1: \documentclass[12pt,preprint]{aastex}
2:
3: \shorttitle{Radio Cores of Bipolar PNs}
4: \shortauthors{Lee, Lim \& Kwok}
5:
6: \begin{document}
7:
8: \title{Optically Thick Radio Cores of Narrow-Waist Bipolar Nebulae}
9:
10: \author{T.-H. Lee}
11: \affil{National Optical Astronomy Observatory, Tucson, AZ 85719\\ and
12: \\ Departemnt of Physics and Astronomy, University of Calgary,
13: Calgary, AB T2N 1N4, Canada} \email{thlee@noao.edu}
14:
15: \author{J. Lim}
16: \affil{Institute of Astronomy \& Astrophysics, Academia Sinica, Taipei
17: 106, Taiwan}
18: \email{jlim@asiaa.sinica.edu.tw}
19:
20: \and
21:
22: \author{S. Kwok}
23:
24: \affil{Department of Physics, Faculty of Science, University of Hong
25: Kong, Hong Kong \\ and \\ Departemnt of Physics and
26: Astronomy, University of Calgary, Calgary, AB T2N 1N4, Canada}
27: \email{sunkwok@hku.hk}
28:
29: \begin{abstract}
30:
31: We report our search for optically thick radio cores in sixteen
32: narrow-waist bipolar nebulae. Optically thick cores are a
33: characteristic signature of collimated ionized winds. Eleven
34: northern nebulae were observed with the Very Large Array (VLA) at
35: 1.3~cm and 0.7~cm, and five southern nebulae were observed with the
36: Australia Telescope Compact Array (ATCA) at 6~cm and 3.6~cm. Two
37: northern objects, 19W32 and M~1-91, and three southern objects,
38: He~2-25, He~2-84 and Mz~3, were found to exhibit a compact radio
39: core with a rising spectrum consistent with an ionized jet. Such
40: jets have been seen in M~2-9 and may be responsible for shaping
41: bipolar structure in planetary nebulae.
42:
43: \end{abstract}
44:
45: \keywords{planetary nebulae: general --- stars: AGB and post-AGB ---
46: stars: winds, outflows --- stars: mass loss --- stars: evolution ---
47: radio continuum: stars}
48:
49: \section{Introduction}
50: \label{sec:intro-jet}
51:
52: Planetary nebulae (PNs) show a remarkably broad range of often
53: intricate morphologies. Butterfly nebulae, classified as bipolar
54: nebulae with a narrow pinched waist at the center \citep{balick02},
55: are probably the most spectacular, and yet at the same time most
56: puzzling. Under the classical interacting stellar winds (ISW) model
57: \citep{kwok78}, a PN is formed when the fast ionized wind from its
58: central star overtakes the material ejected earlier by the slow
59: neutral wind and compresses the material into a dense shell. This
60: produces a round PN if both winds are isotropic. A dense equatorial
61: torus, which blocks the fast wind in that direction, is added to the
62: interacting winds system to explain how asymmetric structure in PNs
63: (e.g., bipolar) can be developed \citep{kah85, bal87}. This,
64: however, can only produce bipolar nebulae with wide waists, not
65: those with narrow ``pinched'' waists. It has been suggested that
66: the creation of the latter nebulae requires a collimated fast wind
67: \citep[e.g.,][]{soker00}. \citeauthor{soker00}'s model invokes a
68: binary star system to produce a collimated fast wind. The system
69: consists of a mass-losing AGB star and a main-sequence or white
70: dwarf companion. The AGB wind captured by the companion forms an
71: accretion disk around the companion that loads and drives a
72: collimated fast ionized wind.
73:
74: The existence of collimated outflows (COFs) has been suggested by
75: numerous observations, usually from the ionized gas images. For
76: example, the highly collimated lobes of some PNs and PPNs
77: (proto-planetary nebulae) imply that their morphology is shaped by a
78: collimated outflow \citep[see review by][and references
79: therein]{kwok03a}. The first discovery of strongly collimated
80: high-velocity bipolar outflows in a PN was by \cite{gieseking85}, in
81: NGC~2392, the Eskimo nebula. They analyzed from spectroscopic
82: observations of this nebula, particularly in the [N~{\footnotesize
83: II}] $\lambda$6583 emission line, they found a set of bipolar jets
84: with a expansion velocity of $\sim$ 200~km s$^{-1}$ and a small
85: angle of collimation less than 10$^{\circ}$. Although unexpected at
86: first, the pervasiveness of COFs in PNs is now widely recognized.
87: COFs have diverse characteristics. They are prominent in emission
88: lines such as [N~{\footnotesize II}] and [S~{\footnotesize II}] and
89: are observed as bipolar, multipolar, point-symmetric and jet-like
90: outflows. The expansion velocities of these COFs range from several
91: tens to a thousand km~s$^{-1}$ \citep[see reviews of][]{lopez02,
92: lopez03}.
93:
94: In addition to ionized gas, COFs have also been seen in neutral gas.
95: Several authors have made detailed studies on CO emission lines of
96: individual PNs and PPNs \citep[e.g.,][]{forveille98, huggins00}.
97: The velocities for these CO outflows are found to be much smaller
98: than those of the ionized COFs. \cite{bujarrabal01} studied the CO
99: emission from a large sample of PPNs as well as two yellow
100: hypergiants and one young PN. They reported collimated fast
101: outflows in 28 of 32 objects, indicating that such outflows are a
102: common phenomenon in the early PN stage.
103:
104: In the model proposed by \cite{soker00}, the collimated winds
105: responsible for shaping butterfly nebulae are formed and collimated
106: close to the central star system. They are difficult to observe in
107: optical wavelengths directly, because dust around the equator limits
108: the depth into the center that optical observations can probe. The
109: dust is optically thin to radio emission, which can therefore probe
110: deep into the center to look for central jets, if the jets are at
111: least partially ionized. Using the VLA, \cite{lim00,lim03} have
112: indeed found a collimated wind in the form of a pair of ionized jet
113: in the core region of the narrow-waist bipolar nebula M~2-9.
114:
115: M~2-9 has a bright central core and two cigar-shaped lobes extending
116: from opposite sides of the core $\sim 20''$ to the north and south.
117: The lobes of the nebula contain three pairs of symmetric knots or
118: bright features: two of them are with mirror symmetry and one pair
119: with point/mirror symmetry. Because of its spectacular morphology,
120: M~2-9 is one of the most well-studied planetary nebulae. Since its
121: discovery in 1947, the core and the outline of the lobes have
122: remained stationary, but the knots and bright features have moved
123: laterally from the west to the east \citep[e.g.,][]{allen72,
124: kohoutek80}. Various models have been proposed to explain the
125: lateral motions of bright features of M~2-9. For example,
126: \cite{allen72} proposed that the changes were the result of a
127: rotating beam of ionizing radiation from the star shining through a
128: matched pair of holes in orbiting clouds. \cite{kohoutek80} noted
129: that it is unlikely such model can result in positional changes only
130: in the east-west direction, and instead suggested that the nebula
131: was rotating as a whole. However, since \cite{goodrich91} concluded
132: that the distance of M~2-9 cannot be only 50 pc as proposed by
133: \cite{kohoutek80}, the unreasonable large velocities of the knots
134: have made their model problematic.
135:
136: Recently in explaining the structural changes of M~2-9,
137: \cite{doyle00} and \cite{livio01} both invoke a symbiotic star
138: system to produce a mirror-symmetric jet. Direct evidence of such a
139: jet is provided by observations of M~2-9 at 6~cm, 3.6~cm, 2~cm, and
140: 1.3~cm by \citep{lim03}. Their observations were largely
141: insensitive to the large-scale free-free emission from the nebula
142: and focused instead on the compact radio core at the center found by
143: \cite{kwok85}. Their results indicate that this optically thick
144: core has a spectral index of $0.67 \pm 0.01$, which is close to the
145: spectral index of 0.6 expected from an isothermal outflow expanding
146: at a constant velocity and opening angle \citep{reynolds86}. More
147: importantly, at the shortest wavelength of 1.3~cm where the jet is
148: best resolved, its bipolar lobes show a mirror-symmetric structure
149: \citep[Figure 1 of][]{lim03}. Both lobes point in the directions of
150: mirror-symmetric brightenings in the nebular walls seen in the {\it
151: HST} image observed just 2 years earlier \citep[N1 and S1 in Figure
152: 2 of][]{doyle00}. The qualitative properties of this jet are in
153: good agreement with those predicted by \cite{livio01}.
154:
155: Motivated by the discovery of the ionized jets in M~2-9, we launched
156: a search for collimated ionized winds in butterfly nebulae using the
157: VLA and the ATCA. We selected a total of 16 objects from Table 1 of
158: \citet{soker00} to search for optically thick radio cores -- a
159: characteristic signature of ionized jets. The selected objects were
160: classified by \cite{soker00} to have either narrow waists (10
161: objects) or possibly narrow waists (6 objects), whose morphology
162: cannot be determined with absolute certainty based on the optical
163: images obtained by \cite{corradi95}. We are interested in finding
164: any central source that is optically thick at short wavelengths
165: where the nebular material is already optically thin. Our goal is
166: to identify suitable candidates for follow-up observations with
167: higher sensitivity and resolution to reveal the structure of the
168: source. We observe each nebula with two wavelengths to calculate
169: its spectral index, which tells us if the emission is optically
170: thick. In this paper, we report the results of our search for the
171: radio cores. The observations and data reduction are described in
172: Section \ref{sec:data-jet}. In Section \ref{sec:results-jet}, we
173: report that radio emission were detected in 9 sources, and that 5 of
174: those detected have optically thick radio cores. Section
175: \ref{sec:discussion} discusses the implication of the high mass loss
176: rates derived from the flux densities of the radio cores. Lastly, a
177: summary is given in Section \ref{sec:sum-jet}.
178:
179: \section{Observations and Data Reduction}
180: \label{sec:data-jet}
181:
182: The survey sample comprises 11 northern objects (Declination $\geq
183: -40^\circ$ for the VLA observation limits), IRAS~07131-0147, M~1-16,
184: NGC~2818, NGC~6302, 19W32, HB~5, NGC~6537, M~3-28, M~1-91, M~2-48,
185: and NGC~7026, observed with the VLA at 1.3~cm and 0.7~cm, and 5
186: southern objects, He~2-25, He~2-36, He~2-84, Th~2-B, and Mz~3,
187: observed with the ATCA at 6~cm and 3.6~cm. The observing
188: wavelengths are the shortest wavelengths feasible at the time of
189: observations, which give the highest angular resolutions possible.
190: The optical properties of the objects are listed in Table
191: \ref{tb:objects-jet}, with columns 1 and 2 giving the common name
192: and the PN G name based on galactic coordinates, columns 3 and 4
193: giving Right Ascension and Declination, column 5 indicating if they
194: appear to have a very narrow waist, column 6 stating if the PN has a
195: compact central source visible in optical images, and columns 7
196: indicating if the nebula shows point or mirror symmetric structure.
197: These properties are mainly from \cite{soker00}, with some update
198: obtained from recent optical images. A question mark means the
199: attribute cannot be determined from available images.
200:
201: \subsection{VLA Observations}
202:
203: The 11 northern objects were observed at 1.3~cm and 0.7~cm with the
204: A-configuration in 2002 March 10. The A-configuration is the most
205: extended configuration, hence it provide the highest angular
206: resolution achieved with the VLA. It allows us to not just look for
207: but resolve if possible a central compact and optically think
208: source, as well as to better discriminate against if not resolve out
209: extended nebular emission. The observations were conducted in
210: fast-switching mode with a calibration cycle time of just 2-3
211: minutes in order to correct for rapid phase variation at such short
212: wavelengths. The integration time for the individual sources is
213: about 20 minutes each at 1.3~cm and 0.7~cm. 3C286 was observed for
214: about 5 minutes in order to calibrate the flux scales for individual
215: sources. The data were reduced and analyzed using the package AIPS
216: following the standard procedures for reducing high-frequency data
217: outlined in Appendix D of the AIPS Cookbook \citep{nrao03}. The
218: calibrated data were then Fourier transformed, weighted and CLEANed
219: to generate the final CLEAN images. Natural weighting was applied
220: in order to obtain images with minimum-noise levels. We achieved an
221: angular resolution of $\sim 0.08''$ at 1.3~cm and $\sim 0.04''$ at
222: 0.7~cm for the final CLEAN maps.
223:
224: \subsection{ATCA Observations}
225:
226: The 5 southern objects were observed at 6~cm and 3.6~cm with the
227: 6-km ``6C'' configuration in 2003 March 15 and 16. These two
228: wavelengths were the shortest wavelengths feasible when the
229: observations took place. With 6-km being the most extended
230: configuration, they provide the highest angular resolution
231: achievable at the time of observation. We observed each source for
232: 15 minutes roughly every 40 minutes over a period of 12 hours in
233: order to optimize the {\it uv}-coverage, obtaining a total
234: integration time of 4-5 hours on each object. The primary
235: calibrator 1934-638 was observed for flux calibration. The data
236: were reduced using the reduction package MIRIAD following the
237: procedures outlined in Miriad Users Guide \citep{sault03}. The
238: final CLEAN images were obtained by performing inverse Fourier
239: transforms from the calibrated visibility datasets to images, and
240: then deconvolve the sidelobes using the CLEAN algorithm. Natural
241: weighting was used and we attained an angular resolution of $\sim
242: 3''$ at 6~cm and $\sim 1.5''$ at 3.6~cm. Where applicable, the
243: self-calibration was performed to improve the image quality. In the
244: end, only He~2-84 has strong enough emission to allow phase
245: self-calibration.
246:
247: \subsection{Optical Observations}
248:
249: Three objects, He~2-36, He~2-84, and 19W32, which have radio
250: emission detected in our study but have no high resolution optical
251: images, were observed with the 2.3m Telescope in Siding Spring
252: Observatory (SSO) in 2003 May 26-30. The Imager instrument and
253: three narrow-band filters, H$\alpha$, [N~{\footnotesize II}] and
254: [O~{\footnotesize III}] were used to take photometric images. The
255: center wavelength and bandwidth of each filter are listed in
256: Table~\ref{tb:filter}, along with the best achieved FWHM during the
257: observations. These filters' bandwidths of $\le$ 10~\AA~allow
258: minimize contaminations from adjacent emission lines. The CCD has
259: dimensions of 1024x1024, with a focal plane scale of 0.59$''$/pixel
260: determined by the telescope optics.
261:
262: The images were reduced with standard techniques using IRAF software
263: \citep{massey97}. The averaged bias frames and the overscan level
264: were subtracted from each target image, which was then trimmed to
265: discard the overscan region and the first and last few columns and
266: rows. The dark current was negligible during the observations, so
267: no dark current subtraction was performed. Finally, the images were
268: divided by the flat-field image for each filter. The exposure
269: frames of each object were then combined to generate the final
270: images in each filter, using a statistical method that removes
271: cosmic ray contamination.
272:
273: \section{Results and Analyses}
274: \label{sec:results-jet}
275:
276: Among the 16 objects observed, we detected radio emission in nine of
277: them, and five of which have a central compact source. In Table
278: \ref{tb:result} we list the detect limits (3$\sigma$) of our
279: observations, and the type of emission that were detected. A more
280: detailed description of the detections and the analysis for the
281: compact sources are given in the following subsections.
282:
283: \subsection{Northern Survey}
284: \label{sec:norsurvey}
285:
286: Of the 11 objects observed by VLA, we found that NGC~6302, Hubble~5
287: and NGC~6537 only exhibit extended radio emission at 1.3~cm that
288: could not be properly mapped due to the lack of short baselines in
289: our observations. Two objects, 19W32 and M~1-91, show only a
290: compact central source at both 1.3~cm and 0.7~cm. No emission was
291: detected for the remaining six objects. This is the first time the
292: centers of all these nebulae have been examined at such high angular
293: resolutions at radio wavelengths. Note that for the six objects with
294: non-detection in our observations, some have been detected at 6~cm
295: by \cite{zijlstra89} and \cite{aaquist90}, whose surveys are more
296: sensitive to the extended emission.
297:
298: Here we present the images for the radio cores of 19W32 and M~1-91,
299: along with their optical images (Figures~\ref{fg:19w32} and
300: \ref{fg:m1-91}). Both objects exhibit a compact stellar-like
301: optical core. We obtained the optical images of 19W32 from SSO 2.3m
302: telescope observations as described above, those of M~1-91 from the
303: IAC (Instituto de Astrof\'{i}sica de Canarias) catalog
304: \citep{manchado96}. All images are oriented with north up and east
305: to the left.
306:
307: \subsubsection{19W32}
308:
309: The H$\alpha$ and [N~{\footnotesize II}] images of 19W32 show a
310: prominent stellar-like optical core and two lobes in opposite
311: directions, extending $\sim 10''$ to the northeast and southwest.
312: Only the H$\alpha$ image is presented in Figure~\ref{fg:19w32}. No
313: [O~{\footnotesize III}] emission was detected for this nebula. The
314: radio continuum images only show a compact core, which coincides
315: with the optical stellar core.
316:
317: \subsubsection{M~1-91}
318:
319: The H$\alpha$, [N~{\footnotesize II}] and [O~{\footnotesize III}]
320: images of M~1-91 shows an optical core and two lobes in opposite
321: directions, extending at least $20''$ to the northeast and
322: southwest. Since the images of the three filters show similar
323: structures, only the H$\alpha$ image is presented in
324: Figure~\ref{fg:m1-91}. M~1-91's morphology is very similar to
325: M~2-9, except that the brightenings of the lobes being point
326: symmetric instead of mirror symmetric. The radio continuum images
327: only show a compact core, which coincides with the optical stellar
328: core. The 1.3~cm emission of M~1-91 shows some extended structures
329: that might align with the point-symmetric structures. This needs
330: further observations to confirm.
331:
332: \subsubsection{Spectral Indices of the Radio Cores}
333: \label{sec:tb}
334:
335: The spectral index, defined as the slope of the radio spectrum
336: $\frac{d\log F_{\nu}}{d\log{\nu}}$, indicates the emission
337: mechanisms and optical depths of the radio emission. Optically
338: thick free-free emission has a rising spectrum, optically thin
339: free-free emission has a flat spectrum ($\sim -0.1$), and optically
340: thin nonthermal emission has a steeply falling spectrum. In the
341: case of a spherically symmetric nebula, the radio emission has a
342: spectral index of 0.6 ($S_{\nu} \propto \nu^{0.6}$) for an
343: isothermal stellar wind with a constant expansion velocity
344: \citep{wright75, panagia75}. For a special case of collimated
345: ionized winds with constant opening angle, the radio emission also
346: has a spectral index of 0.6 \citep{reynolds86}.
347:
348: In order to determine the properties of the radio cores of 19W32 and
349: M~1-91, we calculated the spectral indices of their radio emission.
350: To determine the flux densities of the radio cores observed in 19W32
351: and M~1-91, a two-dimensional Gaussian structure was fit to the
352: measurements by applying the AIPS tasks IMFIT to the clean map and
353: OMFIT to the visibility data. Because of the low S/N ratio in these
354: observations, neither task gives an obviously superior fit.
355: Therefore results from both tasks are given. The entire dataset
356: were used for fitting at both 1.3 and 0.7~cm. We also fit a more
357: restricted dataset at 0.7~cm chosen to have the same sampling range
358: in angular scales as at 1.3~cm to check for a greater inclusion of
359: any extended nebula emission in the larger synthesized beam. We
360: found that both procedures gave the same results within measurement
361: uncertainties ($< 1\sigma$). Thus, in Table~\ref{tb:vla-flux}, only
362: the results obtained by fitting the datasets over their entire {\it
363: uv} range are listed. The rising spectra of both 19W32 and M~1-91
364: suggest that their radio cores are produced by optically thick
365: free-free emission. Fits to the measured visibilities show that
366: these cores are resolved along one dimension; i.e., the minor axis
367: of the synthesized beam. The size of the emitting source will be
368: discussed further in \S\ref{sec:mdot}
369:
370: \subsection{Southern Survey}
371: \label{sec:sousurvey}
372:
373: Of the five objects observed by ATCA, He~2-36 shows only extended
374: emission, He~2-25 and He~2-84 shows only a compact central source,
375: Mz~3 has both, and no emission was detected from Th~2-B.
376: Figures~\ref{fg:he2-25} to \ref{fg:mz3} present the radio and
377: optical images of the four objects with detections. For the three
378: objects with radio core detections, their optical images all show a
379: compact stellar-like cores. The optical images of He~2-25 were
380: obtained from \cite{lee07}, those of Mz~3 from the {\it HST}
381: archive, and those of He~2-36 and He~2-84 from SSO 2.3m telescope
382: observations. As before, images are oriented with north up, east to
383: the left.
384:
385: \subsubsection{He 2-25}
386:
387: The optical images of He~2-25 display a bright central source and a
388: pair of very faint bipolar lobes extending toward north and south
389: (Figure \ref{fg:he2-25}). The radio emission of this object is only
390: detected at 3.6~cm, which shows a central unresolved source
391: coinciding with the optical core.
392:
393: \subsubsection{He 2-36}
394:
395: The optical images of He~2-36 display a bright central source and an
396: equatorial torus with limb brightening arms extending clockwise from
397: the torus edges toward the north and south. The brightening
398: resembles an S-shape point symmetrical distribution. We presented
399: the H$\alpha$ image in two different scales in order to show the
400: detailed structures (Figure \ref{fg:he2-36}). The nebula possesses
401: a more moderate degree of bipolarity (lobes are less collimated)
402: than other butterfly nebulae observed in this work. The radio
403: images only show emission from the equatorial band. Although the
404: radio images show some interesting features, these are not relevant
405: to current investigation.
406:
407: \subsubsection{He 2-84}
408:
409: The H$\alpha$ and [N~{\footnotesize II}] images of He~2-84 show an
410: optical core with an equatorial torus. The two lobes extending at
411: right angles to this torus toward the northeast and southwest appear
412: to have a point symmetrical geometry. The H$\alpha$ image is
413: displayed in two different scales in order to show the detailed
414: structures (Figure \ref{fg:he2-84}). The lobes are very faint and
415: cannot be seen in the [O~{\footnotesize III}] image. The radio
416: continuum images only show a central unresolved bright source, which
417: coincides with the optical core.
418:
419: \subsubsection{Mz 3}
420:
421: The optical images of Mz~3 show a bright core, two approximately
422: spherical bipolar lobes, and even more extensive filamentary
423: nebulosities on either side of the central star can be seen from its
424: H$\alpha$ and [N~{\footnotesize II}] images. Only the H$\alpha$
425: images is presented in Figure \ref{fg:mz3}. The total size of the
426: nebula spans over 50$''$ on the sky. The radio images of Mz~3 show
427: emission from the central source and the two lobes, similar to the
428: radio images of M~2-9 \citep{kwok85}. The radio central source and
429: the two lobes coincide with the optical core and the brightest
430: structures of the optical lobes.
431:
432: \subsubsection{Spectral Indices of the Radio Cores}
433:
434: He~2-25 has a compact radio core detected only at 3.6~cm. The fact
435: that no detection at 6~cm has already indicated that this core is
436: optically thick. Using the detection limit at 6~cm, a lower limit
437: of spectral index is estimated to be 1.25. However, a more precise
438: evaluation of the spectral index of He~2-25 has to wait for
439: observations at shorter wavelengths.
440:
441: He~2-84 and Mz~3 have radio cores detected both at 3.6~cm and 6~cm,
442: suitable for spectral index determination. For He~2-84, we applied
443: the task IMFIT to the clean map and UVFIT to the visibility data to
444: determine the flux densities of its radio core assuming the radio
445: core to have a two-dimensional Gaussian structure. For Mz~3, the
446: flux density was obtained only for the bright compact core in the
447: CLEAN map. This core can be seen in Figure \ref{fg:mz3}. The total
448: flux density was obtained by summing up the flux densities of the
449: CLEAN components in the radio core region, since the convolution of
450: the Clean components with the CLEAN beam might result in the
451: confusion of the core emission and nebular emission. The results
452: are listed in Table~\ref{tb:atca-flux} with 1$\sigma$ uncertainties.
453: Even with only clean components in the core region considered, it
454: seems only part of the diffuse emission is excluded. An analysis of
455: this is given later.
456:
457: The radio core of He~2-84 has a spectral index significantly more
458: negative than the value of $-0.1$ expected for optically thin
459: free-free emission. This suggests that the emission of the observed
460: radio core is dominated by more extended emission that is, as
461: expected, better resolved out at shorter than at longer wavelengths.
462: Therefore the extended emission makes a stronger contribution at
463: 6~cm than at 3.6~cm, resulting in a spectral index that is even more
464: negative than expected for optically thin free-free emission. To
465: illustrate the effects of the extended emission on the flux
466: measurements, Figure \ref{fg:he2-84.uv-amp} plots amplitude vs. {\it
467: uv}-distance for He 2-84 at 6 cm and 3.6 cm. The real component of
468: the amplitudes have been averaged in several {\it uv} bins. The
469: binning size is wider at longer {\it uv}-distance in order to
470: increase the signal to noise ratio at longer baseline. As shown in
471: the figure, the amplitude decreases with increasing {\it
472: uv}-distance which clear demonstrates that an extended component is
473: present.
474:
475: In order to mitigate against contamination from the diffuse nebular
476: emission, we excluded the short baselines ($<$ 50 k$\lambda$) that
477: pick up the extended nebular component. We also excluded the
478: longest baselines at 3.6~cm in order to provide the same angular
479: resolution as at 6~cm. Because of the noise increases caused by
480: removing {\it uv} data, the visibility and map fits did not yield
481: reliable results. Consequently, flux densities were derived by
482: summing the flux densities of the CLEAN components. The results are
483: listed in Table~\ref{tb:atca-uv}, which indeed shows that the
484: extended emission constitutes $\sim$ 90\% of the flux at 6~cm, and
485: $\sim$ 80\% of the flux at 3.6~cm for He~2-84. A similar analysis
486: was also performed on Mz~3, since some contamination from the
487: extended emission is also possible. The results listed in
488: Table~\ref{tb:atca-uv} show that for Mz~3, the flux contributed from
489: the extended emission is relatively small at 6~cm, and there is
490: virtually no contamination at 3.6~cm.
491:
492: The differences between the core flux densities and resultant
493: spectral indices obtained using the different {\it uv} distances
494: suggest that for both He~2-84 and Mz~3, the emission comes from two
495: sources, one compact and one diffuse. In both cases, the more
496: careful analysis incorporating the same {\it uv}-coverage (Table
497: \ref{tb:atca-uv}) indicates that an optically thick radio core has
498: been detected. A separate ATCA investigation of the radio
499: properties of Mz~3 by \cite{bains04} also concluded that Mz~3 has an
500: optically thick core.
501:
502: It is worthwhile mentioning that we did find a compact radio source
503: both in 6~cm and 3.6~cm inside the primary beam of the Th~2-B
504: observation as shown in Figure~\ref{fg:brs}. However, after
505: carefully examining the position of this source with the known
506: objects in the field using
507: SIMBAD\footnote{http://simbad.u-strasbg.fr/}, we determined that
508: this is a background radio source not related to Th~2-B, since it is
509: about 3$'$ away from the optical postion of Th~2-B. This background
510: radio source has a spectral index of $-1$, characteristic of
511: nonthermal radio emission. It is probably a quasar.
512:
513: \section{Discussion}
514: \label{sec:discussion}
515:
516: \subsection{The Mass Loss Rate}
517: \label{sec:mdot}
518:
519: The VLA and ATCA radio flux densities obtained in the previous
520: sections can be used to estimate the mass loss rates of the nebulae,
521: following the procedure in \cite{wright75}. \citeauthor{wright75}
522: gave the mass loss rate for a spherical wind with a spectral index
523: of 0.6 as
524: \begin{equation}
525: \dot{M} = 0.095 \frac{\mu_e v S_{\nu}^{3/4} D^{3/2}}{Z x_e^{1/2}
526: g^{1/2} \nu^{1/2}}~~~~ {\rm M_{\odot}~ yr^{-1}}~,
527: \label{eq:mdot}
528: \end{equation}
529: where {\it v} is the velocity of the mass loss flow in km~s$^{-1}$,
530: $S_{\nu}$ is the flux observed at frequency $\nu$ in Jy, $\nu$ is in
531: Hz, {\it D} is the distance in kpc, {\it Z} is the nuclear charge,
532: $\mu_e$ is the mean atomic weight per electron, $x_e$ is the ratio
533: of electron density to the number density of ionized H, i.e., $n_e =
534: x_e n_p$, and {\it g} is the Gaunt factor, which can be approximated
535: in the radio regime as
536: \begin{equation}
537: g (\nu, T) = \frac{\sqrt{3}}{\pi} \left\{ 17.7 + \ln \left[
538: \frac{ (T_e/{\rm K})^{3/2}}{(\nu/{\rm Hz}) Z} \right] \right\}~,
539: \end{equation}
540: where $T_e$ is the electron temperature \citep[p. 95]{osterbrock89}.
541:
542: The constants $x_e$ and $\mu_e$ can be found by assuming the He to H
543: number ratio {\it y} and the fraction of He that is singly ionized
544: $y'$, since the contributions from heavy elements are negligible due
545: to their low abundance. The singly ionized He to H ratio is $y'y$
546: and the doubly ionized He to H ratio is $(1-y')y$ if there is no
547: neutral He. The ratio of electron density to the number density of
548: ionized H becomes $x_e=1+yy'+2y(1-y')$, and the mean atomic weight
549: per electron is
550: \begin{equation}
551: \mu_e = \frac{1+4y}{1+yy'+2y(1-y')}
552: \end{equation}
553: For $y=0.11$ and $y'=0.5$, $x_e$ and $\mu_e$ have values of 1.165
554: and 1.236, respectively.
555:
556: Assuming $T_e=4000$ K \citep[the temperature of free electrons are
557: mainly contributed by ionization of hydrogen,][]{reid97} and $Z=1$,
558: we calculate the mass loss rate as a function of distance and wind
559: velocity using the flux densities found in our observations. Due to
560: the large uncertainties of the mass loss rate parameters, the
561: differences from the flux densities at different wavelengths for the
562: same object are nebligible. Therefore we give a best estimate of
563: the mass loss rate for each object in Table~\ref{tb:mass}.
564:
565: To calculate the actual mass loss rates, the wind velocities and the
566: distances to the nebulae are needed. Assuming a typical wind
567: velocity of 1000 km s$^{-1}$, and using the distances available from
568: the literature as given in Table \ref{tb:mass}, we obtain $\dot{M}
569: \sim 7 \times 10^{-6}~{\rm M_{\odot}~ yr}^{-1}$ for He~2-25,
570: $\dot{M} \sim 10^{-5}~{\rm M_{\odot}~ yr}^{-1}$ for He~2-84,
571: $\dot{M} \sim 3 \times 10^{-5}~{\rm M_{\odot}~ yr}^{-1}$ for 19W32,
572: $\dot{M} \sim 9 \times 10^{-5}~{\rm M_{\odot}~ yr}^{-1}$ for M~1-91
573: and $\dot{M} \sim 2 \times 10^{-4}~{\rm M_{\odot}~ yr}^{-1}$ for
574: Mz~3, respectively. The mass loss rate for Mz~3 is consistent with
575: that derived by \cite{bains04}. They found $\dot{M} \sim 5-30
576: \times 10^{-5}~{\rm M_{\odot}~ yr}^{-1}$ using the range of $D$
577: available in the literature.
578:
579: We also used the Eq.\ (11) in \cite{wright75} to calculate the size
580: of the emitting source, under the spherical wind assumption. The
581: size equation, when incorporated Eq.\ (\ref{eq:mdot}), reduced to
582: \begin{eqnarray}
583: r(\nu) &=& 5.83 \times 10^{27} S_{\nu}^{1/2} T^{-1/2} \nu^{-1} D
584: ~~~{\rm cm} \nonumber \\
585: &=& 3.9 \times 10^{11} S_{\nu}^{1/2} T^{-1/2} \nu^{-1}
586: ~~~{\rm arcsec}~,
587: \label{eq:size}
588: \end{eqnarray}
589: where $r$ corresponds to the radius of the emitting region where the
590: optical depth $\tau_{\nu}=0.244$. For the unresolved sources,
591: Eq.~(\ref{eq:size}) predicts a source size smaller than the
592: synthesized beam. For the two resolved sources 19W32 and M~1-91, we
593: listed the measured dimensions along the resolved axis and the
594: predicted source sizes in Table \ref{tb:vla-size}. The predicted
595: sizes are about 4 or 5 times smaller than the measured dimensions.
596: It is possible that we underestimated the predicted size by giving
597: the wrong assumption of temperature in Eq.~(\ref{eq:size}).
598: However, to make the predicted size consistent with the measured
599: dimension, the temperature has to be unrealistically small ($\sim$
600: 200 K), in which case the winds are not ionized. Therefore, we
601: concluded that the predicted sizes are smaller is because these
602: sources are elongated, possibly by collimated winds (jets).
603:
604: If we approximate a collimated source as an ellipse with major and
605: minor axes of $a$ and $b$, and the collimated wind has an emission
606: area as the same as a spherical wind, then
607: \begin{equation}
608: a = r \sqrt{2/\theta_0} \ .
609: \end{equation}
610: where $\theta_0=2b/a$ is the opening angle of the jet in radians.
611: Using the measured dimension of the resolved axis as $a$, we can
612: derive the opening angle of the collimated jet for 19W32 and M~1-91
613: as listed in Table \ref{tb:vla-size}. The opening angle can then be
614: used to revise the mass loss rate. As pointed out by
615: \cite{reynolds86}, a collimated wind with a much lower mass loss
616: rate can easily produce the same radio intensity as an uncollimated
617: wind with a much higher mass loss rate, with the relationship for a
618: spectral index of 0.6 is
619: \begin{equation}
620: \frac{\dot{M}{\rm (jet)}}{\dot{M}{\rm (spherical~wind)}} = 0.20
621: \theta_0 (\sin i)^{-1/4}
622: \end{equation}
623: where {\it i} is the inclination of the major axis of the wind to
624: our line of sight. Assuming $i=90^{\circ}$, which gives a lower
625: limit of the ratio, we derived the ratio is $\sim0.02$ in both
626: cases. Thus we revised the mass loss rates of 19W32 and M~1-91 to
627: $\dot{M} \sim 10^{-6}~{\rm M_{\odot}~yr}^{-1}$ for collimated winds,
628: which are about 50 times lower than assuming spherical symmetry.
629: From the analysis of central-star wind P-Cygni profiles observed
630: with the {\it International Ultraviolet Explorer}, the mass loss
631: rates from central stars of PNs are $10^{-9}-10^{-7}$ M$_\odot$
632: yr$^{-1}$ \citep{perinotto83}. These values are consistent with
633: theoretical values expected from winds driven by radiation pressure
634: on ionized gas \citep{pau03}. The mass loss rates we inferred are
635: at least an order of magnitude higher than expected. Such high mass
636: loss rates can be explained if the central stars of these nebulae
637: are symbiotic stars, as in the case suspected for M~2-9 whose
638: inferred mass loss rate is $\sim 6 \times 10^{-6}~{\rm
639: M_{\odot}~yr}^{-1}$ \citep{lim03}. Similarly, \cite{bains04} also
640: favored a binary companion scenario instead of the superwind of a
641: single progenitor to explain the observational properties of Mz~3.
642:
643:
644: \subsection{PNs or Symbiotic Stars?}
645:
646: Symbiotic stars are interacting binary systems, typically consisting
647: of a white dwarf that accretes material from the wind of a cool red
648: giant companion, sometimes forming a disk around the white dwarf.
649: The radiation from the hot star partially ionizes the wind from the
650: cool star. A spectrum of the two together thus shows a combination
651: of absorption lines from the surface of the cool red giant and
652: emission lines from the nebula. First noticed by \cite{merrill19},
653: \cite{plaskett28}, and \cite{merrill32} as stars with combination of
654: spectra, these objects were named symbiotic stars by
655: \cite{merrill58}.
656:
657: The possibility of some well-known PNs can in fact be symbiotic
658: stars was discussed by \cite{kwok03b} and \cite{corradi03}. With
659: the discovery of increasing number of PNs with binary nuclei, the
660: simplistic notion that PNs have single central stars and symbiotics
661: are binaries is no longer viable. It is often difficult to
662: distinguish the two classes of objects by morphology, spectrum, or
663: spectral energy distribution. Only in cases where periodic
664: photometric variability as a result of pulsation, or the
665: observations of photospheric molecular absorption band in the
666: spectrum, that one can be certain that we are dealing with a
667: symbiotic system. However, it is difficult to rule out that certain
668: PNs, including those under study here, are NOT symbiotic stars.
669: Even the evolutionary status of extremely well-studied objects such
670: as M~2-9 is still not certain.
671:
672: If these nebulae with radio cores are indeed symbiotic star systems,
673: the ionized wind we measured would originate from the white dwarf
674: companion. The mass loss rate of this wind does not directly tell
675: the mass loss rate of the AGB star, but they must be related since
676: the accretion disk is material captured from the AGB star wind. If
677: there is a balance between accretion onto the disk and ejection to
678: form a jet, then the mass loss rate of the AGB star must be at least
679: as high as the mass loss rate of the jet. Assuming a spherically
680: symmetric outflow and only a fraction of the slow neutral wind from
681: the AGB star is captured by the WD dwarf, the corresponding mass
682: loss rate of the AGB star would be at least in excess of $10^{-4}$
683: M$_\odot$ yr$^{-1}$. Such values are seldom observed in AGB stars.
684: This problem can be overcome if the AGB star wind is strongly
685: concentrated in the binary orbital plane, as is predicted for
686: symbiotic star systems with relatively small orbital separations.
687:
688: The remaining objects have no detected radio core emission because
689: they are either too distant or too faint. None of those with known
690: distances (0.5 - 7 kpc) are farther than the most distant PN with a
691: detected radio core (M~1-91, 7 kpc). If all these objects have the
692: same radio luminosity, we should be able to detect all of them.
693: Therefore the more likely explanation is that these objects have
694: intrinsically faint radio emission. It is possible that those
695: objects evolved from single stars thus do not produce strong radio
696: emission. Since the mass loss rate from the central stars of PNs
697: are at least an order of magnitude smaller than that of the
698: symbiotic systems, even if these nebulae do have ionized jets from
699: the central stars, their radio emission would be too faint to be
700: detected. Thus radio observations may provide a way to separate
701: symbiotic nebulae from genuine planetary nebulae. This would also
702: mean that there is more than one mechanism to give a nebula a
703: narrow-waist shape. For single stars, one of the more possible
704: mechanisms for shaping nebulae is magnetic fields
705: \citep[e.g.,][]{garcia99}.
706:
707: \section{Summary}
708: \label{sec:sum-jet}
709:
710: We have used the VLA and ATCA to search for signatures of ionized
711: winds -- optically thick radio cores -- in sixteen narrow-waist
712: bipolar nebulae. Eleven nebulae were observed with the VLA at
713: 1.3~cm and 0.7~cm with resolutions of 0.08$''$ and 0.04$''$,
714: respectively. This is the first time that the centers of these
715: objects have been systematically studied at such high angular
716: resolution at radio wavelengths. Two objects, 19W32 and M~1-91,
717: were found to exhibit a compact central source with a positive
718: spectral index. Three more objects examined with the VLA, NGC~6302,
719: Hubble~5, and NGC 6537, exhibit extended radio emission that cannot
720: be mapped properly at such short wavelengths. No emission was
721: detected for six other objects in the VLA survey.
722:
723: Five nebulae were observed with the ATCA at 6~cm and 3.6~cm with
724: resolutions of 3$''$ and 1.5$''$, respectively. He~2-25 was found
725: to have a compact central source only at 3.6~cm. He~2-84 and Mz~3
726: were found to contain a compact central source at both wavelengths.
727: When short baselines that are sensitive to extended nebular emission
728: were excluded, we found that the spectral index of these two radio
729: cores changes with {\it uv}-coverage, suggesting that an optically
730: thick core has been detected. In addition to its compact radio
731: core, Mz~3 also exhibits extended nebulae emission that coincides
732: with the two lobes seen in the {\it HST} optical images. He~2-36
733: also shows extended nebular emission but no compact core. No radio
734: emission was detected for Th~2-B in the ATCA study.
735:
736: For 19W32 and M~1-91, we found that the predicted source sizes are
737: smaller than the measured dimensions, which we attribute to the
738: radio cores being elongated, implying a collimated wind or jet. The
739: possibility that these radio cores represent COFs responsible for
740: shaping the bipolar morphology of the nebulae is particularly
741: intriguing. The derived mass loss rates of the wind from the radio
742: observations are high in comparison to typical fast winds from
743: central stars of planetary nebulae, and one possible solution is
744: that the winds are not spherical symmetric but instead are
745: collimated. If these winds are collimated by a binary system, then
746: the radio imaging offers the best way to test such models. The five
747: objects with radio core detections all exhibit a compact
748: stellar-like optical core. This needs further investigation in
749: order to understand the connection between the optical morphology
750: and the radio ionized jets.
751:
752: We are planning follow-up observations of these radio cores with
753: higher angular resolution in order to study the structures of the
754: cores. We have recently obtained observations of M~1-91 with the
755: VLA A+PT (Pie Town) configuration. The addition of the Pie Town
756: antenna will increase the resolution of the VLA by a factor of two
757: while keeping the full sensitivity of the VLA. The new data will
758: add to our knowledge of the core properties of M~1-91 once the
759: analysis is completed.
760:
761: \acknowledgements
762:
763: We thank the anonymous referee for constructive comments on the
764: manuscript. We are grateful to the local staff of the NRAO,
765: Narrabri and Siding Spring Observatories for their support during
766: observations at the VLA, ATCA and SSO. The National Radio Astronomy
767: Observatory is a facility of the National Science Foundation
768: operated under cooperative agreement by Associated Universities,
769: Inc. The Compact Array is part of the Australia Telescope, which is
770: funded by the Commonwealth of Australia for operation as a National
771: Facility managed by the Commonwealth Scientific and Industrial
772: Research Organisation. The Siding Spring Observatory is operated by
773: the Australian National University, Research School of Astronomy and
774: Astrophysics.
775:
776: THL expresses gratitude to the University of Calgary and the
777: Province of Alberta for various scholarships and a travel grant
778: during her PhD. This work was supported in part by a grant to SK
779: from the Natural Sciences and Engineering Research Council of
780: Canada.
781:
782: \begin{thebibliography}{999}
783:
784: \bibitem[Aaquist \& Kwok(1990)]{aaquist90}
785: Aaquist, O. B. \& Kwok, S. 1990, \aaps, 84, 229
786:
787: \bibitem[Allen \& Swings(1972)]{allen72}
788: Allen, D. A., \& Swings, J. P. 1972, \apj, 174, 583
789:
790: \bibitem[Balick(1987)]{bal87}
791: Balick, B. 1987, \aj, 94, 671
792:
793: \bibitem[Balick \& Frank(2002)]{balick02}
794: Balick, B., \& Frank, A. 2002, \araa, 40, 439
795:
796: \bibitem[Bains et al.(2004)]{bains04}
797: Bains, I., Redman, M. P., Bryce, M., \& Meaburn, J. 2004, \mnras,
798: 354, 549
799:
800: \bibitem[Bujarrabal et al.(2001)]{bujarrabal01}
801: Bujarrabal, V., Castro-Carrizo, A., Alcolea, J., \&
802: S\'{a}nchez Contreras, C. 2001, \aap, 377, 868
803:
804: \bibitem[Corradi(2003)]{corradi03}
805: Corradi, R. L. M. 2003, in ASP Conf. Ser. 303, Symbiotic Stars
806: Probing Stellar Evolution, eds. R. L. M. Corradi,
807: J. Mikolajewska, \& T. J. Mahoney (San Francisco: ASP), 393
808:
809: \bibitem[Corradi \& Schwarz(1995)]{corradi95}
810: Corradi, R. L. M., \& Schwarz, H. E. 1995, \aap, 293, 871
811:
812: \bibitem[Doyle et al.(2000)]{doyle00}
813: Doyle, S., Balick, B., Corradi, R. L. M., \& Schwarz,
814: H. E. 2000, \aj, 119, 1339
815:
816: \bibitem[Forveille et al.(1998)]{forveille98}
817: Forveille, T., Huggins, P. J., Bachiller, R., \& Cox, P. 1998,
818: \apjl, 495, L111
819:
820: \bibitem[Garc\'{i}a-Segura et al.(1999)]{garcia99}
821: Garc\'{i}a-Segura, G., Langer, N., R\'{o}\.{z}yczka, M., \&
822: Franco, J. 1999, \apj, 517, 767
823:
824: \bibitem[Gieseking et al.(1985)]{gieseking85}
825: Gieseking, F., Becker, I., \& Solf, J. 1985, \apjl, 295, L17
826:
827: \bibitem[Goodrich(1991)]{goodrich91}
828: Goodrich, R. W. 1991, \apj, 366, 163
829:
830: \bibitem[Huggins et al.(2000)]{huggins00}
831: Huggins, P. J., Forveille, T., Bachiller, R., \& Cox, P. 2000,
832: \apj, 544, 889
833:
834: \bibitem[Kohoutek(1982)]{kohoutek82}
835: Kohoutek, L. 1982, \aap, 115, 420
836:
837: \bibitem[Kohoutek \& Surdej(1980)]{kohoutek80}
838: Kohoutek, L., \& Surdej, J. 1980, \aap, 85, 161
839:
840: \bibitem[Kahn \& West(1985)]{kah85}
841: Kahn, F., \& West, K. A. 1985, \mnras, 212, 837
842:
843: \bibitem[Kwok(2003a)]{kwok03a}
844: Kwok, S. 2003a, in IAU Symp. 209: Planetary Nebulae: Their
845: Evolution and Role in the Universe, eds. S. Kwok, M. Dopita, \&
846: R. Sutherland (San Francisco: ASP), 3
847:
848: \bibitem[Kwok(2003b)]{kwok03b}
849: Kwok, S. 2003b, in ASP Conf. Ser. 303, Symbiotic Stars Probing
850: Stellar Evolution, eds. R. L. M. Corradi, J. Mikolajewska, \&
851: T. J. Mahoney (San Francisco: ASP), 428
852:
853: \bibitem[Kwok et al.(1978)]{kwok78}
854: Kwok, S., Purton, C. R., \& Fitzgerald, P. M. 1978, \apjl, 219,
855: L125
856:
857: \bibitem[Kwok et al.(1985)]{kwok85}
858: Kwok, S., Purton, C. R., Matthews, H. E., \& Spoelstra,
859: T. A. T. 1985, \aap, 144 ,321
860:
861: \bibitem[Lee et al.(2006)]{lee07}
862: Lee, T.-H., Stanghellini, L., Ferrario, L., \& Wickramasinghe,
863: D. 2007, \aj, 133, in press (March)
864:
865: \bibitem[Lim \& Kwok(2000)]{lim00}
866: Lim, J., \& Kwok, S. 2000, in ASP Conf. Ser. 199, Asymmetric
867: Planetary Nebulae II: from Origins to Microstructures,
868: eds. J. Kastner, S. Rappaport, \& N. Soker (San Francisco: ASP), 259
869:
870: \bibitem[Lim \& Kwok(2003)]{lim03}
871: Lim, J., \& Kwok, S. 2003, in ASP Conf. Ser. 303, Symbiotic
872: Starts Probing Stellar Evolution, eds. R. L. M. Corradi,
873: J. Mikolajewska, \& T. J. Mahoney (San Francisco: ASP), 437
874:
875: \bibitem[Livio \& Soker(2001)]{livio01}
876: Livio, M. \& Soker, N. 2001, \apj, 552, 685
877:
878: \bibitem[L\'{o}pez(2002)]{lopez02}
879: L\'{o}pez, J. A. 2002, Rev. Mex. Astron. Astrofis. Conf. Ser. 13, 139
880:
881: \bibitem[L\'{o}pez(2003)]{lopez03}
882: L\'{o}pez, J. A. 2003, in IAU Symp.\ 209, Planetary Nebulae and their
883: Role in the Universe, eds. S. Kwok, M. Dopita, \& R. Sutherland
884: (San Francisco: ASP), 483
885:
886: \bibitem[Maciel(1984)]{maciel84}
887: Maciel, W. J. 1984, \aaps, 55, 253
888:
889: \bibitem[Manchado et al.(1996)]{manchado96}
890: Manchado, A., Guerrero, M. A., Stanghellini, L., \& Serra-Ricart,
891: M. 1996, The IAC morphological catalog of northern galactic
892: planetary nebulae (La Laguna: Instituto de Astrof\'{i}sica de Canarias)
893:
894: \bibitem[Massey(1997)]{massey97}
895: Massey, P. 1997, A User's Guide to CCD Reductions with IRAF
896:
897: \bibitem[Merrill(1919)]{merrill19}
898: Merrill, P. W. 1919, \pasp, 31, 305
899:
900: \bibitem[Merrill(1958)]{merrill58}
901: Merrill, P. W. 1958, in Etoiles \`{a} Raies d'Emission
902: (cointe-Sclessin: Institut d'Astrophysique), 20, 436
903:
904: \bibitem[Merrill \& Hummason(1932)]{merrill32}
905: Merrill, P. W., \& Hummason, M. L. 1932, \pasp, 44, 56
906:
907: \bibitem[NRAO(2003)]{nrao03}
908: National Radio Astronomy Observatory 2003. The AIPS Cookbook
909: (Charlottesville: NRAO)
910:
911: \bibitem[Osterbrock(1989)]{osterbrock89}
912: Osterbrock, D. E. 1989, Astrophysics of Gaseous Nebulae and
913: Active Galactic Nuclei (Mill Valley: Univ. Science Books)
914:
915: \bibitem[Panagia \& Felli(1975)]{panagia75}
916: Panagia, N., \& Felli, M. 1975, \aap, 39, 1
917:
918: \bibitem[Pauldrach et al.(2003)]{pau03}
919: Pauldrach, A.W.A., Hoffmann, T.L., \& M\'endez, R.H. 2003, in
920: IAU Symp. 209: Planetary Nebulae: Their Evolution and Role
921: in the Universe, eds. S. Kwok, M. Dopita, \& R. Sutherland (San
922: Francisco: ASP), 177
923:
924: \bibitem[Perinotto(1983)]{perinotto83}
925: Perinotto, M. 1983, in IAU Symp. 103, Planetary Nebulae,
926: ed. D. R. Flower (Dordrecht: D. Reidel Publishing Co.), 323
927:
928: \bibitem[Phillips(2004)]{phillips04}
929: Phillips, J.~P.\ 2004, \mnras, 353, 589
930:
931: \bibitem[Plaskett(1928)]{plaskett28}
932: Plaskett, J. S. 1928, Publ. Dom. Astrophys. Obs. Victoria, 4, 119
933:
934: \bibitem[Reid \& Menten(1997)]{reid97}
935: Reid, M. J., \& Menten, K. M. 1997, \apj, 476, 327
936:
937: \bibitem[Reynolds(1986)]{reynolds86}
938: Reynolds, S. P. 1986, \apj, 304, 713
939:
940: \bibitem[Sault \& Killeen(2003)]{sault03}
941: Sault, R. J., \& Killeen, N. 2003, Miriad Users Guide, Australia
942: Telescope National Facility
943:
944: \bibitem[Soker \& Rappaport(2000)]{soker00}
945: Soker, N., \& Rappaport, S. 2000, \apj, 538, 241
946:
947: \bibitem[van der Veen et al.(1989)]{vanderveen89}
948: van der Veen, W. E. C. J., Habing, H. J., \& Geballe, T. R. 1989,
949: \aap, 226, 108
950:
951: \bibitem[Wright \& Barlow(1975)]{wright75}
952: Wright, A. E., \& Barlow, M. J. 1975, \mnras, 170, 41
953:
954: \bibitem[Zijlstra et al.(1989)]{zijlstra89}
955: Zijlstra, A. A., Pottasch, S. R., \& Bignell, C. 1989, \aaps, 79, 329
956:
957: \end{thebibliography}
958:
959: \clearpage
960:
961: \input{tab1}
962: \input{tab2}
963: \input{tab3}
964: \input{tab4}
965: \input{tab5}
966: \input{tab6}
967: \input{tab7}
968: \input{tab8}
969:
970: \clearpage
971:
972: \begin{figure}
973: \plotone{f1a.eps}
974: \plottwo{f1b.eps}{f1c.eps}
975: \caption{Optical and radio images of 19W32. a) SSO H$\alpha$ image in
976: a linear scale. b) VLA 1.3~cm (left) and 0.7~cm (right) radio
977: continuum images with contours superposed. The contours start
978: at $3\sigma$, with the noise level $\sigma = 1.1 \times
979: 10^{-4}$ Jy/beam at 1.3~cm and $\sigma = 3.6 \times 10^{-4}$
980: Jy/beam at 0.7~cm. The radio images are on a much smaller
981: scale, with the axis range only $\sim 1''$. The position of
982: the radio core coincides with the optical stellar core. In
983: this and subsequent figures, the ellipse in the box in the
984: lower-left corner represents the half-power extent of the
985: synthesized beam.}
986: \label{fg:19w32}
987: \end{figure}
988:
989: \begin{figure}
990: \plotone{f2a.eps}
991: \plottwo{f2b.eps}{f2c.eps}
992: \caption{Optical and radio images of M~1-91. a) IAC H$\alpha$ image
993: in a linear scale. b) VLA 1.3~cm (left) and 0.7~cm (right)
994: radio continuum image with contours superposed. The contours
995: start at $3\sigma$, with the noise level $\sigma = 9.95 \times
996: 10^{-5}$ Jy/beam at 1.3~cm and $\sigma = 2.7 \times 10^{-4}$
997: Jy/beam at 0.7~cm. The radio images are on a much smaller
998: scale, with the axis range only $\sim 1''$. The position of
999: the radio core coincides with the optical stellar core.}
1000: \label{fg:m1-91}
1001: \end{figure}
1002:
1003: \begin{figure}
1004: \plottwo{f3a.eps}{f3b.eps}
1005: \caption{Optical and radio images of He~2-25. a) NTT R-band image
1006: \citep{lee07} in a logarithmic intensity scale. b) ATCA
1007: 3.6~cm (right) radio continuum image in greyscale with
1008: contours superposed. The contours start at $3\sigma$ with the
1009: noise level $\sigma = 6.3 \times 10^{-5}$ Jy/beam at 3.6~cm.
1010: No emission is detected at 6~cm.}
1011: \label{fg:he2-25}
1012: \end{figure}
1013:
1014: \begin{figure}
1015: \plottwo{f4a.eps}{f4b.eps}
1016: \plottwo{f4c.eps}{f4d.eps}
1017: \caption{Optical and radio images of He~2-36. a) SSO H$\alpha$ images
1018: show the detailed structures of the center in a linear scale
1019: (left) and the outer lobes in a logarithmic intensity scale
1020: (right). b) ATCA 6~cm (left) and 3.6~cm (right) radio
1021: continuum image in greyscale with contours superposed. The
1022: contours start at $3\sigma$ with the noise level $\sigma = 6.0
1023: \times 10^{-5}$ Jy/beam at 6~cm and $\sigma = 6.1 \times
1024: 10^{-5}$ Jy/beam at 3.6~cm.}
1025: \label{fg:he2-36}
1026: \end{figure}
1027:
1028: \begin{figure}
1029: \plottwo{f5a.eps}{f5b.eps}
1030: \plottwo{f5c.eps}{f5d.eps}
1031: \caption{Optical and radio images of He~2-84. a) SSO H$\alpha$ images
1032: show the detailed structures of the center in a linear scale
1033: (left) and the outer lobes in a logarithmic intensity scale
1034: (right). b) ATCA 6~cm (left) and 3.6~cm (right) radio
1035: continuum image in greyscale with contours superposed. The
1036: contours start at $3\sigma$ with the noise level $\sigma = 5.7
1037: \times 10^{-5}$ Jy/beam at 6~cm and $\sigma = 7.0 \times
1038: 10^{-5}$ Jy/beam at 3.6~cm. The position of the radio source
1039: coincides with the optical stellar core.}
1040: \label{fg:he2-84}
1041: \end{figure}
1042:
1043: \begin{figure}
1044: \begin{center}
1045: \includegraphics[width=40mm]{f6a.eps} \\
1046: \plottwo{f6b.eps}{f6c.eps}
1047: \caption{Optical and radio images of Mz~3. a) {\it HST} H$\alpha$
1048: image in logarithmic scales. b) ATCA 6~cm (left) and 3.6~cm
1049: (right) radio continuum image in greyscale with contours
1050: superposed. The contours start at $3\sigma$ with the noise
1051: level $\sigma = 1.5 \times 10^{-4}$ Jy/beam at 6~cm and
1052: $\sigma = 1.5 \times 10^{-4}$ Jy/beam at 3.6~cm. The radio
1053: core and bipolar lobes coincide with the optical stellar core
1054: and the brightest structures of the optical lobes.}
1055: \label{fg:mz3}
1056: \end{center}
1057: \end{figure}
1058:
1059: \begin{figure}
1060: \plottwo{f7a.eps}{f7b.eps}
1061: \caption{Amplitude vs. {\it uv}-distance for He~2-84 at 6 cm (left)
1062: and 3.6 cm (right). The real component of the ampplitudes
1063: are averaged in several {\it uv} bins. The binning size is
1064: wider in the longer {\it uv}-distance in order to increase
1065: the signal to noise ratio at longer baselines.}
1066: \label{fg:he2-84.uv-amp}
1067: \end{figure}
1068:
1069: \begin{figure}
1070: \plottwo{f8a.eps}{f8b.eps}
1071: \caption{Images of a background radio source near Th~2-B at 6~cm (left)
1072: and 3.6~cm (right).}
1073: \label{fg:brs}
1074: \end{figure}
1075:
1076: \end{document}
1077:
1078: