Interferometric Mapping of Magnetic Fields: The massive star forming region G34.4+0.23 MM
aa r X i v : . [ a s t r o - ph ] J a n Interferometric Mapping of Magnetic Fields: The massive starforming region G34.4+0.23 MM
P. C. Cortes
Departamento de Astronom´ıa, Universidad de Chile, Casilla 36-D Santiago, Chile
R. M. Crutcher
Astronomy Department, University of Illinois at Urbana-Champaign, IL 61801, USA
D. S. Shepherd National Radio Astronomy Observatory, P.O. Box O, 1003 Lopezville Rd, Socorro, NM87801.
L. Bronfman
Departamento de Astronom´ıa, Universidad de Chile, Casilla 36-D Santiago, Chile
ABSTRACT
We report millimeter interferometric observations of polarized continuum andline emission from the massive star forming region G34.4. Polarized thermaldust emission at 3 mm wavelength and CO J = 1 → → II region,which are the central sources in G34.4, suggesting a magnetic field orthogonal tothis axis. This morphology is consistent with a magnetically supported disk seenroughly edge-on. Subject headings:
ISM: magnetic fields ISM:polarization stars: formation The National Radio Astronomy Observatory is a facility of the National Science Foundation operatedunder cooperative agreement by Associated Universities, Inc.
1. Introduction
It is generally accepted that magnetic fields play an important role in the process ofstar formation; magnetic fields are involved in cloud support, fragmentation, and transferof angular momentum. However, the magnetic field is the least observed physical quantityinvolved in such process. Magnetic field observations of molecular clouds are divided intomeasurements of the line-of-sight component of the magnetic field strength through theZeeman effect and observations of the field in the plane of the sky through linear polarizationof dust emission and spectral-line emission. The alignment of dust grains by a magnetic fieldis physically complicated and is still a matter of intense research. It is accepted, though,that aligned dust grains will produce polarized emission perpendicular to the projection ofthe magnetic field onto the plane of the sky. For a recent review of alignment theories seeLazarian (2007).Spectral-line linear polarization has been suggested to arise from molecular clouds underanisotropic conditions, like large velocity gradients (or LVGs) (Goldreich & Kylafis 1981).The prediction suggests that a few percent of linearly polarized radiation should be detectedfrom molecular clouds and circumstellar envelopes in the presence of a magnetic field. Thispolarization will be either parallel or perpendicular to the projection of the field onto theplane of the sky. To obtain a qualitative understanding about this effect, consider the COmolecule emitting unpolarized radiation. Under a magnetic field, a CO molecule will developa small splitting in its rotational, J , energy levels. These magnetic sub-levels will produceradiation components labeled σ for the | M − M ´ | = 1 transitions, and π for the | M − M ´ | = 0 transitions, where both components can be linearly polarized either perpendicular orparallel to the magnetic field. If the gas behaves under isotropic conditions (e.g. no velocitygradients) for any direction the σ and π will populate equally, so the radiation componentsemerging from the radiative decays of these states will combine to give zero net polarization.Now on the other hand, large velocity gradients present in molecular clouds will produceanisotropies in the optical depths for the CO molecular transitions at different directions. Ifthe velocity gradients are smaller in directions parallel to the magnetic field than in thoseperpendicular to the field, the optical depths parallel to the field will be larger than theoptical depths perpendicular to them. Therefore, the escape of radiation involved in de-exciting the upper J state will be then reduced more in directions along the field linesthan in directions perpendicular to them, which will lead to populations of the magnetic σ sub-states that are larger than the populations of the π sub-states due to the differencein the angular distributions of both radiation components. The angular distribution of the σ component peaks in directions along the field lines whereas the π component peaks indirections perpendicular to the magnetic field. In this way, the rate of de-excitation for σ will have a larger decrease, due to photon trapping, than the rate of de-excitation for 3 –the π radiation component. Because, in this picture, the σ component will have a largerpopulation, its emission will be stronger relative to the π component giving raise to a smallamount of linear polarization in the CO emission with the polarization of the σ component,or perpendicular to the magnetic field. The effect was first detected by Glenn et al. (1997) forthe CS molecules while Greaves et al. (1999) detected CO polarized emission for ( J = 2 → J = 3 →
2) transitions.In order to efficiently map polarized emission and infer detailed information about themagnetic field morphology, high resolution observations are required. The BIMA millimeterinterferometer has been used previously to obtain high-resolution polarization maps in sev-eral star forming cores (Lai 2001; Lai et al. 2002, 2003; Cortes et al. 2005; Cortes & Crutcher2006; Cortes et al. 2006). These previous results show fairly uniform polarization morpholo-gies over the main continuum sources, suggesting that magnetic fields are strong, and there-fore cannot be ignored by star formation theory. However, the number of star formationregions with maps of magnetic fields remains small, and every new result is statisticallysignificant. In this work we present polarization maps of the massive star forming regionG34.4, obtained with the BIMA array. We measured continuum polarization at 3 mm andCO J = 1 →
2. Source Description
G34.4 is a newly discovered massive star forming region. It is associated with the IRAS18507+0121 point source, which is located at 3.9 kpc from the Sun, having a v lsr = 57 kms − . Towards the IRAS 18507+0121 point source, Bronfman et al. (1996) detected strongCS( J = 2 →
1) emission with broad line wings that suggested massive star formation.The source is roughly 11 ′ from the H II region complex G34.3+0.2 (Molinari et al. 1996).Observations with the 45-m Nobeyama radio telescope at a resolution of 16 ′′ of HCO + ,H CO + , CS, and C S were presented by Ramesh et al. (1997). Their modeling showed thatthe observed line profiles are representative of a collapsing warm (22 K) core with a mass of800 M ⊙ , hidden behind a ≈ M ⊙ . This region is also associatedwith H O maser emission (Scalise et al. 1989; Palla et al. 1991; Miralles et al. 1994) andCH OH maser emission (Schutte et al. 1993; Szymczak & Kus 2000). Fa´undez et al. (2004)observed this region in the continuum at 1.2 mm with the SEST telescope and derived a massof 2 × M ⊙ . Shepherd et al. (2004) performed the first interferometric observations with 4 –OVRO of the 3-mm continuum, H CO + ( J = 1 → v = 0 , J = 2 → ∼ ′′ (see Figure 1). They also presented near-infrared observations at J, H, and K ′ ( λ c = 1 . µm , respectively). The central source in Figure 2 (labeled G34.4+0.23 MM)showed no trace of NIR emission, but the southern source seemed to be associated with aNIR cluster of young stars and an ultra-compact H II region. Based on the emission fromwarm dust and the lack of NIR emission, the central source was suggested to be a massiveproto-star. From the VLA archive, marginal emission (0.7 mJy) of 6-cm radio continuumwas detected from this source (Shepherd et al. 2004). Rathborne et al. (2005) made a multi-wavelength study of this region. They observed continuum emission at 1.2 mm, 850 µ m,450 µ m, and 350 µ m by using IRAM, JCMT, and the CSO observatories respectively. Theyalso obtained archival data from the SPITZER telescope at 2.4 µ m, 8 µ m, 4.5 µ m, and3.6 µ m and produced combined maps of infrared continuum emission. Their data agreewell with previous observations and positioned the infrared sources at the center of the mmand submillimeter emission. In a recent study, Shepherd et al. (2007) presented a detailedstudy of the G34.4 region. They discovered five massive outflows from two of the existingcloud cores in G34.4. Three outflows are centered near the ultra-compact H II region whilethe remaining two are centered at the MM core. By using mid-IR data from the Spitzertelescope, Shepherd et al. (2007) identified a total of 31 YSO in the G34.4 complex with acombined mass of ∼
127 M ⊙ plus an additional 22 sources that might be cluster membersbased on strong 24 µ m emission.
3. Observation Procedure
We observed G34.4+0.23 MM in May 2004 in the 3-mm continuum and the CO J = 1 → ′′ ) was obtained. The digital correlator was set up to observe both the continuumand the CO J = 1 → − ). The 50 MHz windowfor CO was cut from the continuum window to avoid contamination of the continuum by theCO line and reduced independently. In order to detect circular polarization, a quarter-waveplate was placed in front of the single receiver at each BIMA antenna to select either right(R) or left (L) circular polarization. A second quarter-wave plate grooved orthogonally tothe first was alternately switched into the signal path to observe sequentially both circularpolarizations. Switching between polarizations was sufficiently rapid (every 11.5 seconds)to give essentially identical uv-coverage. Cross-correlating the R and L circularly polarized 5 –signals from the sky gave RR, LL, LR, and RL correlations for each interferometer baseline,from which maps of the four Stokes parameters were produced. The quasars 1751+096 and1743-038 were used as calibrators for G34.4. The instrumental polarization was calibratedby observing the 3C279 quasar, and the “leakages” solutions were calculated from this ob-servation. The calibration procedure is described by Lai (2001). The Stokes images I, U,Q and V were obtained by Fourier transforming the visibility data using a robust weightingscheme (Briggs et al. 1999). Deconvolution in the Stokes I cube was done by applying amaximum entropy algorithm to every channel. The MIRIAD (R. J. Sault, N. E. B. Killeen1998) package was used for data reduction.
4. Observational Results4.1. 3-mm Continuum
The 3-mm continuum results are shown in Figure 2. The beam has a major axis of17.6 ′′ and a minor axis of 15.7 ′′ . The significance level cutoff chosen for the polarizationresults is 3 σ , where σ refers the noise level in the polarized flux image, values below 3 σ areblanked in the data analysis process. The strongest feature of the 3 mm continuum resultis the main compact source centered at ( α, δ ) = (18 : 53 : 18 ,
01 : 25 : 25); this structure isin agreement with the 3 mm continuum result of Shepherd et al. (2004) at higher resolutionand corresponds to their MM (millimeter) core. Some additional structure is seen along thenorth-south axis of the cloud. However, interferometric observations of equatorial sourcescan produce ghost structures along the north-south axis, due to the strong side-lobes thatappear in the synthesized beam at such declinations. In the case of our source, the emissionseen south of the MM core appears to be associated with IRAS 18507+0121 (or the UC H II region with a peak emission centered at ( α, δ )=(18:53:19.5, 01:24:45)), a source that isclearly seen in Figure 1.The polarization observed in our continuum map is within the MM core and the UCH II region, with the highest polarized flux at the MM core. This source has a peak fluxof 273 mJy beam − , an integrated flux of 290 mJy (calculated over a box of 15 ′′ × ′′ centered at the reference position), and a peak polarized flux P of 83 mJy beam − ; the Iand P peaks approximately coincide. The Stokes I image shows some filamentary structurealong the north-south axis between the MM core and the UC H II region, consistent with thegeneral cloud morphology (see Figure 1); while the polarized flux is concentrated around thecores. The average P.A. of the polarized continuum map is -8 ◦ ± ◦ ; suggesting polarizationalong the north-south axis of the filament. The fractional polarization seems uniform overboth cores, with an average value of 0.3 ± .
07. However, interferometric observations that 6 –do not fully sample the u-v plane do not produce reliable fractional polarization results,often overstating the fractional polarization, but not significantly affecting the polarizationposition angle.The Shepherd et al. (2004) continuum observations, centered at 90 GHz, yielded a totalflux of 56 mJy, about 4 times less than our value. Our observations have an rms noiselevel of 6 mJy beam − with an uncertainty in the calibration of 25%, while Shepherd et al.(2004) achieved an rms noise level of 3 mJy beam − with a calibration uncertainty of 15%.Assuming an error of 3 σ and taking into account the uncertainty in the calibration, thelower limit for our measured flux is 200 mJy. By the same argument, the upper limit for theflux measured by Shepherd et al. (2004) is 75 mJy. The remaining ratio of 2.2 between thetwo measurements is due mainly to the different frequencies. Emission from a blackbodywill produce larger values at higher frequencies in the millimeter part of the electromagneticspectrum; for example, the flux at 115 GHz will be about twice that at 90 GHz for a50 K source. The small remaining difference could be explained by missing flux in thehigher resolution interferometric observations of Shepherd et al. (2004) due to missing shortspacings in u-v space. Our compact BIMA D array observations have fairly good short-spacing coverage, with shortest baselines of ∼ ∼ ′′ in size.Emission in the continuum at 3 mm may be contaminated by free-free emission, but itappears not to be the case here. Shepherd et al. (2004) found marginal free-free continuumemission in the 6 cm band that is estimated to be 0.52 mJy at 3 mm. This free-free emissionis unpolarized and negligible when compared with our total flux at 3 mm. However, syn-chrotron emission, which is strongly linearly polarized, could be present. The polarization ofsynchrotron radiation in the case of a homogeneous magnetic field can achieve a level of 72%for a power law emission of n = 0 .
75 (Rohlfs & Wilson 2004). The total level of emissionfound at 6 cm was only 0.7 mJy, which would yield linearly polarized radiation about 0.5mJy. However, due to the ν − n scaling law for synchrotron emission, its contribution to thepolarized flux at the 3 mm band is negligible. Therefore, the contamination in our polarizedflux appears to be minimal. This strongly suggests that we are seeing only polarized emissionfrom dust. J = 1 → J = 1 → J = 1 → − for the red lobe and 38to 54 km s − for the blue lobe). Individual channel maps are shown in Figures 4 and 5 (in-cluding the polarization). Considering the larger beam of our BIMA D-array observations,which will smear out the emission, our results are in agreement with the higher resolutionobservations of Shepherd et al. (2007). Although, we did not resolve all of the individualoutflows detected by their work, it appears that the massive outflow in the MM core (outflowA) is dominating the emission in our observations. Our peak CO flux is calculated aroundthe MM source, centered at ( α, δ )=(18:53:17.3, 01:25:15), and taking a value of ∼
14 Jybeam − . Table 1 presents the average P.A. for all channels associated with polarized COemission. From Table 1 we see polarized emission at higher velocities (105 to 97 km s − )and at more intermediate velocities associated with the MM outflow. The polarization isobserved to be quite uniform, with an orientation mostly along the north-south axis of thefilament, in all velocity channels shown in Figures 4 and 5. The peaks in polarized fluxand polarization coverage is around 64 to 62 km s − , which coincides with the peaks in COemission. This uniform polarization pattern is similar to the polarization orientation of our3 mm continuum results. Taking into consideration that molecular line polarization can beeither parallel or orthogonal to the magnetic field, these results reinforce the polarized dustemission results – a magnetic field orthogonal to the main axis of the filament.
5. Analysis5.1. Core Mass Estimation
The G34+0.23 MM column density and mass are estimated from dust emission. Wefollow the derivation made by Mezger (1994). This derivation uses a parametrized repre-sentation of the dust absorption cross section per H-atom. This parametrization follows τ ν /N H = σ H λ in cm /H-atom, where σ H λ = Z/Z ⊙ b (7 × − λ − µm ) λ µm ≥ . (1) Z/Z ⊙ = 1 is the relative metalicity, λ µ m is the wavelength in µ m, and b is parameter usedto introduce the grain dependence on gas density (Mezger 1994). The b parameter usuallytakes values of b = 1 . H ≤ cm − and b = 3 . N H / cm − = 1 . × ( S ν,int / Jy ) λ µ m( θ s / arcsec ) ( Z/Z ⊙ ) bT e x − x (2) M H / M ⊙ = 4 . × − ( S ν,int / Jy ) λ µ m D ( Z/Z ⊙ ) bT e x − x , (3)where N H = N (H) + 2 N (H ) is the total hydrogen column density, S ν,int is the integratedflux density from the source, θ s = p θ s,min × θ s,max is the angular source size, x = . × λ µm T isthe hcλkT factor for the Planck function, D kpc is the distance to the source in kpc, T is the dusttemperature, and b is taken to be b = 3 . T = 50 K as the most likely dust temperature for the MM core, the valuethat we use here. Using the flux of 290 mJy estimated from our 3 mm continuum emission,we obtain a column density of 6 × cm − and a total mass of 520 M ⊙ . Shepherd et al.(2004) obtained a mass of 250 M ⊙ (their masses varied up to 650 M ⊙ for different values ofdust temperature and emissivity). The general morphology of the G34.4 massive star forming region is of a filament,where all sources are embedded along the main axis of the cloud (see Figure 1 and mapsin Rathborne et al. (2005); Shepherd et al. (2007)). This type of morphology appears tobe widespread in the ISM (Fa´undez et al. 2004). How this elongated cloud morphology isproduced is still an open question; one avenue of research is to understand the dynamicaleffect of magnetic fields on the gas and dust in these regions. Our dust polarized emissionresults show a uniform pattern over both the MM core and the UC H II region, with anaverage P.A. of -8 ◦ ± ◦ aligned with the main axis (north-south) of the filament, betweenthe MM core and the UC H II region (see Figure 2). Polarized emission from dust grainssuggests alignment of grains by a magnetic field; the aligned grains will produce polarizedemission with P.As. orthogonal to the projection of the field onto the plane of the sky.Therefore, our observations suggest a magnetic field morphology with field lines orthogonalto the main axis of the filament between the MM core and the UC H II region.A similar interpretation applies for our CO J = 1 → ◦ ± ◦ is calculatedfor all channels. Figure 6 shows the same plot than Figure 3, but with the polarization mapoverplotted. The image to the left shows an average of channels from 80 to 60 km s − forthe red lobe, while the image to the right shows an average of channels from 54 to 38 km s − ◦ ± ◦ , while in the bluelobe it is -3 ◦ ± ◦ , consistent with the average and overall values shown in Table 1. Figure7 shows panels with spectra from the most intense CO J = 1 → ′′ boxes around the points ( α, δ )=(18:53:17.3,01:25:15) and (18:53:19.2, 01:24:41) which can be easily spotted in Figure 4 at channel mapnumber 30 th , or at V = 62 km s − , corresponding to the main emission peaks on the map.Superposed on to each spectrum are fractional polarization and P.A. values which are alsoconsistent with Table 1 and with the previous CO J = 1 → ◦ . The large fractionalCO ( J = 1 →
0) polarization values obtained in our observations can be explained, mostlikely, by missing flux in the Stokes I emission due to incomplete u-v coverage at the shorterbaselines of the BIMA D array configuration.As with the polarized dust emission results, the polarization line segments seem to bewell aligned with the filament axis, showing little spatial dispersion over all relevant channels.Cortes et al. (2005) showed that strong large velocity gradients will produce polarizationperpendicular to the magnetic field even if a weak continuum source is present, which seemsto be the case in our observations. Therefore, our results for both polarized dust and CO J = 1 → ′ × ′ of the OMC-1 region showsa magnetic field that is not only orthogonal to the main axis of the cloud but also has anhourglass shape. Cortes et al. (2006) made interferometric observations with BIMA towardsthe NGC2071IR star forming region, finding a magnetic field orthogonal to the main axis ofan elongated structure.It has been also suggested that CO line polarization will, most likely, trace a differentenvironment from dust polarization. Cortes et al. (2005) showed that radiation from the CO J = 1 → H ∼
100 cm − which will, most likely, correspond to cloud envelopes,like the proposed cold screen by Ramesh et al. (1997), or to outflow extended regions. In thecase of G34.4, the data support a magnetic field aligned with the main axis of the cloud evenat regions dynamically dominated by the outflows. However, polarized line emission tracingoutflows aligned with magnetic fields has been observed (Girart et al. 1999; Cortes et al.2006). This apparent discrepancy might be explained by the difference in spatial resolutionbetween this work and previous findings; while Cortes et al. (2006) interferometric observa-tions achieved a resolution of 4 ′′ from NGC2071IR (at a much closer distance than G34.4),this work is presenting observations at a coarser resolution of 16 ′′ . A larger beam may smear 10 –out the polarization P.As. erasing the outflow signature which will produce a more uniformpolarization pattern.Taking into account that dust emission will trace regions at higher densities ( n H ≥ cm − ), which in this case corresponds to the MM core and the UC H II region. Both polar-ization results suggest a magnetic field perpendicular to the filament at different densities.Interesting would be to have higher resolution observations to map the polarized emissionfrom the outflows in this region.
6. Summary and Conclusions
The G34.4+0.23 MM massive star forming core was observed in the 3 mm band inpolarized continuum and in CO J = 1 → < φ > = − ◦ ± ◦ , and from the CO polarization < φ > = − ◦ ± ◦ . These results suggest a magnetic field perpendicular to the main axis ofthe filament, between the MM source and the UC H II region, in G34.4. The morphologysuggests a flattened disk with the magnetic field along the minor axis, as predicted by thetheory of magnetically supported molecular clouds. From our 3 mm continuum observations,we estimate a total core mass of 400 M ⊙ , in agreement with previous observations. Bothline and continuum emission agrees in morphology with previous work by Shepherd et al.(2004, 2007). Finally, additional observations, particularly higher resolution interferometricmapping along the filament (including the most northern source seen in Figure 1), will helpto constrain, in greater detail, the morphology of the field and to obtain estimates of itsstrength in the plane of the sky.P. C. Cortes acknowledges support from the ALMA-CONICYT fund for development ofChilean Astronomy through grant 31050003. P. C. Cortes would also like to acknowledge thesupport given by NCSA and the Laboratory for Astronomical Imaging at University of Illinoisat Urbana-Champaign during this research. Finally, P. C. Cortes would like to acknowledgethe contribution by Patricio Sanhueza in making Figure 1. R. M. Crutcher acknowledgessupport from NSF grants AST 05-40459 and 06-06822. L. Bronfman acknowledges supportfrom the Chilean Center for Astrophysics FONDAP 15010003. REFERENCES
Briggs, D. S., Schwab, F. R., & Sramek, R. A. 1999, in Astronomical Society of the Pacific 11 –Conference Series, Vol. 180, Synthesis Imaging in Radio Astronomy II, ed. G. B.Taylor, C. L. Carilli, & R. A. Perley, 127–+Bronfman, L., Nyman, L.-A., & May, J. 1996, A&AS, 115, 81Cortes, P., & Crutcher, R. M. 2006, ApJ, 639, 965Cortes, P. C., Crutcher, R. M., & Matthews, B. C. 2006, ApJ, 650, 246Cortes, P. C., Crutcher, R. M., & Watson, W. D. 2005, ApJ, 628, 780Fa´undez, S., Bronfman, L., Garay, G., Chini, R., Nyman, L.-˚A., & May, J. 2004, A&A, 426,97Girart, J. M., Crutcher, R. M., & Rao, R. 1999, ApJ, 525, L109Glenn, J., Walker, C. K., & Jewell, P. R. 1997, ApJ, 479, 325Goldreich, P., & Kylafis, N. D. 1981, ApJ, 243, L75Greaves, J. S., Holland, W. S., Friberg, P., & Dent, W. R. F. 1999, ApJ, 512, L139Lai, S., Crutcher, R. M., Girart, J. M., & Rao, R. 2002, ApJ, 566, 925Lai, S., Girart, J. M., & Crutcher, R. M. 2003, ApJ, 598, 392Lai, S. P. 2001, PhD thesis, University of Illinois at Urbana - Champaign, Urbana, IL 61801,available at the Astronomy library at the Astronomy buildingLazarian, A. 2007, Journal of Quantitative Spectroscopy and Radiative Transfer, 106, 225Mezger, P. G. 1994, Ap&SS, 212, 197Miralles, M. P., Rodriguez, L. F., & Scalise, E. 1994, ApJS, 92, 173Molinari, S., Brand, J., Cesaroni, R., & Palla, F. 1996, A&A, 308, 573Mooney, T., Sievers, A., Mezger, P. G., Solomon, P. M., Kreysa, E., Haslam, C. G. T., &Lemke, R. 1995, A&A, 299, 869Palla, F., Brand, J., Comoretto, G., Felli, M., & Cesaroni, R. 1991, A&A, 246, 249R. J. Sault, N. E. B. Killeen. 1998, Miriad users guide, BIMARamesh, B., Bronfman, L., & Deguchi, S. 1997, PASJ, 49, 307 12 –Rathborne, J. M., Jackson, J. M., Chambers, E. T., Simon, R., Shipman, R., & Frieswijk,W. 2005, ApJ, 630, L181Rohlfs, K., & Wilson, T. L. 2004, Tools of radio astronomy (Tools of radio astronomy, 4threv. and enl. ed., by K. Rohlfs and T.L. Wilson. Berlin: Springer, 2004)Scalise, E., Rodriguez, L. F., & Mendoza-Torres, E. 1989, A&A, 221, 105Schleuning, D. A. 1998, ApJ, 493, 811Schutte, A. J., van der Walt, D. J., Gaylard, M. J., & MacLeod, G. C. 1993, MNRAS, 261,783Shepherd, D. S., N¨urnberger, D. E. A., & Bronfman, L. 2004, ApJ, 602, 850Shepherd, D. S., Povich, M. S., Whitney, B. A., & Robitaille, T. P. 2007, ApJ, 276, SubmittedSzymczak, M., & Kus, A. J. 2000, A&AS, 147, 181
This preprint was prepared with the AAS L A TEX macros v5.2.
13 –Table 1. Averages of fractional polarization and position angles, per channel, of CO( J = 1 →
0) polarized emission from G34.4Channel Velocity P CO φ CO number [km s − ] [ ◦ ]9 104.8 0.4 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± σ , 3 σ , 6 σ , 10 σ , 15 σ ,20 σ , 25 σ , 30 σ , 35 σ , and 40 σ where σ = 6 mJy beam − . The beam is indicated by the blackcircle at the bottom rigth corner of the map. The gray scale map indicates the polarizedflux ( p U + Q ), also measured in Jy beam − . The line segments show the polarizationP.A. and the fractional polarization, which is indicated by the length of the line. The bar atthe bottom right, below the beam, shows a fractional polarization scale of 0.4. The crossesindicate the location of the MM and the UC H II regions. 16 – ++ Fig. 3.— A composite plot which superpose the averaged blue shifted lobe (in blue contours)and the red shifted lobe (in red contours), plotted as − σ, σ, σ, σ, and 12 σ where σ = 0 . − . The thick lines represent the positive emission, while the negatives are given bythe small gray circle in between both sources. The beam is plotted as a black oval at thebottom right corner. The lobes are calculated using the velocity interval v = 60 to v = 80km s − for the red lobe and v = 38 to v = 54 km s − for the blue lobe. As in previousFigures, the crosses represent the MM core and the UC H II region. 17 –Fig. 4.— Velocity channel maps showing the red-shifted velocity component of the emission.The channel maps have velocity and channel number written at the top left corner of eachmap; the beam is plotted at the bottom right. The Stokes I emission is plotted as contoursof − σ, σ, σ, σ, σ, σ, and 18 σ where σ = 0 . − . Polarized flux is shown asgray scale and the line segments represent fractional polarization and P.A. The bar belowthe beam is the fractional polarization scale of 0.26 and the crosses show the position of theMM core and of the UC H II region. 18 –Fig. 5.— Velocity channel maps showing the blue shifted component of the emission. Thechannel maps have velocity and channel number written at the top left corner of each map;the beam is plotted at the bottom right. As for Figure 4, the Stokes I emssion is shown ascontours of − σ, σ, σ, σ, σ, σ, and 18 σ where σ = 0 . − . Polarized flux isshown as gray scale and the line segments represent fractional polarization and P.A. The barbelow the beam is the fractional polarization scale of 0.26 and the crosses show the positionof the MM core and of the UC H II region. 19 –Fig. 6.— The Figure shows two panels with the red shifted (top panel) and blue shifted(bottom panel) emission as shown by Figure 3, but with polarization maps superposed. TheFigure shows Stokes I as contours of − σ, σ, σ, σ, and 12 σ , where σ = 0 . − .The small bar below the beam represents the fractional polarization scale in the map, cor-responding to 0.22 for the red lobe and 0.25 for the blue lobe. The crosses represent thepositions of the MM core and the UC H II region. 20 – -2 0 2 4 6 8 10 12 14 0 20 40 60 80 100 120 0 0.2 0.4 0.6 0.8 1 F l u x [ Jy ] F r a c t i ona l P o l a r i z a t i on V [km s -1 ] CO(1-0) JyFractional Polarization -2 0 2 4 6 8 10 12 14 0 20 40 60 80 100 120-40-30-20-10 0 10 20 30 40 F l u x [ Jy ] < P . A . > [ deg r ee s ] V [km s -1 ] CO(1-0) JyP.A. 0 2 4 6 8 10 12 14 16 0 20 40 60 80 100 120 0 0.2 0.4 0.6 0.8 1 F l u x [ Jy ] F r a c t i ona l P o l a r i z a t i on V [km s -1 ] CO(1-0) JyFractional Polarization 0 2 4 6 8 10 12 14 16 0 20 40 60 80 100 120-40-30-20-10 0 10 20 30 40 F l u x [ Jy ] < P . A . > [ deg r ee s ] V [km s -1 ] CO(1-0) JyP.A. Fig. 7.— Panels show two different spectra from the most intense CO J = 1 → ′′ boxes centered at ( α, δα, δ