GAS KINEMATICS AND THE DRAGGED MAGNETIC FIELD IN THE HIGH-MASS MOLECULAR OUTFLOW SOURCE G192.16−3.84: AN SMA VIEW

Hauyu Baobab Liu; Keping Qiu; Qizhou Zhang; Josep M. Girart; Paul T. P. Ho

doi:10.1088/0004-637X/771/1/71

1. INTRODUCTION

The ultracompact (UC) H ii region G192.16−3.84 (cf. Molinari et al. 1996) is a well studied B3 star-forming region at a distance of 1.52 kpc (Shiozaki et al. 2011).⁵ The associated 2.1 μm point source (Indebetouw et al. 2003) was found to be embedded in a dense molecular core, which contains 290 M_☉ within a 0.11 pc radius (Shepherd & Kurtz 1999). This source emanates a 50 M_☉ bipolar CO outflow extending 2 farcm 5 (1.1 pc) east and 1 farcm 5 (0.66 pc) west (Snell et al. 1990; Shepherd & Churchwell 1996; Shepherd et al. 1998), which created the biconical cavity (Hodapp 1994; Shepherd et al. 1998; Indebetouw et al. 2003). The bipolar outflow can be further traced by Hα, [S ii], and 4.5 μm emission knots up to ±4 pc away (Devine et al. 1999; Qiu et al. 2008).

Interferometric observations of the centimeter and the millimeter continuum emission toward the 2.1 μm source suggest the existence of circumstellar gas and dust with a total mass of 4–18 M_☉ within a 2'' × 1'' (3040 AU × 1520 AU) region (Shepherd & Kurtz 1999; Shepherd et al. 2001; Shiozaki et al. 2011). At the ≲700 AU scale, the Very Long Baseline Array and very long baseline interferometry (VLBI; Japanese VLBI Network (JVN) and VLBI Exploration of Radio Astrometry (VERA)) observations of the 22 GHz H₂O maser suggested the Keplerian motion of the dense gas and the outward motion of the bipolar outflow (Shepherd et al. 2004; Imai et al. 2006; Shiozaki et al. 2011; see also Codella et al. 1996).

This source was selected for the Submillimeter Array (SMA) observations of dust polarization because it represents one of the clearest cases of a massive disk/outflow system similar to the low-mass star formation (Shu et al. 1987; see also Zhang et al. 1998; Cesaroni et al. 2005; Sridharan et al. 2005; Su et al. 2007; Keto & Zhang 2010 and references therein for another massive case: IRAS 20126+4104 at D = 1.7 kpc, M_* = 7–10 M_☉). In addition, its bright thermal continuum flux at the 0.85 mm wavelength band (2.1 ± 0.63 Jy in the central 15'' area; Shepherd et al. 2004) fulfills the required signal-to-noise ratio. Observations of the dust polarization vectors will provide the complementary aspect of the relative importance of the magnetic field strength.

The observing parameters for the target G192.16−3.84 are introduced in Section 2. The observational results for G192.16−3.84 are presented in Section 3. A brief discussion is provided in Section 4.

2. OBSERVATIONS AND DATA REDUCTION

We performed observations in the 0.88 mm wavelength band in the single receiver polarization mode using the SMA⁶ (Ho et al. 2004; Marrone 2006). The phase referencing and pointing center of the observations is R.A. = $5^{{\rm h}} 58^{{\rm m}} 13\buildrel{\mathrm{s}}\over{.}549$ , decl. = 16°31'58 farcs 30 (J2000). The primary beam size of these observations was 36''. The observations tracked the frequency of 345.796 GHz at window 22 in the upper sideband. The spectral channel spacing was 0.7 km s⁻¹. More details about the observations are summarized in Table 1. The detected spectral lines in these observations are summarized in Table 2.

Table 1. Summary of the SMA Observations

Observing Dates	2011 Nov 10	2012 Jan 6	2012 Feb 2
Array configuration	Compact	Subcompact	Extended
τ_225 GHz	0.05	0.06	0.05
Number of antennas	8	6	7
Time on target loop (hr)	3.5	3.5	3
Flux calibrator	Titan	Ganymede	Callisto
Passband calibrator	Uranus/3C84	3C279	3C279
Gain calibrator	0530+135	0530+135	0530+135
Polarization calibrator	3C84	3C279	3C279
Stokes-Q,U rms (mJy beam⁻¹)	2	3	2

Notes. The listed rms noise levels of the Stokes-Q and Stokes-U images are estimated based on natural weighting images. The Stokes-I image is limited by the dynamic range, and has an higher noise level (see Figure 1).

Download table as: ASCII Typeset image

Table 2. Table of the Detected Molecular Lines

Species	Transition	Frequency	E_u/k	log₁₀A_ij	Note
Species	Transition	(GHz)	(K)	(s⁻¹)	Note
SO₂	4_{3, 1}–3_{2, 2}	332.50524	31	−3.48
	8_{2, 6}–7_{1, 7}	334.67335	43	−3.90
	13_{2, 12}–12_{1, 11}	345.33854	93	−3.62	Blended
	19_{1, 19}–18_{0, 18}	346.65217	168	−3.28
	16_{4, 12}–16_{3, 13}	346.52388	164	−3.47	Blended
SO	9₈–8₇	346.52848	79	−3.27	Blended
	8₈–7₇	344.31061	87	−3.28
CO	3–2	345.79599	33	−5.69
CH₃OH	7_{1, 7}–6_{1, 6}++	335.58200	79	−3.79
H¹³CN	4–3	345.33976	41	−2.69	Blended
H¹³CO⁺	4–3	346.99834	42	−2.49

Notes. We omitted analyzing the blended SO₂ 16_{4, 12}–16_{3, 13} and SO 9₈–8₇ lines.

Download table as: ASCII Typeset image

The absolute flux, passband, and gain calibrations were carried out using the MIR IDL software package. The typical SMA observations may be subject to up to 15% of uncertainty in absolute flux scales. The calibrations of the polarization leakage (i.e., the D-term calibration) were carried out using the MIRIAD software package. We averaged the line-free channels in the lower and upper 4 GHz side bands to generate the continuum channels, and imaged the continuum emission jointly. We carried out the imaging of the Stokes-Q and the Stokes-U components using MIRIAD, and carried out the imaging of the spectral line emission and the Stokes-I component of the continuum emission using the CASA software package.

The minimum and maximum baselines of our observations are ∼8 kλ and ∼260 kλ, respectively. Our SMA observations were therefore sensitive to a maximally detectable scale of ∼15 farcs 8 (Wilner & Welch 1994). Image synthesis using the Briggs Robust 0 weighting for the data collected from these observations yields a synthesized beam of 1 farcs 3 × 0 farcs 87 (P.A. = 90°), and a root-mean-squares (rms) noise level of 110 mJy beam⁻¹ (0.96 K) in each 0.7 km s⁻¹ velocity channel; imaging using the naturally weighted visibilities yields a synthesized beam of 1 farcs 9 × 1 farcs 4 (P.A. = 96°), and an rms noise level of 78 mJy beam⁻¹ (0.30 K) in each 0.7 km s⁻¹ velocity channel.

3. RESULTS

3.1. Geometry

3.1.1. The 0.88 mm Continuum Emission

The SMA Stokes-I image (Figure 1) recovers 1.2 Jy of flux in the central 15'' area. The previous James Clerk Maxwell Telescope measurement of 2.1 ± 0.63 Jy in the central 15'' area (Shepherd et al. 2004) indicates that the SMA Stokes-I image is subject to 43 $^{+13}_{-25}$ % of missing flux (see also Section 2). The averaged intensity of the missing flux over the central 15'' area is 1.3–7.4 mJy beam⁻¹, which is lower than 62% of the first contour level in Figures 1 and 2. We therefore argue that the observed geometry from the 0.88 mm continuum Stokes-I emission is only minimally affected by the missing flux issue. According to the spectral energy distribution reported by Shepherd et al. (2004), the 0.88 mm continuum emission is mainly contributed by the dust thermal emission, as well as the free–free emission from the ionized gas surrounding the central star. The ≲ 1'' scale UC H ii region has already been detected in centimeter continuum emission (Hughes & MacLeod 1993; Shepherd & Kurtz 1999; Shepherd et al. 2001), peaking at $05^{{\rm h}}58^{{\rm m}}13\buildrel{\mathrm{s}}\over{.}531$ , $16^{\circ }31^{{\rm m}}58\buildrel{\mathrm{s}}\over{.}29$ (J2000). Based on the measured spectral index of 0.3 ± 0.07 (Shepherd et al. 2004) and the 1.5 mJy total flux at the 3.6 cm band (Shepherd & Kurtz 1999), we estimate that the free–free continuum emission contributes to 3.6–6.0 mJy at the 0.88 mm wavelength band. Compared with the first contour level of the dust thermal emission at 0.88 mm in Figures 1 and 2, the free–free continuum emission is negligible.

farcs — **Figure 1.** The naturally weighted SMA 0.88 mm continuum image (θ_maj × θ_min = 222 × 174). Contours show the Stokes-I image. Contour levels are 12 mJy beam⁻¹ (3σ) × [1, 2, 4, 8, 16, 32]. The polarization intensity is presented in gray scale. Gray-scale bar is in Jy beam⁻¹ units. The synthesized beam of the Stokes-I,Q,U observations is shown in the lower left. We rotate the polarization vectors by 90° and overplot them at the positions with >2σ detection of the polarization intensity (green: between 2σ and 3σ; yellow: >3σ; σ = 1.4 mJy beam⁻¹). With 2–3σ detections of the polarized emission, assuming the noise characteristic is not largely deviated from the thermal noise, the 1σ uncertainty σ_χ of the polarization position angle χ is ∼11°–7°. The lengths of the polarization vectors are proportional to the polarization fraction. The scale bar in the lower left represents the 10% polarization fraction.
Download figure:
Standard image High-resolution image

**Figure 2.** The velocity integrated intensity map (i.e., moment 0 maps) of CO 3–2, overlaid with the 0.88 mm continuum image (robust 0 weighted). Dark contours show the robust 0 weighted 0.88 mm continuum image (θ_maj × θ_min = 13 × 086), with the levels of 14 mJy beam⁻¹(3σ) × [−1, 1, 2, 4, 8, 16, 32]. The cyan, blue, and red contours show the integrated CO 3–2 emission in the labeled velocity ranges (naturally weighted; θ_maj × θ_min = 18 × 14). The cyan contour levels are 2.0 Jy beam⁻¹ km s⁻¹ × [1, 2, 3]; the blue contour levels are 1.4 Jy beam⁻¹ km s⁻¹ × [1, 2]; the red contour levels are 1.2 Jy beam⁻¹ km s⁻¹ × [1, 2]. The systemic velocity is ∼5.8 km s⁻¹ (see Section 3.2). The dashed circle indicates the 18'' radius of the primary beam. Contours are plotted only in the region encircled by the dashed circle. We plot the spectral profile of the high velocity CO 3–2 components in Figure 4. The velocity integrated intensity map of CO 3–2 (robust 0 weighted; θ_maj × θ_min = 12 × 086) in the intermediate velocity range ([−19, 1] km s⁻¹ and [11, 41] km s⁻¹) is shown by gray scale, while the channel maps in this velocity range is presented in Figure 3.
Download figure:
Standard image High-resolution image

The SMA Stokes-I continuum image (Figures 1 and 2) shows a ∼6'' scale source, of which the emission peak (the G192 envelope, hereafter) at $5^{{\rm h}}58^{{\rm m}}13\buildrel{\mathrm{s}}\over{.}535$ , 16°31'58 farcs 25 (J2000) is connected with an ∼3'' scale elongation toward the southeast. With the ≳ 1'' angular resolution of our observations (Section 2), the peak of the 0.88 mm Stokes-I emission cannot be distinguished from the centimeter continuum peak.

3.1.2. Polarized Dust Emission

We detect significant polarized continuum emission toward the peak of the Stokes-I emission, and detect marginal polarized continuum emission at two positions located in the southeastern elongation. The B-field directions⁷ at the peak of the Stokes-I emission are aligned in the north–south direction. The B-field directions at the other two detections are roughly parallel to the southeastern elongation.

3.1.3. The CO 3–2 Line

The most significant structure detected in CO 3–2 is the biconical cavity walls, centering around the peak of the Stokes-I emission (Figure 2). The channel maps of CO 3–2 show the east–west bipolar cavity walls (Figure 3), which is consistent with the earlier observations of CO 1–0 (Shepherd et al. 1998). There is no zero-spacing information available, and the CO 3–2 maps for the spatially extended cavity walls cannot constrain its column density. The CO 3–2 emission at the higher velocity ranges (e.g., |δv_lsr| ≫ 30 km s⁻¹) shows redshifted (relative to the systemic velocity 5.8 km s⁻¹; see also Section 3.2) gas west of the peak of Stokes-I continuum emission, and shows blueshifted gas around and toward the east of the peak of Stokes-I continuum emission. The spectra of the high velocity CO gas are presented in Figure 4.

**Figure 3.** The velocity channel maps of CO 3–2 in the intermediate velocity range (robust 0 weighted). The red star marks the peak of the 0.88 mm continuum emission (R.A. $5^{{\rm h}}58^{{\rm m}}13\buildrel{\mathrm{s}}\over{.}535$ , decl. 16°31'5825). The gray-scale bar has the unit of Jy beam⁻¹ km s⁻¹.
Download figure:
Standard image High-resolution image

**Figure 3.** The velocity channel maps of CO 3–2 in the intermediate velocity range (robust 0 weighted). The red star marks the peak of the 0.88 mm continuum emission (R.A. $5^{{\rm h}}58^{{\rm m}}13\buildrel{\mathrm{s}}\over{.}535$ , decl. 16°31'5825). The gray-scale bar has the unit of Jy beam⁻¹ km s⁻¹.
Download figure:
Standard image High-resolution image

**Figure 4.** The spectra of CO 3–2 centered at the location of the [−109,−79] km s⁻¹ component (cyan), the [−69,−29] km s⁻¹ component (blue), and the [71, 101] km s⁻¹ component as shown in Figure 2. For each component, the spectrum is averaged in the region enclosed by the first contour levels in Figure 2. For comparison, we overplot the SO₂ 19_{1, 19}–18_{0, 18} spectrum (scaled by a factor of 0.3) as a dotted line, and overplot the manually shifted SO₂ 19_{1, 19}–18_{0, 18} spectrum of which the peak is −93 km s⁻¹ (dotted magenta). We overplot the 1.4 km s⁻¹ velocity resolution CO 3–2 spectrum for the [−109,−79] km s⁻¹ component as a black solid line (1σ ∼ 38 mJy beam⁻¹).
Download figure:
Standard image High-resolution image

The [−109, −79] km s⁻¹ component of the CO 3–2 outflow is potentially contaminated by the CH₃OH 16_{1, 15}–15_{2, 14} (ν: 345.90397 GHz; E_u/k: 333 K) and CH₃OH 16_{−3, 16}–15_{−4, 14} (ν: 345.91919 GHz; E_u/k: 459 K) emission from the G192 envelope. Assuming the averaged gas temperature in the G192 envelope to be equal to the SO₂ rotational temperature 84 K (Section 3.1.4) and a constant CH₃OH abundance, the extrapolation of the detected CH₃OH 7_{1, 7}–6_{1, 6} flux indicates that the CH₃OH 16_{1, 15}–15_{2, 14} transition will contribute to an averaged integrated flux density of only ∼0.2 Jy beam⁻¹ km s⁻¹, which is very dim when compared with the detected integrated flux density of CO in this velocity range (Figure 2). The CH₃OH 16_{−3, 16}–15_{−4, 14} transition should be still dimmer. However, we cannot rule out that the abundance of CH₃OH is enhanced in the warmer center of the hot core. The (CH₃)₂CO 15_{8, 8}–14_{5, 9} EA, 15_{8, 8}–14_{5, 9} AE, and 15_{7, 8}–14_{6, 9} AA transitions, which were detected in Orion-KL (Friedel et al. 2005), can contribute to the emission in the velocity range of ∼[−95.3, −93] km s⁻¹, and the ³⁴SO₂ 17_{4, 14}–17_{3, 15} transition (ν: 345.92928 GHz; E_u/k: 179 K) can contribute to the emission at ∼−116 km s⁻¹. With the observations of only one CO transition, we cannot make sure whether the [−109, −79] km s⁻¹ component is dominantly contributed by the CO 3–2 emission, or is very much contaminated by other molecular lines. We will exclude this component in the following discussion.

The rest of the high velocity CO emission can support the existence of the high velocity bipolar molecular outflow in the east–west direction, emanating from the peak of Stokes-I continuum. The powering source(s) of this high velocity gas remains to be checked by more sensitive observations. If the previous fittings of the inclination of the biconical cavity (i = 63°; Shepherd et al. 1998) can be applied to the |δv_lsr| ≫ 30 km s⁻¹ gas, then the detected deprojected outflow velocities can be up to 210 km s⁻¹.

We do not have data to measure the gas temperature in the high velocity CO 3–2 outflows. However, the [71, 101] km s⁻¹ component is spatially close to the G192 envelope (Figure 2). Considering that the outflowing gas may be partially entrained from the G192 envelope, a comparable initial gas temperature with the G192 envelope could be a reasonable assumption, though the cooling and the shock heating are yet to be considered. For simplicity, we assume the averaged gas temperature in all east–west high velocity CO 3–2 outflow components are comparable with that in the G192 envelope. Table 3 gives the summary of estimated energetics of the |δv_lsr| ≫ 30 km s⁻¹ gas without considering the missing flux, based on the assumption of optically thin, local thermodynamic equilibrium (LTE), the gas temperature of 84 K (Section 3.1.4), [CO]/[H₂] = 10⁻⁴, and the inclination angle i = 63° (Shepherd et al. 1998). The energetics of the molecular outflow on the larger (∼1 pc) scale can be found in Shepherd et al. (1998). Dividing the separations of the [ −69, −29] km s⁻¹ and [71, 101] km s⁻¹ CO 3–2 emission from the G192 envelope by their averaged velocities (i.e., Momentum/Mass) implies dynamic timescales of 910 ± 250 yr and 220 ± 80 yr, respectively. The momentum supply rates from these two high velocity components are 0.14 × 10⁻³ M_☉ km s⁻¹ yr⁻¹ and 0.045 × 10⁻³ M_☉ km s⁻¹ yr⁻¹, respectively. The summed momentum supply rate is ≳ 0.19 × 10⁻³ M_☉ km s⁻¹ yr⁻¹, which is ∼6.2% of the momentum supply rate of the parsec scale CO outflow (Shepherd et al. 1998). Besides the two water maser sources (e.g., Shepherd et al. 2004), we are not sure whether there are other (proto)stars embedded in the G192 envelope. Deep JVLA observations to resolve the radio jet cores may allow us to see whether the high velocity molecular outflows are powered by a single dominant source.

Table 3. The Derived CO 3–2 Parameters for the High Velocity Gas

Component		Mass	Momentum	Energy
Component		(10⁻³ M_☉)	(10⁻³M_☉ km s⁻¹)	(10⁴⁴ erg)
[71, 101]	km s⁻¹	0.069	10	0.16
[ −69, −29]	km s⁻¹	1.2	130	1.4
[ −109, −79]	km s⁻¹	0.28	68	1.7

Notes. The momentum and energy are corrected for the inclination angle of i = 63°. The [ −109, −79] km s⁻¹ component is likely to be contaminated by other molecular lines (see Section 3.1), and therefore our estimates should be considered as an upper limit.

Download table as: ASCII Typeset image

3.1.4. Other Molecular Lines

The rest of the observed molecular lines trace a variety of morphology (Figure 5). The observed CH₃OH transition and the SO₂ transitions mainly trace the G192 envelope and a compact component in the southeast of the G192 envelope. The high excitation SO₂ line emission in the southeastern component may imply either a secondary source or the interaction between the outflow and the ambient gas. The H¹³CO⁺ 4–3 and H¹³CN 4–3 trace both the G192 envelope and the southeastern extension. The SO 8₈–7₇ transition appears to trace the V-shape featured by the CO redshifted gas in the west of the G192 envelope, which can be associated with the innermost part of the cavity wall. Based on the integrated flux of the SO₂ 4_{3, 1}–3_{2, 2}, 8_{2, 6}–7_{1, 7}, and 19_{1, 19}–18_{0, 18} transitions in a R = 1 farcs 5 circular region centered on the G192 envelope, assuming it is optically thin and in LTE, using the expression introduced in Fu et al. (2012), we constrain the rotational temperature in the G192 envelope to be $T_{{\rm rot}}^{{\rm SO}_{2}}\sim 84^{+18}_{-13}$ K.

3.2. Velocity Gradient in Dense Gas

The H¹³CO⁺ 4–3 line generically traces the extended dense gas (Figure 5; Section 3.1). Gaussian fitting of the averaged spectrum of H¹³CO⁺ 4–3 suggests that the systemic velocity of the dense gas is ∼5.8 ± 0.4 km s⁻¹. The comparison of the kinematics traced by the H¹³CO⁺ 4–3 and SO₂ 19_{1, 19}–18_{0, 18} lines can provide clues to how the dense gas continues infalling into the inner disk/envelope.

Figure 6 shows the intensity-weighted averaged velocity maps (i.e., moment 1 maps) of SO₂ 19_{1, 19}–18_{0, 18} and H¹³CO⁺ 4–3. The SO₂ 19_{1, 19}–18_{0, 18} transition traces a north–south velocity gradient in a ≲ 2'' region around the 0.88 mm Stokes-I peak. With the angular resolution of our observations, the resolved direction of the velocity gradient is consistent with the rotating plane of the disk/envelope reported by the water maser observations (Shepherd & Kurtz 1999; Shepherd et al. 2004; Imai et al. 2006). In a slightly bigger area, the H¹³CO⁺ 4–3 line traces a northwest–southeast velocity gradient. The more extended gas shows the velocity close to the 5.8 km s⁻¹ systemic velocity. The motion discussed here is also consistently traced by the other molecular lines presented in Figure 5.

Figure 7 shows the position–velocity (PV) diagrams of the H¹³CO⁺ 4–3 and SO₂ 19_{1, 19}–18_{0, 18} lines. The PV cuts are centered at the peak of the continuum emission [ $5^{{\rm h}}58^{{\rm m}}13\buildrel{\mathrm{s}}\over{.}535$ , 16°31'58 farcs 25 (J2000)]. The comparison of the SO₂ 19_{1, 19}–18_{0, 18} and H¹³CO⁺ 4–3 PV diagrams at P.A. = 0° may imply an accelerated rotation of the warmer gas, which should lie closer to the embedded B3 star. The PV diagrams at P.A. = 130° show a bulk of blueshifted gas in the southeast (0''–5''). With the 0.7 km s⁻¹ velocity resolution of our observations, we cannot resolve a clear trend of motion in that blueshifted gas component.

We jointly analyze the velocity field presented in the PV diagrams (P.A. = 0°) by applying the terminal velocity method (Sofue & Rubin 2001). This method defines the measured terminal velocity v_t by a velocity at which the intensity I equals

$I_{t} = \eta I_{{\rm max}},$

where η is an adjustable parameter that can be optimized by checking the consistency of results from different molecular lines (e.g., Liu et al. 2010). We consider η = 0.2 to be an optimal choice to analyze our data, which suppresses the confusion of gas from outer radii while guaranteeing that all pixels picked up by this method are above the 3σ detection level. We note that relative to the naive choice of η = 1.0, our choice of η = 0.2 would potentially overestimate the rotational velocity. Figure 8 presents the η = 0.2 results up to a radius that the terminal velocity still can be distinguished from the systemic velocity 5.8 km s⁻¹. By assuming the plane of rotation⁸ lies along the axis with P.A. = 0°, the measured terminal velocities within the 0.01 pc (1 farcs 36) radius are consistent with the inclined (i = 63°) Keplerian rotation, which is bound by the gravitational force of the embedded mass M_{* + gas + dust} ∼ 11.2–25.2 M_☉ (Shepherd & Kurtz 1999; Shepherd et al. 2001, 2004; Shiozaki et al. 2011). At the >0.01 pc radii, the steeper decline of the measured terminal velocities than the rotational velocities of the Keplerian models may indicate that the radial infall motion is not negligible in the more extended region (see the discussion in Tobin et al. 2012), although it is also consistent with the changes in the inclination angle. The previous Very Large Array observations of the 22 GHz water maser further traced the [−7, 16] km s⁻¹ Keplerian rotation inward of the ∼0 farcs 25 (360 AU) radius (Shepherd & Kurtz 1999). These data consistently indicate the centripetally accelerated rotation toward the young star. Whether or not a Keplerian disk exists can be checked by the higher angular resolution observations.

**Figure 8.** The detected terminal velocity (relative to the v_lsr = 5.8 km s⁻¹) and the projected Keplerian rotation curve. The origin (i.e., radius 0.0 pc) is defined at $5^{{\rm h}}58^{{\rm m}}13\buildrel{\mathrm{s}}\over{.}535$ , 16°31'5825 (J2000). The black symbols show the terminal velocities traced by the H¹³CO⁺ 4–3 transition. The blue symbols show the terminal velocities traced by the SO₂ 19_{1, 19}–18_{0, 18} transition. We use different symbols to distinguish the values measured from the positive angular offset (i.e., south) and from the negative angular offset (i.e., north). The error bars reflect the uncertainties caused by the limited 0.7 km s⁻¹ velocity resolution. The horizontally hatched region is bound by the two projected (inclination i = 63°) Keplerian rotation curve models, which are bound by the gravitational force of the enclosed stellar and molecular gas mass of 11.2–25.2 M_☉.
Download figure:
Standard image High-resolution image

Both Figures 7 and 8 present the spatial asymmetry of the brightness distribution and the velocity field in H¹³CO⁺ 4–3. The spatial asymmetry of the brightness distribution is consistent with the dust continuum image (Figure 1) that also shows a southeastern extension. However, we cannot rule out the possibility that the observed asymmetry of the velocity field is due to the variation of the excitation condition, abundance, or is caused by the more complicated radiation transfer effects.

4. DISCUSSION

We observed the 0.88 mm polarized thermal dust emission and the molecular line emission which trace a range of excitation conditions toward the B3 star-forming region G192.16−3.84. The kinematics traced by the observed molecular lines as well as the previous detections of the 22 GHz water maser suggest a Keplerian motion of dense gas, continuing from the <0.02 pc radius to inward of the ⩽290 AU radius. The magnetic field lines at 0.02–0.05 pc scale may have already been twisted by the rotational motion, such that the B-field directions are parallel to the plane of rotation. The high velocity molecular outflow is observed to be perpendicular to the rotational plane of the dense gas. Supposedly, the high velocity molecular outflow is driven by the magnetocentrifugal wind over the past ≳10³ yr outflow dynamical timescale. The magnetic field lines at the sub-AU scale may still be organized in a way such that they can drive the east–west bipolar wind. We refer to Hull et al. (2013) for the discussion about the misalignment of magnetic fields and outflows in lower mass protostellar cores.

The lower polarization fraction at the Stokes-I continuum peak (Figure 1) can be explained by the blended magnetic field lines in the inner and outer radii (Tang et al. 2009; Frau et al. 2011; Padovani et al. 2012), though such an interpretation is not unique. This "de-polarization effect" can alternatively be explained by the difference in dust grain properties or grain alignment efficiency (Lazarian 2007), although the significance of this effect is not yet robustly argued by the SMA case studies. We tentatively explain the non detection of the polarized intensity ∼1'' northwest and southeast of the Stokes-I continuum peak by the nearly parallel to line-of-sight orientation of the magnetic field lines dragged by the rotational velocity field. We note that the B-field morphology is essentially unresolved. In an edge-on geometry, if the B field is pulled in by the contraction process, with an hour-glass morphology (e.g., Greaves et al. 1994; Girart et al. 2006, 2009; Rao et al. 2009), most of the unresolved B field lines will be aligned along the line-of-sight, or blended and canceled across the line of sight. The only remaining field detectable in such a case might be the field lines that exit from the top and bottom of the edge-on disk, and hence are aligned parallel to the disk. This would be an alternative to the scenario where the field lines are rendered toroidal by rotational motions. Figure 9 shows the schematic pictures for the proposed scenarios.

**Figure 9.** Schematic pictures illustrating the observed B field (see also Section 4). The shaded areas indicate the inclined thin disk or pseudo disk around the central stars. The B field lines are shown by the thick curves with arrows. The dashed circles labeled with A–E are the observational (synthesized) beams. In beams A and C, the B field lines are nearly parallel to the line-of-sight, thus the dust emission is only weakly polarized. In beam D, the observed polarization percentage can be low because of canceling. In beams B and E, we can detect the B field aligned in the disk plane.
Download figure:
Standard image High-resolution image

Our observations indicate that the B3 (proto)star embedded in G192.16−3.84 can form via a process similar to the scaled-up solar-mass-type star formation. With the sensitivity of the current instruments, observation with a large number of (fainter) samples is not yet possible. The proposed scenario is therefore subjected to the concern of target selection bias. Furthermore, in the environments of the more crowded OB clusters, whether or not the sources like G192.16−3.84 are representative remains an open question.

The SMA data were taken as part of the Large SMA Dust Polarization Survey (PI: Qizhou Zhang). We acknowledge the supports from the SMA staffs. H.B.L. thanks Vivien H.-R. Chen for useful discussions. J.M.G. is supported by the Spanish MINECO AYA2011-30228-C03-02 and the Catalan AGAUR 2009SGR1172 grants.

Facility: SMA - SubMillimeter Array

GAS KINEMATICS AND THE DRAGGED MAGNETIC FIELD IN THE HIGH-MASS MOLECULAR OUTFLOW SOURCE G192.16−3.84: AN SMA VIEW

Article metrics

Permissions

Author affiliations

Dates

ABSTRACT

1. INTRODUCTION

2. OBSERVATIONS AND DATA REDUCTION