1: %\documentclass{aastex}
2: \documentclass[12pt,preprint]{aastex}
3: \usepackage{natbib}
4: \bibliographystyle{apj}
5:
6: %\usepackage{emulateapj5}
7: %\usepackage{latexsym}
8: \shorttitle{}
9: \shortauthors{Paulo C. Cortes}
10:
11: \shorttitle{Interferometric mapping of magnetic fields}
12: \shortauthors{P. C. Cortes}
13:
14: \begin{document}
15:
16:
17: \title{Interferometric Mapping of Magnetic Fields: G30.79 FIR 10}
18: \author{P. Cortes}
19: \affil{Astronomy Department, University of Illinois at
20: Urbana-Champaign, IL 61801}
21: \author{R. M. Crutcher}
22: \affil{Astronomy Department, University of Illinois at
23: Urbana-Champaign, IL 61801}
24:
25: \begin{abstract}
26: We present polarization maps of G30.79 FIR 10
27: (in W43) from thermal dust emission
28: at 1.3 mm and from CO J=$2 \rightarrow 1$ line emission.
29: The observations were obtained using the Berkeley-Illinois-Maryland
30: Association array in the period 2002-2004.
31: The G30.79 FIR 10 region shows
32: an ordered polarization pattern in dust emission,
33: which suggests an hourglass
34: shape for the magnetic field. Only marginal detections
35: for line polarization were made from this region. Application
36: of the Chandrashkar-Fermi method yielded $B_{pos} \approx 1.7$ mG
37: and a statistically corrected mass to magnetic flux ratio
38: $\lambda_{C} \approx 0.9$, or essentially critical.
39: \end{abstract}
40:
41: \keywords{ISM: magnetic fields — ISM:polarization — stars: formation}
42:
43: \section{Introduction}
44:
45: The star formation process
46: involves a number of physical
47: parameters, of which the magnetic field is the least observed.
48: Magnetic field observations are divided into measurements of the Zeeman
49: effect (in order to obtain the magnetic field strength in the line
50: of sight), and linear polarization observations of dust and spectral-line emission.
51: Polarization of dust emission is believed to be perpendicular to the magnetic field
52: under most conditions \citep{Lazarian2003}; hence, polarization of dust emission
53: has been used as a major probe for the magnetic field geometry.
54: In order to efficiently map the polarization of dust emission and infer
55: information about the magnetic field morphology, high resolution
56: observations are required. The BIMA millimeter interferometer
57: has been used previously to obtain high-resolution polarization maps
58: in several star forming cores \citep{Rao1998,Girart1999b,Lai2001,
59: Lai2002,Lai2003}.
60: These results show fairly uniform polarization morphologies
61: over the main continuum sources, suggesting that magnetic fields
62: are strong, and therefore cannot be
63: ignored by star formation theory. However, the number of star formation
64: regions with maps of magnetic fields remains small, and every new
65: result is significant.
66:
67: Spectral line linear polarization has been suggested to arise
68: from molecular clouds under
69: anisotropy conditions \citep{Goldreich1981}. The prediction suggests
70: that a few percent of
71: linearly polarized radiation should be detected from molecular clouds and
72: circumstellar envelopes in the presence of a magnetic field. It is also
73: predicted that the molecular line polarization will be either parallel
74: or perpendicular to the magnetic field, depending on the angles between
75: the line of sight, the magnetic field, and the anisotropic excitation
76: direction \citep{Goldreich1982}. This
77: process is known as the Goldreich - Kylafis effect.
78:
79: In order to use these techniques,
80: we mapped the massive star forming region G30.79 FIR 10 with the BIMA array.
81: We measured continuum polarization at 1.3 mm and
82: CO $J=2\rightarrow 1$ line polarization obtaining high resolution
83: interferometric maps for both measurements.
84:
85: The remainder of this paper is divided in five major sections. Section 2 reviews
86: information about the source, section 3 describes the observation
87: procedure.
88: Section 4 presents the results, section 5 gives the discussion,
89: and section 6 the conclusions and summary.
90:
91:
92: \section{Source Description}
93:
94: G30.79 is a large dust continuum source located within the W 43 region, which is
95: an H II region-molecular cloud complex near $l=31^{\circ}, b=0^{\circ}$.
96: We observed G30.79 FIR 10, which is a massive and dense component in the G30.79
97: complex.
98: \citet{Liszt1995} observed G30.79 in HCO$^{+}$ and $^{13}$CO, concluding
99: that the presence of several rings and shells in the dense molecular gas
100: was a disturbance product of star formation. \citet{Vallee2000}
101: observed the dust continuum emission in this source at 760 $\mu$m with the JCMT
102: telescope. They found linear polarization of about 1.9\% with a position
103: angle (P.A.) of 160$^{\circ}$ at FIR 10.
104: \citet{Mooney1995} also observed this source at 1.3 mm
105: using the IRAM 30-meter antenna, detecting a total flux of 13.6 Jy; their
106: wide field map shows FIR 10 and the extended H II region in the G30.79 complex.
107: H$_{2}$O masers have also
108: been observed toward this region \citep{Cesaroni1988} that are within a half
109: arcsecond of the peak in the \citet{Mooney1995} map and
110: might be a signature of massive star formation.
111: No centimeter radio-continuum emission seems to be associated with
112: FIR 10, which suggests that the source is in an early stage of evolution.
113: \citet{Motte2003} mapped the W43 main complex in dust continuum emission
114: at 1.3 mm and 350 $\mu m$ with the IRAM 30-m telescope and the CSO,
115: respectively.
116: They also mapped the HCO$^{+}$ $J=3\rightarrow2$ line and measured
117: H$^{13}$CO$^{+}$ $J=3\rightarrow2$ towards prominent dust maxima.
118: One of the maxima, W43-MM1, is the compact fragment we mapped with BIMA.
119: \citet{Motte2003} found v$_{lsr}= 98.8$ km s$^{-1}$, $\Delta v=5.9$ km s$^{-1}$
120: (from H$^{13}$CO$^{+}$), T$_{\textnormal{dust}} \sim 19$ K, M$\sim 3600$
121: M$_{\sun}$, and $n(\textnormal H_{2}) \sim 8.8 \times 10^{6}$ cm$^{-3}$.
122: They estimated the virial mass to be M$_{vir} \sim 1000$ M$_{\sun}$,
123: suggesting that this compact fragment should be in a state of gravitational
124: collapse unless there are other sources of support than kinetic energy.
125: The W43 region, is therefore, an excellent region for the study of
126: the earliest stages of massive star formation.
127:
128:
129: \section{Observation Procedure}
130:
131: We observed G30.79 FIR 10 between October 2002 to May 2004,
132: mapping the continuum
133: emission at 1.3 mm and the CO $J=2 \rightarrow 1$ molecular line (at 230 GHz);
134: four tracks with the BIMA array in C configuration were obtained.
135: We set the digital correlator in mode 8
136: to observe both the continuum and the CO $J=2 \rightarrow 1$ line
137: simultaneously. The 750 MHz lower side band was combined
138: with 700 MHz from the upper side band to
139: map the continuum emission, leaving a
140: 50 MHz window for the CO line observation (at a resolution of 1.02 km s$^{-1}$).
141: Each BIMA telescope has a single receiver, and thus the two polarizations
142: were observed sequentially. A quarter wave plate to select either right (R)
143: or left (L) circular polarization was alternately switched into the signal path
144: ahead of the receiver. Switching between polarizations was sufficiently
145: rapid (every 11.5 seconds) to give essentially identical uv-coverage.
146: Cross-correlating the R and L circularly polarized signals
147: from the sky gave RR, LL, LR, and RL for each interferometer baseline,
148: from which maps in the four Stokes parameters were produced.
149: The source 1743-038 was used as phase calibrator for G30.79 FIR 10.
150: The instrumental polarization was calibrated by observing
151: 3C279,
152: and the ``leakages" solutions were calculated from this observation.
153: We used the same calibration procedure
154: described by \citet{Lai2001}.
155:
156: The Stokes images I, U, Q and V were obtained by Fourier transforming
157: the visibility data using natural weighting.
158: The MIRIAD \citep{MIRIAD1995} package was used for data reduction.
159: G30.79 FIR 10 is close to the equator; therefore, we expect strong
160: sidelobes in the beam pattern. We followed \citet{Chernin1995}, who
161: observed NGC2071IR (also a source close to the equator)
162: with the BIMA array, and imaged only out to 20$^{\prime \prime}$
163: radius, due to the strong sidelobes.
164:
165:
166: \section{Observational Results}
167:
168: \subsection{1.3 mm Continuum}
169:
170: Polarized dust emission was detected and mapped toward G30.79 FIR 10
171: (Figure~\ref{G30con}) with a beam
172: size of $5.6^{\prime \prime}\times 3.6^{\prime \prime}$ and a beam
173: P.A. of $4.3^{\circ}$. The
174: Stokes I map shows a peak emission of 1.6 Jy beam$^{-1}$;
175: this peak is consistent with \citet{Mooney1995}
176: observations at the same wavelength.
177: Most of the emission in Figure~\ref{G30con} comes from a compact core
178: $\sim 8^{\prime \prime}$ in radius.
179:
180:
181: The polarized emission shown in Figure~\ref{G30con} has a $3\sigma$
182: level of significance,
183: a peak polarized intensity of 0.03 Jy beam$^{-1}$, and a mean position
184: angle of 4.6$^{\circ} \pm 6.1^{\circ}$.
185: The polarization direction is different in the northern and southern
186: regions;
187: the mean P.A. in
188: the north is -33.2$^{\circ} \pm 6.6^{\circ}$, and in the
189: south is 23.9$^{\circ} \pm 5.5^{\circ}$.
190: The polarization is fairly uniform in fractional polarization and
191: smooth in its direction change from north to south,
192: (see Table \ref{TableG30}).
193: %This
194: %56$^{\circ}$ difference in P.A. suggests a change in the
195: %magnetic field direction (assuming magnetic alignment).
196: \citet{Vallee2000} observed G30.79 FIR 10 using the JCMT
197: telescope at $760$ $\mu$m. They detected continuum polarization at two
198: points in FIR 10. Only one of them is in the general area of our map;
199: however, this point is $\sim 15^{\prime \prime}$ south of the
200: center of our polarization pattern. Hence, these
201: results cannot be compared directly with ours.
202:
203: %Figure 6 shows the angle distribution histogram, 75\% of the P.A.s are
204: %between -30$^{\circ}$ to $40^{\circ}$ which is similar to regions
205: %like NGC 1333 IRAS 4 in which a similar clustering of P.A.s
206: %suggested a poloidal magnetic field \citep{Girart1999b}. Table 2 shows
207: %P.A. at each relevant pixel coordinate.
208:
209:
210:
211:
212: \subsection{CO $J=2 \rightarrow 1$}
213:
214: Polarized emission from the CO $J=2 \rightarrow 1$
215: was detected (Figure~\ref{G30co}). The Stokes I map shows three CO peaks
216: which cover a similar area as the continuum emission.
217: Our peak
218: emission is 1.4 Jy beam$^{-1}$ with a beam size of
219: $5.5^{\prime \prime}\times 3.5^{\prime \prime}$ and a beam P.A. of
220: $4.0^{\circ}$. We only detected a level of polarized emission of
221: at least $3\sigma$
222: at three independant positions in the core. These three points show two
223: distinctive directions which appear to be orthogonal to each other.
224: The polarization in the southern part of the map shows a mean P.A. of
225: $25.5^{\circ} \pm 4.4^{\circ}$, while the north-western part
226: of the core shows a mean P.A. of $-31.7^{\circ} \pm 4.5^{\circ}$.
227: Line polarized emission appears to be parallel to dust polarized
228: emission over the same area in the south; the average
229: P.A. difference is $\sim 2^{\circ}$ which suggests that the magnetic
230: field is perpendicular to the polarization.
231: However, in the north-western region
232: the average difference in P.A. is $\sim 54^{\circ}$. It is posible
233: that our interferometric observations are resolving out the CO emission,
234: which will make the fractional polarization measurements unreliable.
235:
236:
237: \section{Discussion}
238:
239: We mapped polarized dust emission at 1.3 mm from G30.79 FIR 10,
240: and \citet{Vallee2000} detected polarized dust emission at 760 $\mu$m.
241: The emission at both wavelengths seem to be dominated by dust, which is
242: reinforced by the lack of centimeter radio continuum emission from
243: this source. A fully developed
244: H II region is present in the G30.79 region, but its ionization
245: front has not yet reached FIR 10;
246: the H~II region appears to be $\sim 2$ pc away from the FIR 10 source \citep{Mooney1995}.
247: Water masers have also been observed, which is a signpost for high mass
248: star formation. Based on this evidence, we believe that G30.79 FIR 10
249: is in an early stage of evolution.
250:
251: We inferred the source parameters from our data using an area
252: $\theta_{s}=20^{\prime \prime}
253: \times 20^{\prime \prime}$, which corresponds to the area over which we
254: mapped the polarized flux of the core.
255: The dust absorption cross
256: section per H-atom is approximated by a power law \citep{Mezger1990}
257:
258: \begin{equation}
259: \tau_{\lambda}=N_{\textnormal{H}} \sigma_{\lambda}^{\textnormal{H}} \\
260: =N_{\textnormal{H}} (Z/Z_{\sun}) b
261: (7 \times 10^{-21} \lambda^{-2}),
262: \end{equation}
263:
264: \noindent where $\lambda$ is the wavelength,
265: $N_{{\textnormal{H}}}=N({\textnormal{H}}) +
266: 2N({\textnormal{H}}_{2})$
267: is the total hydrogen column density, $Z/Z_{\sun}$ the relative metalicity, and
268: $b$ is a parameter which reflects
269: the variation of dust absorption cross sections
270: \citep{Mezger1990}.
271: Two values are used for $b$; $b=1.9$ reproduces estimates of
272: $\sigma_{400\mu m} \sim 8.3
273: \times 10^{-26}$ cm$^{2}$(H-atom)$^{-1}$, which represents molecular gas
274: of moderate density $n({\textnormal H}_{2}) < 10^{6}$ cm$^{-3}$, and $b=3.4$
275: relates to
276: dust around deeply embedded IR sources at higher densities. Apparently a value
277: of $b=3.4$ was used by \citet{Mooney1995} in their estimates on G30.79 FIR 10.
278: We used the expressions for the cloud mass and the column density given by
279: \citet{Mooney1995}:
280:
281: \begin{equation}
282: \label{cd}
283: (N_{\textnormal{H}}/{\textnormal{cm}}^{-2})=1.93 \times 10^{15}
284: \frac{(S_{\nu ,avg}/Jy)\lambda^{4}_{\mu m}}
285: {(\theta_{s}/arcsec)^{2}(Z/Z_{\sun})bT}
286: \frac{e^{x} - 1}{x}
287: \end{equation}
288: \begin{equation}
289: (M_{\textnormal{H}})=4.1 \times 10^{-10}
290: \frac{(S_{\nu ,avg}/Jy)\lambda^{4}_{\mu m} D^{2}_{kpc}}
291: {(Z/Z_{\sun})bT}
292: \frac{e^{x} - 1}{x},
293: \end{equation}
294:
295: \noindent where $S_{\nu ,avg}$ is the averaged flux from the source,
296: $\theta_{s}=
297: \sqrt{\theta_{s,min} \times \theta_{s,max}}$ is the
298: angular source size, $x=\frac{1.44 \times 10^{4}}
299: {\lambda_{\mu m} T}$ is the $\frac{hc}{\lambda kT}$ factor for the Planck
300: function, and $D_{kpc}$ is the distance to the source in kpc.
301: We obtained
302: $S_{\nu ,avg}$ by averaging the continuum emission over an area of
303: 20$^{\prime \prime}
304: \times$ 20$^{\prime \prime}$, which gives 4.7 Jy. From their 1.3 mm single dish
305: observations, \citet{Mooney1995} obtained 13.6 Jy for a source
306: of $25.4^{\prime \prime} \times 33.2^{\prime \prime}$. Their larger result
307: is due both to the larger area over which they averaged and to loss of flux
308: in the interferometer data due to missing short uv-spacings.
309: We used $T=20$ K, the same dust temperature used by \citet{Mooney1995},
310: who noted the uncertainty of the value, which is based on dust models.
311: \citet{Motte2003} also observed G30.79 FIR 10 (which they called W43-MM1); they
312: fitted a gray body model obtaining a dust temperature of 19 K.
313: A distance of 5.5 kpc to G30.79 FIR 10 was used \citep{Motte2003}.
314: From our date, we find $N_{\textnormal H}=1.3 \times 10^{24}$ cm$^{-2}$ and
315: $M_{\textnormal H} = 3300$ $\textnormal{M}_{\sun}$. The mass is
316: in agreement with \citet{Motte2003} who estimated
317: $M_{\textnormal H}=3600$ $\textnormal M_{\sun}$. However, this value is
318: lower than the values calculated by \citet{Vallee2000},
319: who estimated a mass
320: $M_{\textnormal H}=9 \times 10^{3} M_{\sun}$, and \citet{Mooney1995},
321: who calculated
322: $M_{\textnormal H}=1.1 \times 10^{4} M_{\sun}$. This difference
323: can be explained due to the larger distance assumed by the later investigators.
324:
325: Assuming magnetic alignment, our polarized dust emission
326: map suggests a poloidal morphology for the
327: magnetic field. This is seen by rotating the line segments in Figure
328: \ref{G30con} by 90$^{\circ}$.
329: This poloidal morphology hints at an hourglass shape for the magnetic field
330: in this core. The proposed field morphology is shown by thick dashed lines
331: Figure \ref{G30con}.
332:
333: Using the Chandrasekhar-Fermi effect \citep{Chandrasekhar1953},
334: we can estimate the
335: magnetic field strength from the analysis of the small scale randomness
336: of magnetic field lines. We used the expression in \citet{Crutcher2004a}
337: to estimate the magnetic field strength:
338:
339: \begin{equation}
340: B_{pos}=9.3\frac{\sqrt{n(\textnormal H_{2})}\Delta V}{\delta \phi}
341: \end{equation}
342:
343: \noindent where $B_{pos}$ is the magnetic field strength in $\mu$G,
344: $n(\textnormal{H}_{2})$ is in cm$^{-3}$, $\Delta V$
345: is the FWHM velocity width in km s$^{-1}$, and $\delta \phi$ is the P.A.
346: dispersion in degrees.
347:
348:
349: We used $\Delta V=5.9$ km s$^{-1}$ from the apparently optically thin
350: H$^{13}$CO$^{+}$ measured with a 29$^{\prime \prime}$ beam \citep{Motte2003}.
351: Our $N_{\textnormal H}=1.3 \times 10^{24}$ cm$^{-2}$ implies
352: $N$(H$_{2}$)=6.5$\times 10^{23}$ cm$^{-2}$, which gives
353: $n(\textnormal{H}_{2})=4.8 \times 10^{5}$ cm$^{-3}$ for a core
354: diameter of 16$^{\prime \prime}$ (0.43 pc).
355: The weighted-mean P.A. dispersion $\delta \phi=$21.9$^{\circ}$
356: (after correction for 4.7$^{\circ}$ contribution from measurement
357: uncertainties) was calculated
358: from our polarized dust observations. These values yield
359: $B_{pos}=1.7$ mG. Although, this value may seem extraordinarily high,
360: it is similar to field strengths estimated from maser Zeeman splitting
361: at similar gas densities in other clouds.
362:
363:
364: Using our estimate above for the column density, we can calculate the
365: mass to magnetic flux ratio in term of the critical value \citep{Crutcher2004b}:
366:
367: \begin{equation}
368: \lambda = 7.6 \times 10^{-21} \frac{N(\textnormal H_{2})}{B}
369: \end{equation}
370:
371:
372: \noindent where $N(\textnormal H_{2})$ is in cm$^{-2}$
373: and B in $\mu \textnormal{G}$. We obtained
374: a value of $\lambda \approx 2.8$, which is slightly supercritical.
375: Applying the statistical geometrical correction factor of 1/3
376: \citep{Crutcher2004b}, $\lambda_{C} \approx 0.9$, which is essentially
377: critical.
378:
379: Our CO $J=2 \rightarrow 1$ line polarization detections are too scarce
380: to yield any significant information about
381: the magnetic field. However, comparing the CO polarized emission to the
382: dust polarized emission may help in getting information about the sources
383: of anisotropy in this region.
384: Figure~\ref{G30co} shows that at the south-eastern
385: side of the map the CO polarization is parallel to the polarized
386: dust emission,
387: and in the north-western part of the map they may be roughly orthogonal.
388: Previous polarization observations in sources like
389: DR21(OH) showed CO $J=2 \rightarrow 1$ polarization perpendicular to the
390: polarization of dust emission over the main sources \citep{Lai2003}.
391: A theoretical study by
392: \citet{Cortes2005} suggested that line polarization traces
393: the magnetic field at
394: lower densities than the dust, and the orientation of the line polarization
395: will depend on the degree of anisotropy in the region. \citet{Cortes2005}
396: concluded that
397: the presence of hot continuum sources will preferentially produce
398: CO $J=2 \rightarrow 1$ polarization that will be perpendicular to the dust
399: polarization. In the case of G30.79 FIR10 it is difficult to draw any conclusion
400: without CO $J = 1\rightarrow0$ polarization data and a detailed modeling
401: of the source.
402:
403: \section{Summary and conclusions}
404:
405: We observed G30.79 FIR 10 and successfully mapped
406: CO $J=2 \rightarrow 1$ line and 1.3 mm dust continuum polarized emission
407: with a resolution of $4^{\prime \prime}$.
408:
409: G30.79 FIR 10 is not a well studied region; however, there is
410: evidence that points toward an early stage of development of
411: high-mass star formation. We found
412: a remarkably uniform pattern in our polarized dust emission map,
413: which suggests
414: an hourglass magnetic field morphology. Using the Chandrasekhar-Fermi
415: method, we inferred a plane-of-the-sky magnetic field strength 1.7 mG,
416: which yielded a geometry-corrected mass to magnetic flux ratio of
417: 0.9 to respect to critical. This result is similar to those found
418: in many other regions of low-mass and high-mass star formation.
419:
420: Polarized line emission was also
421: detected in this region; these results pose questions about the
422: sources of anisotropy, which will require more detailed modeling.
423:
424: This research was partially funded by NSF grants AST 02-05810 and 02-28953.
425:
426: \bibliography{biblio}
427:
428: \input{tab1.tex}
429:
430: \newpage
431:
432: \begin{figure}
433: \label{1}
434: \figurenum{1}
435: %\includegraphics[angle=-90,scale=0.7]{f10.ps}
436: \includegraphics[angle=-90,scale=0.7]{f1.ps}
437: \epsscale{0.4}
438: \caption{
439: Polarization map of G30.79 FIR 10 at 1.3 mm. The contours represent
440: the Stokes I emission at
441: -2, 2, 5, 8, 15, 20, 25, 30, 35, 40, and 45 $\times \sigma$;
442: $\sigma=0.034$ Jy/beam.
443: The pixel gray scale shows $3\sigma$ polarized
444: intensity ($\sqrt{Q^{2} + U^{2}}$) for
445: the dust continuum emission, while the black line segments
446: show the fractional polarization and P.A.
447: The scale for the fractional polarization is given by the bar at the
448: left corner that represents 0.1 fractional polarization. The thick dashed
449: lines
450: show the proposed magnetic field morphology.
451: }
452: \label{G30con}
453: \end{figure}
454:
455: \begin{figure}
456: \label{2}
457: \figurenum{2}
458: \includegraphics[angle=-90,scale=0.7]{f2.ps}
459: \epsscale{0.4}
460: \caption{
461: Polarization map of G30.79 FIR 10 for the CO $J=2 \rightarrow 1$ line.
462: The contours represent
463: the Stokes I emission at -0.17, 0.17, 0.35, 0.52, 0.7, 0.88, 1.05, and 1.23
464: Jy/beam. The pixel gray scale shows $3\sigma$ polarized
465: intensity ($\sqrt{Q^{2} + U^{2}}$) for
466: the CO $J=2 \rightarrow 1$ emission, while the black line segments
467: are the polarized line segment map for the CO $J=2 \rightarrow 1$.
468: The scale for the line segments is given by the bar at the
469: left corner that represents 0.31 in fractional polarization.
470: }
471: \label{G30co}
472: \end{figure}
473:
474:
475: \end{document}
476:
477:
478:
479: