1:
2:
3: \documentclass{emulateapj}
4: \slugcomment{{\sc Accepted to ApJ:} July 15, 2007}
5:
6:
7:
8: \renewcommand{\topfraction}{0.3} \renewcommand{\bottomfraction}{0.99} \setcounter{topnumber}{1}
9: \setcounter{bottomnumber}{2}
10: \setcounter{totalnumber}{4} \setcounter{dbltopnumber}{2} \renewcommand{\dbltopfraction}{0.9} \renewcommand{\textfraction}{0.07} \renewcommand{\floatpagefraction}{0.7} \renewcommand{\dblfloatpagefraction}{0.7}
11:
12:
13: \newcommand\um{\ensuremath{\mathrm{\ \mu{m}}}}
14: \newcommand\nm{\ensuremath{\mathrm{\ nm}}}
15: \newcommand\mm{\ensuremath{\mathrm{\ mm}}}
16: \newcommand\cm{\ensuremath{\mathrm{\ cm}}}
17: \newcommand\pc{\ensuremath{\mathrm{\ pc}}}
18: \newcommand\yr{\ensuremath{\mathrm{\ year}}}
19: \newcommand{\LkHa}{LkH$\alpha$ }
20: \newcommand\Exp[1]{\ensuremath{\times10^{#1}}}
21: \newcommand\Msun{\ensuremath{\mathrm{M_\odot}}}
22: \newcommand\Rsun{\ensuremath{\mathrm{R_\odot}}}
23: \newcommand\Lsun{\ensuremath{\mathrm{L_\odot}}}
24:
25: \newcommand\ang{\AA}
26:
27: \newcommand{\Ha}{H$\alpha$}
28: \newcommand{\SII}{[\ion{S}{2}] }
29: \newcommand{\FeII}{[\ion{Fe}{2}] }
30: \newcommand{\OI}{[\ion{O}{1}] }
31:
32:
33:
34:
35:
36:
37: \shorttitle{Bipolar Jet from \LkHa 233}
38: \shortauthors{Perrin and Graham}
39:
40: \begin{document}
41:
42: \title{Laser Guide Star Adaptive Optics Integral Field Spectroscopy\\ of a Tightly Collimated Bipolar Jet\\
43: from the Herbig Ae star \LkHa 233\altaffilmark{1}}
44:
45: \author{Marshall D. Perrin and James R. Graham}
46: \affil{Astronomy Department, University of California,
47: Berkeley, CA 94720-3411}
48: \email{mperrin@astro.berkeley.edu}
49:
50: \altaffiltext{1}{Some of the data
51: presented herein were obtained at the W.M. Keck Observatory, which is
52: operated as a scientific partnership among the California Institute of
53: Technology, the University of California and the National Aeronautics
54: and Space Administration. The Observatory was made possible by the
55: generous financial support of the W.M. Keck Foundation.}
56:
57:
58: \begin{abstract}
59:
60: We have used the integral field spectrograph OSIRIS and laser guide
61: star adaptive optics at Keck Observatory to obtain high angular
62: resolution ($0\farcs06$), moderate spectral resolution ($R\simeq
63: 3800$) images of the bipolar jet from the Herbig Ae star \LkHa 233,
64: seen in near-IR \FeII emission at 1.600 \& 1.644 \micron. This jet is
65: narrow and tightly collimated, with an opening angle of only 9
66: degrees, and has an average radial velocity of $\sim 100$ km s$^{-1}$.
67: The jet and counterjet are asymmetric, with the red-shifted jet much
68: clumpier than its counterpart at the angular resolution of our
69: observations. The observed properties are in general similar to
70: jets seen around T Tauri stars, though it has a relatively large mass
71: flux of $1.2 \pm 0.3\Exp{-7} \Msun$ year$^{-1}$, near the high end of
72: the observed mass flux range around T Tauri stars. We also spatially
73: resolve an inclined circumstellar disk around \LkHa 233, which
74: obscures the star from direct view. By comparison with numerical radiative
75: transfer disk models, we
76: estimate the disk midplane to be inclined $i= 65 \pm 5\degr$ relative to the plane of the sky. Since
77: the star is seen only in scattered light at near-infrared wavelengths,
78: we detect only a small fraction of its intrinsic flux. Because
79: previous estimates of its stellar properties did not account for this,
80: either LkHa 233 must be located closer than the previously believed,
81: or its true luminosity must be greater than previously supposed,
82: consistent with its being a $\sim4 \Msun$ star near the stellar
83: birthline.
84:
85:
86: \end{abstract}
87: \keywords{ ISM: jets and outflows --- ISM: Herbig-Haro objects --- stars: individual (LkHa 233) --- stars:pre-main-sequence }
88:
89:
90: \section{Introduction}
91:
92:
93: \begin{figure*}[t!]
94: \begin{center}
95: \includegraphics[width=6in]{f1.eps}
96: \end{center}
97: \caption{
98: \label{diskjetfig}
99: \textbf{Left:} This $J,H,K_s$
100: composite from the OSIRIS imager shows the X-shaped
101: envelope surrounding \LkHa 233. The dashed rectangle shows the field
102: zoomed
103: in to on the right, while the solid line indicates the outflow axis.
104: \textbf{Right:} A portion of the
105: OSIRIS IFU data cube, centered in wavelength on the \FeII 1.644 $\micron$
106: line (in
107: green) and rotated to show the jet axis vertical. This image reveals a tightly collimated bipolar outflow from
108: \LkHa 233 perpendicular to an edge-on circumstellar disk. The jet is
109: undetectable in the broadband image; only with the spectral
110: resolution of OSIRIS can it be detected. In the right hand panel, the
111: blue-shifted jet points down and the red-shifted jet is upward.
112: In these data we also see scattered light from a roughly edge-on
113: circumstellar disk and a larger bipolar nebula extending
114: perpendicular to the disk plane. The jet has excavated a cavity in
115: this envelope, the limb-brightened edges of which form the well-known
116: "X" centered on \LkHa 233 \citep{aspin85,Perrin2004Sci}.
117: }\end{figure*}
118:
119:
120:
121: Bipolar outflows from young stars feature prominently in many
122: of the most spectacular images of our universe
123: \citep[see,
124: e.g.][]{2007prpl.conf..215B}. But far from being merely aesthetically
125: pleasing, these jets play crucial roles in star formation. Molecular
126: cloud material must shed most of its angular momentum before it can
127: accrete onto a newborn star, and jets have been identified as a key
128: mechanism for this, removing angular momentum from disks and allowing
129: accretion to continue \citep[][ and references
130: therein]{2007prpl.conf..231R}. Recent observations have for the first
131: time provided indications of jet rotation
132: \citep{2002ApJ...576..222B,2004ApJ...604..758C,2005A&A...432..149W,2007astro.ph..3271C},
133: suggesting that outflows may indeed carry angular momentum away
134: from their origin. Jets have also been implicated in regulating the
135: overall efficiency of star formation by injecting turbulence into
136: molecular clouds \citep{2000ApJ...545..364M}.
137:
138: The physical processes which drive these outflows remain poorly
139: understood.
140: Several competing theories have been proposed to explain how gas is accelerated
141: and collimated. The leading contenders all
142: invoke magnetic forces to shape the outflows, with the ``X-wind''
143: model \citep{Shang2007PPV} positing acceleration occurs near the star at the radius
144: of magnetospherical truncation, while the ``disk wind'' model
145: \citep{Pudritz2007PPV} posits acceleration from a wide range
146: of radii across the circumstellar disk. Alternatively, outflows could
147: originate in
148: spherical coronae, which are then collimated by surrounding magnetic
149: fields \citep{1999A&A...348..327S}. Outflows may even evolve through
150: each of these processes in turn \citep{2003Ap&SS.287...25S}.
151: Distinguishing between these theories will require detailed studies of
152: outflows as close to their origins as possible. At present, high
153: angular resolution observations of outflows have mostly focused on
154: older T Tauri stars approaching the main sequence
155: \citep[see][and references therein]{2007prpl.conf..231R}.
156: Observations of more sources at
157: younger ages are needed, especially considering that outflows are
158: intimately related to accretion, which is greatest at the
159: youngest ages \citep{1998ApJ...495..385H}.
160:
161:
162:
163:
164:
165:
166:
167: In addition to age, we also need to understand how outflows vary with
168: stellar mass. Because jets are thought to be driven by magnetic
169: fields, we might suppose that their properties should differ between
170: fully convective T Tauri stars and more massive stars which lack a
171: surface convection zone. Indeed, outflows from high mass YSOs are
172: much less collimated than jets from T Tauri stars, with opening angles
173: of $30-60\degr$ \citep[e.g.
174: ][]{1997ApJ...478..283H,1998ApJ...507..861S,2001Sci...292.1513S}. The
175: Herbig Ae/Be stars \citep{her60,1998ARA&A..36..233W} are intermediate
176: between these two regimes, so studies of their outflows can
177: potentially clarify this transition in jet properties, which in turn
178: may shed light on the underlying physics through which jets are
179: produced. Yet relatively few outflows from Herbig Ae/Be stars have
180: been studied in detail, particularly at high resolution
181: \citep{1994ASPC...62..237M}. In a very few cases, studies
182: of parsec-scale outflows from Herbig Ae stars have
183: suggested that on large
184: spatial scales their velocities and collimation are largely
185: consistent with those observed in T Tauri jets
186: \citep{2004A&A...415..189M}.
187:
188: \begin{figure*}[t]
189: \begin{center}
190: \includegraphics[width=6.5in]{f2.eps}
191: \end{center}
192: \caption
193: {
194: \label{fig1d}
195: One dimensional spectra extracted from the OSIRIS data cube.
196: \textbf{Left:} Square apertures indicate the regions extracted to produce the spectra at right,
197: shown in corresponding colors. The
198: red aperture is centered on knot B.
199: \textbf{Center and Right:} One dimensional spectra extracted from the
200: data cube for those apertures, in OSIRIS' $Hn3$ and $Kn3$ filters; the
201: light grey shading shows the statistical uncertainty. For clarity, the
202: emission line spectra have been scaled up $\times 10$ and $\times20$ relative to
203: the continuum. The outflow is clearly detected in both the
204: 1.600 and 1.644 \micron\ lines of \FeII, and excess emission in
205: Br$\gamma$ is seen at the
206: location of the star. There is a hint of 2.122 \micron\ H$_2$ emission, but it unfortunately
207: falls at the very edge of our bandpass so its reality is uncertain. The
208: four features marked ``X'' are spectral ghosts caused by a grating misalignment
209: during OSIRIS commissioning, which has since been fixed. The
210: observed stellar continuum is $\sim 1/4$ the 2MASS flux levels
211: for \LkHa 233, reflecting the fact that our aperture does not include the whole
212: extended source, possibly stellar variability and/or uncertainties
213: in flux calibration.
214: }
215: \end{figure*}
216:
217:
218:
219:
220: But a lack of high angular resolution studies of Herbig Ae star outflows
221: has thus far prevented detailed
222: comparison with outflows from T Tauri stars. For T Tauri stars, we now
223: have a sophisticated understanding of outflow properties on scales of a
224: few tens of AU, derived from high angular resolution observations.
225: Because stellar outflows are detectable
226: primarily in emission lines from shock-excited ions, line ratios
227: can provide measurements of many physical parameters within the
228: outflow. When both spatially and spectrally resolved data are
229: available, very detailed portraits of outflows have been obtained
230: \citep[e.g.][]{1999A&A...350..917B,2004AJ....127..408B,2007ApJ...660..426H}.
231: Since outflows have characteristic spatial scales of a few
232: AU (corresponding to $0\farcs01-0\farcs1$ at typical distances) and
233: velocities of tens to hundreds of km s$^{-1}$, the necessary
234: observations combine high angular resolution ($< 0\farcs1$)
235: with moderate spectral resolution ($R \sim 2000-4000$). Having both
236: spatially and spectrally resolved data is key to disentangling the
237: outflow structures as close to the disk as possible. In optical
238: wavelengths, such as for the \SII 6717, 6731 \ang\ and \OI 6300, 6364
239: \ang\ emission lines, the necessary angular resolution is available
240: only from space
241: \citep[e.g.,][]{1999ApJ...512..901H,2000ApJ...537L..49B,2002ApJ...580..336W,2004ApJ...604..758C,2007ApJ...660..426H}
242: However, the arrival of integral field spectroscopy with adaptive
243: optics on 8-10 m telescopes has now opened the door to high angular resolution
244: studies of jets from the ground. At the near-infrared wavelengths
245: suitable for adaptive optics, several emission lines of \FeII between
246: $1.2 - 1.6 \mu$m are ideal tracers of shocked outflows
247: \citep{1987ApJ...313..847G}. Furthermore, these lines provide
248: diagnostic potential enabling the determination of physical conditions
249: within the jet \citep{2003A&A...410..155P,2004ApJ...614L..69H}, and
250: ultimately the mass flux, a key parameter in any model of outflow
251: physics.
252:
253: In this paper, we seek to investigate whether the similarity on large
254: spatial scales between
255: outflows from T Tauri and Herbig Ae stars still
256: holds true on finer spatial scales. Toward that
257: end, we present the first high angular resolution integral field
258: spectroscopy of an outflow from a Herbig Ae star, enabling us to
259: investigate its collimation, kinematics, and certain physical
260: parameters on angular scales of $\sim 60$ milliarcseconds. By
261: comparison with the properties of jets seen from T Tauri stars, we
262: hope to clarify the influence of stellar mass upon outflow physics.
263: Also, from a technical perspective, we wish to demonstrate the utilty of
264: laser guide star adaptive optics (LGS AO) for
265: obtaining high angular resolution spectroscopy of an outflow from a
266: fairly faint and obscured source. Because many young stars which drive
267: jets are still deeply embedded in their dusty birthplaces, only a limited number
268: are accessible to traditional natural guide star AO. The
269: growing availability of LGS AO on large
270: telescopes promises to enable high angular resolution studies of
271: a greatly increased target sample, particularly at the youngest ages.
272:
273:
274:
275: We present here observations made with Keck's new integral field spectrograph
276: OSIRIS, using laser guide star adaptive optics, which show in
277: exquisite detail a tightly collimated jet from the Herbig Ae star
278: \LkHa 233 (also V375 Lac, HBC 313). This star, part of the original Ae
279: star sample identified by \citep{her60}, is
280: one of the relatively few Herbig Ae/Be stars known to possess a jet.
281: It is an A4 star, with
282: $\log T_{eff} = 3.93$ \citep{2004AJ....127.1682H}.
283: Its distance has generally been considered to be
284: 880 pc \citep{1978MNRAS.182..687C}.
285: We adopt this distance as well, but some authors have
286: argued for a closer distance; see \S \ref{distance} below.
287: Hernandez et al.\ also derived $M_* = 2.9~
288: M_\odot$, $L_* = 80~ L_\odot$, and age = 2.61 Myr, with $A_V$ = 3.7, $R_V =
289: 5$, but those values are likely not accurate, because
290: they were derived without accounting for the fact that
291: \LkHa 233 is seen only in scattered light. Hence both
292: the extinction to the star and its total luminosity must be higher; see \S
293: \ref{disksection}.
294:
295:
296: \LkHa 233 sits at the center of an
297: X-shaped bipolar nebula visible from optical to near-infrared
298: wavelengths \citep{aspin85,Li94}. Near infrared polarimetry shows a
299: dark lane passing in front of the star, indicating the presence of a roughly
300: edge-on circumstellar disk \citep{Perrin2004Sci}.
301: \citet{corc98} discovered that \LkHa 233 possesses an extended bipolar
302: jet and counterjet visible in the optical \SII 6716, 6731 \ang~lines. The
303: blueshifted jet extends toward position angle (PA) 250\degr, perpendicular to the inferred circumstellar disk
304: and along the symmetry axis of the
305: surrounding nebula.
306: The redshifted counterjet, at 70\degr, was not detected within 0\farcs6 of
307: the star, which Corcoran et al.~interpreted as occultation by
308: the circumstellar disk. A followup study by
309: \citet{2004A&A...415..189M}
310: found that a chain of Herbig Haro objects
311: extends almost 2 pc to either side of \LkHa 233, located
312: approximately along the jet axis but with considerable scatter.
313:
314:
315: In the following section, \S 2, we present our observations and
316: outline the data reduction process. We then describe in \S 3 the observed
317: properties of both the jet and the circumstellar disk, derive
318: estimates for the mass flux in the jet, and re-examine the question of
319: the distance to \LkHa 233. In \S 4 we discuss the implications of
320: these results, with special attention to comparison with the observed
321: properties of jet around other Herbig Ae and T Tauri stars.
322: Conclusions and prospects for future work are given in \S 5.
323:
324:
325:
326:
327:
328:
329:
330:
331: \section{Observations}
332:
333:
334: We observed \LkHa 233 on 2005 Oct 13 UTC using the integral field
335: spectrograph OSIRIS and laser guide star
336: adaptive optics system \citep{2006PASP..118..297W,2006PASP..118..310V} on the W. M. Keck II telescope.
337: The performance of the
338: \LkHa 233 itself served as the tip-tilt
339: reference for the adaptive optics system. The skies were clear and
340: seeing was average, $0\farcs6$ at $K'$, with winds from 15-20 km hr$^{-1}$.
341:
342: OSIRIS is a lenslet-based integral field spectrograph which provides a spectral resolution of
343: 3800 from 1-2.5 microns \citep{2003SPIE.4841.1600L}. We selected its
344: 50 mas pixel$^{-1}$ spectrographic plate scale for our
345: observations. While this scale does not Nyquist-sample the
346: diffraction-limited PSF, it provides a
347: larger field of view and better signal to noise per integration time
348: for extended emission
349: than the finer plate scales. We observed \LkHa 233 using the narrow-band \textit{Hn3}
350: and \textit{Kn3} filters, covering the wavelength ranges
351: 1.594-1.676 and 2.121-2.229 $\micron$, respectively. Total integration
352: time on-source was 1800 s for \textit{Hn3} and 900 s for \textit{Kn3}. Between exposures
353: on \LkHa 233, we obtained equally deep sky observations at a position 20"
354: away.
355:
356:
357: \begin{figure*}[t]
358: \begin{center}
359: \includegraphics[width=6in]{f3.eps}
360: \end{center}
361: \caption{
362: \label{slicesfig}
363: Continuum-subtracted velocity channel maps for \FeII
364: 1.644 \micron\ emission from the jet. Each panel shows the
365: intensity image for a given velocity (relative to the solar
366: system barycenter) noted at the bottom of each panel.
367: These are contiguous spectral channels from OSIRIS,
368: and the field of view shown is $0\farcs9 \times 3\farcs2$.
369: The blueshifted (bottom) jet is smooth and relatively featureless in both
370: intensity and velocity, while the redshifted (top) jet breaks into
371: three distinct clumps, with the central clump having a higher velocity
372: than the other two.
373: }
374: \end{figure*}
375:
376:
377:
378: We reduced the data using an early version of the OSIRIS data
379: reduction pipeline \citep{2004SPIE.5492.1403K}. The data were
380: sky-subtracted, corrected for offsets between the 32 detector readout
381: channels, spatially rectified to extract the individual spectra into a data cube,
382: and corrected for cosmic rays and other
383: blemishes. The A stars BD +9 5007 and HD 7789 served as telluric
384: standards and Point Response Function\footnote{We follow the Spitzer convention of
385: distinguishing between the intrinsic optical Point Spread Function (PSF) and
386: the Point Response Function (PRF), which is the PSF convolved with the
387: detector's pixel response. See
388: \url{http://ssc.spitzer.caltech.edu/postbcd/doc/PRF\_vs\_PSF.pdf}.
389: Because our OSIRIS data used the 50 mas pixel scale, coarser than
390: Nyquist for the Keck AO PSF at 1.6 \micron, the
391: PRF is dominated by the pixel response. By taking the
392: quadrature difference of the observed 61 mas PRF FWHM and 50 mas pixel scale,
393: we estimate that the intrinsic optical PSF had $\sim 35$ mas FWHM,
394: approximately the diffraction limit. Because of the difficulties inherent in trying to
395: accurately measure Strehl ratios on undersampled
396: data \citep{2004SPIE.5490..504R}, we do not attempt to
397: directly measure the Strehl ratio achieved in these observations.}
398: (PRF) references. For BD +9 5007, observed immediately
399: before \LkHa 233, we measure a PRF FWHM of $61 \pm 5$ mas.
400: We also used these stars to establish photometric zero points:
401: we converted their 2MASS $H$ and $K_s$ magnitudes into magnitudes in the OSIRIS \textit{Hn3} and \textit{Kn3} filters via
402: synthetic photometry, and then derived from those magnitudes a zero point for flux-calibrating our
403: \LkHa 233 data.
404:
405:
406: These observations were made during the
407: instrument's commissioning period, and show a variety of instrumental
408: artifacts, most notably ghost images at certain wavelengths caused by
409: crosstalk between lenslets due to a grating misalignment.
410: These ghost images do not occur in the wavelength ranges of interest
411: to us, and so we simply ignore them. The grating has
412: been realigned subsequent to these observations and more recent
413: data do not suffer from these ghosts.
414:
415: OSIRIS also includes an imaging
416: camera whose field of view is offset by 19\farcs4 from the spectrograph
417: field. Our sky observations were obtained such that \LkHa 233 fell on
418: the imager field of view (FOV) while we were obtaining spectrograph skies, and vice
419: versa. In this manner we obtained $Z, J, H,$ and $K$ band observations of \LkHa 233, with
420: total exposure times of 300, 360, 360, and 315 s, respectively. These
421: observations were reduced via the usual steps of flat fielding, sky
422: subtraction, image registration, and summation.
423:
424:
425: During this observing run, we also observed the Herbig Ae/Be stars
426: \LkHa 198, V376 Cas, and Parsamian 21, with identical instrument
427: configuration and exposure times. No outflows were visible around the
428: first two sources. Parsamian 21 has a faint outflow extending
429: to the north, which we postpone discussing for a subsequent paper.
430:
431: \section{Results}
432:
433:
434: \subsection{The Jet}
435: \begin{figure*}[!ht]
436: \begin{center}
437: \includegraphics[width=6in]{f4.eps}
438: \end{center}
439: \caption{
440: \label{plotfig}
441: \textbf{Left:} Continuum-subtracted position velocity diagram for
442: \FeII 1.644 \micron, using a
443: 0\farcs1 wide ``virtual slit'' centered on the jet.
444: \textbf{Center:} Continuum-subtracted image of the jet, integrated in
445: velocity. The upper and
446: lower portions of this image are summed over slightly different
447: wavelength ranges centered on the red- and blueshifted jets, each with
448: bandwidth 0.6 nm. The vertical dotted lines show the boundary of the ``virtual
449: slit'' used to produce the left-hand figure.
450: \textbf{Right:} Measurements of jet FWHM versus distance from the
451: star. This plot was constructed by fitting each row in the
452: continuum-subtracted image with a Gaussian. The FWHM appears narrower
453: at the position of the knots, and broader where the emission is
454: fainter (see \S \ref{jetsection}).
455: }
456: \end{figure*}
457:
458:
459:
460: \label{jetsection}
461:
462:
463:
464: A well-collimated bipolar emission-line jet extends several arcseconds
465: on either side of \LkHa 233 (Figure \ref{diskjetfig}), with the
466: blueshifted jet at a position angle of 249\degr, and the redshifted
467: jet at 69\degr. We also see a roughly edge-on circumstellar disk and
468: bipolar nebula extending perpendicular to the jet. The jet has
469: excavated a cavity in this envelope, the limb-brightened edges of
470: which form the well-known ``X'' centered on \LkHa 233. We detect the
471: jet in the \FeII 1.600 and 1.644 $\micron$ lines, and possibly H$_2$
472: 2.122 \micron\ (Figure \ref{fig1d}). Faint hydrogen Br$\gamma$ 2.16
473: $\micron$ emission is also detected along the course of the jet, but
474: does not show velocity shifts and is spread over a wider spatial
475: extent than the [FeII] emission; this appears to just be Br$\gamma$
476: emission from the star, scattered by dust in the circumstellar
477: envelope. Of the lines observed here, the 1.644 $\micron$ line is the
478: brightest, and we choose to concentrate our analysis on it. In Figure
479: \ref{slicesfig} we show continuum-subtracted images in a range of
480: velocity channels centered on 1.644 $\micron$, and we show in Figure
481: \ref{plotfig} a position-velocity diagram, an summed
482: continuum-subtracted jet image, and a plot of the FWHM.
483:
484: The jet is narrow, tightly collimated, and knotty. Its FWHM increases
485: slowly from 100 AU near its base to $\sim 200$ AU at the edge of our
486: FOV, 1300 AU from the star, giving an opening angle of 9\degr (Figure
487: \ref{plotfig}). Three bright knots are apparent in the redshifted
488: jet, at approximate distances of 0.3, 0.8, and 1.3\arcsec\ from the
489: star, corresponding to 260, 700, and 1150 AU in the plane of the sky.
490: We will refer to these as knots C, B, and A, respectively. The
491: blueshifted jet is noticeably less clumpy, but brightens considerably
492: near its base, reaching its peak intensity at 0\farcs1 (88 AU) from
493: the star.
494: There is an anticorrelation between jet width and intensity (Figure
495: \ref{plotfig}, right), with the bright knots having lower FWHM than
496: the fainter inter-knot emission. This behavior has also been observed
497: for the outflows from RW Aur (Woitas et al. 2002) and HH 23 (Ray et
498: al. 1996).
499:
500: We were unable to find any reported measurements of \LkHa 233's radial
501: velocity in the literature\footnote{We note that Corcoran and Ray
502: (1997,1998) state they measured the systemic velocity using optical Na
503: D lines, but do not anywhere report the actual measured velocity}, nor
504: are there stellar lines suitable for measuring it in our data (since
505: $Br\gamma$ is in emission). Hence we compute velocities relative to
506: the solar system barycenter, without correcting for the radial motion
507: of the star. In this reference frame, the red- and blueshifted jets
508: have average velocities of $+108 \pm3$ and $-89 \pm 3$ km s$^{-1}$,
509: respectively. If the red and blue jets have the same speed relative
510: to the star but opposite directions, then the stellar barycentric
511: radial velocity appears to be $\sim -10$ km s$^{-1}$.
512:
513: A significant difference between the two directions is that the
514: blueshifted jet has a relatively constant velocity along its length,
515: while the redshifted jet shows variation in its velocity. Knot B, at
516: $0\farcs8$, has a higher velocity than the other two clumps by about
517: 30 km s$^{-1}$ (Figure \ref{plotfig}, left). This velocity variation
518: so close to the jet's origin indicates that the amount of acceleration
519: imparted to the jet must fluctuate with time, and must do so in a way
520: that is not symmetric between the jet and counterjet.
521:
522:
523:
524:
525:
526: If we assume the jet is perpendicular to the disk, which has an estimated inclination of $i = 65\degr \pm 5$ (see \S\ref{disksection} below),
527: then the radial velocities we
528: observe are only $40 \pm 9$ percent of the total jet velocity. The implied total jet
529: velocity, 200-300 km s$^{-1}$, is consistent with outflows seen
530: around other Herbig Ae stars \citep{1994ASPC...62..237M}. Based on
531: this, we estimate the
532: proper motion of the jet should be $0\farcs05\pm0.01$ per
533: year. Hence the bright knots in the jet were launched from the star
534: within the last 10 to 20 years, and their motion should be apparent
535: over short timescales.
536:
537:
538:
539:
540:
541:
542: We do not see much evidence for the low velocity component (LVC)
543: reported by \citet{corc98} at about -25 km/s. This may be due to the fact that
544: the \FeII 1.644 $\micron$ line traces higher densities than the \SII
545: 6716, 6731~\AA~ lines, so in the NIR we are only sensitive to dense
546: gas in the core of the jet.
547:
548:
549:
550:
551:
552: The emission lines of \FeII can be used to measure several important
553: physical quantities, such as temperature, electron density,
554: and extinction
555: \citep{2004ApJ...614L..69H}. Most such
556: measurements require the comparison of flux ratios between different
557: lines of \FeII, or between \FeII and other species in the optical.
558: The limited
559: wavelength coverage of our data restricts their
560: diagnostic capabilities, as only the 1.644 and 1.600 $\mu$m lines
561: are present in our wavelength range. The flux ratio of those
562: two lines does provide a measure of the electron density in the shocked gas
563: \citep{2002A&A...393.1035N}. We find that the electron density in \LkHa 233's
564: jet is $n_e \sim 10^4-10^{4.5}$ cm$^{-3}$ and decreases at
565: larger distances from the star; see Figure \ref{electronfig}. This
566: decrease at larger distances has also been seen in many jets from T
567: Tauri stars
568: \citep{2002A&A...393.1035N, 2006A&A...456..189P, 2007AJ....133.1221B, 2007ApJ...660..426H}
569: and is believed to show the slow recombination of hydrogen after the
570: initial shock ionization \citep{1999A&A...342..717B}.
571: However, as \FeII emission traces
572: denser regions of the gas, our measured electron densities
573: are not
574: necessarily indicative of \LkHa 233's outflow as a whole. Future observations
575: which also include the optical \SII doublet are needed to provide complementary
576: measures for gas at lower densities and/or different distances from
577: the shock fronts.
578: \begin{figure}[!t]
579: \begin{center}
580: \includegraphics[width=2in]{f5.eps}
581: \end{center}
582: \caption{
583: \label{electronfig}
584: The ratio of the
585: 1.644 \um\ and 1.600 \um\ lines provides a
586: diagnostic for electron density $n_e$. The solid and dashed lines
587: show the expected ratio as a function of $n_e$ for two different
588: temperatures, from Nisini et al. (2002). The measured line ratios for the
589: three clumps in the redshifted jet of \LkHa 233 are indicated; the
590: length of the crosses shows the uncertainties. The electron density
591: decreases along the jet as one moves away from the star. The values
592: of $\sim 10^4- 10^{4.5}$ cm$^{-3}$ are comparable to electron densities
593: in several T Tauri HH jets measured from the same \FeII line ratio by
594: \citet{2002A&A...393.1035N,2005A&A...441..159N}.
595: }
596: \end{figure}
597:
598:
599:
600:
601:
602: \subsection{The Disk}
603: \label{disksection}
604: \begin{figure*}[!ht]
605: \begin{center}
606: \includegraphics[width=6in]{f6.eps}
607: \end{center}
608: \caption{
609: \label{modelfig}
610: \textbf{Left:} Observed appearance of the disk around \LkHa 233. This
611: $1.6 \um$ image is the median of the continuum from the OSIRIS
612: \textit{Hn3}
613: datacube.
614: \textbf{Center:} A $1.6 \um$ disk model computed with MCFOST,
615: convolved with a measured OSIRIS \textit{Hn3} PRF obtained from observations
616: of a calibration star. The model reproduces the central brightening,
617: sharp V-shaped southern disk face, and less distinct northern disk face.
618: \textbf{Right:} The observed SED for \LkHa 233 (diamonds), plus the
619: computed SED from our model (solid line). The SED is fairly well fit for
620: wavelengths $< 50 \um$; the deficit at longer wavelengths arises
621: because the model presented here includes only the disk and not the
622: surrounding extended nebula, which contains cooler
623: dust at much larger distances from the star (but still within the
624: large IRAS 60 and $100 \um$ beams). The dashed line is the input
625: stellar spectrum, and the solid line shows the resulting spectrum
626: after reprocessing by the disk. See \S 3.2 in the text for a
627: discussion of this model and its limitations.
628: }
629: \end{figure*}
630:
631:
632: In addition to showing the jet, these observations also clearly reveal
633: the circumstellar disk around \LkHa 233, in the form of a
634: bright nebula crossed by a dark lane, the
635: characteristic signature of a roughly edge-on circumstellar disk
636: (Figure
637: \ref{diskjetfig}, right). The presence of this disk was originally
638: suspected based on optical polarimetry \citep{aspin85}.
639: \citet{corc98} were unable to trace the redshifted optical \SII
640: emission any closer than 0\farcs6 to the star, and inferred the
641: presence of a disk with this radius that occulted the jet. We confirm here that
642: interpretation: the circumstellar disk extends from $0\farcs5$ below to
643: $0\farcs6$ above the locus of peak intensity. In contrast to the \SII
644: line, in \FeII 1.644 $\micron$ light the jet can be traced all the way in
645: to the star on both sides (Figure \ref{slicesfig}). There is at most a
646: two-pixel-wide ($0\farcs1$) region at the base of the red jet where
647: the emission is somewhat fainter, possibly due to obscuration. This
648: indicates that the circumstellar dust is not optically
649: thick at these wavelengths, in contrast to the optical. The observed
650: disk morphology is consistent with that seen in visible light in HST
651: WFPC2 observations (Stapelfeldt et al., in prep).
652:
653:
654:
655: The continuum intensity peak is well-resolved in these data, with
656: a FWHM of 0\farcs135 versus 0\farcs06 for a PRF reference star.
657: Therefore we do not observe the star directly, but instead see it in
658: disk-scattered light. The peak intensity is located at the apex of
659: the southwest face of the disk, which is the brighter of the two
660: faces. This is consistent with the southwest side of the system being
661: inclined toward us, as indicated by the blueshifted jet.
662: By comparison with a series of circumstellar disk models computed
663: using the MCFOST Monte Carlo radiative transfer code of
664: \citet{2006A&A...459..797P}, we determined that \LkHa 233's disk is inclined
665: approximately
666: $65 \pm 5$ degrees\footnote{We measure inclinations of the disk midplane relative to the plane of the sky, such that pole-on disks have $i=0\degr$ and edge-on disks have $i=90\degr$}. A numerical model of a
667: circumstellar disk
668: at this inclination extending from 0.6 to 500 AU reproduces fairly well both the
669: observed appearance and SED of \LkHa 233 (Figure \ref{modelfig}). We
670: caution that some parameters, such as the inner disk
671: radius, are not well constrained by these data. Hence
672: this is only one possible disk model which fits the
673: current data,
674: not necessarily a unique best-fit solution.
675:
676: Previous estimates of the stellar properties of
677: \LkHa 233 itself were obtained assuming only regular
678: interstellar extinction, typically around $A_V=2-4$, with reddening $R_V$ between
679: 3 and 5
680: \citep{2004AJ....127.1682H,2006astro.ph..8541M}.
681: However, our results here confirm that \LkHa 233 is seen entirely in
682: scattered light at optical and near-IR wavelengths. The true $A_V$ is much higher than previously estimated,
683: so the observed SED contains only a small
684: fraction of the total luminosity. \LkHa 233 must therefore be
685: substantially more luminous than previously thought. With the
686: stellar temperature fixed at $T_\ast = 8600 K$ (appropriate for an A5
687: star), we were able to
688: simultaneously match both the observed disk morphology and SED by
689: using a stellar radius of $R_\ast = 9 \Rsun$. This implies a total
690: stellar luminosity $L_\ast \sim 400 \Lsun$---about an order of
691: magnitude higher than the 20-80 \Lsun previously found by
692: \citet{2004AJ....127.1682H}.
693: These revised stellar properties place \LkHa 233
694: near the stellar birthline for intermediate mass stars
695: \citep{1990ApJ...360L..47P,2005IAUS..227..196P}, and therefore \LkHa
696: 233 may in fact be an very young star with mass $\sim 4 \Msun$.
697:
698: The key parameters of our model are as follows: $T_{\ast}$ = 8600 K,
699: $R_\ast$ = 9 \Rsun, $r_{out} = 500$ AU, and
700: grain sizes from $0.03-10 \um$ with power law
701: index -3.5.
702: However, we emphasize that the model presented here is \textit{not} a rigorous
703: best fit to the data, merely one plausible fit, and there remain large uncertainties in
704: both disk and stellar properties. In particular, the MCFOST Monte Carlo code
705: used does not yet account for any puffed up inner rim to the disk
706: \citep{2004A&A...417..159D} nor does it include the extended bipolar
707: envelope also present around \LkHa 233 in addition to the disk \citep{Perrin2004Sci}.
708: The lack of the bipolar envelope component results in the present
709: model having an unrealistically large scale height (30 AU at 100 AU)
710: in order to fit the data well.
711: More detailed modeling, quantitatively fit to observations
712: across a wide range of wavelengths, would help clarify the situation,
713: but is beyond the scope of this paper.
714:
715:
716:
717:
718: \subsection{Mass Flux}
719:
720: The observed intensity in the \FeII lines allows us to estimate the mass flux.
721: Currently the physical parameters of the jet, such as the temperature,
722: ionization, and Fe depletion onto grains, are not well known so our
723: estimate will require making several assumptions; future
724: observations will enable a more refined calculation.
725: \citet{2006A&A...456..189P} presented two separate methods for
726: estimating mass flux based on \FeII emission, based on either
727: (A) the jet's apparent size and density, or (B) the observed emission line
728: luminosity.
729:
730:
731:
732: The first method relies on computing the mass outflow rate
733: across a given plane perpendicular to the jet,
734: based on
735: the observed jet cross section, velocity, and density:
736: \[
737: \dot M = \mu m_H n_H \pi r_J^2 v_J
738: \]
739: where $\mu$ is the mean molecular weight, taken to be 1.24, $m_H$ is
740: the mass of hydrogen, $n_H$ is the number density of hydrogen, and
741: $v_J$ is the jet velocity.
742: In the case of \LkHa 233, we do not know $n_H$, but
743: we do know that $n_e \sim 5\Exp{4} \cm^{-3}$, from which we can estimate $n_H$.
744: Typical
745: ionization fractions for HH objects are $x_e = 0.03-0.3$, as determined from both
746: shock models and observational diagnostics
747: \citep{1994ApJ...436..125H,1999A&A...342..717B,2000A&A...356L..41L}, so if we assume
748: $x_e = 0.1$, then $n_H = 5\Exp{5} \cm^{-3}$. Using a rough estimate of
749: $r_J = 90$ AU, we compute $\dot M = 1.0\Exp{-6} \Msun$ year$^{-1}$.
750: This method relies on the assumption that the jet can be treated as a
751: uniformly filled cylinder with density measured by our diagnostic, and thus
752: may be better considered an upper limit on the true $\dot M$.
753:
754: The second method relies on the observed luminosity $L_{[FeII]}$ being
755: proportional to the total number of emitting atoms.
756: Unlike the first method, this second approach does take into account spatial
757: variations in the jet density, but is subject to uncertainties in the
758: absolute flux calibration. In this method we compute $\dot{M}$ using
759: \begin{eqnarray*}
760: \dot M &=& \mu m_H (n_H V) v_t/l_t \\
761: (n_H V) &=& L_{[FeII]} \left( h \nu A_i f_i \frac{Fe+}{Fe}
762: \frac{Fe}{H} \right)^{-1}
763: \end{eqnarray*}
764: where $V$ is
765: the volume filled by the emitting gas, $v_t$ is the tangential
766: velocity of the jet, $l_t$ is the tangential length of the jet volume element,
767: $\case{Fe+}{Fe}$ and $\case{Fe}{H}$ are the ionization fraction and
768: atomic abundance relative to hydrogen,
769: $A_i$ is the Einstein A radiative rate coefficient, and $h$ and $\nu$
770: are as usual the Planck constant and frequency of the radiation.
771: Based on our estimated disk inclination $i=65\degr\pm5$ and observed
772: $v_{radial} = 100$ km s$^{-1}$, we compute $v_t \sim 240\pm40$ km s$^{-1}$, and
773: we take $l_t$ to be the tangential length of each OSIRIS pixel,
774: $44$ AU at 880 pc.
775: Several assumptions are needed
776: in order to transform the
777: observed line luminosity $L_{[FeII]}$ into the density-volume product
778: $n_H V$: without accurate knowledge of the temperature we have to
779: estimate the ionization fraction and upper state fraction for iron. We
780: base our estimates on conditions observed in typical HH
781: objects, for instance by \citet{2006A&A...456..189P} and \citet{2007ApJ...660..426H}, who measured temperatures
782: of $8000-25000$ K based on optical line ratios.
783: Taking those temperatures plus our measured $n_e=10^4 - 10^{4.5}$, we
784: used a 16-level model of the Fe+ ion \cite[][similar to that described by]{2003A&A...410..155P}
785: to estimate the fraction of ions in the upper energy level is $f_i = 0.01 \pm 0.005$.
786: We further assumed
787: that Fe is entirely ionized ($\frac{Fe+}{Fe} = 1$), and that iron
788: has a solar abundance but is 90\% depleted onto solid grains (based on
789: the 87\% depletion reported for the HH 23 jet by Podio et al. 2006).
790:
791:
792: Our spatially resolved IFU data enable us to measure the mass flux rate
793: as it varies along the jet.
794: We first collapse the 2d continuum-subtracted \FeII image perpendicular to the jet to get a total $L_{[FeII]}$ for each row, and
795: then apply the above algorithm to compute $\dot M$.
796: The computed mass flux has a median value of
797: $1.2\Exp{-7} \Msun$ year$^{-1}$, and is fairly constant along most of
798: the jet, varying between $1.0-1.5\Exp{-7} \Msun$ year$^{-1}$. The one
799: exception is the middle knot in the redshifted jet, knot B, where the flux
800: rises to $2.6\Exp{-7} \Msun$ year$^{-1}$.
801: We caution that these values are \textit{approximate} at best, given
802: the lack of knowledge of the temperature and ionization structure, and
803: furthermore represent lower limits, because all of the outflow is not
804: necessarily shocked and luminous, and because we have not corrected for
805: extinction.
806:
807:
808: The two methods give results which differ by an order of magnitude, with the
809: estimate based on method A higher than that of method B.
810: This discrepancy is not entirely unexpected,
811: since the filling factor for \FeII emission is less than unity.
812: Podio et
813: al.\ (2006) found masses computed with method A were often 3-10 times higher
814: than those computed based on emission line fluxes, and they considered
815: the latter method to provide a better estimate of the mass flux
816: transported by the atomic component of the outflow.
817:
818: The values for $\dot M$ for \LkHa 233 are comparable to the
819: high end of the range of mass fluxes seen around T Tauri stars. For
820: instance, if we adopt the lower of the two estimates as the more reliable,
821: the flux in \LkHa 233's jet is
822: roughly comparable to the mass flux computed for RW Aur by Woitas et
823: al.\ (2002), $\sim 10^{-7} \Msun$ year$^{-1}$, though
824: about an order of magnitude higher
825: than the values measured for HH 34 and HH 111 by Podio et
826: al.\ (2006).
827:
828:
829: Outflow rates from young stars are proportional to
830: the accretion rates onto those stars
831: \citep{1990ApJ...354..687C,1995ApJ...452..736H,2006ApJ...646..319E},
832: across a mass range spanning
833: from brown dwarfs \citep{2005ApJ...626..498M} to Herbig Ae/Be stars
834: \citep{1998A&A...331..147C}.
835: Outflow mass fluxes are typically a few percent of the accreting mass flux \citep{1995ApJ...452..736H}
836: As far as we know, there are currently no measurements of the accretion rate onto \LkHa
837: 233, but based on our estimated outflow mass flux, the accretion rate of onto it should be $\sim10^5
838: \Msun$ year$^{-1}$.
839:
840:
841:
842: \subsection{The Distance to \LkHa 233, Revisited}
843:
844:
845: \label{distance}
846:
847: As discussed above, the fact that \LkHa 233 is seen entirely in
848: scattered light implies that it must be substantially more luminous
849: than has previously been thought, assuming it is indeed at a distance
850: of 880 pc. It is worth considering whether it might instead be closer.
851:
852: Establishing accurate distances to Herbig Ae stars is often
853: problematic \citep[e.g.][]{1998AJ....116..890S,2004A&A...419..301M,Perrin2006PDS144}.
854: The 880 pc distance to \LkHa 233 was originally derived by \citet{1978MNRAS.182..687C}, and
855: the majority of subsequent authors have adopted their estimate.
856: Calvet \& Cohen derived that value from the distance modulus to the
857: nearby star HD 213976, a bright B1 star located 6\farcm5 away
858: on the edge of the dark cloud in which \LkHa 233 is embedded.
859: They
860: also estimated photometric distances to two other members of this small cluster,
861: the K stars \LkHa 231
862: and 232, finding $d=600-700$ pc subject to the assumption that those stars are both
863: overluminous by 2 mags.\ in $V$ due to their youth.
864:
865: Is it possible that HD 213976 is in fact not at the same distance as
866: \LkHa 233? Aspin (1985) noted that HD 213976's optical colors imply its $A_V$ is only 0.42, significantly
867: less than the typical $A_V=2.5$ per kpc. This implies both
868: that we look toward it along a galactic sightline with lower than average extinction, and
869: that HD 213976 must be located on the near side of the dark cloud.
870: Hence \LkHa 233 should be at least as far away as HD 213976, if not
871: farther.
872:
873: Some authors \citep[e.g.][]{2003MNRAS.340.1173B} have instead described \LkHa 233 as being
874: a member of condensation A of the Lac OB1 molecular cloud,
875: which is only $418\pm 15$ pc distant \citep{2005AJ....129..856H}.
876: While \LkHa 233's position on the sky does place it on the apparent edge of
877: Lac OB1, we consider it unlikely to actually be a member of that association, for several reasons. First, \LkHa
878: 233's proper motion is very different from that of Lac OB1: $(\mu_l \cos b, \mu_b) = (-9, 21)$ mas/yr
879: \citep{2005A&A...438..769D} versus
880: $(\mu_l \cos b, \mu_b) = (-2.3\pm0.1, -3.4\pm0.1)$ for Lac OB1;
881: \citep{1999AJ....117..354D}.
882: Even more problematic is the fact that Lac OB1 is believed to be 16 Myr old
883: \citep{1999AJ....117..354D}. Because protoplanetary disks dissipate on
884: time scales of 3-10 Myr, the presence of a massive disk around \LkHa
885: 233 is inconsistent with its being that old.
886: For these reasons we prefer to retain the traditional 880 pc distance
887: for \LkHa 233.
888:
889: Because the system's inclination can be constrained based on the disk's
890: appearance, the proper motion of the clumps in the jet can potentially provide a
891: direct measure of the distance. For a disk inclination $i=65\degr\pm 5$ and $d=880 \pc$,
892: the jet proper motion should be $0\farcs06\pm0.01$ year$^{-1}$, while
893: $d=420 \pc$ would imply a proper motion of $0\farcs12\pm0.03$ year$^{-1}$.
894: However, this will be a challenging measurement given uncertainty in the inclination,
895: the possibility of jet precession, variations in the internal working
896: surfaces traced by the clumps, etc.
897:
898:
899:
900:
901: \section{Discussion}
902:
903: The outflow from \LkHa 233 is in general quite similar to
904: those seen around T Tauri stars. The velocity structure and collimation of
905: the jet, the morphology of the knots, and the presence of asymmetries
906: between the jet and counterjet all resemble the properties of outflows
907: seen around lower-mass stars.
908:
909: For instance, we see that for both the red and blue
910: directions, the higher velocity channels are more tightly collimated
911: than the lower velocity channels. Such increased collimation at higher
912: velocities has also been seen in the outflow from DG Tau
913: \citep{2002ApJ...576..222B}. This is in agreement with predictions of MHD models that
914: inner streamlines should dominate the emission
915: \citep[e.g.][]{2004Ap&SS.292..643D}.
916:
917:
918: It is striking how similar \LkHa 233 appears to the T Tauri star HH
919: 30, with its well-known edge-on disk and perpendicular outflow
920: \citep{1996ApJ...473..437B}. \LkHa 233's disk is
921: asymmetric in brightness between the left and right sides, as is
922: HH 30's disk \citep{2004ApJ...602..860W}. The outflows
923: from both stars are knotty, and even the characteristic periods for knot
924: production are similar, 2.5 years for HH 30 and $\approx 5$ years for
925: \LkHa 233.
926: These knots appear to be due to internal shocks within the
927: jet, rather than the jet shocking into the ambient medium.
928: The properties of the knots in \LkHa 233's jet are indeed consistent with internal
929: working surfaces caused by variations in jet outflow velocity. In
930: particular, the brightest knot is also that with the highest velocity,
931: just what would be expected for more rapid material plowing into
932: a slower portion of the jet.
933:
934: It is curious that, at our angular
935: resolution and S/N, only the redshifted jet appears knotty.
936: Similar asymmetries have been seen in outflows from about
937: half of T Tauri
938: stars \citep{1994ApJ...427L..99H}.
939: On the other hand, at
940: least some jets are very symmetric, such as HH 212 in Orion
941: \citep{1998Natur.394..862Z}, which displays a very regular series of knots
942: in nearly perfect pairs on opposite sides of the star. The symmetry of
943: the HH 212 knots and the regularity of knot spacing in many
944: systems (including \LkHa 233) suggest that the production of knots is
945: an inherent part of the jet launch process, possibly due to disk
946: instabilities or recurrent stellar activity cycles. However, the
947: commonness of asymmetric systems indicates that any inherent
948: regularity in the process can easily be disrupted or masked, by
949: collisions with surrounding ambient material, pressure gradients, or asymmetries in
950: magnetic fields very near the star
951: \citep{2006ApJ...650..985W}. But then what causes those asymmetries in
952: magnetic field or pressure?
953: The origin of these various symmetries and asymmetries remains mysterious, and poses a key challenge to
954: theories of jet acceleration and collimation.
955:
956:
957:
958: A particularly useful comparison object is
959: HD 163296, which has perhaps the best-studied outflow from any Herbig Ae star
960: \citep{2000ApJ...542L.115D,2006ApJ...650..985W}. HD 163296, an A2Ve star at $122_{-13}^{+17}$
961: pc \citep{1998A&A...330..145V}, launches a bipolar collimated
962: jet, HH 409, which also is broken into an asymmetric series of knots. The HH 409 jet displays knots on both the jet
963: and counterjet, but on the two sides the knots differ in both and distance
964: from the star. Based on their high spatial resolution observations of HD 163296, Wassell et
965: al.\ concluded that the asymmetries must originate near
966: or within the jet launch region itself, rather than being the result of
967: interactions with the ambient medium. Other authors have reached
968: the same conclusion regarding the asymmetries observed in outflows
969: from T Tauri stars \citep{2002ApJ...580..336W,2006astro.ph..6605P}.
970: Similarly for \LkHa 233 , given that strong
971: asymmetries are present in the outflow within the innermost few hundred
972: AU, it seems
973: unlikely that those asymmetries could be due to the
974: ambient medium. While there is indeed substantial dust and gas
975: near the star (in the observed disk and nebula), we know the outflow
976: from \LkHa 233 has been ongoing for thousands of years based on
977: the length of the observed outflow (at least 1.3 pc; McGroarty et al.\
978: 2004). Therefore any primordial material along the jet axis was
979: long ago swept away, out of the inner few arcseconds where we observe
980: the asymmetries and knots. Explaining the difference in appearance
981: between the red- and blue-shifted jets thus requires the presence of
982: an asymmetric geometry within the inner jet launch region. This
983: asymmetry may take the form of pressure gradients, different magnetic field strengths or differing field geometries
984: between the two directions.
985:
986: The mass flux in the jet of \LkHa 233 is about two
987: orders of magnitude greater than that of HD 163296. The measured
988: electron density $n_e$ for \LkHa 233 is also about a factor of ten higher
989: than that reported for HD 163296, but the value for HD 163296 was
990: measured using ratios of optical \SII transitions which trace lower density parts of
991: the post-shock gas \citep{2006ApJ...650..985W}. Hence it is hard to say how meaningful this difference
992: in $n_e$ is, and the true difference in accretion
993: rate may be somewhat less than this estimate.
994:
995: Outflow mass fluxes are believed
996: to be proportional to accretion rate, and accretion rates are
997: generally larger
998: for younger stars \citep{1998ApJ...495..385H}.
999: Therefore, together with its
1000: higher extinction, greater infrared excess, and larger amount of nebulosity,
1001: the high mass flux of \LkHa 233 suggests that it is at a
1002: younger evolutionary state than HD 163296. This is consistent
1003: with our revised stellar
1004: parameters which place \LkHa 233 near the stellar birthline for
1005: intermediate mass stars.
1006: The outflow velocity is expected to be
1007: of order the escape velocity, $v_{esc} = \sqrt{GM_\ast/R_\ast}$.
1008: For
1009: our revised stellar properties (4 \Msun, 9 \Rsun), the escape velocity
1010: is $v_{esc} \approx 300$ km s$^{-1}$, which compares favorably with the estimated total jet velocity.
1011:
1012:
1013:
1014:
1015: There is a growing body of evidence that outflows from high mass stars
1016: are generally less collimated than those of low mass stars. Given the
1017: tightly collimated nature of \LkHa 233's outflow, the transition to
1018: poorly collimated jets appears to occur at a higher mass.
1019: However, the opening angle of the bipolar cavity in the envelope around \LkHa 233 \citep{Perrin2004Sci}
1020: is 30\degr, much
1021: wider than the jet collimation angle. This, plus the scatter in HH objects
1022: seen by McGroarty et al. (2004), argues either for a change in outflow direction with
1023: time (such as from precession), for a much lower degree of
1024: collimation during earlier stages of the outflow, or for the presence
1025: of a wide-angle wind around the inner jet, unseen in the
1026: \FeII emission which traces relatively dense gas.
1027:
1028:
1029:
1030:
1031:
1032: \section{Conclusions}
1033:
1034: We have used Keck Observatory's laser guide star adaptive optics system and integral field
1035: spectrograph OSIRIS to observe the bipolar outflow of the Herbig Ae
1036: star \LkHa 233. Based on these observations, we believe that:
1037:
1038: \begin{enumerate}
1039: \item {\LkHa 233 has a narrow, collimated bipolar outflow, with the
1040: blueshifted jet at a position angle of 249\degr\ and the
1041: redshifted counterjet at 69\degr. }
1042: \item{At the present angular resolution and sensitivity, the blueshifted
1043: and redshifted jets have similar opening angles
1044: but very different appearances, with the red-shifted jet broken
1045: up into several knots while the blue-shifted jet is smooth. The
1046: knots in the red-shifted jet are consistent with shocked inner
1047: working surfaces within the jet.}
1048: \item {Based on \FeII emission intensity, we estimate the mass flux in this jet to be $(1.2 \pm 0.3) \Exp{-7}
1049: \Msun$ year$^{-1}$. This is toward the top end of mass fluxes
1050: observed around other T Tauri and Herbig Ae stars, and suggests
1051: that \LkHa 233 is a very young source with strong ongoing
1052: accretion.}
1053: \item{The observed properties of \LkHa 233's outflow -- its kinematics, collimation,
1054: asymmetric and knotty appearance, electron density, and estimated mass flux --
1055: are all largely consistent with the observed properties of jets
1056: from T Tauri stars. The transition to
1057: the less-collimated outflows observed around massive YSOs must
1058: therefore occur at some still higher mass threshhold.}
1059: \item {The jet extends perpendicular to a flattened circumstellar
1060: disk, the innermost region of the X-shaped nebula around \LkHa
1061: 233.}
1062: \item{\LkHa 233 is seen only in scattered light from this disk;
1063: the star itself remains obscured at near-infrared wavelengths. Because
1064: of this, previous
1065: estimates of the extinction to the star are low, causing the
1066: total stellar luminosity to be underestimated in prior studies.}
1067: \item{Revised stellar parameters, based on fitting disk models to the
1068: SED, indicate that \LkHa 233 may be a 4 \Msun\ star near the
1069: stellar birthline, but there remain large uncertainties in its true
1070: stellar properties.}
1071: \end{enumerate}
1072:
1073:
1074: The main result of this study is confirmation that the outflow from
1075: \LkHa 233 is quite similar to typical outflows from T Tauri stars, as
1076: has been suggested based on prior observations at lower spatial
1077: resolution \citep{corc98,2004A&A...415..189M}. This supports the
1078: hypothesis that the same physical mechanism is responsible for
1079: creating stellar outflows, across the entire mass range from T Tauri
1080: to Herbig Ae stars. It is worth noting that circumstellar disks
1081: around Herbig Ae stars are also similar in broad outline to those
1082: around T Tauri stars \citep[e.g.][]{grady2005,Perrin2006PDS144,2007prpl.conf..539M} In contrast, there
1083: are suggestions that circumstellar disks around the Herbig Be
1084: and more massive YSOs differ from those around lower masses. In particular, there
1085: is evidence for a transition from magnetic accretion (for T Tauri and
1086: Herbig Ae stars) to direct disk accretion onto the star (for Be stars
1087: and beyond) \citep{2002MNRAS.337..356V,2007MNRAS.377.1363M}. Neither
1088: this transition in accretion mechanisms nor the transition from
1089: collimated to wide-angle outflows are particularly well understood.
1090: But the fact that both transitions occur at approximately the same
1091: stellar mass may be an important clue that these processes are
1092: connected. This may indicate either that the properties of outflows
1093: depend primarily on the properties of the circumstellar disks and how
1094: they accrete onto the star, or that both outflow and disk properties
1095: depend in turn on some third parameter (stellar magnetic fields?)
1096: which undergoes a transition at the Ae/Be boundary.
1097:
1098: Further insight into these matters will depend on obtaining
1099: increasingly detailed spatially resolved spectra of outflows, such as
1100: those provided by integral field spectroscopy. We have shown here the
1101: measured electron density using the 1.644/1.600 $\mu$m ratio, and have
1102: also presented a rough estimate of the mass flux using the 1.644
1103: $\mu$m line. But much more could be done with a more complete
1104: spectrum: Ratios of other lines (most notably 1.644/1.533) provide
1105: alternate
1106: measures of electron density, and comparison with \textit{optical}
1107: \FeII lines can measure the temperature as well
1108: \citep{2005A&A...441..159N,2006A&A...456..189P}. The 1.644, 1.320,
1109: and 1.257 \um\ lines all originate in the same upper state, and thus
1110: their ratios directly constrain the extinction $A_V$. This is a
1111: particularly exciting prospect for the redshifted jet in \LkHa 233: it
1112: is located partially behind the inclined circumstellar disk (occulting
1113: it in the visible but not in the near infrared), and thus it can be
1114: used to measure the extinction \textit{through the disk}. Future
1115: observations that include the complete optical and near-IR spectrum of
1116: \FeII and other shock tracers should enable a complete suite of
1117: diagnostics for temperature, density, ionization, refractory element
1118: depletion, and more. Detailed modeling of \LkHa 233 thus holds great
1119: promise for understanding outflow mechanisms around intermediate-mass
1120: stars.
1121:
1122:
1123:
1124:
1125:
1126:
1127:
1128:
1129: \acknowledgments
1130:
1131:
1132: These observations were made possible by the tremendous efforts of the
1133: OSIRIS instrument team and Keck Observatory staff, whose members are
1134: too numerous to list here in full. Special thanks in particular go to James
1135: Larkin, Shelley Wright, Mike McElwain, Randy Campbell, and Al Conrad.
1136: Conor Laver assisted with the OSIRIS data reduction pipeline. Christophe Pinte and
1137: Gaspard Duchene provided excellent guidance in the use of their
1138: MCFOST Monte Carlo code. MDP thanks Deirdre Coffey and Catherine
1139: Dougados for enlightening discussions and warm hospitality during his
1140: visit to Grenoble. We thank the referee for thoughtful comments that
1141: greatly improved this paper.
1142: The authors wish to recognize that the summit of Mauna Kea has always
1143: held a very significant cultural role for the indigenous Hawaiian community.
1144: We are most fortunate to have the opportunity to observe from this mountain.
1145:
1146: This work has been supported in part by the National Science
1147: Foundation Science and Technology Center for Adaptive Optics,
1148: managed by the University of California at Santa Cruz under
1149: cooperative agreement No. AST-9876783.
1150: MDP was partially supported by a
1151: NASA Michelson Graduate Fellowship, under contract to the Jet
1152: Propulsion Laboratory (JPL). JPL is managed for NASA by the
1153: California Institute of Technology.
1154:
1155:
1156: {\it Facilities:}
1157: \facility{Keck:II}
1158:
1159:
1160: \bibliographystyle{apj-short}
1161: \bibliography{haebes,pol}
1162:
1163:
1164:
1165: \end{document}
1166:
1167: