1: %%
2: %% Millan-Gabet et al. ApJ manuscript
3: %%
4:
5:
6: %\documentclass[12pt,preprint]{/Users/rafael/aastex52/aastex}
7: %\documentclass[12pt,preprint]{/home/rafael/aastex52/aastex}
8: \documentclass[12pt,preprint]{aastex}
9:
10: %%
11:
12:
13: \begin{document}
14:
15:
16: \title{Bright Localized Near-Infrared Emission at 1--4~AU in the AB~Aurigae Disk Revealed by IOTA Closure Phases}
17:
18:
19: \author{
20: R.~Millan-Gabet\altaffilmark{1},
21: J.~D.~Monnier\altaffilmark{2},
22: J.-P.~Berger\altaffilmark{3},
23: W.~A.~Traub\altaffilmark{6,4},
24: F.~P.~Schloerb\altaffilmark{5},
25: E.~Pedretti\altaffilmark{2},
26: M.~Benisty\altaffilmark{3},
27: N.~P.~Carleton\altaffilmark{4},
28: P.~Haguenauer\altaffilmark{6},
29: P.~Kern\altaffilmark{3},
30: P.~Labeye\altaffilmark{7},
31: M.~G.~Lacasse\altaffilmark{4},
32: F.~Malbet\altaffilmark{3},
33: K.~Perraut\altaffilmark{3},
34: M.~Pearlman\altaffilmark{4},
35: N.~Thureau\altaffilmark{2}}
36:
37:
38: %\email{rafael@ipac.caltech.edu}
39:
40: \altaffiltext{1}{Michelson Science Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125, USA,rafael@ipac.caltech.edu}
41: \altaffiltext{2}{University of Michigan Astronomy Department, Ann Arbor, MI 48109-1090, USA.}
42: \altaffiltext{3}{Laboratoire d'Astrophysique de Grenoble, 414 Rue de la Piscine 38400 Saint Martin d'Heres, France}
43: \altaffiltext{4}{Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138, USA}
44: \altaffiltext{5}{University of Massachusetts at Amherst, Astronomy Department, Amherst, MA 01003, USA}
45: \altaffiltext{6}{Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA}
46: \altaffiltext{7}{LETI-CEA, Grenoble, France}
47:
48: \begin{abstract}
49:
50: We report on the detection of localized off-center emission at 1--4~AU
51: in the circumstellar environment of the young stellar object
52: AB~Aurigae. We used closure phase measurements in the near-infrared
53: made at the long baseline interferometer IOTA, the first obtained on a
54: young stellar object using this technique. When probing sub~AU scales,
55: all closure phases are close to zero degrees, as expected given the
56: previously-determined size of the AB~Aurigae inner dust disk.
57: However, a clear closure phase signal of $-3.5 \pm 0.5 \degr$ is
58: detected on one triangle containing relatively short baselines,
59: requiring a high degree of non-point symmetry from emission at larger
60: (AU-sized) scales in the disk. We have not identified any alternative
61: explanation for these closure phase results and demonstrate that a
62: ``disk hot spot'' model can fit our data. We speculate that such
63: asymmetric near-infrared emission detected might arise as a result of
64: localized viscous heating due to a gravitational instability in the AB
65: Aurigae disk, or to the presence of a close stellar companion or
66: accreting sub-stellar object.
67:
68: \end{abstract}
69:
70: \keywords{stars: formation, individual (AB Aurigae)
71: --- planetary systems: protoplanetary disks --- instrumentation:
72: interferometers --- techniques: high angular resolution --- infrared:
73: stars}
74:
75: \section{Introduction} \label{intro_sec}
76:
77: AB~Aurigae (A0~Ve, $d=144$~pc, V~=~7.1, H~=~5.9) is often referred to
78: as the prototype of the Herbig~Ae/Be (HAeBe) class of pre-main
79: sequence stars of intermediate mass. Like their solar-type analogs,
80: the T~Tauri objects, HAeBe objects are known to be surrounded
81: by pre-planetary disks of gas and dust. AB~Aur was one of the first
82: young stellar objects (YSO) to be spatially resolved in the
83: near-infrared (NIR) \citep{rmg1999}, using the technique of long
84: baseline optical interferometry. These and subsequent observations of
85: a relatively large sample of HAeBe (and T~Tauri) objects found
86: that essentially all objects have characteristic NIR sizes much larger
87: than expected from previous-generation disk models that were tuned to
88: fit the spectro-photometric data alone \citep{tuthill1999, rla2000,
89: rmg2001, rla2002, eisner2004, rla2005, eisner2005, jdm2005}. These
90: results have prompted in part a significant revision of disk models,
91: whereby previously ignored detailed physics of the disk inner edge are
92: now incorporated with some success \citep{natta2001, dullemond2001,
93: jdm2002, muzerolle2003,isella2005}. For a recent review of these
94: developments the reader is also referred to \citet{rmg2006}.
95:
96: A common characteristic of all the interferometer studies mentioned is
97: that they were based on single-baseline data, and therefore depended
98: on the interpretation of only visibility amplitudes and with very
99: sparse spatial frequency coverage. One of the unexplored frontiers in
100: optical long baseline interferometry, particularly in the area of YSO
101: science, is the measurement and interpretation of not only visibility
102: amplitudes, but also closure phases (CP), obtained at arrays
103: containing three or more telescopes. Contrary to the visiblity
104: amplitudes or phases measured for baseline pairs, the CP is
105: uncorrupted by atmospheric turbulence \citep{jennison1958,monnier2003},
106: and can be calibrated to higher precision. The CP can have
107: values different from 0 or $180 \degr$ only for a non-point-symmetric
108: sky brightness, and is therefore a highly effective probe of this type
109: of morphology. Moreover, CP and visibility amplitude data
110: may be used together to provide powerful constraints to parametric
111: models describing relatively complex morphologies. In principle, given
112: sufficient spatial frequency coverage, direct image reconstruction
113: becomes possible, as is commonly done at radio wavelengths
114: \citep{readhead1980,cornwell1981}.
115:
116: The new models of HAe objects, and of AB~Aur in particular,
117: successfully reproduce the spectral features and broadband spectral
118: energy distribution (SED) with a model in which no or little optical
119: depth exists between the central star and the dust disk inner edge,
120: such that the frontal stellar heating forms a hot wall of emission
121: from which essentially all the NIR (disk) flux arises. This model is
122: also successful at reproducing the characteristic NIR sizes measured
123: interferometrically. Moreover, the outer regions of circumstellar
124: disks are believed to have scale heights that increase with radius
125: \citep[flaring,][]{kenyon1987}. Thus, non-point-symmetric brightness
126: distributions and non-zero CPs are expected for star-disk systems such
127: as AB~Aur if they have some inclination to the observer's line of
128: sight, as the cooler foreground disk regions occult part of the hotter
129: inner disk regions \citep[see also][]{tuthill1999, malbet2001,
130: isella2005}. The original goal of the work presented here, was to
131: search for such {\it skewed} emission coming from the inner dust disk;
132: and in a forthcoming publication \citep{jdm2006} we discuss in detail
133: the results of our survey of 14 YSOs in the context of hot inner-dust
134: wall models. In this Letter, we report instead on the unexpected
135: detection of a strong CP signal that originates further out in the
136: AB~Aur disk.
137:
138: \section{Observations and Data Analysis} \label{obs_sec}
139:
140: The data presented here were obtained at the upgraded three-telescope
141: Infrared Optical Telescope Array (IOTA) \citep{traub2003}, using the
142: new integrated-optics (IO) H-band combiner IONIC-3
143: \citep{berger2003}. IOTA/IONIC-3 has been in routine operation since
144: December~2002, and first scientific results were published in
145: \citet{jdm2004} and \citet{kraus2005}.
146:
147: The three IOTA telescopes are moveable along two orthogonal linear
148: arms; telescopes A and C can move along a 35~m north-east arm, while
149: telescope B moves along a 15~m south-east arm. This allows an aperture
150: of 35~m~$\times$~15~m to be synthesized, giving a maximum physical
151: baseline $B = 38$~m. For the AB~Aur declination ($\delta = 30 \degr$),
152: the projected baselines are in the range $B_p = 10 - 38$~m, providing
153: spatial resolution $\sim \lambda_0/(2 \, B_p) = 4 - 17$~mas (or 0.6 --
154: 2.4~AU) in H-band ($\lambda_0 = 1.65 \mu m$). The IONIC-3 optical
155: circuit acts to split the light from each telescope before recombining
156: each telescope pair (AB, BC, AC) at three IO couplers, leading to six
157: interferometric channels (two for each baseline).
158:
159: The work presented here is based on observations made over the period
160: December~2002 $-$ December~2004, comprising a total of 18~epochs,
161: 14~two-telescope configurations (visibility amplitude data) and
162: 6~three-telescope configurations (CP data). As is standard practice,
163: observations of AB~Aur were interspersed with observations of nearby
164: calibrator stars to measure slowly varying system visibilities and
165: CPs. We used the calibrator stars HD~32406 (K0III, V~=~6.2, H~=~3.6) and
166: HD~31233 (K0, V~=~7.2, H~=~4.6); of uniform-disk angular diameters $1.2 \pm
167: 0.7$~mas and $0.6 \pm 0.4$~mas respectively. The observations
168: presented here were made using a standard H-band filter ($\lambda_0 =
169: 1.65 \, \mu m$, $\Delta \lambda = 0.30 \, \mu m$) and in un-polarized
170: light.
171:
172: Reduction of the visibility amplitude data was carried out using a
173: custom implementation of the method outlined by \citet{foresto1997}
174: and already described and demonstrated in \citet{jdm2004}. To ensure
175: a good CP measurement, we required that interferograms be detected on
176: $\geq 2$ baselines, a condition nearly always maintained by a
177: real-time fringe packet ``tracker" \citep{pedretti2005}. Based on
178: studies of our absolute night-to-night calibration accuracy, we adopt
179: a typical systematic error of $\sigma_{V^2} = 0.05$ (e.g., 2.5\% error
180: in visibility for unresolved sources); this error has been combined
181: (in quadrature) with the statistical error for plotting and fitting
182: purposes. We followed the method of \citet{baldwin1996} for
183: calculating the complex triple amplitude in deriving the CP,
184: explicitly guaranteeing ``fringe frequency closure" ($\nu_{AB} +
185: \nu_{BC} + \nu_{CA} = 0$). Chromaticity effects are the main
186: limitation to our absolute precision when the calibrator and source
187: are of very different spectral types because of different effective
188: wavelengths. While we minimize these errors by using calibrators of
189: spectral type relatively similar to that of the target, a conservative
190: systematic error of $\sigma_{\mbox{CP}} = 0.5 \degr$ has been adopted
191: here. Further details on our data reduction and calibration methods
192: are given in \citet{jdm2006}. The $V^2$ and CP data have been
193: averaged in the 2-D spatial frequency $(u,v)$ plane using an averaging
194: interval of 4~m, resulting in 75 and 37 independent $V^2$ and CP
195: measurements, respectively. The calibrated $V^2$ and CP data are
196: available from the authors\footnote{The data can also be found in the
197: data archives section of the Optical Long Baseline Interferometry News
198: (OLBIN) web site, {\tt http://www.olbin.jpl.nasa.gov}} in the OI-FITS
199: format \citep{pauls05}.
200:
201: The $V^2$ and CP data are shown in
202: Figures~\ref{v2a_fig}~and~\ref{cp_fig}, along with models discussed in
203: the next section. The main characteristics of our AB~Aur data which
204: underlie the principal results presented in this Letter are as
205: follows: (1) A definite CP signal is detected ($-3 \degr$ to $-4
206: \degr$), which surprisingly appears in a triangle of relatively short
207: baselines (triangle A25B15C10)\footnote{The notation AxxByyCzz
208: indicates an IOTA configuration in which telescope A is located at the
209: station which is xx meters from the array origin, located at the
210: corner of the L-shaped track, and similarly for telescopes B and C},
211: indicating some degree of non-point-symmetry on relatively large
212: spatial scales (1--4~AU, see $\S$~\ref{results_sec}). (2) The CP data
213: are consistent with zero on all other baseline triangles
214: (Figure~\ref{cp_fig}), indicating a high degree of point-symmetry on
215: sub-AU spatial scales for AB~Aur. (3) In spite of including baselines
216: of varying orientations, the curve traced by the $V^2$ {\it vs.}
217: baseline length data (Figure~\ref{v2a_fig}) is relatively smooth,
218: indicating that the NIR brightness is also largely circularly
219: symmetric on small scales (consistent with previous interferometer
220: observations). (4) The $V^2$ {\it vs.} baseline length curve traced
221: by the data does not appear to extrapolate to 1.0 at zero spatial
222: frequency, implying the presence of an extended component contributing
223: a small fraction of the H-band flux ($\sim 8$\% if it were completely
224: resolved by the shortest IOTA baselines, $B_p = 9$~m). We note that of
225: the 14 YSOs in our CP survey \citep{jdm2006}, AB~Aur is unique in
226: showing non-zero CP simultaneous with such high visibilities.
227:
228: \section{Modelling and Results} \label{results_sec}
229:
230: The $(u,v)$ coverage obtained in these observations is relatively
231: dense, unprecedented for YSO observations using an optical
232: interferometer. However, it still contains significant gaps, a
233: pronounced north-south bias, and even for the longest IOTA baselines
234: AB~Aur appears only partially resolved ($V^2_{min} \sim 0.4$); all of
235: which frustrate attempts to directly reconstruct a meaningful image
236: from the $V^2$ and CP data.
237:
238: We have thus turned to {\it parametric imaging} in order to interpret
239: our observations. Motivated by the inner dust-wall models introduced
240: in $\S$~\ref{intro_sec} we begin by modelling the AB~Aur NIR
241: brightness as central star plus a Gaussian ring, of inner diameter
242: $D_{ring}$ and width $W_{ring}$ (expressed as fraction of the inner
243: ring diameter). We use previous results based on SED decomposition
244: which have established that the central star in AB~Aur contributes
245: $\sim 30$\% of the H-band flux and is unresolved (point-like) to the
246: IOTA baselines \citep{rmg2001} -- the exact stellar fractional flux is
247: not relevant to the conclusions of this Letter. Although the
248: visibility amplitude data allow a direct measurement of the ring
249: diameter, they do not constrain the ring width well, and we follow
250: previous workers in adopting a 25\% fractional width for the ring
251: (also not critical to our conclusions). Fits of elliptical rings
252: (representing inclined inner disks) to the $V^2$ data find near-unity
253: major axis/minor axis ratios, i.e., small disk inclinations $\la 30
254: \degr$ consistent with previous results from near-IR interferometry
255: \citep{rmg2001,eisner2004}, scattered light images
256: \citep{grady1999,fukagawa2004} and mm interferometry
257: \citep{corder2005}. We therefore adopt a face-on orientation. From
258: the best fit star$+$ring model shown in Figure~\ref{v2a_fig} it can be
259: seen (as mentioned above) that the curve does not extrapolate to $V(0)
260: = 1.0$, and therefore the model allows for a fraction $f_i$ of the
261: total flux to be completely incoherent (i.e., large compared to the
262: angular resolution of the shortest baselines). The best-fit solution
263: (to $V^2$ only, resulting in $\chi^2_{\nu}[V^2] = 0.9$) has the
264: following parameters: $D_{ring} = 3.1 \pm 0.1$~mas
265: (0.39~$\pm$~0.02~AU), $f_i = 0.08 \pm 0.01$; and is shown with dashed
266: lines in Figures~\ref{v2a_fig}~and~\ref{cp_fig}. This
267: value for the ring diameter is in good agreement with recent
268: interferometer measurements \citep{eisner2004}.
269:
270: The above model describes some of the main features of the AB~Aur
271: system (number of components needed, characteristic size and
272: shape). However, in order to also fit the non-zero CPs, we need to
273: consider non-symmetric models. As described in $\S$~\ref{intro_sec},
274: flared disks with non-zero inclination are expected to evince
275: occultation the near side of the inner disk, resulting in a
276: non-symmetric brightness. However, as just discussed, the inclination
277: of the AB~Aur disk is believed to be low. Moreover, in order to
278: measure the CP resulting from asymmetries in the inner disk, the
279: emission has to be well resolved. As shown in \citet{jdm2006} the CP
280: is strongly suppressed when, as in the AB~Aur case, the inner disk
281: diameter (3.1~mas) is less than 1/2 the fringe spacing corresponding
282: to the longest baseline in a triangle. Here we have confirmed this
283: result by introducing an azimuthal cosine modulation of the Gaussian
284: ring in order to approximate the morphology of a partially occulted
285: inner disk (see \citet{jdm2006} for details of these {\it skewed ring}
286: models), and found that it can produce at most a signal of
287: $\mbox{CP[\mbox{A25B15C10}]} \sim -2 \degr$ when the solutions are
288: constrained to reproduce (even approximately) the $V^2$ data and the
289: zero CP measured in the other baseline triangles.
290:
291: We therefore must add an additional source component at larger spatial
292: scales which introduces additional skewness to the model. In order to
293: quantify the level of asymmetry that is required, we have adopted a
294: simple Gaussian profile which is offset with respect to
295: the central disk emission (``Disk Hot Spot'' model).
296: The new component is thus characterized by its full-width at
297: half-maximum ($W_G$), location (angular separation $r_G$ and position
298: angle $\mbox{PA}_G$) and the fractional flux ($f_G$) it
299: contributes. Fitting this simple model we find that solutions exist
300: that reproduce both the $V^2$ and CP data reasonably well. However, we
301: emphasize that several solutions exist which are statistically
302: equivalent, and although we can assign definite spatial scales and
303: fractional fluxes to the offset Gaussian, we are not able to assign a
304: definite direction to the asymmetry (e.g., there are reasonable solutions
305: with northern as well as southern offsets). We have included
306: the best-fit parameters associated with a few of the best solutions
307: in Table~\ref{fits_tab} for reference.
308:
309: Quantitatively, the solutions found are defined by the following
310: characteristics: (1) when the offset Gaussian feature is allowed to be
311: spatially extended, the angular separations are $r_G \simeq 30$~mas
312: (4.3~AU), width $W_G \simeq 12$~mas (1.7~AU) and fractional flux $f_G
313: \simeq 8$\% (we also note that in this case slightly better fits are
314: found when the inner ring is also allowed to be skewed, rather than
315: symmetric); and (2) for the special case that the offset component is
316: unresolved (i.e., a point source companion), closer separations $r_G
317: \simeq 9$~mas (1.25~AU) and lower fractional fluxes $f_G \simeq 2$\%
318: are found (in this case with no benefit from additional skew in the
319: inner ring). An example fit of the latter case is shown (solid lines)
320: in Figures~\ref{v2a_fig}~and~\ref{cp_fig}. The example
321: shown is typical of the quality of the fits obtained
322: ($\chi^2_{\nu}(\mbox{[$V^2$,CP]} = 1.5$) and corresponds to the
323: parameters: $f_G = 0.02$, $r_G = 9$~mas (1.4~AU) and $\mbox{PA}_G = 22
324: \degr$. As can be seen, although the solution is not entirely
325: adequate on all baseline pairs and triangles, indicating the need for
326: a more complex model, this simple model is capable of reproducing most
327: data well, in particular the observed $\mbox{CP[\mbox{A25B15C10}]} =
328: -4 \degr$.
329:
330: \section{Discussion}
331:
332: The principal result of this Letter is the clear detection of an
333: unexpected localized asymmetry in the circumstellar environment of
334: AB~Aur at spatial scales of 1--4~AU; this is an {\it intermediate} scale
335: between the the inner wall containing the hottest dust and the outer
336: disk.
337:
338: We first considered that the physical origin of this emission could be
339: H-band scattering off the disk atmosphere or ``halo''. The required
340: asymmetry could easily arise from a small photocenter shift due to
341: e.g., forward scattering in combination with a small disk
342: inclination. However, this emission would arise from large angular
343: scales (up to $\sim 1 \arcsec$) and would be completely resolved by
344: even the shortest IOTA baselines, and therefore could not result in a
345: CP signal. Indeed, the model fitting described above shows that the
346: offset Gaussian must be relatively compact ($\lesssim 12$~mas or
347: 1.7~AU), and it is difficult to reconcile the required amount of flux
348: with scattering over such a small area. Due to the low inclination
349: known for the AB~Aur disk, we also do not favor outer disk occultation
350: or opacity effects.
351:
352: A more likely explanation, is that the asymmetry is due to compact
353: thermal emission. The presence and persistence of such a local
354: inhomogeneity in the AB~Aur disk at these spatial scales would however
355: be surprising, given that the dynamical time-scales are very short
356: (1--5~years) compared to the age of the system ($\sim 1-4$~Myr). A
357: possible explanation is that the disk is not steady state, and the
358: feature corresponds to a gravitational instability, possibly
359: associated with an inner extension of the spiral density waves seen in
360: near-IR coronographic and millimiter interferometry images
361: \citep{fukagawa2004,pietu2005}. The existence of a disk instability
362: at few~AU radii and for a disk of this age clearly has important
363: implications for the formation of planets in these disks.
364:
365: Alternatively, we could be detecting a companion embedded in the
366: AB~Aur disk.\footnote{We note that \citet{baines2006} report
367: ``spectro-astrometric'' evidence for binarity in AB~Aur (and several
368: other HAeBe objects). However, they infer large separations of $0.5 -
369: 3.0 \arcsec$; therefore this putative companion can not be related to
370: the CP signatures modelled here.} A stellar companion at the few--AU
371: separation we infer is however unlikely, given that a substantial
372: circumstellar disk has in fact survived around AB~Aur. The intriguing
373: possibility therefore arises that the companion is instead a
374: proto-planetary object forming in the disk.
375:
376: Similar claims (but based on visibility amplitudes rather than the
377: more robust CP) for a disk thermal instability or close companion have
378: been recently made for the active-disk system FU~Orionis
379: \citep{malbet2005}. We also note the possible connection between the
380: extended structure detected here, and the intriguing compact halo
381: (45~mas FWHM) seen in the T~Tauri object DG~Tau in lunar occultation
382: observations by \citep{leinert1991}. As discussed by \citet{jdm2006},
383: disk ``halos'' appear to be common (although not the norm)
384: in young star~$+$~disk systems, and deserve new observational and modelling scrutiny.
385:
386: Our hypothesis is readily testable. If the non-zero CP arises in a
387: ``disk hot spot,'' variability in the CP signal over 1--5~years should
388: reveal the orbital motion that could be followed using short or long
389: baselines from the VLTI AMBER \citep{malbet2004} or MIDI
390: \citep{leinert2003} instruments, respectively. The angular scales
391: involved also make this object a prime target for direct imaging using
392: high resolution techniques on large single-aperture telescopes.
393:
394: \acknowledgments
395:
396: The authors gratefully acknowledge support from SAO, NASA (for third
397: telescope development \& NNG05G1180G), the NSF (AST-0138303,
398: AST-0352723, for work on imaging with IOTA), and the Jet Propulsion
399: Laboratory (JPL awards 1236050 \& 1248252). The IONIC3 instrument has
400: been developed by LAOG and LETI in the context of the IONIC
401: collaboration (LAOG, IMEP, LETI). The IONIC project is funded by the
402: CNRS (France) and CNES (France).
403:
404: \begin{thebibliography}{}
405:
406: \bibitem[Akeson et al.(2000)]{rla2000} Akeson, R. L., Ciardi, D. R.,
407: van Belle, G. T., Creech-Eakman, M. J., and Lada, E. A. 2000, \apj,
408: 543, 313
409:
410: \bibitem[Akeson et al.(2002)]{rla2002} Akeson, R.~L., Ciardi, D.~R.,
411: van Belle, G.~T., \& Creech-Eakman, M.~J.\ 2002, \apj, 566, 1124
412:
413: \bibitem[Akeson et al.(2005)]{rla2005} Akeson, R.~L., Walker, C.H.,
414: Wood, K., Eisner, J.A., Scire, E., Penprase, B., Ciardi, D.~R., van
415: Belle, G.~T., Whitney, B., and Bjorkman, J.E., 2005, \apj, 622, 440
416:
417: \bibitem[Baines et al.(2006)]{baines2006}
418: Baines, D. et al. 2006, \mnras, 367, 737
419:
420: \bibitem[Baldwin(1996)]{baldwin1996}
421: Baldwin, J. E. et al. 1996, \aap, 306, L13
422:
423: \bibitem[Berger et al.(2003)]{berger2003}
424: Berger, J-P. et al. 2003, Proc. SPIE, 4838, 1099
425:
426: \bibitem[Corder, Eisner \& Sargent(2005)]{corder2005}
427: Corder, S., Eisner, J. \& Sargent, A., 2005, \apj, 622, 133L
428:
429: \bibitem[Cornwell \& Wilkinson(1981)]{cornwell1981}
430: Cornwell, T. J. \& Wilkinson, P. N. 1981, \mnras, 196, 1067
431:
432: \bibitem[Coud\'{e} du Foresto, Ridgway \& Mariotti(1997)]{foresto1997}
433: Coud\'{e} du Foresto, V., Ridgway, S. \& Mariotti, J. M. 1997, A\&AS, 121, 379
434:
435: \bibitem[Dullemond, Dominik \& Natta(2001)]{dullemond2001} Dullemond,
436: C. P., Dominik, C., Natta, A. 2001, \apj, 560, 957
437:
438: \bibitem[Fukagawa et al.(2004)]{fukagawa2004}
439: Fukagawa, M. et al. 2004, \apj, 605, L53
440:
441: \bibitem[Grady et al.(1999)]{grady1999}
442: Grady, C. A. et al. 1999, \apj, 523, L151
443:
444: \bibitem[Eisner et al.(2004)]{eisner2004} Eisner, J.A., Lane B.F.,
445: Hillenbrand, L.A., Akeson, R.L. \& Sargent, A.I., 2004, \apj, 613,
446: 1049
447:
448: \bibitem[Eisner et al.(2005)]{eisner2005} Eisner, J.A., Hillenbrand,
449: L.A., White, R.J. Akeson, R.L. \& Sargent, A.I. 2005, \apj, 623, 952
450:
451: \bibitem[Isella \& Natta(2005)]{isella2005}
452: Isella, A. \& Natta, A. 2005, \aap, in press
453:
454: \bibitem[Jennison(1958)]{jennison1958}
455: Jennison, R. C. 1958, \mnras, 118, 276
456:
457: \bibitem[Kenyon \& Hartmann(1987)]{kenyon1987}
458: Kenyon, S. J. \& Hartmann, L. 1987, \apj, 323, 714
459:
460: \bibitem[Kraus et al.(2005)]{kraus2005}
461: Kraus, S. et. al. 2005, \aj, 130, 246
462:
463:
464: \bibitem[Leinert et al.(1991)]{leinert1991} Leinert, Ch. et al. \aap,
465: 250, 407
466:
467: \bibitem[Leinert et al.(2003)]{leinert2003} Leinert, Ch. et al. 2003
468: \apss, 286, 1, 73-83.
469:
470: \bibitem[Malbet et al.(2004)]{malbet2004} Malbet F. et al. 2004, in
471: {\em Proceedings of the SPIE}, 5491, 1722.
472:
473: \bibitem[Malbet et al.(2005)]{malbet2005}
474: Malbet, F. et al. 2005, \aap, 437, 627
475:
476: \bibitem[Malbet, Lachaume \& Monin(2001)]{malbet2001}
477: Malbet, F., Lachaume, R. \& Monin, J. L. 2001, \aap, 379, 515
478:
479: \bibitem[Millan-Gabet et al.(1999)]{rmg1999} Millan-Gabet, R.,
480: Schloerb, F.~P., Traub, W.~A., Malbet, F., Berger, J.~P., \& Bregman,
481: J.~D.\ 1999, \apjl, 513, L131
482:
483: \bibitem[Millan-Gabet, Schloerb, and Traub (2001)]{rmg2001}
484: Millan-Gabet, R., Schloerb, F. P., and Traub, W. A, 2001, \apj, 546,
485: 358
486:
487: \bibitem[Millan-Gabet et al.(2006)]{rmg2006} Millan-Gabet, R. et al.,
488: 2006, in Protostars and Planets V, in press
489:
490: %\bibitem[Monnier(2000)]{jdm2000} Monnier, J. D. 2000,
491: %in principles of Long Baseline Stellar Interferometry,
492: %ed. P. R. Lawson (Pasadena: NASA/JPL/Caltech), 203
493:
494: \bibitem[Monnier(2003)]{monnier2003} Monnier, J. D. 2003, Reports on Progress
495: in Physics, 66, 789
496:
497: \bibitem[Monnier \& Millan-Gabet(2002)]{jdm2002} Monnier, J. D. \& Millan-Gabet,
498: R. 2002, \apj, 579, 694
499:
500: \bibitem[Monnier et al.(2004)]{jdm2004}
501: Monnier, J. D. et al. 2005, \apj, 602, L57
502:
503: %\bibitem[Monnier et al.(2004b)]{jdm2004b}
504: %Monnier, J. D. et al. 2004, Proc. SPIE, 5491, 1370
505:
506: \bibitem[Monnier et al.(2005)]{jdm2005} Monnier, J. D. et
507: al. 2005, \apj, 624, 832
508:
509: \bibitem[Monnier et al.(2006)]{jdm2006} Monnier, J. D. et al. 2006,
510: \apj, in press
511:
512: \bibitem[Muzerolle et al.(2003)]{muzerolle2003} Muzerolle, J., Calvet,
513: N., Hartmann, L., D'Alessio, P. 2003, \apj, 597, 149
514:
515: \bibitem[Natta et al.(2001)]{natta2001} Natta, A.; Prusti, T., Neri,
516: R., Wooden, D., Grinin, V. P., Mannings, V. 2001, \aap, 371, 186
517:
518:
519: %\bibitem[Liu et al.(2005)]{liu2005}
520: %Liu, M. et al. 2005, \apj, 618, L133
521:
522: %\bibitem[Pauls et al.(2004)]{pauls04} Pauls, T., Young, J. S., Cotton,
523: %W. D., Monnier, J. D. 2004, Proc. SPIE, 5491, 1231
524:
525: \bibitem[Pauls et al.(2005)]{pauls05} Pauls, T., Young, J. S., Cotton,
526: W. D., Monnier, J. D. 2005, \pasp, in press
527:
528: %\bibitem[Pedretti et al.(2004)]{pedretti2004}
529: %Pedretti, E. et al. 2004, \pasp, 116, 377
530:
531: \bibitem[Pedretti et al.(2005)]{pedretti2005}
532: Pedretti, E. et al. 2005, Applied Optics, in press
533:
534: \bibitem[Pietu, Guilloteau \& Dutrey(2005)]{pietu2005}
535: Pietu, V., Guilloteau, S. and Dutrey, A., 2005, \aap, 443, 945
536:
537: \bibitem[Readhead et. al.(1980)]{readhead1980}
538: Readhead, A. C. S.; Walker, R. C.; Pearson, T. J.; \& Cohen, M. H. 1980,
539: Nature, 285, 137
540:
541: %\bibitem[Tatulli et al.(2004)]{tatulli2004}
542: %Tatulli, E. et al. 2004, Proc. SPIE, 5491, 117
543:
544: \bibitem[Traub et al.(2003)]{traub2003}
545: Traub, W. A. et al. 2003, Proc. SPIE, 4838, 45
546:
547: \bibitem[Tuthill, Danchi \& Monnier(1999)]{tuthill1999} Tuthill, P. G.; Monnier,
548: John D.; Danchi, William C. 1999, Nature, 398, 487
549:
550: %\bibitem[Vinkovi\'{c} et al.(2006)]{vinkovic2006} Vinkovi\'{c} D.,
551: %Ivezi\'{c} \v{Z}., Jurki\'{c} T., and Elitzur M. 2006, \apj, 636, 348
552:
553: \end{thebibliography}
554:
555: \clearpage
556: \input{tab1}
557:
558: \clearpage
559:
560: \begin{figure}
561: \begin{center}
562: \includegraphics[angle=0,scale=0.7]{f1.eps}
563: \caption{
564: Calibrated $V^2$ data as a function of projected baseline length; and
565: comparison to the two models discussed in the text. For the
566: asymmetric model, the plot shows cuts in the $(uv)$ plane along two
567: distinct directions that encompass most of the AB~Aur data. As
568: discussed in the text, a unique solution for the asymmetric model is
569: not well constrained; in order to illustrate the geometry the inset
570: image ($30 \times 30$~mas) shows the example solution for which the
571: model curves are calculated.
572: \label{v2a_fig}}
573: \end{center}
574: \end{figure}
575:
576: \clearpage
577:
578: \begin{figure}
579: \begin{center}
580: \includegraphics[angle=90,scale=0.7]{f2.eps}
581: \caption{Calibrated CP data as a function of hour angle; and comparison
582: to the models discussed in the text.
583: \label{cp_fig}}
584: \end{center}
585: \end{figure}
586:
587: %%
588:
589: \end{document}
590: