1: \documentclass[12pt,preprint]{aastex}
2:
3:
4: \newcommand{\mysec}[2]{$#1\mbox{$^{\rm ''}\mskip-7.6mu.\,$}#2$}
5: \newcommand{\inthms}[3]{$#1^{\rm h}#2^{\rm m}#3^{\rm s}$}
6: \newcommand{\dechms}[4]{$#1^{\rm h}#2^{\rm m}#3\mbox{$^{\rm s}\mskip-7.6mu.\,$}#4$}
7: \newcommand{\intdms}[3]{$#1^{\circ}#2'#3''$}
8: \newcommand{\decdms}[4]{$#1^{\circ}#2'#3\mbox{$''\mskip-7.6mu.\,$}#4$}
9: \newcommand{\msec}[2]{$#1\mbox{$''\mskip-7.6mu.\,$}#2$}
10: \newcommand{\mmsec}[2]{$#1\mbox{$^s\mskip-7.6mu.\,$}#2$}
11: \newcommand{\mdeg}[2]{$#1\mbox{$^\circ \mskip-7.6mu.\,$}#2$}
12: \newcommand{\mtsec}[2]{$#1\mbox{$^{\rm s}\mskip-7.6mu .\,$}#2$}
13: \newcommand{\val}[3]{$#1\mskip3mu_{-#2}\mskip-30mu^{+#3}$}
14: \newcommand{\vall}[3]{$#1\mskip3mu_{-#2}\mskip-37mu^{+#3}$}
15: \newcommand{\valll}[3]{$#1\mskip3mu_{-#2}\mskip-44mu^{+#3}$}
16: \newcommand{\frii}{FR{\sc \,ii}}
17: \newcommand{\sbeamp}[5]{{$#1\mbox{$''\mskip-7.6mu.\,$}#2$} $\times$ {$#3\mbox{$''\mskip-7.6mu.\,$}#4$}; $+#5^\circ$}
18: \newcommand{\sbeamm}[5]{{$#1\mbox{$''\mskip-7.6mu.\,$}#2$} $\times$ {$#3\mbox{$''\mskip-7.6mu.\,$}#4$}; $-#5^\circ$}
19: \newcommand{\offsetp}[3]{{$#1\mbox{$''\mskip-7.6mu.\,$}#2$}; $+#3^\circ$}
20: \newcommand{\offsetm}[3]{{$#1\mbox{$''\mskip-7.6mu.\,$}#2$}; $-#3^\circ$}
21: \newcommand{\Msun}{M$_{\odot}$}
22:
23:
24: \begin{document}
25:
26:
27: \title{New radio sources and the composite structure of component B\\
28: in the very young protostellar system IRAS~16293--2422\\}
29:
30: \author{Laurent Loinard}
31: \affil{Centro de Radiostronom\'{\i}a y Astrof\'{\i}sica,
32: Universidad Nacional Aut\'onoma de M\'exico,\\
33: Apartado Postal 72--3 (Xangari), 58089 Morelia, Michoac\'an, M\'exico;\\ l.loinard@astrosmo.unam.mx}
34:
35: \author{Claire J.\ Chandler}
36: \affil{National Radio Astronomy Observatory, P.O.\ Box O, Socorro, NM 87801, USA; cchandle@nrao.edu}
37:
38:
39: \author{Luis F.\ Rodr\'{\i}guez and Paola D'Alessio}
40: \affil{Centro de Radiostronom\'{\i}a y Astrof\'{\i}sica,
41: Universidad Nacional Aut\'onoma de M\'exico,\\
42: Apartado Postal 72--3 (Xangari), 58089 Morelia, Michoac\'an, M\'exico;\\ l.rodriguez, p.dalessio@astrosmo.unam.mx}
43:
44: \author{Crystal L.\ Brogan}
45: \affil{National Radio Astronomy Observatory, 520 Edgemont Road, Charlottesville, VA 22903-2475, USA; cbrogan@nrao.edu}
46:
47:
48: \author{David J.\ Wilner and Paul T.P.\ Ho\altaffilmark{1}}
49: \affil{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street,
50: Cambridge, MA 02138;\\ dwilner, pho@cfa.harvard.edu}
51:
52: \altaffiltext{1}{Also at: Academia Sinica, Institute of Astronomy and Astrophysics, Taipei 106, Taiwan}
53:
54:
55: \begin{abstract}
56:
57: In this article, we report high-resolution ($\sim$ \msec{0}{1}--\msec{0}{3}),
58: high-sensitivity ($\sim$ 50--100 $\mu$Jy beam$^{-1}$) Very Large Array
59: 0.7 and 1.3 cm observations of the young stellar system
60: IRAS~16293--2422 in $\rho-$Ophiuchus. In the 0.7 cm image, component A
61: to the south-east of the system looks like its usual binary self. In
62: the new 1.3 cm image, however, component A2 appears to have split into
63: two sub-components located roughly symmetrically around the original
64: position of A2. This change of morphology is likely the result of a
65: recent bipolar ejection, one of the very first such events observed in
66: a low-mass source. Also in component A, a marginal detection of 0.7
67: cm emission associated with the submillimeter component Ab is
68: reported. If confirmed, this detection would imply that Ab is a
69: relatively extended dusty structure, where grain coagulation may
70: already have taken place. With an angular size increasing with frequency,
71: and an overall spectra index of 2, the emission from component B to the
72: north-west of the system is confirmed to be dominated by optically thick
73: thermal dust emission associated with a fairly massive, nearly face-on,
74: circumstellar disk. In the central region, however, we find evidence
75: for a modest free-free contribution that originates in a structure
76: elongated roughly in the east-west direction. We argue that this
77: free-free component traces the base of the jet driving the large-scale
78: bipolar flow at a position angle of about 110$^\circ$ that has long
79: been known to be powered by IRAS~16293--2422.
80:
81: \end{abstract}
82:
83:
84: \keywords{stars: formation --- binaries: general --- astrometry ---
85: radio continuum: stars --- stars: individual (IRAS~16293--2422)}
86:
87: \section{Introduction}
88:
89: A significant fraction of the stars in the Solar neighborhood are
90: known to reside and are suspected to have formed in binary or higher
91: order multiple systems\footnote{The exact fraction may be smaller than
92: initially anticipated for very low-mass stars (Lada 2006), but it is
93: high for the Solar type objects such as those considered here.}
94: (Duquennoy \& Mayor 1991). Yet, while the formation of isolated
95: Solar-type stars is reasonably well understood (Shu, Adams \& Lizano
96: 1987), our comprehension of the formation of multiple stellar systems
97: remains comparatively limited. Several theoretical mechanisms have
98: been put forward, but observational constraints on the earliest
99: evolutionary phases are scarce owing to the difficulty of identifying
100: and studying the various constituents (disks, jets, etc.) of extremely
101: young systems (for a recent review on the properties of young
102: binaries, see Duch\^ene et al.\ 2007). This is because the size scale
103: of, and separation between, these structures is typically of the order
104: of a few to a few tens of astronomical units, and so observations with
105: an angular resolution better than a few tenths of an arcsecond are
106: needed even in the nearest star-forming regions. Unfortunately, data
107: with this kind of angular resolution are currently unavailable at the
108: frequencies (mid-infrared, far-infrared, and submillimeter) best
109: matched to the energy output of deeply embedded sources. Such
110: resolutions can only be reached in the optical and near-infrared (but
111: very enshrouded young stars are not detected at these frequencies),
112: and at radio wavelengths (0.7 cm $\leq$ $\lambda$ $\leq$ 6 cm) thanks
113: to the availability of large radio-interferometers. Most of the
114: structures found in young stellar systems do emit in the radio domain,
115: but different mechanisms are at work in different components (thermal
116: dust emission in accretion disks, thermal bremsstrahlung in jets,
117: gyrosynchrotron in magnetically active young stars,
118: etc.). Consequently, a detailed analysis of each radio source found in
119: a given protostellar system has to be carried out before it can be
120: identified --or at least associated-- with a specific component.
121:
122: \section{IRAS~16293--2422}
123:
124: Located in L1689N, a dark cloud in the $\rho-$Ophiuchus star-forming
125: complex (at $d$ = 120 pc --R.M.\ Torres et al., in prep),
126: IRAS~16293--2422 is a well-studied very young low-mass protostellar
127: system (e.g.\ Ceccarelli et al.\ 2000a). It has been suspected to be
128: multiple since Mundy et al.\ (1986) and Wootten (1989) showed that it
129: was a double source both at millimeter and centimeter
130: wavelengths. Soon after, it was also found to power a multi-lobe
131: outflow system (Mizuno et al.\ 1990), comprising two bipolar and one
132: monopolar structures. The two bipolar flows are fairly compact, and at
133: position angles of approximately 60$^\circ$ and 110$^\circ$ (Mizuno et
134: al.\ 1990, Hirano et al.\ 2001, Castets et al.\ 2001), whereas the
135: monopolar flow is a parsec-scale, blueshifted feature, located
136: eastward of IRAS~16293--2422 at a position angle of almost exactly
137: 90$^{\circ}$ (Mizuno et al.\ 1990). The lack of a redshifted
138: counterpart toward the west is likely related to the location of
139: IRAS~16293--2422 near the western edge of L1689N. While the two bipolar
140: flows are now generally accepted to be driven by two distinct
141: protostellar sources, it is still debated whether the mono-sided
142: structure is driven by yet another young star in the system, or if it
143: is an extension of the compact outflow at P.A.\ 110$^\circ$ (Mizuno et
144: al.\ 1990, Stark et al.\ 2004). In the latter situation, the jet
145: powering the east-west flows must either have undergone significant
146: precession, or have been strongly deflected by a condensation of dense
147: gas along its path.
148:
149: At the best resolution available at centimeter wavelengths,
150: IRAS~16293--2422 is resolved into three radio components (Wootten 1989,
151: Loinard 2002, Chandler et al.\ 2005). Components A1 and A2 --to the
152: south-east of the system; see Fig.\ 1-- are separated from each other
153: by about \msec{0}{34}, and from component B, to the north-west, by
154: about 5$''$. Component B appears to be well resolved in
155: high-resolution 0.7 cm observations (Rodr\'{\i}guez et al.\ 2005), has a
156: spectral index of 2--2.5 across the millimeter and centimeter ranges,
157: and has been interpreted as a nearly face-on, optically thick
158: accretion disk (Rodr\'{\i}guez et al.\ 2005, Chandler et al.\
159: 2005). It has often been associated with the outflow at P.A.\ $\sim$
160: 110$^\circ$ but significant doubts remain about this issue (Chandler
161: et al.\ 2005).
162:
163: Although it has been known for a long time to be at the origin of the
164: outflow at P.A.\ $\sim$ 60$^\circ$ (Mundy et al.\ 1986, 1992), the
165: exact nature of the A1/A2 pair remains unclear. Wootten (1989) argued
166: that A2 is the protostellar source driving the outflow and that A1 is
167: a shock feature along that flow. This conclusion was based --in part--
168: on the near-exact alignment with the flow of the segment joining A1 to
169: A2 at the epoch (1986-1987) of the observations published by Wootten
170: (1989). In that scheme, A1 could either be a knot of ionized gas
171: ejected by A2 some time in the past, or the result of the impact of
172: the jet powered by A2 onto dense circumstellar material. In the former
173: case, one would expect A1 to move steadily away from A2 along the
174: direction of the flow, whereas in the latter, no significant motions
175: are expected between A1 and A2. Multi-epoch radio observations with
176: sufficient angular resolution (better than about \msec{0}{2}) to
177: resolve A1 and A2, however, revealed neither of these behaviors
178: (Loinard 2002, Chandler et al.\ 2005). Instead, they showed that the
179: separation between A1 and A2 has remained constant at about
180: \msec{0}{34} in the last 15 years, whereas the position angle between
181: them has changed from less than 50$^\circ$ in the late 1980s to about
182: 85$^\circ$ in 2003 (Loinard 2002, Chandler et al.\ 2005). This
183: definitely rules out the possibility that A1 is an ejecta from A2,
184: because --as mentioned earlier-- one would then have expected constant
185: position angle and increasing separation rather than the opposite. The
186: possibility that A1 is the result of the interaction between a jet
187: powered by A2 and circumstellar material could still be retained if
188: that jet precessed strongly between the late 1980s and 2003, and were
189: now oriented almost exactly east-west (at P.A.\ $\sim$ 85$^{\circ}$).
190: Recent observations, however, have shown that the 0.7 cm radio emission
191: from A2 (that partly traces free-free radiation from the inner part of
192: the jet) is elongated at P.A.\ $\sim$ 60$^\circ$, very similar to the
193: orientation of the large-scale molecular flow driven from within
194: component A. Thus, it appears that A2 {\em does} drive the large-scale
195: flow at P.A.\ $\sim$ 60$^\circ$, and that A1 is {\em not} a shock
196: feature along that flow.
197:
198: The simplest interpretation of the relative motion between A1 and A2
199: (constant separation and linearly increasing position angle) is that
200: they are two protostellar sources in Keplerian orbit (Loinard
201: 2002). In this hypothesis, the total mass of the system can be
202: estimated from Kepler's third law. Assuming a face-on circular orbit,
203: the mass of the system implied by the motions is 2.7(d/120 pc)$^3$
204: \Msun\ (where $d$ is the distance --Chandler et al.\ 2005). For an
205: elliptical orbit, a somewhat smaller value would be permitted, but the
206: total mass has to be larger than 1.4 (d/120 pc)$^3$ \Msun\ for the
207: system to remain bound. Moreover, this minimum value is only attained
208: for an extremely eccentric orbit observed exactly at periastron. For
209: less special circumstances, a mass larger than 2 to 3(d/120 pc)$^3$
210: \Msun\ is required. Finally, an analysis of the absolute proper
211: motions (Loinard 2002, Chandler et al.\ 2005) shows that, in this
212: orbital motion scenario, the center of mass would have to be very
213: close to A2, so that A2 would have to be significantly more massive
214: than A1. The A1/A2 pair would then be a binary system made of a
215: relatively low-mass protostar ($M$ $\lesssim$ 0.5 \Msun) in orbit
216: around a somewhat more massive one ($M$ $\gtrsim$ 2 \Msun). That
217: binary system would, in turn, be orbited by the protostar associated
218: with component B, and IRAS~16293--2422 would be a hierarchical triple
219: system. Note that if A1 and A2 indeed form a binary system, then the
220: position angle between them should keep increasing monotonically with
221: time. Furthermore, if observations with a quality similar to those
222: used by Loinard (2002) and Chandler et al.\ (2005) are obtained
223: regularly in the coming decade, it will become possible to fit a
224: Keplerian orbit to the data, and determine accurately the mass of
225: both components.
226:
227: An alternative to this orbital motion scenario was recently formulated
228: by Chandler et al.\ (2005). In that proposal, A1 is interpreted as a
229: shock resulting from the impact of a strongly precessing (or wobbling)
230: jet driven by a third --as-yet undetected-- protostar in the
231: system. Of course, in this alternative scheme, the change of position
232: angle with time should decelerate, and eventually reverse its course
233: because the jet must oscillate around an equilibrium value. Note that
234: the required fast rate of precession or wobbling (about 40$^\circ$ in
235: less than twenty years) has never been observed before, and would
236: require the new protostar to be a member of a very tight binary. Given
237: the overall morphology of the system, the most likely situation would
238: be for this new member to be a close companion of A2 --with a
239: separation smaller than 22 AU (\msec{0}{18} if d = 120 pc --Chandler
240: et al.\ 2005). It is interesting to point out that, in that
241: alternative scheme, there would also be two protostellar sources in
242: the A component and, including component B, IRAS~16293--2422 would
243: again be a hierarchical triple system. Thus, the two scenarios put
244: forth so far call for the existence of three protostellar sources in
245: IRAS~16293--2422. Given that three outflow systems are known to be
246: powered from within IRAS~16293--2422 (see above), it is, of course,
247: tempting to associate each one of the sources with one of the flow
248: features. So far, however, the only secure association is that between
249: A2 and the flow at P.A.\ $\sim$ 60$^\circ$.
250:
251: In recent high-resolution, submillimeter observations (Chandler et
252: al.\ 2005), IRAS~16293--2422 was also found to be a triple source (Aa,
253: Ab, and B --see Fig.\ 1), but the exact correspondence between the
254: radio and submillimeter sources in the A component is not immediately
255: obvious\footnote{The situation is clear for component B, where the
256: position of the submillimeter and radio sources are almost identical,
257: and the submillimeter flux is in excellent agreement with that
258: expected from the extrapolation of the centimeter data.}. Component Aa
259: is located just about halfway between A1 and A2 (Fig.\ 1) and it is
260: unclear if it is another protostar in the system, or a blend of
261: roughly equal amounts of emission associated with A1 and A2. If it
262: was a new source, its position roughly \msec{0}{2} from A2 would, of
263: course, make it a potential candidate for the companion of A2 driving
264: the precessing jet that impacts circumstellar material at
265: A1. Component Ab, on the other hand, is located about \msec{0}{6} to
266: the north-east of component A, and is not detected at any other
267: wavelengths. Since the mechanism at work is almost certainly dust
268: thermal emission, Ab must be a dense dusty structure, and Chandler et
269: al.\ (2005) argued that it may be associated with a fourth protostar
270: in the system. The absence of centimeter emission, however, shows that
271: it is not associated with a strong ionized wind. Moreover, for thermal
272: dust emission, the spectral energy distribution is expected to be a
273: power law with an index $\alpha$ between 2 and 4. From the 0.5 Jy flux
274: detected at 305 GHz, one would expect a 0.7 cm flux of about 10 mJy if
275: the spectral index is 2, and 0.2 mJy if $\alpha$ is 4. A 10 mJy source
276: would have been easily detected in the 0.7 cm observations reported by
277: Rodr\'{\i}guez et al.\ (2005) unless the emission were fairly extended
278: ($\gtrsim$ 0.5$''$) and significantly filtered out. Only a very
279: marginal detection (2$\sigma$) would have been obtained in the second
280: case ($\alpha$ = 4), but the source would have been detected above
281: 5$\sigma$ for any spectral index smaller than 3.5 (again, unless the
282: emission were extended). Thus, the non-detection of Ab in the 0.7 cm
283: data reported by Rodr\'{\i}guez et al.\ (2005) implies either that the
284: emission is fairly extended, or that the spectral index is larger than
285: 3.5, and the dust consequently largely unprocessed (or a combination
286: of both). Together with the lack of free-free emission, these
287: characteristics argue in favor of either an extremely young
288: protostar, or even a starless condensation.
289:
290: IRAS~16293--2422 has long been known to exhibit a very active
291: chemistry (e.g.\ van Dishoeck et al.\ 1995, Cazaux et al.\ 2003),
292: and is one of the sources where the hot corinos (Ceccarelli et al.\
293: 2007) that have been argued to trace the earliest stages of low-mass
294: star-formation have first been identified. The earliest observations
295: (e.g.\ Ceccarelli et al.\ 2000b and references therein) did not
296: resolve the various components of IRAS~16293--2422, but several recent
297: high-resolution interferometric millimeter observations (Kuan et
298: al.\ 2004, Bottinelli et al.\ 2004, Chandler et al.\ 2005) have
299: demonstrated that the organic and highly deuterated molecules
300: considered signposts of hot corinos exist in both component A and
301: component B. It appears, therefore, likely that each component harbors
302: at least one very young protostar.
303:
304: \bigskip
305:
306: In summary, although it has been extensively studied and several
307: characteristics of IRAS~16293--2422 are now well-established, many
308: other questions remain open about the very nature of several of the
309: sources in the system, and their exact relation to the large-scale
310: outflows in the region. It is important to address these issues,
311: because IRAS~16293--2422 is a rare example of a very young multiple
312: system, that offers a unique opportunity to study the earliest stages
313: of the formation of multiple systems.
314:
315: \section{Observations}
316:
317: In this article, we will present and analyze
318: two new, high-resolution, high-sensitivity, radio continuum
319: observations of IRAS~16293--2422 obtained at 0.7 and 1.3 cm (Tab.\
320: 1). The data were collected with the {Very Large Array} (VLA) of the
321: {National Radio Astronomy Observatory} (NRAO\footnote{NRAO is a
322: facility of the National Science Foundation operated under cooperative
323: agreement by Associated Universities, Inc.}) in the most extended (A)
324: and second most extended (B) configurations of the array at 1.3 cm and
325: 0.7 cm, respectively. The standard VLA continuum frequency setups were
326: used: two frequencies (22.4851 and 22.4351 GHz at 1.3 cm, and 43.3149
327: and 43.3649 GHz at 0.7 cm) were observed simultaneously, and in both
328: circular polarizations, with 50 MHz of bandwidth each. The absolute
329: flux density was set using observations of 3C 286. For improved flux
330: accuracy, we did not assume 3C 286 to be a point source, but instead
331: used a model image provided by NRAO. The phase calibrator was PKS
332: J1625--2527 whose absolute position is expected to be accurate to
333: about 2 milli-arcseconds. The fast-switching technique --that
334: consists of rapidly alternating observations of the source and the
335: phase calibrator with a cycle time of 2 minutes-- was used for both
336: observations; this ensured optimal imaging fidelity. The 0.7 cm data
337: were restored with natural weighting of the visibilities, whereas the
338: 1.3 cm data were imaged with a robust weighting scheme intermediate
339: between uniform and natural (see Tab.\ 1 for the corresponding
340: synthesized beams).
341:
342: To complement these new data, we will also make use of observations
343: previously published in the literature. In particular, we will use the
344: recent 0.7 cm observations obtained in 2003.47 and reported by
345: Rodr\'{\i}guez et al.\ (2005), and the 3.6 cm high-resolution
346: observations obtained in 2003.65, and reported by Chandler et al.\
347: (2005). For these two datasets, new images were made from the
348: calibrated visibilities. The 0.7 cm data from 2003.47 were
349: initially imaged with natural weighting of the visibilities.
350: For the study of component A (which is faint at 0.7 cm) as well
351: as for accurate comparison with the new 1.3 cm data, this image
352: was also smoothed to the resolution of the 1.3 cm image (Tab.\ 1)
353: The 3.6 cm data were imaged with a robust weighting scheme
354: intermediate between natural and uniform.
355:
356: Our new B-array 0.7 cm observations were found to produce fluxes
357: systematically 30\% lower than the older A-array observations.
358: This was originally identified by comparing the total flux of
359: component B (which has never been found to be variable in any
360: previous observations) in the two datasets. A similar difference
361: was found when comparing the total flux corresponding to similar
362: baselines and hour angles in our A- and B-configuration observations.
363: This effect clearly cannot be due to missing flux related extended
364: emission, since this would have produced more flux in the
365: B-configuration observations than in the A-array data, rather
366: than the observed opposite. Instead, the flux deficit is likely
367: related to the relatively poorer weather conditions during our B-array
368: observations. To account for this effect, the new B-configuration
369: 0.7 cm data were multiplied by 1.3 before imaging. The only instance
370: when this will be important will be in Sect.\ 4.3, when we estimate
371: the flux of the possible 0.7 cm counterpart of Ab.
372:
373:
374: \section{Results and discussion}
375:
376: \subsection{Component A}
377:
378: In our new 0.7 cm observation (Fig.\ 2b), component A looks like its
379: usual binary self, with sub-components A1 and A2 clearly resolved. In
380: the new 1.3 cm image, however, it has clearly become triple (Fig.\ 2a,
381: particularly the inset). This is the first time that component A
382: appears triple rather than double in a radio image. While the
383: easternmost source is clearly A1, the relation between the two western
384: sources and A2 is {\it a priori} quite unclear. Since they are
385: identified here for the first time, we have very little information on
386: the spectral energy distributions of the two western sources. The only
387: statement that can safely be made on their spectral properties is that
388: they are both continuum (rather than spectral) features because they
389: are detected with identical fluxes at both observed frequencies
390: (22.4851 and 22.4351 GHz --see Sect.\ 2). It is also noteworthy that
391: the total flux of component A in our new 1.3 cm image (5.2 $\pm$ 0.3
392: mJy) is comparable to the 1.3 cm flux of component A in all previous
393: observations (see Fig.\ 7, left panel in Chandler et al.\ 2005). Thus,
394: while the {\em morphology} of A2 has undergone significant changes,
395: its {\em flux} has not. This suggests, in particular, that the current
396: overall spectral index of component A is similar to that found in
397: previous observations. In the centimeter regime, that spectral index
398: is about 0.53 (see Fig.\ 2, left panel in Chandler et al.\ 2005)
399: suggesting that free-free is the dominant emission process.
400:
401: To examine further the relation between the three sources associated
402: with component A in our new 1.3 cm image and the usual sub-components
403: A1 and A2, it is useful to superimpose our new 1.3 cm map onto one of
404: the previous high resolution radio images of component A. For this
405: purpose, we will use the smoothed version of the A-array 0.7 cm image
406: (Sect.\ 3), because it has been obtained relatively shortly before the
407: 1.3 cm data presented here, and because it provides a good compromise
408: between angular resolution and sensitivity. Since the absolute positions
409: of the various components of IRAS~16293--2422 change with time
410: (Loinard 2002, Chandler et al. 2005), one has to properly register the
411: images before the superposition. The absolute proper motion of A2 is
412: known to be very similar to that of component B (Chandler et al.\ 2005).
413: From the four recent observations considered in this paper, the proper
414: motion of component B can be calculated to be
415:
416: \[ \mu_\alpha \cos \delta = -7.6 \pm 1.2 \mbox{~mas yr$^{-1}$~~~~;~~~~~} \mu_\delta = -25.8 \pm 1.1 \mbox{~mas yr$^{-1}$} \]
417:
418: \noindent The latter value is somewhat larger than those quoted by
419: Loinard et al.\ (2002) and Chandler et al.\ (2005). The reasons for
420: this discrepancy are not entirely clear, but may be related to the
421: fact that the older data used by Loinard et al.\ (2002) and Chandler
422: et al.\ (2005) are of lower quality than the more recent observations
423: considered here. In any case, we will use the values quoted above,
424: since they are more appropriate for the comparisons that will
425: be made in this paper between recent observations. Given that the 0.7
426: cm image was obtained 2.64 years before the present 1.3 cm data, it was
427: shifted by $-$20.1 mas and $-$68.0 mas in right ascension and declination,
428: respectively, before the superposition. The result of the comparison
429: (Fig.\ 3) confirms that the easternmost 1.3 cm component is nearly exactly
430: coincident with the 0.7 cm source A1, and can indeed be identified with
431: it. Neither of the two western 1.3 cm sources, however, appears to coincide
432: with source A2. Instead, they are located roughly symmetrically on each
433: side of A2, and we shall refer to them as A2$\alpha$ and A2$\beta$ in the
434: rest of the paper (Fig.\ 2). The line joining A2$\alpha$, A2, and A2$\beta$
435: (thick dotted line on Fig.\ 3) is at a position angle of about
436: 62$^\circ$, remarkably similar to the direction of the large-scale
437: flow known to be driven by A2. This strongly suggests that A2$\alpha$
438: and A2$\beta$ are the result of a recent bipolar ejection by the
439: protostar within A2, the position of which is likely near the cross in
440: Fig.\ 3. While bipolar ejections are not uncommon in young stellar
441: objects (e.g. Mart\'{\i} et al.\ 1995, Curiel et al.\ 2006), they
442: rarely lead to such dramatic morphological changes. In particular, in
443: our case, it seems that very little emission is left at the very
444: position of the protostar, the emission concentrating instead in the
445: two bipolar lobes. This usually does not happen in bipolar ejections
446: by protostars. Also, while bipolar ejections have been observed from
447: relatively massive young stellar system, this is --to our knowledge--
448: the first time that it is seen in such a low-mass source. It is
449: interesting to note that a situation almost exactly opposite of that
450: reported here was recently found in a radio source in Orion. Source
451: ``n'', which had been bipolar in all previous VLA observations, was
452: indeed found to be single in the 2007 VLA image to be published by L.\
453: G\'omez et al.\ (in prep).
454:
455: A2$\alpha$ and A2$\beta$ were clearly not present in previous radio
456: images so if they are indeed ejecta, they must be very recent
457: ones. Water masers associated with component A are found to expand at
458: a velocity projected on the plane of the sky of 40--65 km $^{-1}$
459: (Wootten et al.\ 1999). If this value can be taken as representative
460: of the projected velocity of the jet powered by A2, then it would have
461: taken just about a year for A2$\alpha$ and A2$\beta$ to reach the
462: present position. This is consistent with their absence from any
463: previous observation. Note, also, that if A2$\alpha$ and A2$\beta$ are
464: ejecta moving at about 50 km s $^{-1}$ away from the protostar
465: associated with A2, then their proper motions should be easily
466: detectable. An observation of IRAS~16293--2422 in the upcoming A
467: configuration of the VLA (around 2007.8), should show A2$\alpha$
468: and A2$\beta$ displaced from their current position by about
469: \msec{0}{1}.
470:
471: \subsection{Relative motion between A1 and A2}
472:
473: The relative motion between A1 and A2 in the last 15 years or so has
474: been investigated in detail by Loinard (2002) and Chandler et al.\
475: (2005). As mentioned earlier, these studies have led to the conclusion
476: that the separation between A1 and A2 has remained constant at about
477: \msec{0}{34}, whereas their relative position angle has changed by
478: about 40$^\circ$ between the late 1980s and 2003. The new 0.7 cm
479: observation reported here can be used to further monitor the evolution
480: of the relative position of A1 and A2. The value of the position angle
481: found in that observation (82 $\pm$ 3$^\circ$ --Fig.\ 4) is similar to
482: those reported for the 2003 observations by Chandler et al.\
483: (2005). Obtaining information from the new 1.3 cm observation is
484: somewhat more difficult because of the change of structure undergone
485: by component A2. In Fig.\ 4, we report the separation and position
486: angle between A1 and both A2$\alpha$ and A2$\beta$. As discussed
487: earlier, however, the protostar in A2 is unlikely to be coincident
488: with either of these two sources. Instead, it must be near the cross
489: in Fig.\ 3, a location roughly equivalent to the mean of the positions
490: of A2$\alpha$ and A2$\beta$. The red circles in Fig.\ 4 indicate the
491: separation and position angle between A1 and the cross in Fig.\
492: 3. Taken together, these new observations suggest that the separation
493: between A1 and the protostar associated with A2 has remained constant
494: around \msec{0}{34} in the last few years, whereas their relative
495: position angle is between 80 and 90$^\circ$. Unfortunately, this does
496: not shed much new light on the relative motion between A1 and A2. In
497: particular it is insufficient to discriminate between the orbital path
498: scenario (where one would expect the position angle to keep increasing
499: linearly) and the precessing/wobbling scheme (where the change of
500: position angle should decelerate and eventually reverse its course).
501: More observations in the coming decade will be needed to settle this
502: issue.
503:
504: \subsection{0.7 cm emission from component Ab?}
505:
506: We mentioned in Sect.\ 1 that the submillimeter source Ab, located
507: \msec{0}{6} to the north-east of Component A has, to date, never been
508: detected at other wavelengths. To further examine this issue, we
509: searched both of our new images for a counterpart of Aa. While there
510: is clearly no detection in the present 1.3 cm data, a possible source
511: is detected at the 4$\sigma$ level in our new B-configuration 0.7 cm
512: dataset (Fig.\ 5). The best 2-dimensional Gaussian fit to this
513: structure yields a peak flux density of about 0.76 mJy beam$^{-1}$, and
514: an integrated flux of 1.8 $\pm$ 0.6 mJy. Note that the noise level in
515: the 0.7 cm A array observations published by Rodr\'{\i}guez et al.\
516: (2005) was 0.1 mJy beam$^{-1}$, so this source should have been detected
517: in that dataset also if it had been present. The fact that it was not,
518: implies either that the present marginal detection is not real, or that
519: the 0.7 cm emission associated with Ab is extended, and was more heavily
520: filtered out in the A array observations published by Rodr\'{\i}guez et
521: al.\ (2005) than in the present B array data. Deep, C-configuration VLA
522: observations at 0.7 cm ought to settle this issue. If the marginal
523: detection reported here is real, then the spectral index of the
524: emission associated with component Ab must be smaller than about 3
525: independently of the extent of the emission. Thus, component Ab would
526: be a relatively extended structure containing somewhat processed
527: dust. Combined with the lack of centimeter emission tracing winds,
528: this would favor a scenario where Ab is a starless condensation rather
529: than a protostar.
530:
531: %% Add 30% or not? Add 20 or 30% quadratically??? Talk to Claire
532:
533: \subsection{Component B: disk and jet combined}
534:
535: Component B is clearly resolved in our new 1.3 cm observations (Fig.\
536: 6.c), and its morphology is very similar to that in the A-configuration
537: 0.7 cm image smoothed to a similar resolution (Fig.\ 6.a). It is slightly
538: more compact, however, at 1.3 cm (\msec{0}{19} $\times$ \msec{0}{15}) than
539: at 0.7 cm (\msec{0}{21} $\times$ \msec{0}{20}). This is consistent
540: with the trend noticed by Chandler et al.\ (2005) between archival 1.3
541: cm and 3.6 cm data --in the 3.6 cm image shown in Fig.\ 1, the source
542: is found to have a deconvolved size of \msec{0}{16} $\times$
543: \msec{0}{10}-- and suggests that the (linear) angular size of the
544: emission increases roughly as $\nu^{+0.3}$. This increase of the angular
545: size with frequency is opposite to what is expected from optically thick
546: free-free emission. Combined with the measured spectral index in the
547: centimeter regime of about 2 (Chandler et al.\ 2005), this strongly
548: suggests that the dominant radiation mechanism is optically thick thermal
549: dust emission as proposed before (Rodr\'{\i}guez et al.\ 2005, Chandler
550: et al.\ 2005). Clearly, to make the emission optically thick, the
551: circumstellar disk associated with component B must be quite dense and
552: massive (Chandler et al.\ 2005). Since the emission is well-resolved in
553: our data, we can investigate if and how the spectral index depends on
554: radius. We find that the spectral index in the very inner region (at R $<$
555: \msec{0}{1}) is 1.4 $\pm$ 0.2, whereas it is 2 $\pm$ 0.2 in the region
556: \msec{0}{1} $<$ R $<$ \msec{0}{2} (Fig.\ 6.d). The spectral index outside
557: of R = \msec{0}{2} becomes difficult to calculate reliably because the
558: emission fades quickly, but it is clearly larger than 2. The lower value
559: of the spectral index near the center of component B suggests that the
560: optically thick thermal dust emission in these regions coexists with a
561: lower spectral index component. The most likely candidate for this
562: additional component would be free-free radiation from the base of an
563: ionized wind. Such a wind would indeed be expected to exist in an
564: actively-accreting protostar like that believed to reside at the center
565: of component B.
566:
567: It is interesting in this context, to compare the structure of the
568: inner and outer regions of component B. The image most suitable for
569: that purpose is the full-resolution 0.7 cm image obtained in the A
570: configuration of the VLA (Fig.\ 6.b), because it provides the
571: best compromise between sensitivity and angular resolution. It is
572: noteworthy in that image that the outer isophotes are elongated
573: roughly in the north-south direction (at a position angle marginally
574: positive), whereas the inner isophotes are elongated roughly in the
575: east-west direction (at P.A.\ $\sim$ 110$^\circ$) very similar to that
576: of one of the large-scale outflows known to originate from within
577: IRAS~16293--2422. This would be consistent with the idea that
578: component B is at the origin of that flow; the radio source associated
579: with it being the superposition of a compact thermal jet at P.A.\
580: $\sim$ 110$^\circ$, and a somewhat extended disk roughly perpendicular
581: to the jet. It is puzzling in that scenario, however, that no high-velocity
582: emission is detected towards component B in any molecular transition
583: (Chandler et al.\ 2005). If a thermal jet is associated with the
584: protostar near the center of component B, one would indeed expect to
585: find entrained high velocity gas associated with it. Future,
586: high-resolution, millimeter and submillimeter spectral observations
587: might help resolve this puzzle.
588:
589:
590: \section{Conclusions and perspectives}
591:
592: The new observations presented in this paper provide a number of
593: interesting clues on the nature and properties of the various radio
594: sources in IRAS~16293--2422.
595:
596: \begin{enumerate}
597:
598: \item Component A2 is definitely confirmed to be at the origin of the
599: compact bipolar flow at P.A.\ $\sim$ 60$^\circ$. Indeed, in the new
600: 1.3 cm image, A2 has undergone a dramatic morphological change, being
601: now composed of two emission peaks distributed roughly symmetrically
602: around A2 at a position angle of almost exactly 60$^\circ$. We argue
603: that this change reflects a recent bipolar ejection, an interpretation
604: that will be confirmed if forthcoming observations reveal the
605: appropriate proper motions.
606:
607: \item Unfortunately, the new 0.7 and 1.3 cm data do not allow us to
608: characterize much better the relative motion between A1 and A2. It
609: remains unclear, therefore, whether the changes in relative position
610: observed in the last two decades reflect an orbital motion, or the
611: precession or wobbling of a jet. It will be important to settle that
612: issue to decide where the third star in the system is, and if that
613: star might be at the origin of the third outflow structure known to
614: exist in IRAS~16293--2422. Further observations in the next ten years
615: or so ought to provide a definite answer.
616:
617: \item A marginal detection of 0.7 cm emission near the position of the
618: submillimeter source Ab is reported. It would be very important to
619: further characterize the very nature of this source, and its relation
620: to the other members of the system. If the present marginal detection
621: is real, it should be easily confirmed by future deep 0.7 cm
622: observations in the C configuration of the VLA.
623:
624: \item The radio emission from component B at the northwest of the
625: system is confirmed to be dominated by optically thick thermal dust
626: emission, suggesting that the radio source in that component traces an
627: optically thick accretion disk. At the very center of that structure,
628: however, the lower observed value of the spectral index suggests the
629: existence of a modest contribution from a thermal jet. The
630: orientation of the inner 0.7 cm isophotes further suggests that this
631: free-free component traces the base of the compact bipolar flow at
632: P.A.\ $\sim$ 110$^\circ$. This association would appear reasonable
633: since the flow at P.A.\ $\sim$ 110$^\circ$ has a dynamical age of only
634: a few thousand years (Mizuno et al.\ 1990), and component B is known
635: to harbor organic and deuterated molecular species expected to be
636: present only in the youngest protostellar sources (Cazaux et al.\
637: 2003, Kuan et al.\ 2004, Bottinelli et al.\ 2004, Chandler et al.\
638: 2005).
639:
640: \end{enumerate}
641:
642:
643: \acknowledgements
644: L.L., L.F.R., and P.A.\ acknowledge the financial support of DGAPA,
645: UNAM and CONACyT, M\'exico. D.J.W. acknowledges partial support from
646: NASA Origins of Solar Systems Program Grant NAG5-11777.
647:
648: \begin{thebibliography}{}
649:
650: \bibitem[Bottinelli et al.(2004)]{2004ApJ...617L..69B} Bottinelli, S., et
651: al.\ 2004, \apjl, 617, L69
652:
653: \bibitem[]{749}
654: Castets, A., Ceccarelli, C., Loinard, L., Caux, E., \& Lefloch, B.\
655: 2001, A\&A, 375, 40
656:
657: \bibitem[Cazaux et al.(2003)]{2003ApJ...593L..51C} Cazaux, S., Tielens,
658: A.~G.~G.~M., Ceccarelli, C., Castets, A., Wakelam, V., Caux, E., Parise,
659: B., \& Teyssier, D.\ 2003, \apjl, 593, L51
660:
661: \bibitem[Ceccarelli et al.(2007)]{2007prpl.conf...47C} Ceccarelli, C.,
662: Caselli, P., Herbst, E., Tielens, A.~G.~G.~M., \& Caux, E.\ 2007,
663: Protostars and Planets V, 47
664:
665: \bibitem[]{761}
666: Ceccarelli, C., Castets, A., Caux, E., Hollenbach, D., Loinard, L.,
667: Molinari, S., \& Tielens, A.G.G.M.\ 2000, A\&A, 355, 1129
668:
669: \bibitem[Ceccarelli et al.(2000)]{2000A&A...357L...9C} Ceccarelli, C.,
670: Loinard, L., Castets, A., Tielens, A.~G.~G.~M., \& Caux, E.\ 2000, \aap,
671: 357, L9
672:
673: \bibitem[]{769}
674: Chandler, C.J., Brogan, C.L., Shirley, Y.L., \& Loinard, L.\ 2005, ApJ, 532,
675: 371
676:
677: \bibitem[Curiel et al.(2006)]{2006ApJ...638..878C} Curiel, S., et al.\
678: 2006, \apj, 638, 878
679:
680: \bibitem[]{776}
681: Duch\^ene, G., Delgado-Donate, E., Haisch, K.E., Jr., Loinard, L.;
682: Rodr\'{\i}guez, L. F.\ 2007, Protostars \& Planets V, B. Reipurth, D.
683: Jewitt, and K. Keil (eds.), University of Arizona Press, Tucson, 279
684:
685: \bibitem[]{781}
686: Duquennoy, A., \& Mayor, M.\ 1991, A\&A, 248, 485
687:
688: \bibitem[]{784}
689: Hirano, N., Mikami, H., Umemoto, T., Yamamoto, S., \& Taniguchi, T. 2001, ApJ, 547, 899
690:
691: \bibitem[Kuan et al.(2004)]{2004ApJ...616L..27K} Kuan, Y.-J., et al.\ 2004,
692: \apjl, 616, L27
693:
694: \bibitem[]{790}
695: Lada, C.J.\ 2006, ApJ, 640, L63
696:
697: \bibitem[]{793}
698: Loinard, L.\ 2002, RMAA, 38, 61
699:
700: \bibitem[Marti et al.(1995)]{1995ApJ...449..184M} Marti, J., Rodriguez,
701: L.~F., \& Reipurth, B.\ 1995, \apj, 449, 184
702:
703: \bibitem[]{799}
704: Mizuno, A., Fukui, Y., Iwata, T., Nozawa, S., \& Takano, T. 1990, ApJ, 356, 184
705:
706: \bibitem[]{802}
707: Mundy, L. G., Wilking, B. A., \& Myers, S. T. 1986, ApJ, 311, L75
708:
709: \bibitem[]{805}
710: Mundy, L. G., Wootten, H. A., Wilking, B. A., Blake, G. A., \& Sargent, A. I. 1992, ApJ, 385, 306
711:
712: \bibitem[]{808}
713: Rodr\'{\i}guez, L.F., Loinard, L., D'Alessio, P., Wilner, D.J., \& Ho, P.T.P.\
714: 2005, ApJ, 621, L133
715:
716: \bibitem[]{812}
717: Shu, F.H., Adams, F.C., \& Lizano, S.\ 1987, ARAA, 25, 23
718:
719: \bibitem[]{815}
720: Stark, R., et al. 2004, ApJ, 608, 341
721:
722: \bibitem[van Dishoeck et al.(1995)]{1995ApJ...447..760V} van Dishoeck,
723: E.~F., Blake, G.~A., Jansen, D.~J., \& Groesbeck, T.~D.\ 1995, \apj, 447,
724: 760
725:
726: \bibitem[]{822}
727: Wootten, A. 1989, ApJ, 337, 858
728:
729: \end{thebibliography}
730:
731: \clearpage
732:
733: \begin{table}[!t]
734: \caption{Observing logs}
735: \vspace{0.4cm}
736: \centering
737: \begin{tabular}{lcrclcll}
738: \hline \hline
739: Date & Project & \multicolumn{1}{c}{$\nu$} & Conf. & Weighting/ &
740: Synthesized beam & Figs & Refs\tablenotemark{a}\\%
741: & & (GHz) & & Smoothing\\%
742: \hline
743: 2003.47 & AL592 & 43.2299 & A & Nat. & \sbeamp{0}{09}{0}{05}{1.8} & 6b & 1,2 \\%
744: 2003.47 & AL592 & 43.2299 & A & Nat./Smo. & \sbeamm{0}{13}{0}{06}{0.5} & 3, 6a & 3 \\%
745: 2003.65 & AL589 & 8.4601 & A & Rob. & \sbeamp{0}{39}{0}{19}{6.8} & 1 & 2 \\%
746: 2005.20 & AC778 & 43.2299 & B & Nat. & \sbeamm{0}{30}{0}{17}{1.9} & 2b, 5 & 3\\%
747: 2006.11 & AL672 & 22.4601 & A & Rob. & \sbeamm{0}{13}{0}{06}{0.5} & 2a, 3, 6c & 3\\%
748: \hline \hline \\
749: \end{tabular}
750: \tablenotetext{a}{1=Rodr\'{\i}guez et al.\ (2005); 2=Chandler et al.\ (2005); 3=This work}
751: \end{table}
752:
753: \clearpage
754: \begin{figure}[!t]
755: \centerline{\includegraphics[height=0.6\textwidth,angle=-90]{f1.eps}}
756: \caption{3.6 cm VLA A-array image of IRAS~16293--2422 obtained in
757: 2003.65 (Chandler et al.\ 2005) and restored with a weighting scheme
758: intermediate between natural and uniform. The first contour and the
759: contour interval are 0.1 mJy beam$^{-1}$, while the noise level is
760: 0.02 mJy beam$^{-1}$. The synthesized beam, shown at the bottom-right
761: corner, is \sbeamp{0}{39}{0}{19}{6.8}. The radio components A1, A2,
762: and B are labeled, and the position of the submillimeter sources Aa,
763: Ab, and B are indicated. The errors on the submillimeter source positions
764: are significantly smaller than the crosses indicating their locations.}
765: \end{figure}
766:
767: \clearpage
768: \begin{figure*}[!t]
769: \centerline{\includegraphics[width=0.85\textwidth,angle=-90]{f2.eps}}
770: \caption{(a) 1.3 cm continuum observation of IRAS~16293--2422 obtained
771: in 2006.11 restored with a weighting scheme of the visibilities
772: intermediate between natural and uniform. The first contour is at 0.2
773: mJy beam$^{-1}$ and the contour step is 0.3 mJy beam$^{-1}$, while the
774: noise is 0.04 mJy beam$^{-1}$. The synthesized beam, shown near the
775: top-left corner, is \sbeamm{0}{13}{0}{06}{0.5}. (b) 0.7 cm image of
776: IRAS~16293--2422 obtained in 2005.20 in the B configuration of the
777: VLA when restored with natural weighting. The first contour is at 1.5
778: mJy beam$^{-1}$ and the contour step is 0.3 mJy beam$^{-1}$, while the
779: noise is 0.18 mJy beam$^{-1}$. The synthesized beam, shown near the
780: top-left corner, is \sbeamm{0}{30}{0}{17}{1.9}. In both panels, the
781: positions of the submillimeter sources Aa and Ab are indicated,
782: and insets provide zooms on component A. Note that component A is
783: clearly triple in the 1.3 cm image, while it remained double in the
784: 0.7 cm map.} \end{figure*}
785:
786: \clearpage
787: \begin{figure}[!t]
788: \centerline{\includegraphics[width=0.35\textwidth,angle=270]{f3.eps}}
789: \caption{Comparison between the smoothed 0.7 cm image of component A
790: obtained in 2003 (black contours and grey scale) and our new 1.3 cm image
791: (red contours). The angular resolution of the two images is identical
792: (\sbeamm{0}{13}{0}{06}{0.5}) and is shown at the bottom-left. Translational
793: shifts have been applied to account for the overall proper motion of the
794: region. The first contour and the contour interval for the 1.3 cm image
795: are the same as in Fig.\ 2; they are at 0.6 and 0.15 mJy beam$^{-1}$,
796: respectively, for the 0.7 cm image. The dotted line is at a position
797: angle of 62$^\circ$}
798: \end{figure}
799:
800: \clearpage
801: \begin{figure*}[!t]
802: \centerline{\includegraphics[width=0.4\textwidth,angle=270]{f4.eps}}
803: \caption{Evolution of the separation (left) and relative position
804: angle (right) of the A1/A2 pair. For the 1.3 cm data obtained in
805: 2006.11, three points are shown: in black the parameters for
806: A2$\alpha$ and A2$\beta$, and in red those corresponding to the
807: position were the protostar is inferred to be. The dotted line in the
808: left panel shows a constant separation of \msec{0}{34}, whereas the
809: dotted line in the right panel shows the best linear fit to all the
810: data points but those corresponding to the 2006.11 image.}
811: \end{figure*}
812:
813: \clearpage
814: \begin{figure}[!t]
815: \centerline{\includegraphics[width=0.35\textwidth,angle=270]{f5.eps}}
816: \caption{0.7 cm image of component A. The first contour is at 0.7 mJy
817: beam$^{-1}$, and the contour interval is 0.2 mJy beam$^{-1}$, while
818: the noise is 0.18 mJy beam$^{-1}$. The synthesized beam, shown at the
819: bottom-left corner is \sbeamm{0}{30}{0}{17}{1.9}. Note the existence
820: of a positive signal near the position of Ab.}
821: \end{figure}
822:
823: \clearpage
824: \begin{figure*}[!t]
825: \centerline{\includegraphics[width=0.8\textwidth,angle=0]{f6.eps}}
826: \caption{Comparison between high angular resolution images of
827: component B in IRAS~16293--2422 obtained at 0.7 cm (panels a--b) and
828: 1.3 cm (panel c). In panel (b), we show the full-resolution 0.7 cm
829: combined A+B configuration image. The first contour is at 0.6 mJy
830: beam$^{-1}$ and the contour interval is 0.2 mJy beam$^{-1}$, while the
831: noise level is 0.12 mJy beam$^{-1}$. The synthesized beam, shown at
832: the bottom-right corner is \sbeamp{0}{09}{0}{05}{1.8}. In panel
833: (c) we show the full resolution 1.3 cm A configuration image. The
834: first contour is at 0.2 mJy beam$^{-1}$ and the contour interval is
835: 0.15 mJy beam$^{-1}$. In panel (a) we show the 0.7 cm image smoothed
836: to the resolution of the 1.3 cm image. The first contour is at 0.8 mJy
837: beam$^{-1}$ and the contour interval is 0.4 mJy beam$^{-1}$. Finally,
838: in panel (d), we show the spectral index between 0.7 and 1.3 cm. The
839: grey scale goes from 1 (white) to 3 (black). The coutours are at 1.4,
840: 1.6, 1.8, 2, 2.2, 2.4, and 2.6. The resolution in panels (a), (c) and
841: (d), shown at the bottom-left of each panel is \sbeamm{0}{13}{0}{06}{0.5}}.
842: \end{figure*}
843: \clearpage
844: \end{document}
845:
846: