A Tail Structure Associated with a Protoplanetary Disk around SU Aurigae

Recent observations by the Atacama Large Millimeter and Submillimeter Array (ALMA) revealed that there are a few protoplanetary disks that have a tail-like structure or a tidal stream associated with them in the Taurus-Auriga star-forming region. It is considered as a signature that the disks may be gravitationally influenced by internal or external agents and lose a large amount of disk mass during their evolution. Such events affect the development of planetary systems and may produce a wide variety of planet configurations reported in exoplanet observations.

Previous observations in optical wavelength toward a protoplanetary disk around a young T-Tauri star SU Aurigae (hereafter SU Aur) suggested that reflection nebulae were associated with SU Aur. Such a reflection nebula usually seen in young stellar objects (YSOs) can be attributed to an outflow cavity as a result of stellar emission reflection at the cavity wall (e.g., Tamura et al. 1991; Lucas & Roche 1998a, 1998b; Padgett et al. 1999). However, recent observations in optical and near-infrared wavelengths suggested a different origin for SU Aur, and revealed a protoplanetary disk around SU Aur and an extended tail-like structure associated with the disk (Grady et al. 2001; Chakraborty & Ge 2004; Bertout & Genova 2006; Jeffers et al. 2014; de Leon et al. 2015). Possible explanations for the origin of the tail are discussed in de Leon et al. (2015) and they suggest that tidal interaction with a brown dwarf is the the most plausible mechanism. As a result, the tail-like structure may be a consequence of the fact that the disk can obtain or lose some amount of mass via mass transfer from or onto the disk.

In the standard model of planet formation, kilometer-sized objects, so-called planetesimals, are formed by the coagulation of dust grains, and planetesimals are responsible for forming more massive bodies, planetary embryos, or protoplanets in a protoplanetary disk (Hayashi et al. 1985). If the tail-like structure is a tidal stream, a large number of small mass objects including (proto)planets that reside in a protoplanetary disk known as a planet nursery must be released with gas and dust in the ambient, which might be a clue for the origin of the isolated planetary mass objects. In the observations by the Institut de Radio Astronomie Millimetrique (IRAM) Plateau de Bure Interferometer (PdBI) and ALMA, the RW Aurigae system is also an established example in which a recent stellar flyby is believed to induce a tidal stream (Cabrit et al. 2006; Rodriguez et al. 2018) and the detailed smooth particle hydrodynamics (SPH) simulations reproduced the observations of the system in detail (Dai et al. 2015).

As part of the Strategic Explorations of Exoplanets and Disks with Subaru project (Tamura 2009), we conducted H-band polarimetry observations toward a protoplanetary disk around SU Aur in the Taurus-Auriga star formation region, which is at a distance of 158 ± 2 pc (Gaia Collaboration et al. 2018). SU Aur is a 6.8 ± 0.1 million year old (Bertout et al. 2007) G1—2-type star (Herbig 1952; Calvet et al. 2004; Ricci et al. 2010), whose stellar mass is 1.7 ± 0.2 M_⊙ (Calvet et al. 2004; Bertout et al. 2007), dust disk mass is 9 × 10⁻⁴ M_⊙ (Ricci et al. 2010), and mass accretion onto the central star is (0.5–3) × 10⁻⁸ M_⊙ yr⁻¹ (Calvet et al. 2004). It is surrounded by a parental molecular cloud and its extinction (Av) is 0.9 mag (Bertout et al. 2007). The H-band scattered light observation found that small dust grains, typically 0.1–1 μm in size, form a long tail-like structure of more than 350 au in length extending from the disk body to the west (de Leon et al. 2015). Soon after the Subaru observation, CO observations by ALMA revealed the same tail-like structure more than 1000 au long extending from the northern part of the disk body to the southwest direction.

In this paper, we describe the details of the CO (J = 2 − 1) ALMA observations and data reduction in Section 2. In Section 3, the details of the morphological and kinematical structures of the disk and tail-like structure are shown. Several new findings by observations are also provided. In Section 4, possible mechanisms for generating the tail-like structure are discussed based on several theoretical works in the last decade. Finally, we discussed the origin of the tail-like structure. Section 5 summarizes our findings and implications.

The ALMA archival data (2013.1.00426.S) are used in the paper. Observations were performed on 2015 July 19 and August 8 at Band 6, containing 36 and 40 12 m antennas, respectively. The baseline length of the observed array configuration spanned from 15 m to 1.57 km, resulting in a maximum recoverable size of 11 arcsec. J0423−0120 was used as a bandpass calibrator and J0510+1800 (J0510+180) is used to calibrate for the amplitude and phase. The 2 side band (2SB) receivers were tuned to observe the CO (J = 2 − 1) rest frequency of 230.538 GHz and the nearby continuum at 1.3 mm in wavelength. The spectral resolution was set up to be 30.518 kHz, corresponding to 0.04 km s⁻¹. The average precipitable water vapor during observation was about 1.38 mm. The Band 6 data were imaged using the CLEAN task provided in the 4.3.1 version package of the Common Astronomy Software Applications (CASA), and self-calibration was applied to correct for short-term phase variation in the continuum data. We checked the rms noise level at each round by generating the image and stopped self-calibration with amplitude and phase correction when the rms noise level reached the minimum. Then we apply the calibration table to the CO (J = 2 − 1) line data and generate the images with a robust parameter of 1.5 in Briggs weighting. As a result of all calibrations, the synthesized beam of 0235 × 0356 is obtained.

Finally, the sensitivity of the primary beam is scaled by the primary beam correction task so that the sensitivity becomes uniform in the entire field of view (FoV). Note that the sensitivity of the FoV boundary generally becomes quite low in comparison with the one at the FoV center. The emission becomes noisy and blurred along the boundary. The primary beam correction task increases the emission at the FoV boundary by simply rescaling the sensitivity in the entire FoV to equate with the sensitivity level at the FoV center, and noise also simultaneously increases by the same amount. Therefore, the signal-to-noise (S/N) ratio still remains the same.

The moment 0 (integrated intensity) map in CO (J = 2 − 1) of the protoplanetary disk around SU Aur is presented in Figure 1(a). The figure shows a disk component and a tail-like structure more than 1000 au long extending from the northern part of the disk body to the southwest direction. The tail-like structure of small dust grains that extends at least 350 au in the same orientation was also observed in the H-band by the Subaru telescope (de Leon et al. 2015). Thus, the tail-like structure in gas significantly extends beyond the structure in the 0.1–1 μm-sized dust grain. The disk component has a peak intensity of 400 mJy beam⁻¹ km s⁻¹ in the CO (J = 2 − 1) emission line. The resultant rms is 23 mJy beam⁻¹ km s⁻¹ in the integrated time on source of 14.5 minutes after data flagging, providing an S/N of approximately 14. Note that the fan-shaped emission seen in southwest region in the moment maps could possibly be artificial. The sensitivity of the primary beam decreases as a function of Gaussian distribution, and thus the sensitivity at the FoV boundary is half of the value at the FoV center in ALMA.¹⁰ Since the tail-like structure reaches close to the FoV circular boundary, just outside of the imaging area, the emission may spread out in a fan-shaped form in the southwest direction due to the smearing effect.

Zoom In Zoom Out Reset image size

No signature of collimated jet, outflow, or outflow cavity that were previously reported in the optical and infrared observations (Nakajima & Golimowski 1995; Chakraborty & Ge 2004) are detected in the CO and millimeter continuum observations, at least within a 15 arcsec region from a central star. Figure 1(b) displays the moment 1 (velocity distribution) map. It shows disk rotation and the tail-like component with its V_LSR at roughly 6.5 km s⁻¹ uniformly extending to the southwest direction. The velocity gradient is twisted from the southeast–northwest direction in the very inner part to the northeast–southwest direction in the outer part, indicating that the disk could be twisted or warped although the possibility of infall around the disk cannot be fully discarded (Facchini et al. 2018). Figure 1(c) represents a moment 2 (velocity dispersion) map. It also shows both disk and tail-like components with similar morphology with the moment 0 and 1 maps. An interesting feature in the moment 2 map is the high-velocity dispersion region displaying in a boomerang shape near the center of the disk. It extends along the tail-like structure then disappears down the tail. The reflection nebulae that emission scattered at the outflow cavity wall (Tamura et al. 1991; Fuente et al. 1998; Padgett et al. 1999) previously reported is probably part of the tail-like structure obtained in the CO ALMA observation.

In order to directly compare the dust distribution with CO, we provide a superposed image of the H-band and 1.3 mm continuum emission that traces 0.1–1 μm and millimeter-sized dust, respectively, on the CO map as shown in Figure 2. Since the disk component and its boundary are not clear in the intensity map, the moment 1 (velocity distribution) map that shows the disk component more clearly by rotational motion is overlaid for reference. Unlike the CO tail-like feature, the compact structure is detected in the 1.3 mm continuum emission, whose peak intensity is 850 mJy beam⁻¹, as displayed in the white contour. H-band emission, on the other hand, shows a tail-like structure in the same orientation as the one seen CO emission. The results support that 0.1–1 μm-sized dust is well coupled with gas, while millimeter-sized dust is not. It should be noted that the possible reason for the nondetection of millimeter-sized dust in the tail-like structure could be the sensitivity limit since the dust mass in the tail-like structure is estimated to be ∼6 × 10⁻⁸ M_⊙ (de Leon et al. 2015), too small of an amount of dust to detect in the observation. The tail-like structure in millimeter-sized dust cannot be discarded at this moment.

Zoom In Zoom Out Reset image size

We generate a velocity profile of the disk component that covers a region with a radius of 1.2 arcsec, corresponding to ∼172 au, from the central star (Figure 3(a)). It shows a double-peaked profile separated into the blueshift (gas approaching) and redshift (gas receding) components at 5.8 ± 0.3 km s⁻¹. It is of high interest that the redshifted emission is more dominant than the blueshifted component, indicating that some additional velocity component is associated with the disk other than disk rotation. Figures 3(b) and (c) show the profiles of the velocity and velocity dispersion curves along the tail-like structure shown in Figures 1(b) and (c), respectively. The point of the origin on the distance is indicated by a cross in the figures. It is noteworthy that the velocity structure provided by the first and second moment maps (Figures 1(b) and (c)) shows the disk rotation smoothly varied to the linear motion forming in the collimated tail-like structure and maintains an almost constant velocity of ∼6.5 km s⁻¹ without any discontinuity. Thus, the tail-like component is redshifted by 0.5–1 km s⁻¹ from the systemic velocity of 5.8 km s⁻¹. The tail component has a velocity dispersion of ∼0.7 km s⁻¹ whereas the disk component has ∼2 km s⁻¹.

Zoom In Zoom Out Reset image size

Figure 4 shows a velocity channel map with ∼0.3 km s⁻¹ intervals for the CO (J = 2 − 1) emission line. The interesting tail-like feature of importance is detected from 5.14 to 8.16 km s⁻¹ and most of the tail-like structure is redshifted from the systemic velocity of 5.8 km s⁻¹. It is even more noteworthy that the tail-like structure can be seen as two components, a long tail-like structure more than 1000 au extending to the west seen between 5.14 and 6.94 km s⁻¹, and a relatively small elongated structure a few 100 au long extending to the (south)east direction between 7.25 and 8.16 km s⁻¹. The two detected components are morphologically consistent with our H-band result previously observed by the Subaru telescope (see Figures 1 and 3 in de Leon et al. 2015).

Zoom In Zoom Out Reset image size

4.1. Kinematic Structure of the Star and Disk System

The moment maps and several channels in the velocity channel map reveal that SU Aur has a distinctive tail-like structure that is kinematically connected with a disk component. It is beneficial to see non-keplerian motion in the disk+tail system for discussing the velocity structure and its origin. To investigate the interaction between the disk and the tail-like structure, it is important to differentiate their motions and analyze each component individually.

We made a simple keplerian disk model in the same orientation with the observed disk component and then computed the residual after subtracting the model from the observed moment 1 map. In the model, the disk is treated as gas moving in the keplerian velocity along the circular orbit around the central star. A detailed description can be found in the Appendix in Isella et al. (2007). We also assume the disk has flat geometry in the model. Figure 5(a) shows the same velocity distribution map as in Figure 1(b). Figure 5(b) shows the residual after subtracting the keplerian disk model shown in Figure 5(c) from the observed velocity distribution shown in Figure 5(a). Note that we focus on the deviation of the disk rotation and subtraction applied only to the disk component. Thus, the velocity of the tail-like component remains the same as the original. As a result of subtraction, the motion in the southern part of the disk is almost canceled by the model counterpart, resulting in approximately zero velocity. Therefore, it supports that the keplerian rotation in the southern part of the disk remains almost intact. On the other hand, the velocity component is not sufficiently subtracted in the northern part of the disk and some velocity component still remains by 4 km s⁻¹ from the systemic velocity. It implies that gas in the northern part of the disk does not have keprarian rotation and the redshift component dominates in the region. The result is consistent with the observationally derived velocity profile of the disk given in Figure 3(a).

Zoom In Zoom Out Reset image size

The detected tail-like structure that is connected with the northern part of the disk can also be traced by the channel map shown in Figure 5. Since the motion of the tail-like structure is smoothly connected with only northern part of the disk, it must be responsible for the velocity deviation from keplerian motion as shown in the residual mentioned above. Another feature of interest is that the tail-like structure has two velocity components. The channel map shows that a redshift component between 5.14 and 6.94 km s⁻¹ extends more than 1000 au to the west from the disk, while the other seen between 7.25 and 8.16 km s⁻¹, extending a few hundred astronomical units to the east from the disk. Similar structure is seen in the RW Aur binary or triple star system (Ghez et al. 1993) as tidal sculpting probably caused by gravitational interaction due to a stellar encounter. They indicate that the tidal streams should be a result of stripping the primary protoplanetary disk around RW Aur A by the flyby of a companion RW Aur B (Cabrit et al. 2006; Rodriguez et al. 2018). The fact that keplerian and non-keplerian motion simultaneously exist in the disk implies that the disk must be influenced by some kind of internal or external agent that significantly destroys keplerian rotation. We discuss several possible mechanisms that may have impacted the SU Aur system in the following section.

4.2. Origin of the Tail-like Structure

As a result of the ALMA millimeter observations, characteristic appearance of SU Aur has become clear and a tail-like structure is associated with the disk around SU Aur. The geometric-kinematic analysis of the observed tail-like structure reveals the positional relation of the disk+tail system and mass flow motion along the tail-like structure. We consider three possible scenarios: (1) collision with a (sub)stellar intruder or a gaseous blob from the ambient cloud, (2) interaction with the external cloud channeling its material toward the disk in a filamentary-like stream, and (3) ejection of a planetary or brown dwarf mass object due to gravitational instability via multibody gravitational interaction. Among these three possibilities, the collision with a (sub)stellar intruder or a gaseous blob is probably the most plausible explanation based on the physical properties of SU Aur. Follow-up observations by a shock tracer such as the SO emission line would be productive for confining the origin of the tail-like structure.

E.A. is supported by MEXT/JSPS KAKENHI grant No. 17K05399. M.T. is supported by MEXT/JSPS KAKENHI grant No. 18H05442, 15H02063, and 22000005. E.V. acknowledges financial support from the Russian Foundation for Basic Research (RFBR), Russian-Taiwanese project #19-52-52011. H.B.L. is supported by the Ministry of Science and Technology (MoST) of Taiwan (grant No. 108-2112-M-001-002-MY3 and 108-2923-M-001-006-MY3). This paper makes use of the following ALMA data: ADS/JAO.ALMA#2013.1.00426.S. ALMA is a partnership of ESO (representing its member states), NSF (USA) and NINS (Japan), together with NRC (Canada), MOST, and ASIAA (Taiwan), and KASI (Republic of Korea), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO, and NAOJ. The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc. Data analysis was carried out on the Multi-wavelength Data Analysis System operated by the Astronomy Data Center (ADC), National Astronomical Observatory of Japan. We appreciate Munetake Momose for technical advice in the data analysis. We would like to thank the anonymous referee who helped improve the manuscript.