A Very Compact Extremely High Velocity Flow toward MMS 5/OMC-3 Revealed with ALMA

Figure 1 presents 1.3 mm continuum images obtained from multiangular resolution. The low-resolution 1.3 mm continuum image presented in Figure 1(a) shows that the 1.3 mm continuum emission is elongated in the east–west direction. This elongation is consistent with the axis of the EHV flow (see Sections 3.3 and 3.4). The 1.3 mm continuum emission enhanced along the jet axis is likely due to the hot dust associated with the jet ejected from the central region. The total flux of the 1.3 mm continuum emission in the low-resolution image in Figure 1(a) is 154 mJy, including the contribution from the free–free jet. Assuming that the spectral index of the free–free jet is ∼0.6 (Reynolds 1986; Anglada et al. 1998), the expected flux density of the free–free emission at 1.3 mm becomes 0.94 mJy, based on the 3.6 cm flux density of 0.12 mJy with 3σ upper limit (Reipurth et al. 1999). Hence, the flux density attributed to the free–free emission is 0.6% of the 1.3 mm total flux. The peak position is measured as R.A. = 5^h35^m22464, decl. = −5°01'14304. This position is offset by 0.12 arcsec with respect to the location of HOPS 88, which is identified as a Class 0 source by observations with the Hershel and Spitzer space telescopes (Furlan et al. 2016). The positional offset is comparable to the positional uncertainty of the Hershel observation. Thus, we consider that this continuum source is likely associated with HOPS 88. In this paper, we refer to the position of the identified 1.3 mm continuum source as the location of the protostar.

Zoom In Zoom Out Reset image size

In the high-resolution image presented in Figure 1(b), we can confirm a compact structure associated with HOPS 88. This compact component contains ∼9.5% of the total flux measured from the low-resolution image in Figure 1(a). In order to characterize the morphology of the compact component, the two-dimensional Gaussian fitting tool in the Common Astronomy Software Application (task "IMFIT") was used. The total flux, peak flux, and deconvolved size are measured to be 57 ± 3 mJy, 37 ± 1 mJy beam⁻¹, and (0.14 ± 0.02) arcsec × (0.12 ± 0.02) arcsec (P.A. = 144° ± 43°), which corresponds to ∼56 × ∼45 au in the linear size scale, respectively. The residual after extraction of the two-dimensional Gaussian fit is less than 10% as compared with the peak flux. The mass of the circumstellar material traced by the 1.3 mm continuum emission (M_1.3mm) can be estimated as

$$\begin{eqnarray}&&{M}_{1.3\mathrm{mm}}=\displaystyle \frac{{F}_{1.3\mathrm{mm}}{d}^{2}}{{\kappa }_{1.3\mathrm{mm}}{B}_{1.3\mathrm{mm}}({T}_{\mathrm{dust}})},\end{eqnarray} \tag{ 1 }$$

where F_1.3mm is the total integrated 1.3 mm flux of the two-dimensional Gaussian fit, d is the distance to the source, κ_1.3mm is the dust mass opacity at λ = 1.3 mm, T_dust is the dust temperature, and B_1.3mm(T_dust) is the Planck function at a temperature of T_dust. We adopt κ_1.3mm = 0.009 cm² g⁻¹ from the dust coagulation model of the MRN (Mathis et al. 1977) distribution with thin ice mantles at a number density of 10⁶ cm⁻³ computed by Ossenkopf & Henning (1994). We assume a gas-to-dust mass ratio of 100. Here, T_dust = 21.5 K is adopted (Sadavoy et al. 2016). Given the measured total flux of 57 ± 3 mJy from the two-dimensional Gaussian fitting, the mass is estimated to be M_1.3mm ∼ 1.8 × 10⁻² M_⊙.

Figure 2 shows the dense gas distribution traced by C¹⁸O J = 2–1 and ${{\rm{N}}}_{2}{{\rm{D}}}^{+}$ J = 3–2. The C¹⁸O J = 2–1 traces a centrally condensed compact structure associated with the protostar. The deconvolved size and P.A. are measured to be (4.7 ± 0.4) arcsec × (3.8 ± 0.3) arcsec and 100° ± 25° based on the two-dimensional Gaussian fitting.

Zoom In Zoom Out Reset image size

${{\rm{N}}}_{2}{{\rm{D}}}^{+}$ J = 3–2 emission shows a large-scale structure extending along the northwest–southeast direction. Note that the gas distribution is consistent with the elongation of the OMC-3 filament (e.g., Chini et al. 1997; Lis et al. 1998; Johnstone & Bally 1999). In contrast to the C¹⁸O J = 2–1 emission, ${{\rm{N}}}_{2}{{\rm{D}}}^{+}$ J = 3–2 emission is very weak in the vicinity of the protostar and no emission with 5σ level is detected toward the 1.3 mm continuum emission and the C¹⁸O emission peak. The ${{\rm{N}}}_{2}{{\rm{D}}}^{+}$ molecule is known to be abundant only in the cold and dense environments where molecules such as CO are frozen out onto the surface of the dust grains, forming icy mantles (Fontani et al. 2012; Giannetti et al. 2014). The CO molecule is frozen out onto grain mantles when the gas temperature is T ≤ 19 K (Qi et al. 2011). Thus, after a protostar emerges, CO appears in the gas phase due to heating by radiation from the central protostar. The CO in the gas phase significantly decreases ${{\rm{N}}}_{2}{{\rm{H}}}^{+}$ formation and accelerates ${{\rm{N}}}_{2}{{\rm{H}}}^{+}$ destruction (Bergin et al. 2001). The anticorrelation of gas-phase CO and ${{\rm{N}}}_{2}{{\rm{H}}}^{+}$ has been confirmed by numerous observations of the protostellar environment (Caselli et al. 1999; Bergin et al. 2002; Jørgensen et al. 2004). Since MMS 5 harbors a protostar, it is natural to expect warm gas around the protostar, which explains the anticorrelation between ${{\rm{N}}}_{2}{{\rm{D}}}^{+}$ and C¹⁸O gases. The systemic velocity measured from the optically thin ${{\rm{N}}}_{2}{{\rm{D}}}^{+}$ J = 3–2 emission line is 11.0 km s⁻¹, which is consistent with the central velocity of the dip in the C¹⁸O J = 2–1 line profile. This appears to indicate that both gases have an identical center of mass and that ${{\rm{N}}}_{2}{{\rm{D}}}^{+}$ J = 3–2 encloses C¹⁸O J = 2–1 around MMS 5.

Figure 3 presents the position–velocity (PV) diagram cutting along P.A. = 169°, which is the direction perpendicular to the outflow axis. Figure 3 shows that the C¹⁸O J = 2–1 spreads to ±1.3 km s⁻¹ with respect to v_sys. The blueshifted emission is twice as bright as the redshifted emission. There exist intensity peaks in both the blue- and redshift components with the positional offset of ∼1 arcsec with respect to the position of the 1.3 mm continuum peak, indicating the velocity gradient along the major axis of the envelope. The feature observed in Figure 3 can be interpreted as a rotational motion within the envelope. In Figure 3, we overlaid two rotation curves expected from the Keplerian rotation (${v}_{\phi }\propto {r}^{-0.5}$) and the angular momentum conservation (${v}_{\phi }\propto {r}^{-1}$). The rotation curve expected from the angular momentum conservation (${v}_{\phi }\propto {r}^{-1}$) seems to fit well and follow the emission peaks in the PV diagram. In addition, a high-velocity gas of $| {v}_{\mathrm{LSR}}$ $-{v}_{\mathrm{sys}}|$ ≥ 1 km s⁻¹ is detected toward the center (within ≤0.5 arcsec from the center) and the compact component is particularly clear in the redshifted component. However, the C¹⁸O component in Figure 3 does not contradict the velocity curve expected from both the angular momentum conservation (${v}_{\phi }\propto {r}^{-1}$) and Keplerian rotation (${v}_{\phi }\propto {r}^{-0.5}$) assuming the central protostellar mass of 0.1 M_⊙. High angular resolution and high-sensitivity observations are required in order to determine the origin of the high-velocity gas within the region of 200 au from the protostar or that at the emission peak.

Zoom In Zoom Out Reset image size

Figure 4 shows the CO J = 2–1 line profile toward the peak position of MMS 5. The figure shows low- and high-velocity components. The main component ranges between velocities of ±7 km s⁻¹ with respect to the systemic velocity. This component has a strong absorption at the center. The velocity center of the dip is v_LSR = 11.0 km s⁻¹, which is consistent with the systemic velocity. In addition, we have detected EHV flows in CO J = 2–1 toward MMS 5. The detected velocity ranges of the blueshifted and redshifted components are −64 to −99 km s⁻¹ and 43–78 km s⁻¹, respectively. In other words, the EHV flow is accelerated up to the range of 80–100 km s⁻¹ with respect to the systemic velocity.

Zoom In Zoom Out Reset image size

The CO J = 2–1 channel maps shown in Figure 5 indicate that the molecular outflow associated with MMS 5 is elongated along the east–west direction. In the velocity range of $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}|$ = 0–10 km s⁻¹ (Figure 5(a)), the extended CO emission is detected rather uniformly. The complication of the emission is attributed to contamination from the ambient molecular cloud. In the velocity range of $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}|$ = 10–50 km s⁻¹ (Figures 5(b)–(e)), CO emission delineates a V-shaped structure. The size of this structure extends up to ∼24 arcsec, which corresponds to ∼0.05 pc (P.A. = 79° ± 2°), and the opening angle of the V-shape is measured to be ∼40°. The outflow is elongated in an approximately east–west direction, velocity of ∼10 km s⁻¹ and size of ∼0.1 pc, which are consistent with those estimated in previous studies (Aso et al. 2000; Williams et al. 2003; Takahashi et al. 2008). At $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}|$ = 50–100 km s⁻¹ (Figures 5(f)–(j)), which corresponds to the velocity of the EHV flow, a geometrically collimated structure appears. The length and width of the collimated structure measured from the redshifted flow are ∼7000 and ∼1200 au, respectively. The synthesized beam size (≤600 au) is smaller than the measured outflow width; hence the collimated component is spatially resolved. The blueshifted component also shows some degree of collimation. The emission is not detected at the protostar position.

Zoom In Zoom Out Reset image size

In channel maps presented in Figure 6, we detected collimated SiO J = 5–4 emission associated with MMS 5. The redshifted emission is mainly detected in the velocity range of $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}|$ = 50–70 km s⁻¹. Faint SiO J = 5–4 redshifted emission with signal-to-noise ratio (S/N) ≃ 5 is also detected in the low-velocity range of $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}| \leqslant 50$ km s⁻¹. For the blueshifted emission, there is no detection (S/N < 3) in the velocity range of $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}| \leqslant 80$ km s⁻¹. In the velocity range of $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}|$ = 80–90 km s⁻¹, the blueshifted EHV component is detected in Figure 6(i). The extension of the SiO J = 5–4 redshifted emission is 4 arcsec, which corresponds to ∼1600 au (P.A. = 96° ± 1°), and the width along the minor axis is ∼0.7 arcsec, which corresponds to ∼270 au. Thus, the EHV component is spatially resolved with our synthesized beam. The comparison of the high– and low angular resolution images indicates recovering ∼91% of the total integrated flux. Therefore, the detected SiO J = 5–4 emission is concentrated to the collimated structure captured with the high angular resolution data. In the western part of the outflow and jet (blueshifted component), a chain of knots is observed in the near-infrared H₂ v = 1–0 line (Yu et al. 2000; Stanke et al. 2002). The direction of the sequence of H₂ knots coincides with the jet direction in our observations. However, the chain of H₂ knots is shifted to the south more than 5 arcsec compared with the location of the SiO and CO jets in our observation. In addition, the width of the H₂ knots (∼15 arcsec) closest to the MMS 5 (i.e., location of the protostar) is much wider than that of the CO and SiO jets (∼5 arcsec) in our observation. These differences clearly mean that the origin of the H₂ knots is not related to the CO and SiO jets associated with MMS 5.

Zoom In Zoom Out Reset image size

In Figure 7(a), we compared the spatial distribution of the integrated intensity SiO J = 5–4 emission with that of CO J = 2–1 emission. The SiO collimated structure (redshifted emission) is observed at the root of a V-shaped structure detected in CO J = 2–1 emission. The SiO J = 5–4 collimated structure, which comes only from the EHV velocity component (80–90 km s⁻¹), is more compact than the V-shaped CO J = 2–1 structure by a factor of 7.5, in which the CO J = 2–1 emission is obtained from the relatively low-velocity regime ($| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}|$ = 10–50 km s⁻¹). The collimated high-velocity structure detected in CO (11,000 au) is larger than that detected in SiO (2500 au) but still more compact than the V-shaped outflow (14,000 au). It is natural to consider that the SiO EHV component and CO low-velocity component trace physically different flows. In order to distinguish between the two detected components, hereafter we use the terms "jet" and "outflow." The jet is defined as a geometrically collimated emission and has the EHV range of ejection speed of ≳50–100 km s⁻¹ (Kwan & Tademaru 1988; Hirth et al. 1994a, 1994b). In contrast, the outflow is defined as a bipolar structure that has a wide opening angle and has a relatively low velocity of ≲50 km s⁻¹ (Lada 1985).

Zoom In Zoom Out Reset image size

The V-shaped CO outflow has P.A. = 79°, while the CO and SiO jets have P.A. ∼ 90°. Thus, the P.A. of the jet and outflow is not the same as that seen in Figure 7(a). The difference in P.A. between the jet and outflow is about 18° measured in the redshifted side (for detailed discussion, see Section 4.1). As seen in Figures 7(b) and (c), the jet component is detected in both CO J = 2–1 and SiO J = 5–4 emissions. Within both CO and SiO jets, several clumpy structures are detected. They appear more or less periodically with a spacing of ≃200 au. The interpretation of the clumpy structures is discussed in Section 4.2.

Figures 5 and 6 show asymmetry in the outflow and in the jet. For example, in the red lobe side, the northern emission is stronger than the southern emission. As for the jet, the red component is stronger than the blue component. We cannot figure out the mechanism to explain these asymmetric structures only from the present observational data. However, we suggest some possibilities for the asymmetry: a nonuniform density distribution of the surrounding medium caused by turbulence and an initial asymmetric structure of the star-forming core. In the latter part of this paper, we focus on the properties of the jet and outflow, while we do not discuss the asymmetry.

Figure 8(a) shows the PV diagram cutting along the major axis of the outflow (P.A. = 79°). In this diagram, we found three components. The first exists in the high-velocity range between −110 and −70 km s⁻¹ for the blueshifted component and between 50 and 70 km s⁻¹ for the redshifted component with respect to the systemic velocity. The spatial extension of this component is about 18 arcsec (∼7000 au) with respect to the protostar. This component corresponds to the jet. The second component shows a parabolic structure in the PV diagram (as delineated by the white line in Figure 8(a)), having the velocity range $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}| \ \leqslant 60\,\mathrm{km}\,{{\rm{s}}}^{-1}$. Although the spatial extension of this component is almost the same as that in the jet, this component corresponds to the V-shaped outflow. The parabolic structure seems to interact with the jet around the terminal velocity (v_jet ∼ 60 km s⁻¹) in the redshifted component. The third component consists of very low-velocity gas ($| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}| \ \leqslant 10\,\mathrm{km}\,{{\rm{s}}}^{-1}$) and has a larger spatial extension than the first and second components. The third component shows the same velocity gradient as in the outflow, which might be related to the swept-up gas by the outflow. We will not discuss this component any further in this paper.

Zoom In Zoom Out Reset image size

In order to more precisely derive the inclination angle of the low-velocity outflow with a relatively wide opening angle detected in CO J = 2–1, we use a simplified analytical model (wind-driven model) introduced by Lee et al. (2000). In cylindrical coordinates, the structure of the outflow shell is described as

$$\begin{eqnarray}&&z={{CR}}^{2},\end{eqnarray} \tag{ 2 }$$

where z and R are the height from the disk midplane and the distance from the outflow (or z-) axis, respectively. The outflow velocities in the radial (v_R) and z-direction (v_z) are given by

$$\begin{eqnarray}&&{v}_{R}={v}_{0}\,R,\end{eqnarray} \tag{ 3 }$$

and

$$\begin{eqnarray}&&{v}_{z}={v}_{0}\,z,\end{eqnarray} \tag{ 4 }$$

respectively. In Equations (2)–(4), C and v₀ are free parameters representing the spatial and velocity distributions of the outflow shell. We also introduce the inclination angle, i, of the outflow shell (see Figure 21 of Lee et al. 2000). The parameter C is determined by the spatial structure of the emission from the outflow. After C is determined, the parameters v₀ and i can be estimated to consider the inclination effect in the PV diagrams, in the same manner as in the previous study (Lee et al. 2000).⁶

Hereafter, we use the following observational data sets to derive the physical parameters of the outflow/jet. For the CO emission, we use both blueshifted and redshifted emissions to estimate the physical parameters and drive their mean values. Since SiO emission is only detected in the redshifted component, the physical parameters of the SiO jet are calculated only using the one-side component. The curvature parameter is fit to the outflow shells as C = 0.07 arcsec⁻¹. Then, based on the measured C, the outflow structures give v₀ = 6.7 km s⁻¹ arcsec⁻¹, and the inclination angle of the outflow axis with respect to the plane of the sky is i = 50°.

Using the derived inclination angle of the outflow shell, we estimated the sizes, velocities, and dynamical timescales of the outflow and jet that are listed in Table 3. The dynamical timescale is estimated by the intrinsic length scale L and the expansion speed v as t_dyn ∼ L/v ∼ (L_obs/v_obs) tan 50°, where L_obs is the projected length of the outflow/jet and v_obs is the line-of-sight velocity of the outflow/jet. The projected maximum size of CO J = 2–1 and SiO J = 5–4 emission, L_obs, was measured at the 10σ contour in their channel maps (Figures 5 and 6). The maximum redshifted gas velocity measured in Figure 8 was used as v_jet,obs of the jet, while the mean velocity of CO outflow (∼30 km s⁻¹) measured in Figure 8(a) was used for v_out,obs of the outflow. The outflow has a dynamical timescale of t_dyn ∼ 1300 (L_out,obs/9300 au)(v_out,obs/30 km s⁻¹)⁻¹ yr. The jet dynamical timescale is estimated to be t_dyn ∼ 470 (L_jet,obs/7000 au)(v_jet,obs/70 km s⁻¹)⁻¹ yr, using CO J = 2–1. Thus, the dynamical timescale of the outflow is a factor of ∼3 longer than that of the jet. In addition, the molecular outflow (∼14,000 au) is larger than that of the jet (∼11,000 au). Using the timescales and size scales, we discuss the driving mechanism of this outflow/jet system in Section 4.1.

Figures 8(b) and (c) show zoomed images of PV diagrams of the redshifted jet component obtained in the CO and SiO emissions, respectively. The high-velocity component located in the velocity range of $| {v}_{\mathrm{LSR}}-{v}_{\mathrm{sys}}|$ = 55–70 km s⁻¹ traces the EHV jet. Within the EHV jet of SiO in Figure 8(c), we can see several bright compact knots. As shown by the white lines in Figure 8(c), each bright knot has a velocity gradient inside when cutting along the jet direction. This is consistent with the result of the jet simulation by Stone & Norman (1993), who considered the case of the episodic jet. They obtained a similar sawtooth velocity field, in which the line-of-sight velocity is decelerated in a bright knot while increasing the distance from the central source. Santiago-García et al. (2009) reported the similar sawtooth structure in their observations. Our result presented in Figure 8(c) shows a similar pattern of the velocity gradient (both the size scale and the velocity change). This sawtooth structure implies that the observed EHV jet in the SiO emission traces the unsteady or episodic gas ejection.

Two scenarios are proposed to explain the driving mechanism of the outflow and the jet, namely, (1) nested disk wind and (2) entrainment scenarios. In the nested disk wind scenario, different types of flow are driven from different radii of a rotationally supported disk (Tomisaka 2002; Banerjee & Pudritz 2006; Machida et al. 2008; Tomida et al. 2013) in which the low- and high-velocity flows are driven near the disk outer and inner edges, respectively (Machida 2014). On the other hand, in the entrainment scenario, only the high-velocity jet appears near the protostar and the ambient (infalling) material is accelerated or entrained by the jet (Arce et al. 2007; and references therein). In both scenarios, the high-velocity component (i.e., jet) is considered to be driven by the Lorentz and centrifugal forces (Kudoh & Shibata 1997; Spruit et al. 1997; Tomisaka 2002; Machida et al. 2008; Seifried et al. 2012). The low-velocity outflow is also magnetically accelerated by the rotation of the outer disk region in the nested disk wind scenario, while it is entrained by the jet driven near the protostar in the entrainment scenario.

In the nested disk wind scenario, since the low-velocity outflow appears before the emergence of high-velocity flow and protostar, the low-velocity outflow is expected to have a longer dynamical time than the high-velocity jet. In addition, the low-velocity component (i.e., the outflow) should precede the high-velocity component (i.e., the jet) even if only in the very early phase of the star formation (Machida 2014). However, the high-velocity jet overtakes the low-velocity outflow after a short time because the jet velocity is much higher than the outflow velocity. Thus, the nested disk wind scenario predicts that the jet length is shorter than the outflow length for a very short duration immediately after the protostar formation. Instead, in the entrainment scenario, since the jet entrains the surrounding gas and makes the low-velocity outflow, the jet length should be comparable to or larger than the outflow length at any time after protostar formation. Thus, only the observation of low- and high-velocity components in a very early phase of the star formation could determine their driving mechanism.

In our observation, the outflow and jet images have an angular resolution of 0.2 arcsec, which corresponds to ≃80 au. The outflow and jet launching points are expected to be located at ≳2 and ≲0.5 au from the central star (Tomisaka 2002; Machida 2014). Recent ALMA studies have shown that an outflow with a wide opening angle appears in the outer-disk region, ∼100 au (Alves et al. 2017), and a jet with good collimation is driven near the disk's inner edge, ∼0.05 au (Lee et al. 2017). The velocities of these flows are orders of 10 and 100 km s⁻¹, respectively. Although we are not able to spatially resolve the outflow and jet launching points (or radii) due to the limited angular resolution, the observed gas morphologies and gas velocities share similar characteristics. This indicates that MMS 5 outflow and jet might be launched from the different radii.

As described in Section 3.5, the dynamical timescale of the jet is shorter than that of the outflow by a factor of 3. The outflow and the jet extend up to ∼14,000 and ∼11,000 au, respectively. The size of the jet is shorter than that of the outflow. This is consistent with the nested disk wind scenario.

Another possible piece of evidence to support the nested disk wind scenario is the axis difference between the outflow and the jet (δθ ∼ 17°; Figure 7), which could be explained by considering different launching radii. The axis difference between the outflow and the jet can be explained by recent MHD simulations (Matsumoto et al. 2017; Lewis & Bate 2018), in which a warped disk forms in a weakly turbulent cloud and the direction of low-velocity outflow changes with time (see also Matsumoto & Tomisaka 2004). In these studies, the outflows are not always aligned with a large-scale disk, including both Keplerian and pseudo disks. The disk normal on the scale of ≳100 au is roughly aligned with the outflow axis, while that on the scale of ≲100 au is misaligned with the outflow axis (Figure 9 of Matsumoto et al. 2017). Since the protostar was not resolved and no high-velocity jet appears in Matsumoto et al. (2017), we cannot confirm a difference between the outflow and jet axes. However, the inclination of the disk and thus the direction of gas flow ejected from the disk would depend on their scale. Thus, it is possible to expect a difference between the jet and the outflow axes in the nested disk wind scenario.

In our observation, the outflow dynamical timescale is approximately three times longer than that of the jet (see Table 3), hence the outflow is considered to be more evolved. We also consider the possibility that a disk precesses in a short timescale of ≲10³ yr. Since the outflow traces the mass ejection history for the last ∼10³ yr (dynamical timescale), the change of the outflow axis can be interpreted as the change of the normal direction of the disk (i.e., precession) where the outflow is driven. On the other hand, since the gas associated with the jet traces only the mass ejection history around the protostar in the recent ∼10² yr, the direction of the jet indicates the instantaneous directions of the angular momentum and the poloidal magnetic field of the disk in the vicinity of the protostar, which would not be related to a long-term precessing motion observed in the long-lived outflow.

In the nested disk wind scenario, the jet appears immediately after protostar formation. Thus, the protostellar age tends to be the same as the dynamical time of the jet. However, when the jet age is very young, there is also the possibility that we only detected the recently driven jets and missed the preexisting jets. Although we cannot deny the possibility of the disappearance of preexisting long-lived jets, we could not find any signature of the high-velocity component (jet) larger than the low-velocity component (outflow). In summary, although the jet entrainment scenario could not completely be ruled out, there is useful evidence to support the nested disk wind scenario such as the difference in size, the dynamical timescale, and the axis between the outflow and the jet. Our observation implies that the outflow and jet are driven by a different radius as expected in the nested disk wind scenario.

As shown in Figure 7, we found several knots in the high-velocity jet (or EHV flow). These types of knots are confirmed in other EHV flows observed by molecular line emission (Santiago-García et al. 2009; Kwon et al. 2015). Usually these knots are explained by episodic mass ejection from the region near the protostar (e.g., Stone & Norman 1993). The CO jet observed in MMS 5 is comparable to the smallest known EHV flows (e.g., ∼5000 au in Hirano et al. 2010), and the SiO jet is smaller than these flows by a factor of 3–5. Thus, our target is one of the youngest among known EHV flows.

In order to analyze these knots, we superimposed the SiO integrated intensity map on the CO integrated intensity map in Figure 9(a). We identified six knots from both SiO and CO emissions and estimated the spacing between neighbor knots, as listed in Table 4. The average and median spacings are measured to be Δθ ≃ 0.46 and 0.36 arcsec, which correspond to ΔL = Δθ · d/cos i ≃ 280 and 220 au, respectively, with i = 50°. The intrinsic jet velocity is estimated to be v = v_jet,obs/sin i ≃ 70 km s⁻¹/sin 50° ≃ 90 km s⁻¹. Considering that the knots are caused by the episodic mass ejection, the protostar ejects gas every ∼9–12 yr.

Zoom In Zoom Out Reset image size

In addition, as presented in Figure 9(b), the knots appear to wiggle within the jet. The wiggling is fitted by a simple sine curve in Figure 9(c), which may be precession of the jet or disk around the protostar. Note that, as seen in Figure 9, the knots are roughly aligned along the jet axis except for knot E. Note also that although we fitted the wiggling as a simple sine curve for simplicity, we need further high-resolution observations to more precisely determine the mechanism. Assuming that the sine curve covers one cycle of the precession, the period of the cycle can be estimated as P ∼ 50 yr (∼jet one cycle length/jet velocity v_jet). Assuming the Keplerian disk around the protostar, we can estimate the typical radius inducing the precession. Using the Keplerian rotation period P = 2π/Ω and Keplerian angular velocity ${\rm{\Omega }}={({{GM}}_{* }/{r}^{3})}^{1/2}$, we can estimate the precession radius as

$$\begin{eqnarray}&&{r}_{\mathrm{disk}}=6.3\,{\left(\displaystyle \frac{{M}_{* }}{0.1{M}_{\odot }}\right)}^{1/3}\,{\left(\displaystyle \frac{P}{50\mathrm{yr}}\right)}^{-2/3}\,\mathrm{au},\end{eqnarray} \tag{ 5 }$$

to fit the rotation period as P ∼ 50 yr. When the protostellar mass of 0.1 M_⊙ is assumed,⁷ the precession radius becomes approximately 6 au. We cannot at this time specify the cause of the existence of the precession motion at the radius r_disk. However, there are some possibilities or expectations. The radius r_disk may correspond to the size of the rotationally supported disk. Alternatively, the gravitational instability or magnetic dissipation may be induced at r_disk (Machida 2014). Moreover, a binary companion may exist around r_disk. We need further high-resolution observations to identify the cause of the precession.

Finally, we estimated the jet launching radius based on the jet velocity. We assume that the jet velocity approximately corresponds to the Keplerian rotation velocity at the launching radius (Kudoh & Shibata 1997),

$$\begin{eqnarray}&&{v}_{\mathrm{Kep}}{}^{2}=\displaystyle \frac{{{GM}}_{* }}{{r}_{\mathrm{jet}}}.\end{eqnarray} \tag{ 6 }$$

From Equation (6), assuming v_Kep is equal to the intrinsic jet velocity v_jet, the launching radius is estimated as

$$\begin{eqnarray}&&{r}_{\mathrm{jet}}=1\,\left(\displaystyle \frac{{M}_{* }}{0.1\,{M}_{\odot }}\right)\,{\left(\displaystyle \frac{{v}_{\mathrm{jet},\mathrm{obs}}}{70\mathrm{km}{{\rm{s}}}^{-1}}\right)}^{-2}\,{\left(\displaystyle \frac{\sin 50^\circ }{\sin i}\right)}^{2}\,{R}_{\odot }.\end{eqnarray} \tag{ 7 }$$

Although there are uncertainties in deriving Equation (7), such as the inclination angle and the protostellar mass, our result indicates that the jet appears near the surface of the protostar.