1: \documentclass[preprint2]{aastex}
2: %\documentclass[12pt,preprint]{aastex}
3: \shorttitle{Polarimetry of R CrA Region}
4: \shortauthors{Choi et al.}
5:
6: % For preprint only
7: \usepackage{times}
8: \frenchspacing
9:
10: % Custom macros
11: \newcommand{\Jypb}{{Jy beam$^{-1}$}}
12: \usepackage{epsfig}
13: \newcommand{\ploteps}[2]{\centerline{\ifdim#2=0mm\epsfig{file=#1}%
14: \else\epsfig{file=#1,width=#2}\fi}}
15: %%% parameters for floats
16: \renewcommand{\floatpagefraction}{0.8}
17: \renewcommand{\dblfloatpagefraction}{0.8}
18: \renewcommand{\topfraction}{0.9}
19: \renewcommand{\dbltopfraction}{0.9}
20: \renewcommand{\bottomfraction}{0.9}
21: \renewcommand{\textfraction}{0.1}
22: \setcounter{topnumber}{4}
23: \setcounter{dbltopnumber}{2}
24: \setcounter{bottomnumber}{4}
25: \setcounter{totalnumber}{6}
26:
27: \slugcomment{To appear in the Astrophysical Journal}
28:
29: \begin{document}
30: \fontsize{10}{10.6}\selectfont
31: \title{Centimeter Polarimetry of the R Coronae Australis Region}
32: \author{\sc Minho Choi\altaffilmark{1},
33: Ken'ichi Tatematsu\altaffilmark{2},
34: Kenji Hamaguchi\altaffilmark{3,4},
35: and Jeong-Eun Lee\altaffilmark{5}}
36: \altaffiltext{1}{International Center for Astrophysics,
37: Korea Astronomy and Space Science Institute,
38: Hwaam 61-1, Yuseong, Daejeon 305-348, South Korea;
39: minho@kasi.re.kr.}
40: \altaffiltext{2}{National Astronomical Observatory of Japan,
41: 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan.}
42: \altaffiltext{3}{Center for Research and Exploration
43: in Space Science and Technology
44: and X-ray Astrophysics Laboratory,
45: NASA Goddard Space Flight Center,
46: Greenbelt, MD 20771.}
47: \altaffiltext{4}{Department of Physics, University of Maryland,
48: Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250.}
49: \altaffiltext{5}{Department of Astronomy and Space Science,
50: Astrophysical Research Center
51: for the Structure and Evolution of the Cosmos,
52: Sejong University, Gunja 98, Gwangjin,
53: Seoul 143-747, South Korea.}
54: \setcounter{footnote}{5}
55:
56: \begin{abstract}
57: \fontsize{10}{10.6}\selectfont
58: Circularly polarized 3.5 cm continuum emission
59: was detected toward three radio sources in the R CrA region
60: using the Very Large Array.
61: The Class I protostar IRS 5b persistently showed polarized radio emission
62: with a constant helicity over 8 yr,
63: which suggests that its magnetosphere has a stable configuration.
64: There is a good correlation between the Stokes $I$ and Stokes $V$ fluxes,
65: and the fractional polarization is about 0.17.
66: During active phases the fractional polarization
67: is a weakly decreasing function of Stokes $I$ flux,
68: which suggests that IRS 5b is phenomenologically similar
69: to other types of flare stars such as RS CVn binaries.
70: The variability timescale of the polarized flux is about a month,
71: and the magnetosphere of IRS 5b must be very large in size.
72: The Class I protostar IRS 7A was detected once
73: in circularly polarized radio emission,
74: even though IRS 7A drives a thermal radio jet.
75: This detection implies
76: that the radio emission from the magnetosphere of a young protostar
77: can escape the absorption by the partially ionized wind
78: at least once in a while.
79: The properties of IRS 7A and IRS 5b suggests
80: that Class I protostars have
81: organized peristellar magnetic fields of a few kilogauss
82: and that the detectability of magnetospheric emission
83: may depend on the evolutionary status of protostar.
84: Also reported is the detection of circularly polarized radio emission
85: toward the variable radio source B5.
86: \end{abstract}
87:
88: \keywords{ISM: individual (R Coronae Australis IRS 5) --- ISM: structure
89: --- stars: flare --- stars: formation}
90:
91: \section{INTRODUCTION}
92:
93: Magnetic fields play an important role in star formation
94: (Andr{\'e} 1996; Feigelson et al. 2007; Bouvier et al. 2007).
95: This is especially true
96: in the protostellar accretion/evolution processes,
97: and peristellar magnetic fields are expected or even required
98: in many theoretical models:
99: Protostellar outflows are generated
100: through magnetocentrifugal ejection mechanism (e.g., Pudritz et al. 2007).
101: The magnetorotational instability may provide the viscosity
102: required to explain the dynamics of accretion disks (Balbus \& Hawley 1991).
103: Magnetospheres may force the protostellar spin
104: to be coupled with the rotation of circumstellar disk,
105: and the inner edge of accretion disks may be truncated
106: by the magnetosphere near the corotation radius
107: (Camenzind 1990; K{\"o}nigl 1991).
108: The transfer of material from the accretion disk to the protostar
109: may be done through magnetospheric accretion (Bouvier et al. 2007).
110: Despite all these expectations,
111: observational evidence of peristellar magnetic fields of protostars
112: is very difficult to come by.
113:
114: Much of our knowledge about the peristellar magnetic fields
115: of young stellar objects comes from the observations of T Tauri stars,
116: especially radio continuum observations at centimeter wavelengths
117: (Andr{\'e} et al. 1992; Andr{\'e} 1996).
118: The key characteristics are the variability and circular polarization
119: of emission from nonthermal electrons.
120: The variability timescale is usually hours to days,
121: much longer than solar-type flares,
122: which implies that large-scale magnetic fields are involved.
123: Moderate degrees of circular polarization were observed,
124: suggesting that the emission mechanism is
125: gyrosynchrotron radiation from mildly relativistic electrons.
126: The detection of circular polarization at $\sim$5 GHz implies
127: a dipole-like large-scale magnetosphere
128: with a surface field strength of $\sim$1 kG (G{\"u}del 2002).
129:
130: X-ray surveys of pre-main-sequence stars suggest
131: that the level of activity decays with stellar age
132: (Preibisch \& Feigelson 2005),
133: and we expect that Class I protostars may be
134: magnetically more active than T Tauri stars.
135: However, the free-free absorption by thermal wind
136: is supposed to obscure the magnetosphere of protostars in most cases
137: (Andr{\'e} et al. 1992; Andr{\'e} 1996).
138: The youngest object detected in circularly polarized radio emission
139: is IRS 5 in the R CrA region (Feigelson et al. 1998).
140: In fact, IRS 5 is the only known protostellar object
141: persistently displaying circularly polarized emission
142: (Forbrich et al. 2006; Miettinen et al. 2008).
143: Therefore, IRS 5 provides a wonderful opportunity
144: to study the magnetic activity of protostars.
145:
146: To obtain high-quality images of the R CrA region,
147: we observed in the centimeter continuum with the Very Large Array (VLA)
148: of the National Radio Astronomy Observatory.
149: The main results were presented in Paper I (Choi et al. 2008).
150: In this paper, we present the polarimetry of the R CrA region.
151: We describe the data in \S~2.
152: In \S~3 we briefly report the results of the polarimetry.
153: In \S~4 we discuss the star forming activity of the radio sources
154: showing circularly polarized centimeter continuum emission.
155: A summary is given in \S~5.
156:
157: \section{DATA}
158:
159: We analyzed 14 data sets
160: to measure the circularly polarized flux densities
161: of the centimeter continuum sources in the R CrA region.
162: Details of the observations are described in Paper I.
163: See Table 1 of Paper I for the parameters of the VLA observing runs
164: and the observing track numbers.
165: (The track numbers are preceded by `Tr' throughout this paper.)
166: Maps of circularly polarized (Stokes $V$) intensity
167: were made using a CLEAN algorithm.
168: All the images presented in this paper
169: are corrected for the primary beam response,
170: and the lowest contour level roughly corresponds to the 3 $\sigma$ value,
171: where $\sigma$ is the rms of noise.
172:
173: \input{fig1.tex}
174:
175: The accuracy of polarization calibration of VLA data
176: is usually better than 0.5\% for compact sources.
177: For circular polarization measurements,
178: the beam squint can induce spurious signals,
179: which may affect sources located far away from the pointing center.
180: For snapshot images,
181: this effect can be as large as 20\% at the edge of the field of view.
182: For the maps made from the data of a whole track,
183: the circular polarization may be accurate to a few percent.
184: This issue will be discussed again at the end of \S~3.
185:
186: \section{RESULTS}
187:
188: Stokes $V$ flux was detected
189: toward three sources in the 3.5 cm maps
190: while none of the radio sources
191: showed detectable Stokes $V$ flux in the 6.2 cm maps.
192: The polarized emission of IRS 5 was reported previously
193: (Feigelson et al. 1998; Forbrich et al. 2006; Miettinen et al. 2008),
194: and we can now identify the source as IRS 5b.
195: Detections of polarized emission from IRS 7A and B5
196: are reported here for the first time.
197:
198: \input{fig2.tex}
199: \input{tab1.tex}
200:
201: IRS 5b showed detectable polarization in most observing runs,
202: but IRS 5a was never detected in the Stokes $V$ flux maps (Figs. 1 and 2).
203: The Stokes $V$ flux densities from the archival data sets
204: are usually, but not always, consistent with the values
205: reported by Feigelson et al. (1998) and Forbrich et al. (2006).
206: Table 1 lists the total flux densities,
207: both Stokes $I$ and Stokes $V$ fluxes
208: (hereafter $I$ and $V$ fluxes, respectively),
209: and the fractional circular polarization, $\Pi_c = |V|/I$, of IRS 5b.
210:
211: The $V$ flux from IRS 7A was detected
212: during the observations on 1998 July 19 (Tr 5).
213: The source was unresolved, and the peak position was
214: $\alpha_{2000}$ = 19$^{\rm h}$01$^{\rm m}$55.31$^{\rm s}$,
215: $\delta_{2000}$ = --36\arcdeg57$'$21.7$''$,
216: which agrees well (within 0.4$''$) with the position of IRS 7A
217: determined from a higher-resolution map (see Table 2 of Paper I).
218: The total flux was $V$ = 0.19 $\pm$ 0.02 mJy.
219: The fractional polarization cannot be exactly estimated
220: because IRS 7A has both compact and extended components
221: in the Stokes $I$ map.
222: Taking the intensity of the Stokes $I$ map with uniform weighting
223: as an upper limit on the $I$ flux from the central compact source,
224: the fractional polarization was $\Pi_c \gtrsim$ 0.07.
225:
226: Circularly polarized emission from B5 was detected
227: during the observations on 2005 February 3 (Tr 13).
228: The source was unresolved, and the peak position was
229: $\alpha_{2000}$ = 19$^{\rm h}$01$^{\rm m}$43.28$^{\rm s}$,
230: $\delta_{2000}$ = --36\arcdeg59$'$12.3$''$,
231: which agrees well (within 0.1$''$)
232: with the peak position of the Stokes $I$ source from the same data set.
233: The total flux was $V$ = 0.26 $\pm$ 0.04 mJy,
234: and the fractional polarization was $\Pi_c$ = 0.12 $\pm$ 0.02.
235:
236: The other radio sources in the R CrA region
237: were undetected in the $V$ flux.
238: These unpolarized sources are useful
239: in checking the quality of the polarization calibration.
240: If there had been a problem in the polarization calibration,
241: strong sources would show
242: a noticeable amount of apparent polarization systematically.
243: IRS 7B, IRS 1, and IRS 2 are suitable for this purpose.
244: Their $|V|$ intensity was always less than 0.05 m\Jypb,
245: and their $\Pi_c$ was less than 2 $\sigma$ level.
246: Therefore, we conclude that the polarization calibration was good
247: and that the detected $V$ flux densities of IRS 5b, IRS 7A, and B5
248: do not include any significant contribution from instrumental effects.
249:
250: \section{DISCUSSION}
251:
252: \subsection{IRS 5}
253:
254: IRS 5 is an interesting example of protobinary system.
255: It is one of the youngest binaries with X-ray emission from both members,
256: indicating that both have high-energy magnetic activity
257: (Hamaguchi et al. 2008).
258: They displayed a concurrent enhancement of radio fluxes
259: and were suggested to be interacting (Paper I).
260: The projected separation of the binary is 0.9$''$,
261: which corresponds to 160 AU at a distance of 170 pc (Knude \& H{\o}g 1998).
262: IRS 5 as a whole is a Class I source
263: (Wilking et al. 1997; Nutter et al. 2005; Paper I),
264: but the accretion activity is weak.
265: Nisini et al. (2005) derived an accretion rate
266: of $\sim$3 $\times$ 10$^{-8}$ $M_\odot$ yr$^{-1}$ for IRS 5a
267: (the rate for IRS 5b may be even smaller),
268: which is much smaller than the accretion rate
269: of other Class I objects such as IRS 1 and IRS 2
270: (2 $\times$ 10$^{-6}$ $M_\odot$ yr$^{-1}$
271: and 3 $\times$ 10$^{-7}$ $M_\odot$ yr$^{-1}$, respectively).
272: IRS 5a/b may be Class I protostars
273: with the accretion activity halted temporarily.
274: (See White et al. 2007 for a discussion
275: on the accretion rate of Class I protostars in general.)
276: Nisini et al. (2005) suggested a spectral type of K5--K7V for IRS 5a,
277: and the spectral type of IRS 5b is probably later than that.
278:
279: \subsubsection{IRS 5b}
280:
281: While IRS 5b and 5a are comparable in the $I$ flux,
282: all the detectable $V$ flux can be attributed to IRS 5b (Figs. 1 and 2),
283: and separating them is important
284: in the analysis of the source properties.
285: Figure 3 shows the $I$ and $V$ light curves of IRS 5b,
286: and Figure 4 shows the relation between $I$ and $V$.
287: The flux uncertainty shown in Figure 3
288: is the statistical uncertainty listed in Table 1,
289: but one should also consider the uncertainty in the absolute flux scale,
290: usually about 1\%--2\% (VLA Calibrator Manual).%
291: \footnote{See http://www.vla.nrao.edu/astro/calib/manual.}
292: However, the flux scale cancels out
293: in the calculation of the ratio of flux densities,
294: and only the statistical uncertainty needs to be considered
295: for the fractional polarization.
296:
297: \input{fig3.tex}
298: \input{fig4.tex}
299:
300: The light curves of IRS 5b (Fig. 3) shows
301: that $V$ flux tends to be enhanced during the flare events.
302: Such a trend is not obvious if IRS 5a/b are unresolved:
303: Forbrich et al. (2006) thought
304: that the $V$ flux appeared uncorrelated to the $I$ flux
305: because they were distracted by the flux variability of IRS 5a.
306: Miettinen et al. (2008) also reported
307: that the fractional polarization can be as high as 0.33,
308: but their angular resolution was not high enough
309: to make a clear interpretation.
310: By separating IRS 5b from 5a,
311: it is possible to make a detailed interpretation of the polarimetry results,
312: as described below.
313:
314: \subsubsubsection{Helicity}
315:
316: The $V$ flux was always positive (right-handed)
317: during the 8 yr of monitoring.
318: The helicity of emergent radiation is related
319: to the configuration of magnetic fields
320: (Morris et al. 1990).
321: The constant helicity suggests
322: that the magnetic field geometry responsible for the polarization
323: is very stable and has a large scale (Mutel et al. 1987).
324: Here, the large-scale geometry means
325: a star-sized global magnetosphere or an even larger structure.
326: VLBI observations revealed
327: that some RS CVn binaries have circumbinary halo
328: and that some T Tauri stars have radio structures
329: as large as 15 times the stellar radius
330: (Mutel et al. 1985; Phillips et al. 1991).
331: Dipole-like magnetospheric models are often employed
332: to explain the radio observations of RS CVn binaries, T Tauri stars,
333: and other radio stars (G{\"u}del 2002).
334: In the case of IRS 5b,
335: the structure of the magnetosphere can be more complicated
336: because it may possess a circumstellar disk.
337:
338: \subsubsubsection{The Active Phase in 1998 September--October}
339:
340: As mentioned above,
341: there is an overall correlation between the $I$ and $V$ light curves.
342: During the active phase in 1998 (Tr 7--12),
343: the signal-to-noise ratio is high enough
344: that a more detailed description of the light curves is possible.
345: While the $I$ curve shows a relatively rapid fluctuation,
346: the $V$ curve varies slowly and shows only one peak (Fig. 3).
347: The difference in the fluctuation timescale suggests
348: that the $I$ flux may respond to the energy injection relatively quickly
349: and the $V$ flux may come from a population of electrons
350: having a longer relaxation timescale.
351: This behavior can be explained by a core-halo model
352: (Mutel et al. 1985; see the discussion below).
353:
354: \subsubsubsection{$I$-$V$ Correlation}
355:
356: When only IRS 5b is considered,
357: the correlation between $I$ and $V$ fluxes is quite good (Fig. 4).
358: The $V$ flux is roughly proportional to the $I$ flux.
359: The average value of $\Pi_c$ is 0.17, and the standard deviation is 0.04.
360: This good correlation strongly implies
361: that the $I$ and $V$ fluxes come from the same source of energy
362: and that the radiation mechanisms in the active phase
363: and in the quiescent phase are basically the same.
364:
365: If the quiescent radio emission is caused
366: by frequent injections of energetic electrons
367: through microflares (G{\"u}del 2002),
368: radio luminosity is expected to be related to X-ray luminosity.
369: The 3.5 cm flux density of IRS 5b in the quiescent phase
370: is $\sim$0.5 mJy (Paper I),
371: and the radio luminosity is
372: $L_R \approx$ 1.7 $\times$ 10$^{16}$ erg s$^{-1}$ Hz$^{-1}$.
373: The X-ray luminosity of IRS 5b is
374: $L_X$ = 1.9 $\times$ 10$^{30}$ erg s$^{-1}$ (Hamaguchi et al. 2008).
375: Then the ratio is $L_X/L_R$ = 1.1 $\times$ 10$^{14}$ Hz,
376: which is consistent with the empirical value of 10$^{15\pm1}$ Hz
377: for active stars (G{\"u}del 2002).
378: (See also the $L_R$-$L_X$ relationship
379: discussed by Feigelson et al. 1998 and Forbrich et al. 2006.)
380: Therefore, the quiescent radio emission of IRS 5b
381: is phenomenologically similar to what is observed in active late-type stars.
382:
383: Figure 4 shows
384: that the relation between $\Pi_c$ and $I$ is unclear
385: when all data are considered.
386: If the attention is restricted to the active phase only ($I >$ 0.8 mJy),
387: the scatter of $\Pi_c$ is smaller,
388: and $\Pi_c$ seems to be a weakly decreasing function of $I$.
389: This trend is clearer if only the active phase of 1998 is considered.
390: Such a behavior was also found in some radio active stars:
391: the fractional circular polarization of RS CVn binaries and Algol
392: decreases with increasing luminosity (Mutel et al. 1987, 1998).
393: This anticorrelation was explained using a model
394: in which the quiescent emission comes
395: from an optically thin region (halo)
396: of low Lorentz factor ($\gamma \lesssim$ 5)
397: while the intense flare emission comes from an optically thick region (core)
398: of more energetic electrons (Mutel et al. 1985, 1987).
399:
400: \subsubsubsection{Strength of Magnetic Fields}
401:
402: The emission mechanism of circularly polarized radio emission
403: at centimeter wavelength
404: is nonthermal gyrosynchrotron emission,
405: and the detection of $V$ flux at centimeter wavelength implies
406: that the strength of the magnetic fields in the source region
407: is a few hundred gauss or larger (G{\"u}del 2002).
408: The turnover frequency, $\nu_{\rm peak}$,
409: is a good indicator of the field strength.
410: For example, White et al. (1989) derived a simple relation
411: for a dipole source with typical parameters of M dwarf stars,
412: \begin{equation}
413: \nu_{\rm peak} = 4.2 (B_0/1000)^{0.76}\ {\rm GHz},
414: \end{equation}
415: where $B_0$ is the surface magnetic field strength in G.
416: In the quiescent phase, IRS 5b is brighter than IRS 5a,
417: and the spectral index of IRS 5b in the 4.9--8.5 GHz range
418: is flat or slightly negative (Paper I).
419: Then the turnover frequency of IRS 5b
420: may be roughly in this frequency range,
421: and the inferred field strength is $B_0 \approx$ 2 kG.
422:
423: \subsubsection{IRS 5a}
424:
425: While IRS 5a is more active in X-rays than IRS 5b,
426: as it frequently showed solar type X-ray flares (Hamaguchi et al. 2008),
427: it never showed polarized radio emission.
428: The nondetection of circular polarization of IRS 5a
429: is not simply owing to the relatively low $I$ flux,
430: because $V$ flux was undetected even in the strong flare event of Tr 4.
431: The 3 $\sigma$ upper limit on the polarization fraction is 0.03.
432: Therefore, while both radio polarization and X-ray emission
433: are related to magnetic activity,
434: there is no simple correlation
435: between polarized radio flux and X-ray flares.
436:
437: IRS 5a and 5b are quite similar in many aspects.
438: They have comparable X-ray luminosities and almost identical X-ray spectra
439: in the quiescent phase (Hamaguchi et al. 2008).
440: Since they are in a binary system,
441: they must have a similar age and are in the same environment.
442: Then what causes the difference in the polarization of radio emission?
443: A possible answer is the source geometry.
444: It is well known that the fractional circular polarization of RS CVn binaries
445: is dependent on the inclination angle of the system
446: (Mutel et al. 1987, 1998; Morris et al. 1990).
447: If this dependency is applicable to the IRS 5 system, it suggests
448: that the magnetosphere of IRS 5b may have a pole-on configuration
449: while that of IRS 5a may be equator-on.
450:
451: \subsubsection{Peculiarity of IRS 5}
452:
453: IRS 5b is the only known protostellar object
454: persistently displaying circularly polarized radio emission.
455: While this gives us a good opportunity
456: to study the magnetic activity of protostars,
457: it raises a question: why IRS 5b is an exception to the rule
458: that the protostellar magnetosphere
459: is hidden behind the free-free photosphere
460: (the surface where the optical depth of free-free radiation equals unity)?
461: That is, IRS 5b has either an unusually small free-free photosphere
462: or an unusually large magnetosphere.
463: Possible answers may include
464: a small mass-loss rate (decreased ionized wind),
465: a circumstellar disk with an inner hole (lack of disk-driven warm wind),
466: and/or large-scale magnetic fields generated by disk dynamo
467: (a disk-scale magnetosphere)
468: (Andr{\'e} et al. 1992, Tout \& Pringle 1996).
469:
470: Another unusual trait of IRS 5 is
471: that it is the only known protostellar binary system
472: with both members detected in X-rays (Hamaguchi et al. 2008).
473: An interesting issue is the effect on the ionization of disk material
474: because an accretion disk in such a system
475: is illuminated by two X-ray sources.
476: In terms of energy, only a small fraction of the X-ray photons from a star
477: can illuminate the disk around its companion.
478: (For example, a face-on disk with a radius about a third of binary separation
479: would intercept less that 3\% of the luminosity of the companion star.)
480: However, this extra X-ray source can irradiate the disk from above
481: if the binary system is not coplanar,
482: and the ionizing photons can penetrate into the disk midplane
483: relatively easily.
484: In addition, such an effect can generate some asymmetry in the disk.
485: Study of the dynamics and chemistry of accretion disks
486: in such an environment would be useful
487: because many stars form as members of multiple systems.
488: A circumstellar disk in a multiple system
489: may go through the X-ray irradiation
490: not only from the host star but also from its companion
491: at some point during the protostellar evolution.
492:
493: \subsection{IRS 7A}
494:
495: The evolutionary status of IRS 7A is unclear (see \S~5.1.1 of Paper I)
496: because the spectral energy distribution (SED) is poorly constrained.
497: IRS 7A is definitely not a Class II object (T Tauri star).
498: The X-ray characteristics of IRS 7A is similar to that of Class I objects
499: (Hamaguchi et al. 2005).
500: The mid-IR SED of IRS 7A is similar
501: to that of the Class 0/I transitional object IRS 7B
502: (also see the color-color diagram in Paper I),
503: but IRS 7A is less embedded than the nearby Class 0 source SMA 2
504: (Groppi et al. 2007).
505: IRS 7A displays an outflow activity (Paper I),
506: suggesting that accretion is also active.
507: Considering all these facts,
508: IRS 7A is most likely a protostar in the early part of the Class I stage.
509: Then it is the youngest known object
510: detected in circularly polarized radio emission,
511: or a protostar showing an evidence of magnetosphere
512: with an organized field configuration
513: and a field strength of the order of 10--100 G.
514:
515: \input{fig5.tex}
516:
517: Circularly polarized flux from IRS 7A was detected
518: only once out of the 13 observing tracks reported here (Fig. 5).
519: In addition, Suters et al. (1996) and Forbrich et al. (2007)
520: also made Stokes $V$ maps of the R CrA region
521: but did not report any detection at the position of IRS 7A.
522: Therefore, IRS 7A must be displaying
523: a detectable circular polarization only rarely.
524: However, at the epoch of the detection (Tr 5),
525: the $I$ flux was not particularly high.
526: In fact, the $I$ flux was on the low side of the scatter
527: (see Table 4 of Forbrich et al. 2006).
528: Even if we consider the intensity
529: at the position of IRS 7A in uniform-weight maps
530: (to minimize the blending of flux from the extended outflow),
531: the $I$ intensity at Tr 5 is on the weaker side.
532: This difference between $I$ and $V$ fluxes obviously shows
533: that most of the $I$ flux comes from a thermal source (Paper I)
534: and only a fraction of $I$ flux can be directly associated with the $V$ flux.
535: Therefore, if the angular resolution had been high enough
536: to separate the thermal and nonthermal sources,
537: the fractional circular polarization of the nonthermal emission
538: must have been much higher than 0.07 (see \S~3).
539:
540: In contrast to the radio $V$ flux,
541: the X-ray emission of IRS 7A was detectable most of the time
542: (Hamaguchi et al. 2005; Forbrich et al. 2006, 2007).
543: Then an active magnetosphere probably exists all the time,
544: and nonthermal radio emission may persist at some level.
545: If so, why is the circular polarization usually undetectable?
546: In fact, the nondetection is not surprising
547: because the radio emission from the peristellar region of protostars
548: is supposed to be obscured owing to the free-free absorption
549: by the partially-ionized thermal winds (Andr{\'e} et al. 1992).
550: Then the question becomes how the radiation from the magnetosphere
551: survived the absorption at the epoch of detection?
552: Several possibilities can be suggested.
553: First, the magnetosphere momentarily grew very large.
554: Either the whole magnetosphere expanded
555: or a portion of it stretched out.
556: Second, the free-free photosphere shrank temporarily.
557: This rarefication of wind can happen
558: if the mass-loss rate (hence the accretion rate) decreases for some reason.
559: Third, the thermal wind was far from spherical
560: and allowed an occasional direct view of the magnetosphere.
561: Such a favorable line-of-sight can be achieved
562: if the wind is collimated and changes directions.
563: In principle, these possibilities can be tested using the VLBI technique
564: although the transient nature of the phenomenon will be a major difficulty.
565:
566: \subsection{Magnetic Fields of Protostars}
567:
568: If we take IRS 7A and IRS 5b as indicative examples
569: of the evolution of magnetically active Class I objects,
570: for lack of other examples, a plausible picture emerges:
571: Most X-ray emitting Class I protostars may have magnetospheres.
572: Younger ones, such as IRS 7A, may hide
573: the existence of organized magnetic fields most of the time
574: because the mass-loss rate is usually high
575: and the wind is optically thick.
576: Older protostars, such as IRS 5b, have a decreased level of mass loss,
577: the wind becomes optically thin,
578: and the magnetosphere becomes easily visible.
579: (See the discussion by Andr{\'e} et al. 1992.)
580: The obvious next question is
581: whether such magnetospheres exist around Class 0 protostars.
582: Since there is no bona fide Class 0 object with detectable X-ray emission,
583: the answer to this question may be hard to get.
584: (This issue may be somewhat controversial
585: because IRS 7B is around the verge of the Class 0 protostar category.
586: See Hamaguchi et al. 2005
587: for more discussions on the implications of X-ray emission from IRS 7B.)
588: A good starting point may be monitoring observations
589: of X-ray-emitting young protostars such as IRS 7B.
590: Another possible strategy is monitoring observations
591: of Class 0 protostars that are weak in free-free emission.
592: In this respect, the detection of circularly polarized 6 cm emission
593: from the very low luminosity object L1014-IRS
594: warrants further investigations of similar objects (Shirley et al. 2007).
595:
596: The protostellar magnetosphere plays an important role
597: in the circumstellar accretion
598: and the accretion of matter in the inner disk.
599: Theoretical models and observational characterization
600: are relatively well developed for T Tauri stars,
601: but not much is known about protostellar magnetospheres.
602: Results of radio polarimetry, as reported in this paper,
603: are beginning to provide valuable information on this important issue.
604: The radio properties of IRS 7A and IRS 5b imply
605: that the magnetic activity of Class I protostars
606: are qualitatively similar to those of T Tauri stars.
607: Future telescopes with better sensitivity and higher angular resolution
608: should be able to help us understand
609: the role of magnetic processes in protostellar evolution,
610: and monitoring observations are crucial
611: because of the time variability of these sources.
612:
613: \subsection{Frequency Dependence of Polarization}
614:
615: All the three sources displaying circularly polarized emission at 3.5 cm
616: did not show any detectable polarization at 6.2 cm.
617: Such a frequency dependence is consistent with the observed properties
618: of other magnetically active stars such as RS CVn systems:
619: the fractional polarization tends to increase with frequency
620: in the 5--15 GHz range
621: (White \& Franciosini 1995; Garc{\'\i}a-S{\'a}nchez et al. 2003).
622: However, it is not easy to explain this behavior
623: using simple theoretical models of gyrosynchrotron radiation,
624: and inhomogeneous models were suggested
625: (Jones et al. 1994; White \& Franciosini 1995).
626: It is unclear how magnetically active protostars would behave
627: over a wide range of frequency,
628: and multifrequency observations of IRS 5b will be interesting.
629:
630: \subsection{B5}
631:
632: B5 is a radio source with rapid flux variability
633: (Brown 1987; Suters et al. 1996).
634: It is not an extragalactic source,
635: but its nature is poorly known (Paper I).
636: Feigelson et al. (1998) suggested that B5 is a brown dwarf,
637: but we cannot rule out the possibility
638: that it may be a background object
639: located much further away than the Cr A cloud.
640:
641: \input{fig6.tex}
642:
643: Circularly polarized emission from B5 was never reported before
644: and detected only once (Tr 13) out of the 13 observing tracks (Fig. 6).
645: The $I$ flux during Tr 13 was larger than the flux at the other epochs
646: by a factor of $\sim$3 (see Fig. 2 of Paper I).
647: The fractional polarization during Tr 13
648: was also higher than the other epochs
649: because polarized flux must have been detectable most of the time
650: if $\Pi_c$ were constant.
651: This enhancement of $I$ and $\Pi_c$ during Tr 13 suggests
652: that B5 seems to have gone through an outburst
653: of highly polarized radio emission.
654: Interestingly, the quiescent emission is also nonthermal
655: because the spectral index during Tr 1 is highly negative
656: (see Table 3 of Paper I).
657: Then the radio emission structure of B5 must be peculiar:
658: the quiescent emission may be
659: synchrotron radiation from a large-scale structure,
660: and the intense flare emission may be gyrosynchrotron radiation
661: from a compact structure, probably a peristellar magnetosphere.
662: Since the 6.2 cm source is extended
663: and has a position offset with respect to the 3.5 cm source (Paper I),
664: the large-scale structure is probably an outflow.
665:
666: If B5 is a brown dwarf as suggested by Feigelson et al. (1998),
667: it joins the small list of substellar objects
668: detected in circularly polarized radio emission
669: (Berger et al. 2001; Berger 2006).
670: These objects show much higher radio luminosity
671: than what is expected from the X-ray luminosity,
672: which may explain the nondetection of B5 in X-rays.
673: It is not well understood
674: how such cool dwarfs can generate large-scale magnetic fields
675: (Berger 2006).
676:
677: On the other hand, the peculiar emission structure
678: suggests a quite different possibility:
679: B5 may be an exotic object
680: containing a magnetically active compact source
681: that is ejecting material at a highly relativistic speed.
682: But it is certainly not an X-ray binary
683: because no X-ray detection was ever reported.
684: The nature of B5 is still an open question.
685:
686: \section{SUMMARY}
687:
688: The R CrA region was observed using the VLA in the 3.5 and 6.2 cm continuum
689: to image the young stellar objects
690: and study the star forming activity in this region.
691: In addition, archival VLA data from recent observations
692: made in extended configurations were analyzed.
693: The results based on the imaging of Stokes $I$ intensity
694: were reported in Paper I,
695: and the results of Stokes $V$ imaging are reported in this paper.
696: A total of 3 sources was detected in the 3.5 cm Stokes $V$ flux
697: while none was detected in the 6.2 cm circularly polarized continuum.
698: The main results are summarized as follows:
699:
700: 1.
701: Circularly polarized emission from IRS 5 was detected most of the time,
702: and the exact source of the $V$ flux was identified as IRS 5b.
703: When IRS 5a and 5b are separated,
704: there is a good correlation between the $I$ and $V$ fluxes of IRS 5b.
705: The circular polarization is present at a $\sim$17\% level,
706: even in the quiescent phase,
707: as long as the sensitivity is high enough to detect the $V$ flux.
708:
709: 2.
710: The $V$ flux from IRS 5b was always positive
711: over the 8 yr of monitoring,
712: suggesting that the magnetosphere has a stable configuration.
713: During the active (flaring) phases,
714: when the signal-to-noise ratio is high enough,
715: the fractional polarization is a weakly decreasing function of $I$ flux.
716: The active phases of IRS 5b last about a month or longer,
717: and the light curve of $V$ flux varies more slowly than that of $I$ flux,
718: which suggests
719: that the source of the circularly polarized emission is huge in size.
720: These properties suggests
721: that IRS 5b is phenomenologically similar to other types of flare stars
722: and has a very large magnetic structure.
723:
724: 3.
725: IRS 5a was never detected in the circularly polarized radio continuum
726: even during flares.
727: The cause of the difference between IRS 5a and 5b is unclear,
728: but a geometrical effect is possible.
729:
730: 4.
731: Circularly polarized flux from IRS 7A was detected once
732: out of the 13 observing tracks analyzed.
733: The fractional circular polarization was at least 7\%.
734: Since IRS 7A drives a thermal radio jet,
735: it is unclear how the polarized flux
736: was not completely absorbed by the thermal wind.
737:
738: 5.
739: Considering the properties of IRS 7A and IRS 5b, we suggest
740: that most (X-ray emitting) Class I protostars may have magnetospheres
741: and that the detection of radio emission from the magnetosphere
742: becomes easier as the protostar evolves and the mass-loss rate decreases.
743: Future surveys of similar Class I sources
744: with the Expanded Very Large Array
745: will be very exciting
746: because the continuum sensitivity will be improved by an order of magnitude,
747: and it may be possible to test the hypothesis above.
748:
749: 6.
750: Circularly polarized emission from B5 was detected once,
751: and the fractional circular polarization was 12\%.
752: B5 was in an outburst when the $V$ flux was detected.
753: While the detection of circular polarization shows
754: that there is an organized magnetic field structure,
755: the nature of B5 is still unclear.
756:
757: \enlargethispage{-15\baselineskip}
758:
759: \acknowledgements
760:
761: We thank Y. L. Shirley and J. Cho for helpful discussions and suggestions.
762: M. C. was supported by the Korea Science and Engineering Foundation (KOSEF)
763: through the grant of the basic research program R01-2007-000-20196-0.
764: K. H. is supported by the NASA Astrobiology Program under CAN 03-OSS-02.
765: J.-E. L. gratefully acknowledges the support by KOSEF
766: under a cooperative agreement with the Astrophysical Research Center
767: for the Structure and Evolution of the Cosmos (ARCSEC).
768: The National Radio Astronomy Observatory is
769: a facility of the National Science Foundation
770: operated under cooperative agreement by Associated Universities, Inc.
771:
772: \begin{references}
773: \reference{} Andr{\'e}, P. 1996, in ASP Conf. Ser. 93,
774: Radio Emission from the Stars and the Sun,
775: ed. A. R. Taylor \& J. M. Paredes (San Francisco: ASP), 273
776: \reference{} Andr{\'e}, P., Deeney, B. D., Phillips, R. B.,
777: \& Lestrade, J.-F. 1992, ApJ, 401, 667
778: \reference{} Balbus, S. A., \& Hawley, J. F. 1991, ApJ, 376, 214
779: \reference{} Berger, E. 2006, ApJ, 648, 629
780: \reference{} Berger, E., et al. 2001, Nature, 410, 338
781: \reference{} Bouvier, J., Alencar, S. H. P., Harries, T. J.,
782: Johns-Krull, C. M., \& Romanova, M. M. 2007,
783: in Protostars and Planets V, ed. B. Reipurth, D. Jewitt,
784: \& K. Keil (Tucson: Univ. Arizona Press), 479
785: \reference{} Brown, A. 1987, ApJ, 322, L31
786: \reference{} Camenzind, M. 1990, Rev. Mod. Astron., 3, 234
787: \reference{} Choi, M., Hamaguchi, K., Lee, J.-E., \& Tatematsu, K. 2008,
788: ApJ, in press (Paper I)
789: \reference{} Choi, M., \& Tatematsu, K. 2004, ApJ, 600, L55
790: \reference{} Feigelson, E. D., Carkner, L., \& Wilking, B. A. 1998,
791: ApJ, 494, L215
792: \reference{} Feigelson, E., Townsley, L., G{\"u}del, M., \& Stassun, K. 2007,
793: in Protostars and Planets V, ed. B. Reipurth, D. Jewitt,
794: \& K. Keil (Tucson: Univ. Arizona Press), 313
795: \reference{} Forbrich, J., Preibisch, T., \& Menten, K. M. 2006,
796: A\&A, 446, 155
797: \reference{} Forbrich, J. et al. 2007, A\&A, 464, 1003
798: \reference{} Garc{\'\i}a-S{\'a}nchez, J., Paredes, J. M.,
799: \& Rib{\'o}, M. 2003, A\&A, 403, 613
800: \reference{} Groppi, C. E., Hunter, T. R., Blundell, R., \& Sandell, G. 2007,
801: ApJ, 670, 489
802: \reference{} G{\"u}del, M. 2002, ARA\&A, 40, 217
803: \reference{} Hamaguchi, K., Choi, M., Corcoran, M. F., Choi, C.-S.,
804: Tatematsu, K., \& Petre, R. 2008, ApJ, in press
805: \reference{} Hamaguchi, K., Corcoran, M. F., Petre, R., White, N. E.,
806: Stelzer, B., Nedachi, K., Kobayashi, N.,
807: \& Tokunaga, A. T. 2005, ApJ, 623, 291
808: \reference{} Jones, K. L., Stewart, R. T., Nelson, G. J.,
809: \& Duncan, A. R. 1994, MNRAS, 269, 1145
810: \reference{} Knude, J., \& H{\o}g, E. 1998, A\&A, 338, 897
811: \reference{} K{\"o}nigl, A. 1991, ApJ, 370, L39
812: \reference{} Miettinen, O., Kontinen, S., Harju, J., \& Higdon, J. L. 2008,
813: A\&A, 486, 799
814: \reference{} Morris, D. H., Mutel, R. L., \& Su, B. 1990, ApJ, 362, 299
815: \reference{} Mutel, R. L., Lestrade, J. F., Preston, R. A.,
816: \& Phillips, R. B. 1985, ApJ, 289, 262
817: \reference{} Mutel, R. L., Molnar, L. A., Waltman, E. B.,
818: \& Ghigo, F. D. 1998, ApJ, 507, 371
819: \reference{} Mutel, R. L., Morris, D. H., Doiron, D. J.,
820: \& Lestrade, J. F. 1987, AJ, 93, 1220
821: \reference{} Nisini, B., Antoniucci, S., Giannini, T.,
822: \& Lorenzetti, D. 2005, A\&A, 429, 543
823: \reference{} Nutter, D. J., Ward-Thompson, D., \& Andr{\'e}, P. 2005,
824: MNRAS, 357, 975
825: \reference{} Phillips, R. B., Lonsdale, C. J., \& Feigelson, E. D. 1991,
826: ApJ, 382, 261
827: \reference{} Preibisch, T., \& Feigelson, E. D. 2005, ApJS, 160, 390
828: \reference{} Pudritz, R. E., Ouyed, R., Fendt, C., \& Brandenburg, A. 2007,
829: in Protostars and Planets V, ed. B. Reipurth, D. Jweitt,
830: \& K. Keil (Tucson: Univ. Arizona Press), 277
831: \reference{} Shirley, Y. L., Claussen, M. J., Bourke, T. L., Young, C. H.,
832: \& Blake, G. A. 2007, ApJ, 667, 329
833: \reference{} Suters, M., Stewart, R. T., Brown, A., \& Zealey, W. 1996,
834: AJ, 111, 320
835: \reference{} Tout, C. A., \& Pringle, J. E. 1996, MNRAS, 281, 219
836: \reference{} White, R. J., Greene, T. P., Doppmann, G. W., Covey, K.~R.,
837: \& Hillenbrand, L. A. 2007,
838: in Protostars and Planets V, ed. B. Reipurth, D. Jweitt,
839: \& K. Keil (Tucson: Univ. Arizona Press), 117
840: \reference{} White, S. M., \& Franciosini, E. 1995, ApJ, 444, 342
841: \reference{} White, S. M., Kundu, M. R., \& Jackson, P. D. 1989,
842: A\&A, 225, 112
843: \reference{} Wilking, B. A., McCaughrean, M. J., Burton, M. G., Giblin, T.,
844: Rayner, J. T., \& Zinnecker, H. 1997, AJ, 114, 2029
845: \end{references}
846:
847: %\TimeStamp
848:
849: \end{document}
850:
851: