1: \documentclass[12pt,preprint]{aastex}
2: \begin{document}
3:
4: \title{Resolved Mid-Infrared Emission Around AB Aur and V892 Tau with
5: Adaptive Optics Nulling Interferometric Observations$^{1}$}
6: \footnotetext[1]{The results presented here made use of the of MMT Observatory,
7: a jointly operated facility of the University of Arizona and the Smithsonian
8: Institution.}
9: \author{Wilson M. Liu$^{2}$,
10: Philip M. Hinz$^{2}$,
11: William F. Hoffmann$^{2}$,
12: Guido Brusa$^{2,3}$,
13: Doug Miller$^{2}$,
14: and Matthew A. Kenworthy$^{2}$}
15: \altaffiltext{2}{Steward Observatory, University of
16: Arizona, 933 N. Cherry Ave., Tucson, AZ, USA 85721, email: wliu@as.arizona.edu}
17: \altaffiltext{3}{Osservatorio Astrofisico di Arcetri, Largo Enrico Fermi 5,
18: 50125 Florence, Italy}
19:
20: \begin{abstract}
21: We present the results of adaptive optics nulling interferometric observations
22: of two Herbig Ae stars, AB Aur and V892 Tau. Our observations at 10.3 $\mu$m
23: show resolved circumstellar emission from both sources. Further analysis
24: of the AB Aur emission suggests that there is an inclined disk surrounding
25: the star. The diameter of the disk is derived to be 24 to 30 AU with
26: an inclination of 45$\degr$ to 65$\degr$ from face-on, and a
27: major-axis PA of $30\degr \pm 15\degr$ (E of N). Differences in the
28: physical characteristics between the mid-IR emission
29: and emission at other wavelengths (near-IR and millimeter), found in
30: previous studies, suggest a complex structure for AB Aur's circumstellar
31: environment, which may not be explained by a disk alone. The similarity
32: in the observed size of AB Aur's resolved emission and that of another
33: Herbig Ae star, HD 100546, is likely coincidental, as their respective
34: evolutionary states and spectral energy distributions suggest significantly
35: different circumstellar environments.
36: \end{abstract}
37: \keywords{stars: individual (AB Aur, V892 Tau), stars: circumstellar matter,
38: stars: pre-main sequence, instrumentation: adaptive optics,
39: techniques: interferometric}
40:
41: \section{Introduction}
42: Over the past several years, numerous circumstellar disks have been
43: observed surrounding pre-main sequence (PMS) stars, including Herbig Ae (HAE)
44: stars. HAE stars, the evolutionary precursors to intermediate-mass
45: main-sequence stars like Vega, show infrared (IR) emission in
46: excess of what is expected from their photosphere. This emission was
47: thought to originate from optically thick, geometrically thin circumstellar
48: disks \citep{hill92,la92}, a model later modified to include disk flaring
49: in order to explain features in the observed spectral energy distributions
50: of the disks \citep{kh87,cg97}. Alternative models also include dusty
51: envelopes of material or a combination of a disk and an envelope
52: \citep{hart93,miro99}. Given the variety of models used to explain
53: the IR excess around PMS stars, observations of their circumstellar
54: environments are necessary to determine which of these models is most
55: representative of their true environment.
56:
57: The HAE star AB Aurigae (A0; $d=144$ pc \citep{vdA98}) has been the subject
58: of many studies based upon observations at several different wavelengths.
59: The star has an estimated age of 2-5 Myr \citep{ms97,vdA98}. Near-IR (NIR)
60: emission from the AB Aur circumstellar region has been observed, probing
61: thermal emission in the inner AU of the disk using long-baseline
62: interferometry \citep{eis03,m-g01}. Both studies suggest the
63: presence of a slightly inclined distribution of dust with an empty or
64: optically thin inner region (i.e., a ring-like structure).
65: Another recent study in the NIR has detected scattered light from the disk
66: at greater separations (out to 580 AU) and finds the disk to have a small
67: inclination \citep{f04}.
68: Observations in the mid-IR (MIR) suggest evidence for resolved circumstellar
69: material at 12 and 18$\mu$m at several tens of AU from the star
70: \citep{cj03,marsh}. Longer wavelength
71: observations of AB Aur in the millimeter were shown to have spatially
72: resolved molecular line emission at a few hundred AU \citep{ms97}.
73: Reflection nebulosity has also been detected in the optical by
74: \citet{grady} which finds material out to 1300 AU and a disk inclination
75: of less than $45\degr$.
76:
77: V892 Tau is a HAE star located in the Taurus-Auriga star forming
78: region, at a distance of about 140 pc \citep{elias}. NIR speckle
79: interferometry of the star revealed an elongated structure with a PA of
80: $90\degr$. The source of emission is speculated to be either a highly inclined
81: disk or a bipolar outflow \citep{haas97}.
82:
83: In this Letter, we present results of nulling interferometric observations
84: in the MIR of these two HAE stars, AB Aur and V892 Tau, in which
85: we have clearly resolved emission from both sources. From
86: these observations, we infer and discuss the physical properties of the
87: circumstellar material in the AB Aur system.
88:
89: \section{Observations and Data Reduction}
90: Observations were made in 2002 November and 2004 February at the 6.5 m MMT
91: Telescope on Mt. Hopkins, Arizona. The BracewelL Infrared Nulling
92: Cryostat (BLINC; \citet{hinz_phd}) provided supression of starlight, and the
93: Mid-Infrared Array Camera (MIRAC; \citet{hoff}) provided the final stop
94: for the two beams of the interferometer. The 2002 run consisted of
95: preliminary observations of both science targets at 10.3$\mu$m
96: (10\% bandpass). Four sets of 500 frames (each frame
97: with 0.5 s integration) were taken for each target, resulting in 2000 frames
98: for each source. Images of a point-soure calibrator, $\alpha$ Ari, were
99: taken before and after the science observations. Frames were sky subtracted
100: using off-source fields taken in between each set of frames. Photometry was
101: extracted from each of the science frames, and each set of frames was
102: examined for the best instrumental null (Instrumental Null = Nulled Flux/
103: Unnulled Flux). Source nulls were derived for each set of frames
104: (Source null = Instr. Null - Calibrator Null; see \citet{liu04} for a
105: full description of nulling calibration and diagnostics, as well as a
106: typical image from the BLINC-MIRAC camera).
107:
108: Followup observations were made in 2004 at the MMT, also at 10.3 $\mu$m,
109: utilizing BLINC-MIRAC with the telescope's adaptive optics (AO) secondary
110: mirror. Since the deformable mirror is the telescope's
111: secondary mirror, there is no need for an intermediate set of reimaging and
112: correcting optics between the secondary mirror and science camera.
113: This has the benefit of optimizing throughput and
114: decreasing background emissivity in the MIR by avoiding the use of extra
115: warm optics. The adaptive secondary also provides two major benefits
116: particular to our nulling observations: 1) The wavefront is stabilized,
117: allowing us to precisely tune the interferometer for the
118: best possible suppression of starlight; 2) Fewer data frames need to be
119: taken since the wavefront stabilization allows us to precisely tune the
120: destructive interference. These observations included 15 sets of 10 frames
121: (150 frames total) for AB Aur in destructive interference, and 3 sets of 10
122: frames for V892 Tau in destructive interference. Frames of the objects
123: in constructive interference, as well as off-source sky frames were taken
124: in between each set of nulled images.
125:
126: In order to probe the spatial distribution and orientation of the
127: circumstellar emission, the followup observations of AB Aur were taken at
128: several different rotations of the interferometer baseline relative to
129: the sky (see \citet{liu03} for a full description of this technique).
130: This was also attempted V892 Tau, however, poor observing conditions
131: on the night this object was observed allowed only limited followup.
132:
133: \section{Results and Discussion}
134: The preliminary observations in 2002 show evidence for resolved emission
135: in both sources. Source nulls derived from these non-AO observations
136: show the resolved flux surrounding each object to be at a level of 10 to
137: 20\% of its full 10.3 $\mu$m flux (3-6 Jy for both sources). Followup
138: observations in 2004 are described below.
139:
140: \subsection{V892 Tau}
141: Due to the limited observations of V892 Tau in 2004, we make no
142: conclusions about the spatial distribution of emission around this object.
143: However, from our observations we are able to confirm the presence of
144: extended emission in the MIR first uncovered by our preliminary observations.
145: The extended emission is verified to be at a level of $11\% \pm 6\%$ (about 3 Jy;
146: $2 \sigma$ error)
147: of the full 10.3 $\mu$m flux of the star. The resolved emission is detected
148: at a position angle of 164$\degr$, and assuming a Gaussian flux distribution
149: is responsible or the emission, has a diameter of about 20 AU. Follow up
150: observations are needed to probe the geometry of the resolved emission.
151:
152: \subsection{AB Aur}
153: Observations of AB Aur in 2004 were taken at five different rotations of
154: the interferometer baseline spanning $115\degr$ in increments of $30\degr$.
155: This rotation allows us to probe a range of position angles around
156: the star for resolved emission. The source null for each rotation and
157: corresponding position angle is listed in Table \ref{tab-abaur_source}
158: for each of three different calibrations of the data (see next paragraph).
159: Plots of the data are shown in Figure \ref{fig-abaur_source}. If
160: the emitting dust is in an inclined disk, one would expect the dependence
161: of the source null vs. position angle to be sinusoidal
162: \footnote{The transmitted signature of the nulling
163: interferometer is an interference pattern with interference fringes along the
164: baseline. If these fringes are parallel to the major axis of a disk, more
165: of the disk's light will be nulled, resulting in a lower percentage of
166: remaining light. When the fringes are aligned orthogonally to the major axis,
167: the value of the source null will be higher (i.e., there is more light
168: remaining). Therefore, there should be a sinusoidal variation in null with
169: respect to rotation of the baseline.}.
170: From a fit to the data, of the form $N = a + b * sin(PA + \theta)$,
171: with the period fixed at 180$\degr$ \citep{liu03}, we derive physical properties
172: for an inclined disk around AB Aur. For the fit, the values of $a$, $b$, and
173: $\theta$ are physically related to the size, inclination, and PA of the
174: major-axis, respectively.
175:
176: Observations of the calibrator star, $\beta$ Gem, were taken before and
177: after the observations of AB Aur. The calibrator shows
178: a significant change in the level of null we were able to achieve during
179: the observations, due probably to changes in observing conditions and
180: the effectiveness of AO wavefront correction. As a result, we calibrate
181: our data (calculate the source nulls) for AB Aur in three different
182: ways to determine how this affects the results. The different calibrations
183: are as follows: 1) We use the first calibrator measurement to calculate
184: all the source nulls; 2) We use a linear fit (in time) between the two
185: calibrator measurements to calculate the source nulls; 3) We use the last
186: calibration to calculate all the source nulls.
187:
188: All three methods of calibration yield a result in which the dust distribution
189: is significantly more resolved in one PA ($\approx 30\degr$) than another
190: offset by $90\degr$ ($\approx 120\degr$). This result is suggestive of a
191: flattened or elongated structure as the source of MIR excess emission.
192: Assuming two simple brightness distributions (a Gaussian disk and a ring),
193: physical parameters are derived from the fits to the source null vs.
194: position angle (Table \ref{tab-sinefit}). For each model we have
195: derived the size and inclination of the material needed to reproduce the
196: null vs. position angle profile we have observed. This is repeated for each
197: of the three calibrations, allowing us to assess the error introduced
198: by the calibration issues described above. All three calibrations yield
199: similar results and indicate that the 10.3 $\mu$m emission originates from a
200: separation 12 to 17 AU from the star. The presumed disk has a significant
201: inclination, 45 to 65 degrees from face-on and the position angle (PA) of the
202: major axis of the disk is $30\degr \pm 15\degr$. If the actual distribution
203: of warm dust is a combination of a flattened structure and a uniform symmetric
204: component (such as a disk plus envelope), then the disk component would
205: need to be more inclined to account for the amplitude in null variation.
206:
207: Previous studies have also observed AB Aur at MIR
208: wavelengths. \citet{cj03} used the Keck I telescope to observe the star
209: at 11.7 and 18.7 $\mu$m. They find that it is marginally resolved
210: at the longer wavelength. At 18.7 $\mu$m, they find an angular diameter of
211: about 1$\arcsec$ at the half-maximum flux level. This suggests that the
212: 18 $\mu$m emission is originating from a separation of about 70 AU,
213: several times greater than the 10 $\mu$m emission. A study by \citet{marsh}
214: finds evidence for resolved emission at 11.7 and 17.9 $\mu$m and
215: derives diameters of 40 and 80 AU for the emission, respectively.
216: Taking into account the derived size scales from this and both previous MIR
217: studies (separations of 12-17 AU, 20 AU, and 40-70 AU for the 10.3, 11.7,
218: and 18 $\mu$m emission, respectively), we note that the wavelength vs.
219: separation profile agrees with the radial temperature profile expected for
220: a flared disk, T $\sim$ r$^{-1/2}$ \citep{cg97}, assuming that the emission
221: is primarily thermal in nature. We also note that the NIR
222: study of \citep{m-g01} finds that the thermal 2$\mu$m disk size is about
223: 0.7 AU, which is also roughly consistent with a continuous flared disk.
224:
225: Studies at other wavelengths include NIR studies (e.g.
226: \citet{f04,eis03}), and millimeter observations \citep{natta01,ms97},
227: \citet{ms97} found that AB Aur is "marginally resolved" in molecular line
228: emission at 3 mm. They find a major axis PA of $79\degr$ with an inclination
229: of $76\degr$ from face-on. The significant inclination of
230: the disk agrees with the derived inclination of this study. However
231: more recent studies in the millimeter \citep[and references therein]{natta01}
232: cite a much smaller inclination ($< 30\degr$). A recent NIR
233: study by \citet{f04} using AO coronagraphic observations finds
234: a scattered light disk in H-band with a PA of $58\degr$ and a significantly
235: smaller inclination of $30\degr$ from face-on. \citet{eis03} used the
236: Palomar Testbed Interferometer to obtain K-band observations of inner
237: ($<0.5$ AU) disk surrounding AB Aur and find the inclination to be small,
238: within $30\degr$ of face-on, in agreement with \citet{f04}. The observations
239: of this study suggest a greater inclination for the MIR emission. One also
240: notes that the major axis PA derived for the MIR emission in this study
241: differs significantly (50-70\degr) from previous studies both in the NIR and
242: millimeter. This points to a difference in geometry for the dust between
243: the inner (a few AU) and outer (hundreds of AU) system. The biggest
244: difference between this study and those at other wavelengths is in the
245: inclination of the disk. Previous studies at several different wavelengths
246: all agree on a significantly smaller inclination than found by this study.
247: This, in combination with the discrepancies in the PA of the disk suggest
248: that the structure may be more complex than a disk alone, where emission
249: at different wavelengths are dominated by material with a different
250: distribution.
251:
252: Comparing AB Aur to a sample of 14 Herbig Ae stars observed in the MIR
253: by \citet{leinert} may be helpful in placing AB Aur into context and gaining
254: insight into their circumstellar environments. The \citet {leinert} study finds
255: a correlation between the size of resolved emission and the SED classifications of
256: \citet{meeus}. They find that the Meeus et al. Type I sources, characterized
257: by a rising MIR SED, tend to have spatially larger circumstellar emission
258: regions. By contrast, Type II sources have flat or declining MIR SEDs have
259: smaller resolved sizes. The resolved size of the AB Aur emission found
260: in this study would classify AB Aur as a Type I source, consistent with the
261: initial classification by \citet{meeus} by SED alone. While it appears that
262: the each type shows similar physical characteristics, there also
263: seems to be evidence that a simple, all encompassing physical model may not
264: be an ideal explanation for each Herbig type. For example, it is interesting to
265: note some differences, from nulling observations and other previous
266: studies, between AB Aur and another resolved Type I Herbig Ae star, HD 100546.
267: The primary difference highlighted by nulling observations is that AB Aur's
268: radial wavelength (temperature) profile seems to be consistent with a continuous
269: disk, whereas HD 100546's relative 10 $\mu$m vs. 20 $\mu$m emission region sizes
270: suggest an inner clearing \citep{liu03}. Other studies have found difference
271: in the age (10 Myr for HD 100546 vs. 2-5 Myr for AB Aur; \citet{vdA98,ms97}) and
272: evolutionary states \citep{bouwman} of the two stars. It seems, therefore,
273: that although the two stars show similarities in the size of their resolved
274: emission at 10 $\mu$m, the emission arises from physically different
275: distributions of circumstellar dust.
276:
277: \section{Ongoing Work: Surveys of Intermediate-Mass Stars}
278: The observations of AB Aur and V892 Tau presented here are part of a
279: survey of 14 nearby HAE stars for resolved circumstellar material in the
280: MIR. Nulling interferometric observations
281: with the BLINC-MIRAC instrument from the MMT and Magellan I (Baade) 6.5-m
282: telescopes are now complete. Results and an analysis of the full sample will
283: be presented in an upcoming paper. Also currently underway is a survey of
284: nearby intermediate-mass main-sequence stars for second-generation exozodiacal
285: dust, again utilizing nulling interferometry with AO. Completed
286: observations include those of Vega, presented in \citet{liu04}.
287:
288: \section{Acknowledgements}
289: W.L. was supported under a Michelson Graduate Fellowship. The authors thank
290: the staff at the MMT for excellent support. We thank B. Duffy for technical
291: support with the BLINC-MIRAC instrument. We thank the anonymous referee
292: for helpful comments. BLINC was developed under a grant
293: from NASA/JPL, and MIRAC is supported by the NSF and SAO. The MMT AO system
294: was developed with support from AFOSR.
295:
296:
297: \begin{thebibliography}
298:
299: \bibitem[Bouwman et al.(2003)]{bouwman}
300: Bouwman, J., de Koter, A., Dominik, C., \& Waters, L.~B.~F.~M.\ 2003, \aap,
301: 401, 577
302: \bibitem[Chen \& Jura(2003)]{cj03} Chen, C.~H.~\& Jura, M.\
303: 2003, \apj, 591, 267
304: \bibitem[Chiang \& Goldreich(1997)]{cg97} Chiang, E.~I.~\&
305: Goldreich, P.\ 1997, \apj, 490, 368
306: \bibitem[Hartmann, Kenyon, \& Calvet(1993)]{hart93} Hartmann,
307: L., Kenyon, S.~J., \& Calvet, N.\ 1993, \apj, 407, 219
308: \bibitem[Eisner et al.(2003)]{eis03} Eisner, J.~A., Lane,
309: B.~F., Akeson, R.~L., Hillenbrand, L.~A., \& Sargent, A.~I.\ 2003, \apj,
310: 588, 360
311: \bibitem[Elias(1978)]{elias} Elias, J.~H.\ 1978, \apj, 224,
312: 857
313: \bibitem[Fukagawa et al.(2004)]{f04} Fukagawa, M., et al.\
314: 2004, \apjl, 605, L53
315: \bibitem[Grady et al.(1999)]{grady} Grady, C.~A., Woodgate,
316: B., Bruhweiler, F.~C., Boggess, A., Plait, P., Lindler, D.~J., Clampin, M.,
317: \& Kalas, P.\ 1999, \apjl, 523, L151
318: \bibitem[Haas, Leinert, \& Richichi(1997)]{haas97} Haas, M.,
319: Leinert, C., \& Richichi, A.\ 1997, \aap, 326, 1076
320: \bibitem[Hillenbrand et al.(1992)]{hill92}
321: Hillenbrand, L.~A., Strom, S.~E., Vrba, F.~J., \& Keene, J.\ 1992, \apj,
322: 397, 613
323: \bibitem[Hinz(2001)]{hinz_phd} Hinz, P.~M.\ 2001, Ph.D.~Thesis, University of
324: Arizona
325: \bibitem[Hoffmann et al.(1998)]{hoff} Hoffmann, W.~F., Hora,
326: J.~L., Fazio, G.~G., Deutsch, L.~K., \& Dayal, A.\ 1998, \procspie, 3354,
327: 647
328: \bibitem[Kenyon \& Hartmann(1987)]{kh87} Kenyon, S.~J.~\&
329: Hartmann, L.\ 1987, \apj, 323, 714
330: \bibitem[Lada \& Adams(1992)]{la92} Lada, C.~J.~\& Adams,
331: F.~C.\ 1992, \apj, 393, 278
332: \bibitem[Leinert et al.(2004)]{leinert} Leinert, C., et al.\
333: 2004, \aap, 423, 537
334: \bibitem[Liu et al.(2004)]{liu04} Liu, W.~M., et al.\ 2004,
335: \apjl, 610, L125
336: \bibitem[Liu et al.(2003)]{liu03} Liu, W.~M., Hinz, P.~M.,
337: Meyer, M.~R., Mamajek, E.~E., Hoffmann, W.~F., \& Hora, J.~L.\ 2003, \apjl,
338: 598, L111
339: \bibitem[Mannings \& Sargent(1997)]{ms97} Mannings, V.~\&
340: Sargent, A.~I.\ 1997, \apj, 490, 792
341: \bibitem[Marsh et al.(1995)]{marsh} Marsh, K.~A., Van Cleve,
342: J.~E., Mahoney, M.~J., Hayward, T.~L., \& Houck, J.~R.\ 1995, \apj, 451,
343: 777
344: \bibitem[Meeus et al.(2001)]{meeus} Meeus, G., Waters,
345: L.~B.~F.~M., Bouwman, J., van den Ancker, M.~E., Waelkens, C., \& Malfait,
346: K.\ 2001, \aap, 365, 476
347: \bibitem[Millan-Gabet et al.(2001)]{m-g01}
348: Millan-Gabet, R., Schloerb, F.~P., \& Traub, W.~A.\ 2001, \apj, 546, 358
349: \bibitem[Miroshnichenko et al.(1999)]{miro99} Miroshnichenko, A.,
350: Ivezi{\' c} , {\v Z}., Vinkovi{\' c} , D., \& Elitzur, M.\ 1999,
351: \apjl, 520, L115
352: \bibitem[Natta et al.(2001)]{natta01} Natta, A., Prusti, T.,
353: Neri, R., Wooden, D., Grinin, V.~P., \& Mannings, V.\ 2001, \aap, 371, 186
354: \bibitem[van den Ancker et al.(1998)]{vdA98} van den Ancker, M.~E.,
355: de Winter, D., \& Tjin A Djie, H.~R.~E.\ 1998, \aap, 330, 145
356:
357: \end{thebibliography}
358:
359: \clearpage
360: \begin{deluxetable}{cccccc}
361: \tablecaption{Source Nulls for AB Aur \label{tab-abaur_source}}
362: \tablewidth{0pt}
363: \tablehead{
364: \colhead{Set \#} &
365: \colhead{Rotation} &
366: \colhead{Maj. Axis Pos.($\degr$ E of N)} &
367: \colhead{Cal. 1} &
368: \colhead{Cal. 2} &
369: \colhead{Cal. 3}
370: }
371: \startdata
372: 1 & -160 & 4 & 26.5\% & 18.9\% & 16.9\%\\
373: 2 & -135 & 170 & 20.7 & 19.1 & 11.1\\
374: 3 & -105 & 131 &14.9 & 8.1 & 5.3\\
375: 4 & -75 & 107 & 13.1 & 9.7 & 3.5\\
376: 5 & -45 & 77 & 18.9 & 12.7 & 9.3\\
377: \enddata
378: \end{deluxetable}
379:
380: \clearpage
381: \begin{deluxetable}{ccccc}
382: \tablecaption{Derived Parameters for AB Aur Disk \label{tab-sinefit}}
383: \tablewidth{0pt}
384: \tablehead{
385: \colhead{Cal. Set} &
386: \colhead{Guassian FWHM (AU)} &
387: \colhead{Incl.($\degr$)} &
388: \colhead{Ring Diameter (AU)} &
389: \colhead{Incl.($\degr$)}
390: }
391: \startdata
392: 1 & $30 \pm 3$ & $47 \pm 5$ & $34 \pm 3$ & $45\pm 5$\\
393: 2 & $27 \pm 3$ & $52 \pm 5$ & $30 \pm 3$ & $50\pm 5$\\
394: 3 & $24 \pm 2$ & $64 \pm 6$ & $28 \pm 3$ & $63\pm 6$\\
395: \enddata
396: \end{deluxetable}
397:
398:
399: \clearpage
400: \begin{figure}
401: \epsscale{0.85}
402: \plotone{f1.eps}
403: \caption{Source nulls vs. rotation of interferometer baseline derived for
404: AB Aur for each of three calibrations described above. The variation in
405: null is consistent with the presence of an inclined disk.}
406: \label{fig-abaur_source}
407: \end{figure}
408:
409: \end{document}
410: