1: \documentclass[12pt,preprint]{aastex}
2: \shorttitle{NGC7538}
3: \shortauthors{Sandell et al.}
4: %\slugcomment{corrected to agree with proofs, May 7/03}
5:
6: \begin{document}
7:
8: \def\arcmin{{$^{\prime}$}}
9: \def\arcsec{{$^{\prime\prime}$}}
10: \def\ptsec{$''\mskip-7.6mu.\,$}
11: \def\psec{$^s\mskip-7.6mu.\,$}
12: \def\Msun{\,{\rm M$_{\odot}$}}
13: \def\Lsun{\,{\rm L$_{\odot}$}}
14: \def\ltsim{$\stackrel{<}{\sim}$}
15: \def\gtsim{$\stackrel{>}{\sim}$}
16: \def\jtra#1#2{${\rm J}\!\!=\!\!{#1}\!\!\to\!\!{#2}$}
17:
18: %** added
19: \def\degr{$^{\circ}$}
20: %\def\arcmin{''}
21:
22: \title{NGC\,7538\,S - a High-Mass Protostar with a Massive Rotating Disk}
23:
24:
25: \author{G\"oran Sandell\altaffilmark{1}, Melvyn Wright\altaffilmark{2}, James R. Forster\altaffilmark{2}}
26:
27: \altaffiltext{1}{Universities Space Research Association, NASA Ames Research Center, MS 144-2,
28: Moffett Field, CA 94035, U.S.A.}
29: \email{gsandell@mail.arc.nasa.gov}
30: \altaffiltext{2}{Radio Astronomy Laboratory, University of California, Berkeley
31: 601 Campbell Hall, Berkeley, CA 94720, U.S.A}
32:
33: \begin{abstract}
34:
35: We report the detection of a massive rotating disk around the high-mass
36: Class 0 candidate NGC\,7538\,S. The disk is well-resolved with BIMA
37: ($\theta_A$ =3\ptsec7) in 3.4~mm continuum and in H$^{13}$CN \jtra10. It
38: is seen nearly edge on and has a size of $\sim$30,000 AU. A
39: young, powerful outflow perpendicular to the rotating disk is mapped in SiO
40: \jtra21 and HCO$^+$ \jtra10. The dynamical age of the outflow is $\le$
41: 10,000 yr. The velocity gradient seen in H$^{13}$CN is consistent with
42: Keplerian rotation. Assuming that the gas is gravitationally bound, the
43: mass of the central object is $\sim$ 40 \Msun. The mass of the
44: continuum ``disk'' is $\geq$ 100 \Msun\ and has a luminosity of
45: $\sim$10$^4$ \Lsun. H$^{13}$CN gives a mass $\sim$400 \Msun\ for the
46: rotating disk, and $\sim$1000 \Msun\ for the extended
47: (20\arcsec {}) envelope. Our observations confirm that this is an extremely
48: massive protostar in its earliest stages.
49:
50:
51: \end{abstract}
52:
53: \keywords{ISM: clouds -- (stars:) circumstellar matter -- stars:
54: formation -- stars: pre-main sequence -- submillimeter}
55:
56: \section{Introduction}
57:
58: The formation of high mass stars is still poorly understood, although
59: one expects a high mass protostar to form an accretion disk and drive
60: an outflow similar to what one sees in low mass protostars. Recent
61: surveys show that outflows are very common in high-mass star forming
62: regions \citep[and references therein]{Beuther02} but that accretion disks
63: have been far more elusive, even though they will have to be larger and
64: have larger velocity gradients than accretion disks around low mass
65: stars, which show velocity gradients of only a few tenths of
66: km~s$^{-1}$, see e.g. \citet{Mannings97}. Although there have been a
67: number of papers reporting rotating disks around high-mass stars, most
68: of these show marginal, barely resolved systems. The best case to date,
69: the young high-mass star IRAS 20126$+$4104 \citep{Zhang98,Cesaroni97},
70: has a bolometric luminosity of 1.3 10$^4$ \Lsun\ and drives a massive
71: molecular outflow.
72:
73: In this letter we present the first results from high spatial
74: resolution BIMA observations of NGC\,7538\,S. \citet{Sandell03a}
75: showed that NGC\,7538\,S is a cold, very massive elliptical dust
76: source ($\sim$400\Msun) associated with H$_2$O and OH maser emission
77: but only weak free-free emission, which has all the characteristics of
78: a high-mass Class 0 source. The protostellar source is located in the
79: molecular cloud south-east of the large \ion{H}{2} region NGC\,7538,
80: which is known to contain several centers of active and on-going
81: high-mass star formation \citep{Werner79}. The distance to NGC\,7538 is
82: assumed to be 2.8 kpc. NGC\,7538\,S is about 80\arcsec\ to the south
83: of the well studied ultracompact \ion{H}{2} region IRS\,1.
84:
85: NGC\,7538\,S is more isolated than IRS\,1 and therefore easier to
86: study. We confirm that NGC\,7538\,S is a high mass protostar and
87: resolve the disk around the protostar both in continuum and in
88: molecular lines. In this letter we present data that suggest that
89: the protostar is surrounded by a massive, rotating accretion disk,
90: which drives a very young and powerful outflow perpendicular to the
91: disk.
92:
93:
94: \section{Observations and Data Reduction}
95:
96: The observations of NGC\,7538\,S were made with the BIMA array in
97: 2001/2002 using two frequency settings in the B and C-array
98: configuration. The correlator was split into four 25 MHz bands
99: resulting in a velocity resolution of $\sim$0.34 km~s$^{-1}$. In
100: this letter we base most of our discussion on the frequency setting
101: that included H$^{13}$CN \jtra10, and HCO$^+$ \jtra10, which was
102: observed in both B- and C-arrays. The other frequency setting included
103: the molecular transitions HN$^{13}$C \jtra10, SiO v=0 \jtra21 and
104: H$^{13}$CO$^+$ \jtra10. Only one C-array track was obtained in this
105: setting. The images therefore have poorer spatial resolution and
106: sensitivity, but are consistent with the deeper images seen in
107: H$^{13}$CN and HCO$^+$.
108:
109: The data were reduced and imaged in a standard way using MIRIAD
110: software \citep{Sault95}. Phase calibration was applied using
111: observations of the quasar 0102$+$584 at intervals of 30 minutes. The
112: phase calibrator was observed using an 800 MHz bandwidth for 3 min.
113: 3C454.3 was observed for 10 minutes as a bandpass calibration. The
114: flux density scale was checked from observations of Mars. The
115: uncertainty in the absolute amplitude scale $\sim$15 \%, but the
116: relative amplitude of spectral lines within the same receiver tuning is
117: within $\sim$5 \%.
118:
119: The data were imaged with weighting inversely proportional to the
120: variance in order to obtain the best signal to noise ratio. Averaging
121: 4 spectral channels we obtained an RMS noise level of 24 mJy (0.3 K)
122: with a synthesized beam FWHM 3\ptsec75 $\times$ 3\ptsec59 and peak
123: sidelobe level of -7 \%. Spectral windows which did not contain any
124: significant spectral line emission were averaged to provide a continuum
125: image with an RMS $\sim$3 mJy. The continuum emission was subtracted
126: from the spectral line channels, and the images deconvolved using the
127: CLEAN algorithm.
128:
129:
130: \section{The continuum ``disk''}
131: \label{sect-cont}
132:
133: The single dish sub-mm continuum maps \citep{Sandell03a} show
134: NGC\,7538\,S as an elliptical source embedded in a narrow dust filament.
135: With BIMA we filter out the extended cloud emission and detect only the
136: sub-mm source. At 3.4~mm the continuum source is well resolved with a
137: peak on the northeastern side of an extended elliptical source (Fig 1).
138: A two-component Gaussian fit shows that the northeastern peak is an
139: unresolved point-like source with a flux density of 20 mJy offset
140: 3\ptsec7, 0\ptsec5 from the nominal sub-mm position. The position of the
141: 3.4~mm point source coincides with the OH and H$_2$O maser and the 6~cm
142: VLA position \citep[O. Kameya 2003, private communication]{Argon00,Kameya90}. At
143: 6~cm the VLA source has a flux density of 2.8~mJy. If the 6~cm emission
144: is due to an ionized wind, i.e. follows a $\nu^{0.6}$ frequency
145: dependence, about half of the 3.4~mm flux is an excess, presumably due
146: to hot dust close to the protostar. However, the free-free emission
147: could also have a steeper frequency dependence. In the following we
148: assume that the point-source emission is all due to free-free emission.
149:
150: The extended elliptical source has a size of 13\ptsec7 $\times$
151: 8\ptsec1 and a position angle (P.A) of 71\degr, offset $+$1\ptsec7, $-$0\ptsec4 from the
152: nominal sub-mm position with an integrated flux density of 190
153: mJy, i.e. it agrees within errors both in size and position with the
154: sub-millimeter source. In this respect this source
155: differs from IRS\,1 and IRS\,9, which both appear unresolved at 3~mm
156: \citep{Tak00} and from nearby low mass protostars, which are only
157: marginally resolved with a $\sim$0\ptsec5 beam \citep{Looney00}.
158:
159:
160: To estimate a mass from our continuum observations we adopt a
161: single-temperature optically thin thermal dust model. The total mass of
162: gas and dust, M, can then be expressed as ${\rm M = S_\nu
163: D^2/(\kappa_\nu B_\nu(T_d))}$, where ${\rm B_\nu(T_d)}$ is the Planck
164: function, $\kappa_\nu$ is the dust mass opacity, $T_d$ is the dust
165: temperature, $D$ is the distance, and S$_\nu$ is the integrated flux
166: density at the frequency $\nu$. We assume a gas to dust ratio of 100
167: and adopt a dust mass opacity at 87 GHz, $\kappa_{87} $= 0.0073 $cm^2
168: g^{-1}$, corresponding to a dust emissivity, $\beta$ = 1. This value was
169: also used by \citet{Looney00} in their BIMA study of low mass
170: protostellar disks. With a dust temperature,T$_d$ = 35~K
171: \citet{Sandell03a}, we find the total mass of the continuum disk to be
172: 100\Msun. We note, however, that the isothermal greybody fits by
173: \citet{Sandell03a} predicted a steeper emissivity law, $\beta$ = 1.6,
174: which would result in a five times smaller mass opacity and hence a more
175: massive disk.
176:
177:
178: \section{A rotating protostellar disk ?}
179: \label{sect-moldisk}
180:
181: The map of integrated H$^{13}$CN \jtra10 emission (Fig. 1) shows a
182: bright elongated source surrounded by a more extended envelope
183: superposed on the 3.4~mm continuum. The elongated source in the center
184: of the map has the same orientation and extent as the source we
185: see in dust continuum. For a two-component Gaussian fit, we find the
186: bright emission offset 2\ptsec0, 1\ptsec0 from the nominal sub-mm
187: position, i.e. within errors coincident with the 6~cm VLA and the OH/H$_2$O
188: maser position at +3\ptsec7, +1\ptsec0. The fitted size: 11\ptsec1
189: $\times$ 6\ptsec8 with a P.A. = 55 $\pm$ 5\degr, is slightly smaller
190: than the continuum disk. The position velocity plot taken
191: along the major axis of the bright H$^{13}$CN emission shows a clear
192: velocity gradient across the source (Fig 2). The emission is
193: red-shifted to the southwest and blue-shifted to the northeast with the
194: center of symmetry coinciding within 2\arcsec\ with the position of the
195: protostar. A rotation curve derived from these data is slightly
196: asymmetric with the central velocity close to $-$56 km s$^{-1}$. It
197: reaches a peak of $\pm$1.35 km~s$^{-1}$ at a radius of 5\arcsec\
198: (14000 au) in both directions.
199: The observed velocity gradient corresponds to an enclosed
200: mass of $\sim$ 30 / ${\rm sin}^2i$ \Msun\ for a rotating disk
201: to be gravitationally bound. With an assumed inclination, i =
202: 60\degr\ (see below) the mass is $\sim$ 40
203: \Msun.
204:
205: H$^{13}$CO$^+$ \jtra10 and HN$^{13}$C \jtra10 were observed
206: with only one C-array track and therefore lack the spatial resolution to
207: resolve the disk. Both molecules show strong extended emission towards
208: the protostar and we can therefore use all three molecules to get an
209: estimate of the mass of the disk and the surrounding molecular
210: envelope. For estimating column densities and masses we assume LTE and
211: that the excitation temperature for all molecules is the same as that
212: for the dust, i.e. 35~K. The permanent dipole moments are taken from
213: \citet{Blake87}, except for HCO$^+$, for which we adopt $\mu$ = 4.07
214: Debye \citep{Haese79}. For total mass estimates we use abundance
215: ratios to H$_2$ similar to those in the OMC-1 extended ridge
216: \citep{Blake87}, i.e. [HCO$^+$]/[H$_2$] = 2 10$^{-9}$, [HCN]/[H$_2$] =
217: 5 10$^{-9}$, and [HNC]/[H$_2$] = 5 10$^{-10}$, and the isotope
218: ratio, [$^{12}$C]/[$^{13}$C] = 85 \citep{Wilson94}. For H$^{13}$CN we
219: get a peak column density of 7 10$^{13}$ cm$^{-2}$ for the disk and
220: 5 10$^{13}$ cm$^{-2}$ for the surrounding envelope, resulting in a
221: disk mass $\sim$ 400 \Msun\ and an envelope $\sim$ 1000 \Msun. For
222: H$^{13}$CO$^+$ and HN$^{13}$C we do not have enough spatial resolution
223: to separate the disk and the envelope, but we derive combined masses
224: $\sim$ 1000 and 3000\Msun, for H$^{13}$CO$^+$ and HN$^{13}$C,
225: respectively.
226:
227:
228: \section{The high velocity outflow }
229: \label{sect-flow}
230:
231: We see clear high velocity wings in both HCO$^+$ \jtra10 and SiO v=0
232: \jtra21. Towards the protostar the HCO$^+$ shows dominantly
233: blue-shifted high velocity emission extending 29 km s$^{-1}$ from the
234: systemic velocity ($\sim$ $-$56 km $s^{-1}$) with a sharp, deep
235: absorption on the red-side at nearby cloud velocities, suggesting
236: infall of the surrounding cooler envelope. The SiO line wings at the
237: position of the disk appear more symmetric, presumably
238: due to the broader beam and the absence of self-absorption. Both SiO
239: and HCO$^+$ show the same distribution of the high velocity gas. Here
240: we only present the results from our HCO$^+$ imaging, which has better
241: signal-to-noise and higher spatial resolution.
242:
243: HCO$^+$ shows a strong, compact, jet-like bipolar high velocity outflow
244: at a pa of 147\degr $\pm$ 5\degr\ (Fig 3). The symmetry axis of the
245: outflow passes through the VLA source. The H$^{13}$CN map shows that
246: the surrounding cloud core is more dense to the north than to the
247: south. This is also evident from the outflow, which is much more
248: compact to the northwest than to the southeast. In the northwest the
249: outflow is blue-shifted and terminates sharply at $\sim$7\arcsec\ from
250: the protostar. To the southeast the outflow is mostly red-shifted and
251: well collimated (aspect ratio $\sim$ 2:1) to $\sim$16\arcsec\ from the
252: protostar, where it expands to become a more wide angle flow,
253: presumably breaking out of the cloud at $\sim$ 30\arcsec\ from the
254: protostar. The strongest blue-shifted emission is seen at the tip of
255: the northwestern outflow lobe with outflow velocities extending to 27
256: km s$^{-1}$ from the systemic cloud velocity. The southeastern outflow
257: lobe shows both red- and blue-shifted emission with outflow velocities of
258: $\sim$24 km~s$^{-1}$ in the red and blue-shifted emission up to 18
259: km~s$^{-1}$ in the dense, well collimated part of the outflow. At the
260: tip of the outflow, the outflow velocities are lower, presumably
261: because the high velocity gas is too tenuous to be excited in HCO$^+$.
262:
263: Since there is both blue- and red-shifted emission to the southeast ,
264: and little overlap between the outflow lobes at the star, the outflow
265: must have fairly high inclination. In the following we assume an
266: inclination of 60\degr, which is roughly consistent with the observed
267: aspect ratio of the disk, i.e. we see the disk almost edge on. The
268: northwestern blue-shifted outflow lobe is still confined by the
269: surrounding cloud, while the southeast outflow appears to have broken
270: through the dense cloud core. Since the extent of the outflow to the
271: southeast is less well defined we derive outflow characteristics mainly
272: from the northwest blue-shifted outflow lobe.
273:
274: The blue-shifted outflow lobe gives a dynamical time scale of
275: $\sim$2,000 yr, while the red-shifted emission gives $\sim$10,000 yr,
276: corrected for a 60\degr\ inclination. Note that the velocities are
277: probably much higher in the red-shifted outflow, since it is expanding
278: through a less dense cloud. We assume HCO$^+$ to be optically thin
279: with an excitation temperature of 30~K, the same as \citet{Shepherd96}
280: used in their study of outflows from high mass stars. If we further
281: assume a normal HCO$^+$ abundance in the outflow (Section
282: ~\ref{sect-moldisk}), we find a mass of 9 \Msun\ for the
283: northwest blue-shifted outflow and 8 \Msun\ for the redshifted gas
284: in the southeast outflow lobe. The mass weighted momentum flux, $
285: {\rm F = mv/t_{dyn}}$ and the mechanical luminosity, $L_{mec} =
286: mv^3/2r$, where r is the extent of the outflow, is 90 10$^{-3}$ {\rm
287: $M_{\odot} km s^{-1} yr^{-1}$, and 130 \Lsun\ for the blue outflow
288: alone. The mass loss rate, based only on the confined blue-shifted
289: outflow, could be as high as $\sim$5~10$^{-3}$ {\rm
290: $M_{\odot}~yr^{-1}$.
291:
292: \section{Discussion and Conclusions}
293: \label{sect-discussion}
294:
295: The new data we present in our letter reveal a well resolved elliptical
296: source mapped in several optically thin lines as well as in 3.4 mm
297: continuum. The peak of the line emission is centered within 1\arcsec\ -
298: 2\arcsec\ of the VLA 6cm continuum, OH, and H$_2$O maser emission
299: position. The 3.4 mm continuum emission has an extent similar to the
300: elongated source seen in optically thin high-density tracers with a
301: peak (hotspot) within 1\arcsec\ of the VLA and maser position.
302: Follow-up observations at 1~mm in both line and continuum
303: (Sandell, Wright, \& Forster 2003, in preparation) show the disk more clearly and show that CH$_3$CN
304: \jtra{12}{11}, which traces hot gas, is centered within 0\ptsec2 of
305: this position. FIR observations \citep{Werner79,Thronson79} show a FIR
306: source with a luminosity of $\sim$ 10$^4$ \Lsun, with a narrow FIR
307: spectral energy distribution, strongly suggesting that the FIR
308: luminosity is produced by a single central source. The FIR source has
309: no near or mid-IR counterpart suggesting that it is heavily obscured.
310: The FIR observations do not have enough spatial resolution or
311: positional accuracy to confirm that the luminosity originates from the
312: the OH/H$_2$O masers, but all 1665 MHz OH masers require a luminous
313: source for their excitation; therefore we conclude that the source
314: powering the OH maser is the origin for the bulk of the observed FIR
315: luminosity.
316:
317: The H$^{13}$CN map (Fig 2) shows a clear velocity gradient along the
318: major axis of the elliptical source. If we interpret this as a rotating
319: disk gravitationally bound by the central source, we find that the
320: center of symmetry is within $\le$ 2\arcsec\ of the maser source.
321: Since this position also lies on the symmetry axis of the outflow,
322: defined by our high spatial resolution HCO$^+$ map, there is no doubt
323: that this is the center of the mass and the protostar. From the
324: observed velocity gradient we derive an enclosed mass of 40 \Msun
325: (Section ~\ref{sect-moldisk}). Since we have resolved the emission of
326: the disk from the envelope, we can use the integrated emission of the
327: disk to derive a mass. Assuming normal abundances and a kinetic
328: temperature of 35~K for the disk we find a mass $\sim$ 400 \Msun from
329: the H$^{13}$CN, and $\sim$ 100 \Msun from the 3.4~mm continuum
330: emission. Considering the uncertainties in molecular abundances and
331: dust opacity, these estimates are in reasonable agreement. The mass of
332: the central object and the disk are of the same magnitude.
333:
334: The bipolar HCO+ high velocity outflow (Section ~\ref{sect-flow}) may be
335: driven by the rotating protostellar disk. The outflow is extremely
336: young; the dynamical timescale from the blue-shifted outflow indicates
337: an age of $\leq$ 2000 yr, with an upper limit of 10,000 yr derived from
338: the red-shifted outflow. \citet{Shepherd96} show that there is a clear
339: correlation between mass loss rate and bolometric luminosity of the
340: central source over a wide range of luminosities. If we use the
341: correlation derived by \citet{Shepherd96}, our observed mass loss rate
342: corresponds to a luminosity of close to 10$^5 $\Lsun, while the
343: observed luminosity is only 10$^4 $\Lsun. Our observations therefore
344: suggest an elevated mass loss rate in the very earliest stages of the
345: formation of a massive star. Since we find the rotating accretion disk
346: to be more massive than the central protostar, the disk will be highly
347: unstable and is likely to feed both the outflow and the central
348: accreting protostar. Model calculations by \citet{Yorke95} for a
349: somewhat less massive central star (10 \Msun{}), find that the
350: accretion disk evolves into a thin flared disk with a ratio of
351: M$_\star$/M$_{disk}$ $\sim$ 4, for all the models that they computed.
352: Even though they ignored magnetic fields, these models should still
353: give an idea of how the disk is expected to evolve.
354:
355: There is only one near-IR source in the vicinity of NGC\,7538\,S, IRS\,11
356: (Fig 3). This IR source is associated with nebulosity and a knot of
357: vibrationally excited H$_2$ emission, suggesting that it is most likely
358: a young star. There are several additional H$_2$ knots in the vicinity
359: of NGC\,7538\,S; two appear to be associated with the molecular
360: outflow. The only other evidence for young objects in this region is
361: an unexplained high velocity HCO$^+$ emission feature 10\arcsec\ southwest
362: of the protostar, three vibrationally excited H$_2$ knots $\sim$ 20\arcsec\
363: east of it, and an H$_2$O maser $\sim$ 30\arcsec\ west of it. The high
364: velocity HCO$^+$ could be part of a wide angle disk-wind or due to a
365: young low luminosity source in the outskirt of the disk. This could
366: explain the observed asymmetry in the continuum disk (more extended to
367: the southwest), but it cannot produce the observed velocity gradient.
368: The eastern H$_2$ knots and the western H$_2$O maser are unrelated to
369: the protostellar source.
370:
371: It is therefore possible that the extended disk/envelope system
372: harbours yet another protostar. Such a protostar would have much lower
373: luminosity and mass, since the massive protostar can explain most, if
374: not all of the observed luminosity. Although high-mass stars usually
375: form in clusters, all the evidence we have so far, suggest that
376: NGC\,7538\,S is an isolated high mass star. If the central protostar is
377: a single star, it may evolve into a late O or early B-star.
378:
379: \acknowledgements
380:
381: We thank Dr. Osamu Kameya for sharing his unpublished VLA results with
382: us. The BIMA array is operated by the Universities of California
383: (Berkeley), Illinois, and Maryland with support from the National
384: Science Foundation.}
385:
386:
387:
388:
389: \begin{thebibliography}{}
390:
391: \bibitem[Argon, Reid and Menten(2000)]{Argon00}
392: Argon, A. L., Reid, M. J., and Menten, K. M. 2000, \apjs, 129, 159
393:
394: \bibitem[Beuther et al.(2002)]{Beuther02}
395: Beuther, H., Schilke, P., Menten, K. M., Motte, F., M.,
396: Shridharan, T. K., and Wyrowski, F. 2002, \apj, 566, 945
397:
398: \bibitem[Blake et al.(1987)]{Blake87}
399: Blake, G. A., Sutton, E. C., Masson, C. R., Phillips, T. G. 1987, \apj, 315, 621
400:
401: \bibitem[Cesaroni et al.(1997)]{Cesaroni97}
402: Cesaroni, R., Felli, M., Testi, L., Walmsley, C. M., and Olmi, L.
403: 1997, \aap, 325, 725
404:
405: \bibitem[Davis et al.(1998)]{Davis98}
406: Davis, C.J., Moriarty-Schieven, G., Eisl\"offel, J., Hoare, M. G.,
407: and Ray, T.P. 1998, \aj, 115, 1118
408:
409:
410: \bibitem[Haese and Woods(1979)]{Haese79}
411: Haese, N. N., Woods, R. C. 1979, Chem. Phys. Lett., 61, 396
412:
413: \bibitem[Kameya et al.(1990)]{Kameya90}
414: Kameya, O., Morita, K. -I., T.I., Kawabe, R., and Ishiguro, M. 1989, \apj,
415: 355, 562
416:
417: \bibitem[Looney, Mundy and Welch(2000)]{Looney00}
418: Looney, L. W., Mundy, L. G., and Welch, W.J. 2000, \apj, 529, 477
419:
420:
421: \bibitem[Mannings and Sargent(1997)]{Mannings97}
422: Mannings, V., and Sargent, A. 1997, \apj, 490, 792
423:
424: \bibitem[Sandell and Sievers(2003)]{Sandell03a}
425: Sandell, G., and Sievers, A. 2003, \apj, submitted
426:
427: \bibitem[Sault, Teuben and Wright(1995)]{Sault95}
428: Sault, R. J., Teuben, P. J., and Wright, M. C. H., 1995, in ASP Conf.
429: Ser. 77: Astronomical Data Analysis Software and Systems IV, Eds.
430: R.A. Shaw, H.E. Payne, and J.J.E. Hayes (Astronomical Society of the
431: Pacific: San Francisco), p. 433
432:
433: \bibitem[Shepherd and Churchwell(1996)]{Shepherd96}
434: Shepherd, D. S., and Churchwell, E. 1996, \apj, 472, 225
435:
436: \bibitem[van der Tak et al.(2000)]{Tak00}
437: van der Tak, F. F. S., van Dishoeck, E. F., Evans II, N. J., and
438: Blake, G. A. 2000, \apj, 537
439:
440:
441: \bibitem[Werner et al.(1979)]{Werner79}
442: Werner, M. W., Becklin, E. E., Gatley, I., Matthews, K., Neugebauer,
443: G., and Wynn-Williams, C.G. 1979, \mnras, 188, 463
444:
445: \bibitem[Thronson and Harper(1979)]{Thronson79}
446: Thronson, H. A., Jr, and Harper, D. A. 1979, \apj, 230, 133
447:
448: \bibitem[Wilson and Rood(1994)]{Wilson94}
449: Wilson, T. L., and Rood, R. T. 1994, \araa, 32, 192
450:
451: \bibitem[Yorke, Bodenheimer and Laughlin(1995)]{Yorke95}
452: Yorke, H. W., Bodenheimer, P., and Laughlin, G. 1995, \apj, 443, 199
453:
454: \bibitem[Zhang et al.(1998)]{Zhang98}
455: Zhang, Q., Hunter, T. R., and Sridharan, T. K. 1998, \apj, 505, L151
456:
457: \end{thebibliography}{}
458:
459:
460: \plotone{f1.eps}
461: \figcaption[]{Color image of 3.4~mm continuum emission of NGC\,7538\,S
462: overlaid with contours of H$^{13}$CN \jtra10 emission integrated over
463: the velocity range $-$70 to $-$40 km~s$^{-1}$. The maps are offset
464: ($-$3\ptsec7,$-$0\ptsec5) from the 3.4~mm peak: $\alpha$(2000.0) =
465: 23$^h$ 13$^m$ 44\psec98, $\delta$(2000.0) = $+$ 61\degr\
466: 26\arcmin\ 49\ptsec2. The intensity scale for the continuum is in
467: Jy~beam$^{-1}$. The contours for the H$^{13}$CN emission go from $-$6 K km
468: s$^{-1}$ (dotted contours) and thereafter from 6 K km s$^{-1}$ with
469: steps of 6 K km s$^{-1}$. The beam FWHM is plotted in blue.}
470:
471: \plotone{f2.eps}
472: \figcaption[]{Position velocity plot of the H$^{13}$CN \jtra10 emission
473: along the disk plane at pa = 60.0 \degr. Positive offsets are to the
474: northeast. The velocity gradient is seen in all three hyperfine lines.
475: The velocity scale is relative to the strongest hyperfine component, F
476: = 2 - 1, centered at a systemic velocity of $\sim$ $-$56 km s$^{-1}$.
477: The line drawing at the bottom of the plot shows the position of three
478: hyperfine components and their relative intensity. The disk emission is
479: red-shifted to the southwest and blue-shifted to the northeast (see
480: text).}
481:
482: \plotone{f3.eps}
483: \figcaption[]{Contour map of blue- and red-shifted high velocity
484: HCO$^+$ \jtra10 integrated over 20 km s$^{-1}$ overlaid on a greyscale
485: image of integrated H$^{13}$CN. The red circle marks the position of
486: the protostar and the white line shows the symmetry axis of the
487: outflow. The star symbol marks the position of IRS\,11 and the filled
488: squares are H$_2$ knots from \citet{Davis98}. The contours are
489: logarithmic and start at 4.6 K km s$^{-1}$ for HCO$^+$ with a step of
490: 10$^{0.2N}$, where N= 1,2.. The beam FWHM is plotted in green.}
491:
492:
493: \end{document}
494:
495: