Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 343, 571-584 (1999)

Previous Section Next Section Title Page Table of Contents

5. Discussion

5.1. The jet-driven outflow paradigm

5.1.1. Entrainment mechanism

Several mechanisms have been proposed to account for the formation of molecular outflows. The two most appealing class of models involve, respectively, a shock propagating in a collimated protostellar jet and a wide-angle wind driven by the central protostar (see Cabrit et al. 1997 for a recent review).

The overall structure of HH 211 is extremely reminiscent from the jet-driven outflow paradigm: in both lobes, and especially in the apparently unperturbed eastern lobe, the CO cavities are in the wake of shocks which are placed exactly at the terminal ends of the protostellar jet. These morphological coincidences strongly support shock-entrainment models (Raga & Cabrit 1993, Masson & Chernin 1993, Chernin et al. 1994) as the formation mechanism of the HH 211 molecular outflow. A further argument comes from the agreement between the observations and the shape predicted by our simple modelling of bow-shock propagation. Recent high-angular resolution interferometric studies of extremely young flows driven by low-mass protostars resulted in similar conclusions (e.g. Bachiller et al. 1995, Gueth et al. 1998).

However, the other class of dynamical models that have been proposed to explain the formation of molecular outflows, namely wide-angle winds collimated by a stratified ambient medium (e.g. Shu et al. 1991, Masson & Chernin 1992, Li & Shu 1996), cannot be formally ruled out by these observations. These models are indeed able to explain some of the characteristics of HH 211. For instance, in the model developed by Li & Shu (1996), the wide-angle wind driven by the protostar has a strongly density-enhanced region on the axis, which would thereby be observed as a collimated jet. However, the terminal part of the HH 211 lobes show cavities precisely situated in the wake of strong bow-shocks, which suggests that the shocks do play an important dynamical role in the formation processes of the outflow. These coincidences might be difficult to explain in a wind-driven shell picture (in the Li & Shu's model, the cavities do not close back on the axis), while there are naturally predicted by jet-driven flow models. It may also be possible that both entrainment mechanisms occur, with variable efficiencies during the protostellar evolution (Stahler 1993), as for instance the central "jet" in the wind-model of Li & Shu should also be able to induce shock-entrainment.

The simulations presented by Cabrit et al. (1997) show that wind-driven and bow-shocks models have similar kinematical predictions, which both are in reasonable agreement with the position-velocity diagram of HH 211 (Fig. 5). Consequently, and pending detailed channel maps simulations, the kinematical properties of HH 211 can hardly be used to distinguish between the different flow formation mechanisms. A potential problem of wind-driven shells is however the lack of low-velocity component when the flow is not in the plane of the sky (Cabrit et al. 1997). As for the propagating bow-shock, note that the simple modelisation that we have developed (Sect. 4) allows to derive the overall shape of the excavated cavity, but the cumulative effects on the velocity distribution at the bow surface are too complex to be analysed in this simple framework.

In any case, detailed numerical simulations are required to further investigate the entrainment mechanisms in young, embedded molecular outflows. Suttner et al. (1997) and Smith et al. (1997) presented three-dimensional simulations of the propagation of a dense jet in a molecular cloud, which can directly be compared with the HH 211 case: the main morphological and kinematical properties of the CO and H2 emission can be roughly reproduced. The results obtained by Cabrit et al. (1999) also show that the shape of the low-velocity CO cavity can be explained by simulations of the propagation of a time-variable jet into a stratified ambient medium (see also Cabrit et al. 1997). Our observations do however point out the need for channel maps simulations to allow reliable comparisons to be performed.

5.1.2. Nature of the HH 211 CO jet

There are several possibilities to explain the nature of the high-velocity CO jet of HH 211 (see also Richer et al. 1992, who discussed the similar feature observed, though with lower angular resolution, in the NGC 2024-FIR5 outflow).

First, we may observe the protostellar jet itself , in which case the jet would be molecular and rather dense ([FORMULA], see Sect. 3.3). Current models of jet formation involve a magnetized accretion disk surrounding the central source, and can be classified in two categories, depending on whether the ejected material arises from the surface of the rotating disk (e.g. Ouyed & Pudritz 1997) or from the interacting zone between the stellar magnetosphere and the disk (the so-called X-wind model; e.g. Shu et al. 1995). In both cases, the outflowing material is accelerated up to the Alfvén surface and is further collimated into a jet (e.g. Heyvaerts & Norman 1989); in the X-wind case, this jet is only the denser on-axis region of a wide-angle wind (Li & Shu 1996; see also discussion in the previous section). Our observations cannot put constraints on the actual ejection zone, too small compared to the angular resolution. The HH 211 CO jet appears however to be collimated at a distance of [FORMULA] AU from the protostar, and has a rather large transverse size (up to 1000 AU). This last point could be difficult to explain in the framework of current jet models. As for the observed apparent acceleration, we note that the length of the jet ([FORMULA] AU) precludes the view that we are observing the true (magneto-centrifugal) acceleration of the gas in the vicinity of the protostar (see e.g. Pelletier & Pudritz 1992). The jet simulations presented by Ouyed & Pudritz (1997) do predict an apparent acceleration beyond the Alfvén surface, but it is not clear whether such behaviour persists on large scales. Wind-driven flow models (Li & Shu 1996) can explain the apparent overall acceleration of the outflowing material with distance from the exciting source, but the kinematical properties of the apparent central "jet" have still to be specified.

Another, attractive possibility to explain the HH 211 high-velocity CO observations is a turbulent mixing layer of ambient material, entrained through Kelvin-Helmholtz instabilities along the sides of an underlying jet (Cantó & Raga 1991, Taylor & Raga 1995). Raga & Cabrit (1993) discussed how the swept-up gas re-expands in the wake of a leading bow-shock to fill the cavity: this mechanism could allow material to be present and then become entrained along the jet sides. Note that the HH 211 cavity is not as prominent in the CO [FORMULA] images (Fig. 2) as in the CO [FORMULA] maps (Fig. 3), which actually suggests that the cavity could be filled in with low excitation material. In the case of the formation of a fully turbulent jet, due to the growth of the mixing layer, Stahler (1994) showed that velocity dispersion combined with opacity effects can produce an apparent acceleration with distance from the protostar, as observed in the HH 211 case. An argument in favour of such a turbulent entrainment arises from the fact that the two lobes present slightly different kinematical behaviours: the redshifted lobe shows higher velocities and a steeper slope in the position-velocity diagram than the eastern one (Fig. 5). These properties may derive from an intrinsic velocity difference, but also from more gas being present at a given velocity. This is actually what can be qualitatively expected for HH 211 if entrainment mechanisms occur along the jet sides: the amount of accelerated gas depends in that case on the density of the surrounding material and the redshifted lobe of HH 211 is actually propagating in a dense molecular filament (Sect. 3.2).

Finally, the sub-structures observed within the HH 211 jet (Sect. 3.3) suggest that at least part of the high-velocity CO emission could be associated with the wake of several small bow-shocks propagating into the underlying protostellar jet. Such shocks are created by internal working surfaces at the position of velocity discontinuities (Raga et al. 1990, Raga & Cabrit 1993). Their turbulent wakes present an apparent acceleration, with the fastest velocities just behind the working surface (Raga & Cabrit 1993).

Clearly, higher angular resolution images are needed to properly assess the actual nature of the HH 211 CO jet. Detailed numerical simulations of the propagation of protostellar jets/winds in a molecular medium (see previous section) are also required to understand the morphological and kinematical properties of the observed gas.

5.1.3. Shocks

An interesting point revealed by the observations presented in this paper is the structure of the terminal shock of the eastern (blueshifted) lobe. The CO and H2 emission are perfectly coincident, which suggests that the shock is essentially radiative. Moreover, both CO and H2 present a linear extension ahead of the bow. This feature has already been observed in high-angular resolution interferometric maps of other extremely young molecular outflows, either through SiO [FORMULA] emission (L 1448, Dutrey et al. 1997; L 1157, Gueth et al. 1998) or CH3OH [FORMULA] A+ emission (NGC 1333/IRAS 2, Bachiller et al. 1998). It could thus be a rather common feature in bow-shocks propagating in protostellar jets. Gueth et al. (1998) and Bachiller et al. (1998) discussed the possible origin of such elongated structures ahead of the bows. For instance, numerical simulations of MHD jets show that bow-shocks can develop elongated conical structures in their front part, the so-called "nose cone" (e.g. Clarke et al. 1986, Ouyed et al. 1997), which could be identified with the observed features. Raga et al. (1998) have shown that a similar feature can also be produced in a shock propagating in a non-top hat jet.

5.1.4. Protostellar condensation

The HH 211-mm protostellar condensation is resolved and extends in a direction roughly perpendicular to the jet axis (the orientation difference is [FORMULA]). This situation is obviously reminiscent of the classical disk/jet picture. If the observed elongated structure is indeed a (thin) circumstellar disk, an inclination angle of [FORMULA] on the plane of the sky can be derived from the measured axis ratio ([FORMULA] AU). However, the HH 211-mm condensation has a size (and a mass) much larger than the typical sizes (masses) of T Tauri disks (e.g. Dutrey et al. 1996) and corresponds to a much earlier evolutionary stage of the close protostellar environment. It seems thus more likely that we are actually observing a flattened envelope or torus surrounding the central protostar. Such a structure could for instance have been created by a strong perturbation of the protostellar envelope by the powerful ejection of matter, which can empty the polar directions. The Class 0 object L 1157-mm presents evidence of such an interaction (Gueth et al. 1997). However, a flattened protostellar envelope can also be produced by an intrinsic evolution of the collapsing process. Recent theoretical works and numerical simulations have shown that gravitational infall is not isotropic, providing the collapsing cloud has a small magnetic field (Galli & Shu 1993a, 1993b), was initially rotating (Yorke et al. 1993, 1995) or was initially already flattened (Hartmann et al. 1996).

5.2. Perturbations of the flow structure

Even if the overall structure of the HH 211 molecular outflow is rather simple and allows a detailed analysis, several complications due to either internal evolution or external perturbations are showing up.

Multiple ejection events.  Several ejection events seem to have taken place in HH 211. In addition to the two strong, symmetrically placed shocks, the H2 emission in HH 211 reveals smaller and fainter shocks in the eastern lobe, and a more diffuse emission in the western one (see the discussion by McCaughrean et al. 1994). H2 emission is also detected at the very extremity of the western lobe (Fig. 1). Finally, the sub-structures observed in the CO jet (Sect. 3.3) are most probably the signature of a varying ejection phenomenon. Hence, the picture emerging from these observations includes several shocks which propagate in a protostellar jet, only the most important ejection event having resulted in fully-developed bow-shocks, which have then evacuated cavities in the interstellar medium.

Jet bending.  The western and eastern lobes of the jet are almost perfectly linear, but not along the same axis (see Fig. 4 and first panel of Fig. 8): there is a slight but significant misalignment of about [FORMULA]. Similar morphologies have already been observed in protostellar jets (see e.g. Fendt & Zinnecker 1998) as well as in jets driven by active galaxies (e.g. Killeen et al. 1986). Fendt & Zinnecker (1998) listed the possible origins of such jet bendings: Lorentz forces on the magnetic jet, motion of the jet source in a binary system, or dynamical pressure of the external medium. In the HH 211 case, we can probably rule out the second mechanism, as the angular resolution of the observations would allow to resolve the binary system (whose separation is of the order of the shift of the jet extremity). The third mechanism corresponds to the case where the central exciting source has an important proper motion in a direction roughly perpendicular to the jet axis: the ejected material can be deflected by the dynamical pressure of the ambient medium (note that the exact effect, in terms of kinematics and morphology, depends on the actual interaction mechanism). In that case, a rough estimate of the protostar velocity can be obtained by assuming the ejected material has completely lost the transverse velocity of the exciting source: from the apparent displacement of the protostar ([FORMULA] AU) and assuming a timescale of 1000 years, we then derive a velocity of [FORMULA] km  s -1.

Other misalignments.  Direct evidences for precession of the ejection direction in young molecular outflows have already been reported for several sources (e.g. Eislöffel et al. 1996, Gueth et al. 1996). The HH 211 case is far less clear, but some of its characteristics suggest that a modification in the ejection and/or propagation direction could have occurred: in addition to the jet bending mentioned above, several misalignments can be noted. The mean axis of the high-velocity jet is indeed different from the low-velocity cavity axis (see Fig. 4 and last panel of Fig. 3). Moreover, the eastern (blueshifted) part of the jet presents a clear bend (Figs. 4 and 8). Whereas its terminal part is aligned with the axis of the bow-shock and with the linear extension ahead of the bow, the main, longer and brighter part of the jet is oriented along a different axis. The location of the bend corresponds to a small H2 shock. This bend could result either from a continuous precession of the jet (the angular resolution limitation would avoid to see the small oscillations of the jet near the protostar) or to a deflection of the jet against a dense clump of the interstellar medium (see Raga & Cantó 1995). The western (redshifted) part of the jet does not present such a clear bending, but the strong perturbation of the terminal part of the lobe (see Sect. 3.1) precludes any detailed morphological analysis at this position. Finally, we note that faint CO emission is detected at roughly the systemic velocity, in two positions placed symmetrically about the protostar position, but along a southeast-northwest axis which differs by more than [FORMULA] from the HH 211 jet axis (see arrows in Fig. 3, panel at velocity 8.2 km  s -1). The nature of this emission remains difficult to assess. It could trace a previous ejection event, or simply correspond to inhomogeneities in the ambient medium.

Interaction with the interstellar medium.  A last point worth mentioning is the perturbation of the flow due to the interaction with the ambient interstellar medium. Despite its extreme youth, HH 211 illustrates this point, which has already been noted for several other sources. As discussed in Sect. 3.1, the western lobe of HH 211 seems to be perturbed by the presence of a dense filament in the molecular cloud. Interaction with an external medium could also be invoked to explain the incomplete southern flank of the eastern cavity. Clearly, the observed outflow structures depend on the intrinsic flow formation mechanisms as well as on interaction with the close interstellar environment.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: March 1, 1999