that have been
846: put forth in the context of classical TTS are changes in the magnetic field structure and/or heating process 
847: of the corona as a result of accretion. 
848: The absence of magnetic activity on 2M\,1207 is further surprising, as accreting substellar 
849: objects with similar magnetic field strength, 
850: e.g. V410\,Anon\,13, have been shown to be X-ray emitters 
851: \citep{Muzerolle00.2, Guedel06.1}.
852: For the substellar regime, the relation between accretion and X-ray luminosity has not yet been established 
853: on a statistical sample. If such a relation, 
854: as suggested by 2M\,1207 and TWA-5\,B, is confirmed, this would provide one more piece of evidence 
855: for the TTS-like character of brown dwarfs. 
856: X-ray and radio studies of the remaining two TWA brown dwarfs, both of which are non-accreting, should be useful.
857: 
858: 
859: \section{Conclusions}\label{sect:conclusions}
860: 
861: A substantial number of high-resolution spectroscopic observations has been collected for the brown 
862: dwarf 2M\,1207. These data have allowed us to study its accretion variability on
863: timescales from hours to $\sim 3$\,yrs.  
864: While small variations are obviously present throughout the rotation ($\sim 1$\,d), major 
865: changes in the structure of accretion may take place over months or years. 
866: A summary of all H$\alpha$ width measurements suggests a trend towards increasing accretion rate 
867: during the last three years. However, it is not clear if this trend reflects 
868: erratic changes or a systematic evolution. 
869: All in all, our dedicated monitoring has shown that 
870: the accretion pattern on 2M\,1207 is probably more complex than expected.
871: It comprises, longterm changes of the accretion rate and (maybe independently) of the flow structure
872: (months to years), variations on the timescale of the rotation rate (hours to days), and possibly
873: on even shorter timescales as suggested by our high temporal resolution data. 
874: 
875: 2M\,1107 is among the lowest mass dwarfs with measured mass accretion rate. The values derived from
876: the various observations obtained in a time-span of $\sim 3$\,yrs, differ by nearly $2$\,dex, similar
877: to (and possibly explaining) the scatter observed in the relation between accretion rate and stellar mass. 
878: Whether accretion variability is decisive for shaping the $\dot{M}_{\rm acc} - M$ relation will be established in
879: future systematic spectroscopic studies across a wide mass range. 
880: 
881: %% TABLE COLLECTION
882: 
883: %\begin{table}
884: %\begin{center}
885: %\caption{Spectral regions used for extraction of line and continuum counts.\label{tab:lines}}
886: %\begin{tabular}{lrrr}\hline
887: %Ident. & \multicolumn{1}{c}{left cont.} & \multicolumn{1}{c}{right cont.} & \multicolumn{1}{c}{line}  \\ 
888: %               & \multicolumn{1}{c}{[\AA]}          & \multicolumn{1}{c}{[\AA]}           & \multicolumn{1}{c}{[\AA]} \\ \hline
889: %H$\alpha$      & $6530 - 6545$  &  $6580 - 6595$  &  $6555 - 6570.8$ \\
890: %H$\beta$       & $4849 - 4855$  &  $4868 - 4873$  &  $4857 - 4867$ \\
891: %H$\gamma$      & $4332 - 4336$  &  $4346 - 4350$  &  $4338 - 4344$ \\
892: %\hline
893: %\end{tabular}
894: %\end{center}
895: %\end{table}
896: 
897: \begin{table}
898: \begin{center}
899: \caption{Summary of H$\alpha$ width measurements for 2M\,1207.\label{tab:w10}}
900: \begin{tabular}{lrl} \hline
901: Date                  & $W_{\rm 10\,\%}$ & Reference \\ 
902:                       & [km/s]       &      \\ \hline
903: Apr 2002              & $\sim 300^\dagger$  & \protect\cite{Gizis02.2} \\
904: May 8, 2003           & $170...200$  & \protect\cite{Mohanty03.2} \\
905: Jan 29 - Feb 1, 2005  & $209...215$  & \protect\cite{Scholz05.4} \\ 
906: Mar 17 - 19, 2005     & $253...308$  & \protect\cite{Scholz05.4} \\
907: Mar 27 - 30, 2005     & $279...304$  & \protect\cite{Scholz05.4} \\
908: Feb 21 - 22, 2006     & $261...281$  & present work \\
909: Apr 11, 2006          & $253$        & present work \\
910: May 8 - 9, 2006       & $281...322$  & present work \\
911: May 16, 2006          & $320$        & \protect\cite{Whelan07.1} \\
912: %\begin{tabular}{lrrl} \hline
913: %Date                  & $W_{\rm 10\,\%}$ & $W_{\rm EQ}$ & Reference \\ 
914: %                      & [km/s]       & [\AA]       &      \\ \hline
915: %Apr 2002$^\dagger$    &              & $\sim 300$  & Gizis (2002) \\
916: %May 8, 2003           & $170...200$  & $27.7$      & Mohanty et al. (2003) \\
917: %Jan 29 - Feb 1, 2005  & $209...215$  & ? & Scholz et al. (2005) \\ 
918: %Mar 17 - 19, 2005     & $253...308$  & ? & Scholz et al. (2005) \\
919: %Mar 27 - 30, 2005     & $279...304$  & ? & Scholz et al. (2005) \\
920: %Feb 21 - 22, 2006     & $261...281$  & ? & present work \\
921: %Apr 11, 2006          & $253$        & ? & present work \\
922: %May 8 - 9, 2006       & $281...322$  & $140...390$ & present work \\
923: %May 16, 2006          & $320$        & ? & Whelan, priv.comm. \\
924: \hline
925: \multicolumn{3}{c}{$^\dagger$ low-resolution data, no $10$\,\% width measured.} \\
926: \end{tabular}
927: \end{center}
928: \end{table}
929: 
930: \begin{table}
931: \begin{center}
932: \caption{Equivalent widths and FWHM for detected emission lines.\label{tab:eqw}}
933: \begin{tabular}{lrrr}\hline
934: Identification & $W_{\rm EQ}^{(a)}$              & typ.error$^{(a,b)}$         & FWHM$^{(c)}$ \\ 
935:                & \multicolumn{1}{c}{[\AA]} & \multicolumn{1}{c}{[\AA]}   & \multicolumn{1}{c}{[km/s]} \\ \hline
936: %\multicolumn{3}{c}{Original binning} \\
937: H$\beta$              & $4.5...47.1$  & $\sim 6$    & $180$ \\
938: H$\gamma$             & $0.3...13.5$  & $\sim 2$    & $180$ \\
939: %H$\delta$             & $0...10.0$  & $\sim 1.5$  & $168$ \\
940: %H$\epsilon$+Ca\,II\,H & $0...10.7$    & $\sim 1.5$  & $144$ \\
941: %Ca\,II\,K             & $0...3.1$     & $\sim 1$    & $18$ \\
942: H$\delta$             &    ...       & ...  & $168$ \\
943: H$\epsilon$+Ca\,II\,H &    ...       & ...  & $144$ \\
944: Ca\,II\,K             &    ...       & ...  & $18$ \\
945: He\,I\,$\lambda5876$  & $1.8 ... 9.1$ & $\sim 2$    & $30$ \\
946: H$\alpha$             & $145 ... 394$ & $\sim 45$   & $210$ \\
947: He\,I\,$\lambda6678$  & $0.9 ... 2.4$ & $\sim 2$    & $18$ \\
948: Ca\,II\,$\lambda8498$ & $0 ... 0.6$   & $\sim 0.35$ & $12$ \\
949: Ca\,II\,$\lambda8542$ & $0 ... 0.3$   & $\sim 0.25$ & $12$ \\
950: Ca\,II\,$\lambda8662$ & $0 ... 0.4$   & $\sim 0.3$  & $12$ \\ 
951: %${\rm [O\,I]}\,\lambda6300$ & $1.3 ... 5.7$  & $\sim 1.5$ & $$ \\ 
952: %${\rm [O\,I]}\,\lambda6363$ & $0.1 ... 1.6$  & $\sim 0.5$ & $$ \\ 
953: \hline
954: \multicolumn{4}{l}{$^{(a)}$ No equivalent widths are given for the higher Balmer lines} \\
955: \multicolumn{4}{l}{and for Ca\,II\,H+K because of the low continuum} \\
956: \multicolumn{4}{l}{$^{(b)}$ Errors for the equivalent width are estimated} \\ 
957: \multicolumn{4}{l}{ by assuming a $15$\,\% uncertainty in the} \\
958: \multicolumn{4}{l}{ continuum flux.} \\
959: \multicolumn{4}{l}{$^{(c)}$ measured in an average spectrum that} \\
960: \multicolumn{4}{l}{represents the typical line profile for 2M\,1207} \\
961: \multicolumn{4}{l}{during the observation.} \\
962: % FOR THE FWHM SEE FILES: eqw_l_'linefeature'_skyauto6_rebin1.dat
963: \end{tabular}
964: \end{center}
965: \end{table}
966: 
967: 
968: \begin{table}
969: \begin{center}
970: \caption{Activity and accretion measures of all known brown dwarfs in TWA.\label{tab:bds}}
971: \begin{tabular}{lrrrrrc} \hline
972: Designation & IR excess & $W_{\rm H\alpha;10\,\%}$ & $L_{\rm x}$    & $L_{\rm R}$ & FUV & Refs \\ 
973:             &           & [km/s]                   & [erg/s]        & [$10^{14}$\,erg/s/Hz]  &     & \\ \hline
974: TWA\,5B     &           & $162$                    & $4\,10^{27}$    & $< 2.0$  &         & (-,1,2,3,-) \\ 
975: 2M\,1207    & $\surd$   & $170...320$              & $<1.2\,10^{26}$ & $< 0.98$ & $\surd$ & (4,5,6,3,7) \\
976: 2M\,1139    & $-$       & $111$                    & ?               & ?        &         & (4,1,-,8,-) \\
977: SSSPM\,1102 &           & $194$                    & $<5.3\,10^{26}$ & $< 1.6$  &         & (-,1,9,3,-) \\
978: \hline
979: \multicolumn{7}{l}{(1) - \cite{Mohanty03.2}, (2) - \cite{Tsuboi03.1}, (3) - \cite{Osten06.3},} \\
980: \multicolumn{7}{l}{(4) - \cite{Riaz06.1}, (5) - see references in this paper, (6) - \cite{Gizis04.1},} \\
981: \multicolumn{7}{l}{(7) - \cite{Gizis05.1}, (8) - \cite{Burgasser05.1}, (9) - this paper.} \\
982: \end{tabular}
983: \end{center}
984: \end{table}
985: 
986: 
987: 
988: 
989: %% FIGURE COLLECTION
990: 
991: %
992: % OUTPUT FROM    plot_timeseries_line.pro
993: %
994: \begin{figure}
995: \begin{center}
996: \parbox{15cm}{
997: \parbox{5cm}{\includegraphics[width=5cm, angle=0]{./f1a.eps}}
998: \parbox{10cm}{\includegraphics[width=10cm, angle=0]{./f1b.eps}}
999: }
1000: \caption{Temporal evolution of parameters characterizing the H$\alpha$ emission of 2M\,1207 during May 8-9, 2006.
1001: \label{fig:ha_timeseries}}
1002: \end{center}
1003: \end{figure}
1004: 
1005: %
1006: % OUTPUT FROM    plot_all_spec.pro   AND   calc_varprofile.pro
1007: %
1008: \begin{figure}
1009: \begin{center}
1010: \parbox{17cm}{
1011: \parbox{5.5cm}{\includegraphics[width=5.5cm, angle=0]{./f2a.eps}}
1012: \parbox{5.5cm}{\includegraphics[width=5.5cm, angle=0]{./f2b.eps}}
1013: \parbox{5.5cm}{\includegraphics[width=5.5cm, angle=0]{./f2c.eps}}
1014: }
1015: \parbox{17cm}{
1016: \parbox{5.5cm}{
1017: \includegraphics[width=5.5cm, angle=0]{./f2d.eps}
1018: }
1019: \parbox{5.5cm}{
1020: \includegraphics[width=5.5cm, angle=0]{./f2e.eps}
1021: }
1022: \parbox{5.5cm}{
1023: \includegraphics[width=5.5cm, angle=0]{./f2f.eps}
1024: }
1025: }
1026: \caption{{\em top} -- Normalized average H$\alpha$ line profiles of 2M\,1207 obtained with UVES in May 2006; {\em bottom} - Normalized variance line profiles and zero-variability level.\label{fig:Ha_profiles}}
1027: \end{center}
1028: \end{figure}
1029: 
1030: %
1031: % OUTPUT FROM    plot_mike_spec.pro 
1032: %
1033: \begin{figure}
1034: \begin{center}
1035: \includegraphics[width=5.5cm, angle=0]{./f3.eps}
1036: \caption{H$\alpha$ line profiles of 2M\,1207 obtained with MIKE in Feb and Apr 2006. The data were normalized to the average continuum level near the line. \label{fig:Ha_profiles_mike}}
1037: \end{center}
1038: \end{figure}
1039: 
1040: %
1041: % OUTPUT FROM    plot_correlations.pro
1042: %
1043: \begin{figure}
1044: \begin{center}
1045: \includegraphics[width=7cm, angle=0]{./f4a.eps}
1046: \includegraphics[width=7cm, angle=0]{./f4b.eps}
1047: \caption{Equivalent widths of Gaussian fits to the emission ($W_{\rm EQ; em}$) and absorption ($W_{\rm EQ; abs}$) 
1048: component of the H$\alpha$ line, and normalized total observed flux 
1049: ($F_{\rm obs} = F_{\rm em} - F_{\rm abs}$) versus H$\alpha$ $10$\,\% width.\label{fig:Ha_flux}}
1050: \end{center}
1051: \end{figure}
1052: 
1053: %
1054: % OUTPUT FROM    calc_timelags.pro
1055: %               
1056: \begin{figure}
1057: \begin{center}
1058: \includegraphics[width=15cm, angle=0]{f5.eps}
1059: \caption{Mean of the standard deviation of fractional profile changes $D(\lambda)$ in the H$\alpha$ line as a function of time lag. Numbers on top of the panel indicate the number of data points in each of the time lag bins.\label{fig:ds}}
1060: \end{center}
1061: \end{figure}
1062: 
1063: %
1064: % OUTPUT FROM   calc_timelags.pro
1065: %
1066: \begin{figure}
1067: \begin{center}
1068: \parbox{17cm}{
1069: \parbox{5.5cm}{
1070: \includegraphics[width=5.cm, angle=0]{./f6a.eps}
1071: }
1072: \parbox{5.5cm}{
1073: \includegraphics[width=5.cm, angle=0]{./f6b.eps}
1074: }
1075: \parbox{5.5cm}{
1076: \includegraphics[width=5.cm, angle=0]{./f6c.eps}
1077: }
1078: }
1079: \caption{Correlation matrix of H$\alpha$ for different subset of exposures (denoted $i$) obtained on May 8-9, 2006.
1080: {\em left} -- $i=1,...,19$; {\em middle} -- $i=20,...38$; {\em right} -- $i=7,...,10$.}
1081: %: contours of r=minlevel,0.5,0.65,0.75,0.85,0.95,0.99,0.99. From top left to bottom right: n1, n2a, n2b, all.}
1082: \label{fig:corr_matrix_Ha}
1083: \end{center}
1084: \end{figure}
1085: 
1086: %
1087: % OUTPUT FROM    plot_balmer_spec.pro
1088: %               
1089: \begin{figure}
1090: \begin{center}
1091: \parbox{17cm}{
1092: \parbox{8cm}{
1093: \includegraphics[width=7cm, angle=0]{./f7a.eps}
1094: }
1095: \parbox{8cm}{
1096: \includegraphics[width=7cm, angle=0]{./f7b.eps}
1097: }
1098: }
1099: \caption{Contemporaneous Balmer emission line profiles (H$\alpha$ to H$\epsilon$) for two epochs during the UVES run (8-frame averages). \label{fig:balmer}}
1100: \end{center}
1101: \end{figure}
1102: 
1103: %
1104: % OUTPUT FROM    plot_HeID3.pro
1105: %               
1106: \begin{figure}
1107: \begin{center}
1108: \includegraphics[width=8cm, angle=0]{./f8.eps}
1109: \caption{Two spectra in the region of He\,I\,$\lambda5876$ showing the change from the typical profile with extended red wing (solid line) to the weak, symmetric line (dotted line).\label{fig:heI}}
1110: \end{center}
1111: \end{figure}
1112: 
1113: %
1114: % OUTPUT FROM    plot_m_mdot.pro
1115: %               
1116: \begin{figure}
1117: \begin{center}
1118: \includegraphics[width=15cm, angle=0]{./f9.eps}
1119: \caption{Relation between mass and mass accretion rate for sub-stellar objects; data from the literature cited by \protect\cite{Muzerolle05.1}\label{fig:mdot}}
1120: \end{center}
1121: \end{figure}
1122: 
1123: 
1124: 
1125: 
1126: %% If you wish to include an acknowledgments section in your paper,
1127: %% separate it off from the body of the text using the \acknowledgments
1128: %% command.
1129: 
1130: %% Included in this acknowledgments section are examples of the
1131: %% AASTeX hypertext markup commands. Use \url without the optional [HREF]
1132: %% argument when you want to print the url directly in the text. Otherwise,
1133: %% use either \url or \anchor, with the HREF as the first argument and the
1134: %% text to be printed in the second.
1135: 
1136: \acknowledgments
1137: 
1138: This paper is based on data collected under the ESO program 077.C-0323. 
1139: BS acknowledges financial support from ASI contract ASI-INAF I/023/05/0.
1140: BS wishes to thank S. Randich and S. Mohanty for useful discussions on the UVES 
1141: pipeline analysis and brown dwarf emission line spectra. RJ and AS acknowledge 
1142: support from an NSERC grant to RJ. 
1143: 
1144: %% To help institutions obtain information on the effectiveness of their
1145: %% telescopes, the AAS Journals has created a group of keywords for telescope
1146: %% facilities. A common set of keywords will make these types of searches
1147: %% significantly easier and more accurate. In addition, they will also be
1148: %% useful in linking papers together which utilize the same telescopes
1149: %% within the framework of the National Virtual Observatory.
1150: %% See the AASTeX Web site at http://www.journals.uchicago.edu/AAS/AASTeX
1151: %% for information on obtaining the facility keywords.
1152: 
1153: %% After the acknowledgments section, use the following syntax and the
1154: %% \facility{} macro to list the keywords of facilities used in the research
1155: %% for the paper.  Each keyword will be checked against the master list during
1156: %% copy editing.  Individual instruments or configurations can be provided 
1157: %% in parentheses, after the keyword, but they will not be verified.
1158: 
1159: {\it Facility:} \facility{VLT:Kueyen (UVES)}
1160: 
1161: 
1162: %% The reference list follows the main body and any appendices.
1163: %% Use LaTeX's thebibliography environment to mark up your reference list.
1164: %% Note \begin{thebibliography} is followed by an empty set of
1165: %% curly braces.  If you forget this, LaTeX will generate the error
1166: %% "Perhaps a missing \item?".
1167: %%
1168: %% thebibliography produces citations in the text using \bibitem-\cite
1169: %% cross-referencing. Each reference is preceded by a
1170: %% \bibitem command that defines in curly braces the KEY that corresponds
1171: %% to the KEY in the \cite commands (see the first section above).
1172: %% Make sure that you provide a unique KEY for every \bibitem or else the
1173: %% paper will not LaTeX. The square brackets should contain
1174: %% the citation text that LaTeX will insert in
1175: %% place of the \cite commands.
1176: 
1177: %% We have used macros to produce journal name abbreviations.
1178: %% AASTeX provides a number of these for the more frequently-cited journals.
1179: %% See the Author Guide for a list of them.
1180: 
1181: %% Note that the style of the \bibitem labels (in []) is slightly
1182: %% different from previous examples.  The natbib system solves a host
1183: %% of citation expression problems, but it is necessary to clearly
1184: %% delimit the year from the author name used in the citation.
1185: %% See the natbib documentation for more details and options.
1186: 
1187: \bibliographystyle{apj}
1188: \bibliography{apj-jour,uves}
1189:  
1190: 
1191: %% Use the figure environment and \plotone or \plottwo to include
1192: %% figures and captions in your electronic submission.
1193: %% To embed the sample graphics in
1194: %% the file, uncomment the \plotone, \plottwo, and
1195: %% \includegraphics commands
1196: %%
1197: %% If you need a layout that cannot be achieved with \plotone or
1198: %% \plottwo, you can invoke the graphicx package directly with the
1199: %% \includegraphics command or use \plotfiddle. For more information,
1200: %% please see the tutorial on "Using Electronic Art with AASTeX" in the
1201: %% documentation section at the AASTeX Web site,
1202: %% http://www.journals.uchicago.edu/AAS/AASTeX.
1203: %%
1204: %% The examples below also include sample markup for submission of
1205: %% supplemental electronic materials. As always, be sure to check
1206: %% the instructions to authors for the journal you are submitting to
1207: %% for specific submissions guidelines as they vary from
1208: %% journal to journal.
1209: 
1210: %% This example uses \plotone to include an EPS file scaled to
1211: %% 80% of its natural size with \epsscale. Its caption
1212: %% has been written to indicate that additional figure parts will be
1213: %% available in the electronic journal.
1214: 
1215: %\begin{figure}
1216: %\epsscale{.80}
1217: %%\plotone{f1.eps}
1218: %\caption{Derived spectra for 3C138 \citep[see][]{heiles03}. Plots for all sources are available
1219: %in the electronic edition of {\it The Astrophysical Journal}.\label{fig1}}
1220: %\end{figure}
1221: 
1222: \clearpage
1223: 
1224: %% Here we use \plottwo to present two versions of the same figure,
1225: %% one in black and white for print the other in RGB color
1226: %% for online presentation. Note that the caption indicates
1227: %% that a color version of the figure will be available online.
1228: %%
1229: 
1230: %\begin{figure}
1231: %%\plottwo{f2.eps}{f2_color.eps}
1232: %\caption{A panel taken from Figure 2 of \citet{rudnick03}. 
1233: %See the electronic edition of the Journal for a color version 
1234: %of this figure.\label{fig2}}
1235: %\end{figure}
1236: 
1237: %% This figure uses \includegraphics to scale and rotate the still frame
1238: %% for an mpeg animation.
1239: 
1240: %\begin{figure}
1241: %%\includegraphics[angle=90,scale=.50]{f3.eps}
1242: %\caption{Animation still frame taken from \citet{kim03}.
1243: %This figure is also available as an mpeg
1244: %animation in the electronic edition of the
1245: %{\it Astrophysical Journal}.}
1246: %\end{figure}
1247: 
1248: %% If you are not including electonic art with your submission, you may
1249: %% mark up your captions using the \figcaption command. See the
1250: %% User Guide for details.
1251: %%
1252: %% No more than seven \figcaption commands are allowed per page,
1253: %% so if you have more than seven captions, insert a \clearpage
1254: %% after every seventh one.
1255: 
1256: %% Tables should be submitted one per page, so put a \clearpage before
1257: %% each one.
1258: 
1259: %% Two options are available to the author for producing tables:  the
1260: %% deluxetable environment provided by the AASTeX package or the LaTeX
1261: %% table environment.  Use of deluxetable is preferred.
1262: %%
1263: 
1264: %% Three table samples follow, two marked up in the deluxetable environment,
1265: %% one marked up as a LaTeX table.
1266: 
1267: %% In this first example, note that the \tabletypesize{}
1268: %% command has been used to reduce the font size of the table.
1269: %% We also use the \rotate command to rotate the table to
1270: %% landscape orientation since it is very wide even at the
1271: %% reduced font size.
1272: %%
1273: %% Note also that the \label command needs to be placed
1274: %% inside the \tablecaption.
1275: 
1276: %% This table also includes a table comment indicating that the full
1277: %% version will be available in machine-readable format in the electronic
1278: %% edition.
1279: 
1280: %% If you use the table environment, please indicate horizontal rules using
1281: %% \tableline, not \hline.
1282: %% Do not put multiple tabular environments within a single table.
1283: %% The optional \label should appear inside the \caption command.
1284: 
1285: %% If the table is more than one page long, the width of the table can vary
1286: %% from page to page when the default \tablewidth is used, as below.  The
1287: %% individual table widths for each page will be written to the log file; a
1288: %% maximum tablewidth for the table can be computed from these values.
1289: %% The \tablewidth argument can then be reset and the file reprocessed, so
1290: %% that the table is of uniform width throughout. Try getting the widths
1291: %% from the log file and changing the \tablewidth parameter to see how
1292: %% adjusting this value affects table formatting.
1293: 
1294: 
1295: %% Tables may also be prepared as separate files. See the accompanying
1296: %% sample file table.tex for an example of an external table file.
1297: %% To include an external file in your main document, use the \input
1298: %% command. Uncomment the line below to include table.tex in this
1299: %% sample file. (Note that you will need to comment out the \documentclass,
1300: %% \begin{document}, and \end{document} commands from table.tex if you want
1301: %% to include it in this document.)
1302: 
1303: %% \input{table}
1304: 
1305: %% The following command ends your manuscript. LaTeX will ignore any text
1306: %% that appears after it.
1307: 
1308: \end{document}
1309: 
1310: %%
1311: %% End of file `sample.tex'.
1312: