Protostars and Planets V 2005 8398.pdf

GW ORIONIS: A T-TAURI MULTIPLE SYSTEM OBSERVED WITH AU-SCALE RESOLUTION. J. P.Berger, Laboratoire d’Astrophysique de Grenoble,, BP-53, F-38041 Grenoble Cedex, France,[email protected], J. Monnier, E. Pedretti, University of Michigan, , Ann Arbor, MI 48109-1090. USA , R. Millan-Gabet, California Institute of Technology, Pasadena, CA 91125, USA, F. Malbet, K. Perraut, P. Kern, M. Benisty, P. Haguenauer, Laboratoire d’Astrophysique de Grenoble, F-38041 Grenoble Cedex, France, P. Labeye, CEA-LETI 38054 Grenoble, Cedex, France, W. Traub, N. Carleton, M. Lacasse, Harvard Smithsonian Center for Astrophysics, Cambridge, MA 02138, USA, S. Meimon, ONERA, Chatillon, France, C. Brechet, E. Thiebaut, CRAL, Lyon, France, P.Schloerb, University of Massachusetts at Amherst, Astronomy Department, Amherst, MA 01003, USA.

GW Orionis is a well known single-line spectroscopic bi- 3 instrument which allows to measure simultaneously 3 vis- nary classified as a T Tauri . The measured period is ≈ 242 ibilities and one closure phase (Monnier et al. 2004). The days. The are separated by ≈ 1.1AU and have a nearly addition of several measurements at different hour angle and circular . The analysis of the residuals have revealed the two IOTA configuration allowed a map of the (u,v) plane suf- signature of a putative third companion with ficient to carry out the first reconstruction of an image of a T ≈ 1000 days. Tauri star with resolution (see Figure 1). The combination of spectroscopic measurements and spec- A detailed analysis of visibilities and closure phases is tral energy distribution modelization has lead Mathieu et al.(1991) however preferable if one is to quantify the system parameters to describe GW Orionis as a primary star surrounded with a cir- with a certain accuracy (see Figure 2). cumstellar disk while the companion tidally carves a gap in the These observations leads us to conclude that, indeed, there disk creating a circumbinary dust distribution. The constrain is a third component to the system at a distance of 20mas. and on the gap comes mainly from the dip in the SED. Artymowicz responsible for roughly 20% of the H band flux. Strong resid- and Lubow (1994) used these observations to test the scenario uals in the visibility+closure phase fitting are a indication that of gap opening in a circumbinary disk. we might be also resolving the second star (the one detected by spectroscopy). Two observations epochs show a clear motion of the third component. These observations should lead to a major rethinking of the GW Orionis system. In particular the dust distribution geometry has to be rethought. The major interest of such a work will be to combine spectroscopic, interferometric and photometric data in order to constrain the model. A second step of this work will certainly be to assess the long-term stability of such a complex system by complementing the observations and modelling the dynamical evolution. Figure 1: First attempt at reconstructing an image of GW Ori- onis using IOTA/IONIC-3 in Dec 2004.

The complexity of the system is further enhanced by the presence of periodic dimmings in the UBVR photometry. Shevchenko et al. reported that this phenomenon had a pe- riod of 242 days but lasted only from 1987 to 1992 after when the dimmings disappeared. On the other side DeWarf et al. re- port UBV photometry carried between 1992 and 1997 which show cyclic variability of the photometry of period ≈ 1100 days which they attribute to the third companion marginally detected by Mathieu et al.(1991). Marginal detection of a period of 242 days is also reported. Figure 2: Visibilities and closure phase measured on GW Ori We present the first direct confirmation of the presence in Nov 2003. of a third companion to the system. We used the infrared long-baseline interferometer IOTA and its focal instrument IONIC-3 to observe GW Orionis at References two epochs. IOTA offers the possibility to move three 40cm telescopes on a grid of baselines going from 5m to 35 meters Mathieu et al., 1991, AJ, 101,2184 and organised on a L-shaped configuration. In the best case Shevchenko et al., 1998, AL, 24,528 the observer can therefore synthetize an aperture of roughly DeWarf et al.,2002, BAAS, 34,1134 15×38 meters leading to a resolution of 5 mas in the maximum Artymowicz and Lubow, 1994, ApJ, 651 baseline configuration in the H band. Monnier et al. 2004, ApJ, 602, L57-60 The observations we report here were made with the IONIC-