Estimate of the envelope mass of the star-planet system GQ LupA,b

Coordinator: K. Schreyer, R. Neuhauser, G. Wuchterl, E. W. Guenther, K Tachihara

We intent to determine the gas content towards the classical T Tauri star-planet system GQLup. Using VLT, Subaru & HST, a co-moving companion 0.7 to west of GQ Lup was found.In terms of theoretical calculations, the mass of the object was determined to be only fewJupiter masses based on spectroscopic results. Thus, this object would be one of the firstplanet directly imaged. Due to the rather high luminosity of the planetary candidate GQLup b, the chemical composition of the star, and the near- & mid-infrared excess, we expectthat the system is very young 2 Myr and a sufficient amount of gas is still present in theenvelope for that we intend to search for answering fundamental questions of planet formationprocesses.

Program is available and data products can be downloaded

Scientific Background

By means of direct imaging and spectroscopy we have recently identified an object of only afew MJupiter orbiting the classical T Tauri star GQ Lupi (Neuhauser et al. 2005, Fig. 1). The object is located at a projected separation of 0.7 arcsec, corresponding to at least 100 AU (D= 140 pc). The fact that GQ Lupi has a disk and is orbited by an object with a mass of onlya few MJupiter allows for the first time to study the formation of planets directly.
The crucial question is, how, and why this object formed. The answer to this question mustcertainly be connected to the properties of the disk and its envelope which we would like tostudy in more detail in this project. The previous knowledge on the GQ Lupi system canbe summarized as follows: The IRAS maps (Fig. 4) and the 13 CO and HCO+ large-scalesurveys by Tachihara et al. (1996; 2001) show the presence of a gaseous and dusty envelopewhich is not well studied yet. By combining the (v×sini) and the photometric rotation period(unpublished) for the star itself, we derived an inclination of i = 35 ± 10o for the rotation axisof the disk. We can safely assume that the inclination of the disk is likely to be about thesame, as the inclination of the rotation axis of the star, like in Eps Eri or the solar system.
Since the projected separation of GQ Lupi b is 100 AU, the general question we would liketo answer is whether objects formed in the disk at such a large distance, or whether it formedcloser in, and was then kicked outwards. In the first case, the disk has to extend up to 100AU and more, and a significant amount of gas and dust has to be present in the disk at sucha large distance from the star. In the second case, the orbit has to be highly eccentric. In thiscase the disk is truncated at the distance of the periastron of the planet, because of the violentintegration of GQ Lupi b with the disk at periastron. If we assume that the specific angularmomentum of GQ Lupi b is the same as that of Jupiter, the eccentricity would have to be0.95 so that the apastron is at 100 AU. This value it not unreasonable as other exo-planetshave orbits of similar eccentricity (e.g. HD 80606 b has an eccentricity of 0.927; Naef et al.2001, A&A 375, L27). If the eccentricity is 0.95, the periastron distance of GQ Lupi b will beonly a few AU, and the disk truncated at this radius. Even if the companion will later turnout to be a low-mass brown dwarf, the study of its physical relationship to its disk and thedisk envelope is very important.
Using the IRAS data and the flux measurement at 1.3 mm, Nuernberger et al. (1997) estimatedthe total mass of the disk as 4.3 10 M-3 . However, based on the large-scale IRAS and COmaps as well as on the modelling of the spectral energy distribution (Fig. 2) using e.g. the 2Dradiative transfer code (e.g. Manske & Henning 1999), we do not assume that the whole massis located inside a disk of 100 AU. We propose that a sufficient amount of material is stillplaced in the envelope for which we intend to search.

Scientific Aims
In order to study the morphology of the envelope gas around this star-planet system (R 10 000 AU), we propose to observe the CO 3­2 line at 345 GHz with a raster map of 25 pointscentered on the source. With these observations, we are sensitive to spatial scales of the halfbeam size of 9 ...10 (= 1300 ...1500 AU).
The first question we would like to answer is the mass of the envelope. With the previousmeasurements, the dust-mass of the disk was constrained only by one measurement at 1.3mm.As pointed out by Lineweaver & Grether (2004) about 90% of the stars do not have Jupiter-like planets. The most natural explanation for this that in 90% of the stars, the disk-massis simply to small to form planets. However, the disks are feeded by their envelopes. Thus,if the envelope less massive the disks does likely not develop Jupiter-like planets, too. SinceGQ Lupi apparently did form a massive companion, the disk and the envelope must be moremassive than that of other young stars, or at least the mass distribution must be different.
However, this is in contradiction with the previous estimate of the dust-mass. However, themeasurement at 1.3mm by Nuernberger et al. (1997) was perfomed only in the OFF-ON mode (SEST beamsize 24 ) thus that no map with reasonable spatial resolution is available. Thuswe can really find out if it was possible that the GQ Lupi b formed at 100 AU or not.
In the evolutionary sequence of planetary systems, the disks lose first their gas contents andthan their dust (becoming invisible in the mm continuum emission). With the proposed COobservations we intend to find out if there is gas still present in the area of the disk on scales of(1500 AU) or not. If there gas we can conclude that the disk is very young and the formationof planets is more rapid as previously thought. The line profile will indicate the origin of theemission. Based on the fact that the star is visible, the density of the envelope gas should notbe higher than 104 cm-3 making a large amount of self-absorption improbable. Therefore, thedetection of a double-line profile as expected for Keplerian motion in the disk (see example inFig. 3) can indicate that gas is still present in the disk. Using an available 2D line radiativetransfer code (Semenov et al. 2005), we can estimate the mass of the disk itself, and we can getfurther constraints for the inclination. Otherwise, the detection of "only" a single Gaussianline profile towards the central source will indicate that the (inner, not detectable) disk is stillsurrounded by a sufficient amount of gas placed in a low-density envelope.
In addition, this system would allow to study the density gradient in the envelope which isalready more evolved than deeply embedded young stellar objects. Furthermore, we wouldlike to answer whether it is possible to detect relics of an molecular outflow or if an outflowis still present? This would give a convincing argument that this system is really very youngand the planet formation occurs very fast.
Again these observations helps us understanding whether it is possible that GQ Lup b formedat 100 AU, or not. By studying the properties of the disk and the closer envelope of GQLupi, we will thus learn much more how planets are formed in circumstellar disks of youngstars. We know that with the beamsize of 18 we will not be able to resolve spatial detailsof the disk itself. But these measurements will be a first step to study the properties of theenvironment of such a kind of object.

Technical detatils and time estimate
We propose to observe the spectral line CO 3-2 at 345 795.990 MHz with a raster map of 25points centered at the source itself (see Fig. 3). The source velocity is vlsr 6.5 km s-1 knownfor the surrounding Lupus 1 cloud emission (Tachihara et al. 1996, 2001). We expect a typicallinewidth of v = 1...3 km s-1 .
We propose to observe in total 4...5 min on each ON position. The number ONs perOFF and the OFF integration time will depend on the weather conditions. In addition, themapping strategy: short ONs integration times and more recurrences of the raster map will aswell depend on the weather. It would be possible to start with an inner 3 x 3 raster map witha point separation of 18 (beamsize), and later it can be zoomed out with larger separationsas shown in Fig. 3. A suitable OFF position would be located 2000 to the west.
The pure ON integration time would be 25 x 5 min = 125 min. Assuming three ONs per oneOFF ( 10 min OFF integration time), we need seven OFFs in total 70 min. Thus, we wouldneed 3.25 h pure integration time. Adding 100% of the pure integration time for pointing,calibration and telescope movement, we estimate in total 6.5 h observing time for thisproject.

Lineweaver, C. H. & Grether, D., 2004, DPS 36.3508
Manske, V. & Henning, Th., 1999, A&A 349, 907
Neuhauser, R., Guenther, E.W., Wuchterl, G., Mugrauer, M., Bedalov, A., Hauschildt, P.H., 2005, A&A 435,L13
Nuernberger, D., Chini, R., Zinnecker, H., A&A 324, 1036
Schreyer, K., Henning, Th., van der Tak, F. F. S., Boonman, A. M. S., van Dishoeck, E. F., 2002, A&A 394,561
Semenov, D., Pavlyuchenkov, Ya., Schreyer, K., et al., 2005, ApJ 621, 853
Tachihara, K., Dobashi, K., Mizuno, A., Fukui, Y., 1996, PASJ 48, 489
Tachihara, K., Toyoda, S., Onishini, T., Mizuno, A., Fukui, Y., 2001, PASJ 53, 1081
Wolf, S., & D'Angelo, G., 2005, ApJ 619, 1114