0810.1271/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: \slugcomment{Accepted by the Astrophysical Journal}
3: % 
4: % Revision History:
5: %
6: %\tighten
7: %\eqsecnum
8: 
9: %\received{}
10: %\accepted{}
11: %\journalid{}{}
12: %\articleid{}{}
13: 
14: \newcommand{\vdag}{(v)^\dagger}
15: % \newcommand{\sun}{\ifmmode\odot\else$\odot$\fi}
16: % \newcommand{\degns}{\ifmmode^\circ\else$^\circ$\fi}
17: % \newcommand{\deg}{\ifmmode^\circ\else$^\circ$\fi\ }
18: \newcommand{\water}{\hbox{H$_2$O }}
19: % \newcommand{\H2O}{\hbox{H$_2$O}}
20: % \newcommand{\h2o}{\hbox{H$_2$O}}
21: % \def\solar{\ifmmode_{\mathord\odot}\else$_{\mathord\odot}$\fi~}
22: \newcommand{\arcs}{\char'175~}
23: \newcommand{\arcsns}{\char'175}
24: % \newcommand{\arcm}{\char'023~}
25: \newcommand{\kms}{~km~s$^{-1}$}
26: \newcommand{\kmss}{~km~s$^{-1}$~}
27: % \newcommand{\app}{$\approx$~}
28: % \newcommand{\eg}{e.g.,~}
29: % \newcommand{\ie}{i.e.,~}
30: % \newcommand{\etal}{et~al. }
31: % \newcomannd{\etalc}{et~al., }
32: % \def\NH3{\hbox{NH$_3$}}
33: \newcommand{\vlsr}{\hbox{V$_{lsr}$}}
34: % \def\Blos{B$_{\rm los}$ \hskip 1pt}
35: % \def\DnuD     {\hbox{$\Delta\nu_D$}}
36: % \def\DnuB     {\hbox{$\Delta\nu_B$}}
37: % \def\ts{\thinspace}
38: % \def\wff{W\thinspace{44} }
39: % \def\wffns{W\thinspace{44}}
40: % \def\wte{W\thinspace{28} }
41: % \def\wtens{W\thinspace{28}}
42: % \def\icfft{IC\thinspace{443} }
43: % \def\icfftns{IC\thinspace{443}}
44: % \def\j18{J1801$-$231}
45: \def\i16{IRAS~16342}
46: \begin{document}
47: 
48: \title{The Motion of Water Masers in the Pre-Planetary Nebula
49: IRAS~16342$-$3814}
50:  
51:  
52: \author{M. J Claussen\altaffilmark{1},\,\,R. Sahai\altaffilmark{2},\,\, 
53: and M. R. Morris\altaffilmark{3}}
54: \altaffiltext{1}{National Radio Astronomy Observatory (NRAO)
55: Array Operations Center, P.O. Box O, Socorro, NM 87801, USA;
56: mclausse@nrao.edu}
57: \altaffiltext{2}{Jet Propulsion Laboratory, MS183-900, California Institute
58: of 
59: Technology, Pasadena, CA 91109}
60: \altaffiltext{3}{Division of Astronomy and Astrophysics, University of
61: California
62: at Los Angeles, 405 Hillgard Avenue, Los Angeles, CA 90095}
63: \newcount\h \newcount\m \newcount\t
64: \h=\the\time \m=\h \divide\h by 60
65: \t=\h \multiply\t by 60 \advance\m by-\t
66: \t=\m \divide\m by 10 \multiply\m by 10 \advance\t by -\m \divide\m by 10
67: 
68: 
69: % \date{\ifDRAFT \fbox{{\em DRAFT}} at \number\h:\number\m\number\t\
70: %\fi\today}
71: 
72: \begin{abstract}
73: 
74: We present high angular resolution observations, using the 
75: Very Long Baseline Array (VLBA) of the NRAO, of the high-velocity
76: water masers toward the ``water-fountain'' pre-planetary
77: nebula, IRAS~16342-3814.  The detailed structure of the water masers
78: appears to be that of bow shocks on either side of a highly collimated
79: jet.  The proper motions of the water masers are approximately equal
80: to the radial velocities; the three-dimensional velocities
81: are approximately $\pm$180\kms, which leads to a very short dynamical
82: time-scale of $\sim$ 100 years.  Although we do not find direct 
83: evidence for precession of the fast collimated jet, there may be
84: indirect evidence for such precession.
85: 
86: \end{abstract}
87: 
88: \keywords{masers ---  }
89: 
90: \section{Introduction}
91: It is well accepted that asymptotic giant branch (AGB) stars evolve into
92: planetary nebulae (PN), but the mechanisms and details are still
93: uncertain. What seems to be clear is that the period of transition, namely
94: the pre-planetary nebula (PPN) phase, is probably rather short, 
95: %({\bf sahai -- i am not so sure that this is fully supported by
96: %observations -- there are at least a few PPNs which appear to ve bery
97: %long lived, like Frosty})
98: and thus observational evidence of what transpires during this interesting
99: phase is somewhat lacking. The development of a bipolar reflection nebula
100: --- two opposing lobes, usually with an equatorial ``waist'' of high
101: extinction --- around a star which has probably evolved off the AGB may be
102: a common phase leading to the establishment of asymmetric planetary
103: nebulae (Sahai et al. 2007). The mechanism for how this bipolarity is 
104: established during this evolutionary phase is still an ongoing subject 
105: of intense discussion (e.g., Balick \& Frank 2002). Mechanisms which have 
106: been considered include: (1) shaping of the outflowing wind by the 
107: gravitational field of a binary companion (Morris 1981, 1987; Mastrodemos 
108: \& Morris 1999), (2) dynamical forcing by a strong, stellar magnetic field 
109: (Garcia-Segura et al. 2005), or (3) sculpting of a prior, more spherically 
110: symmetric outflow by wandering and/or episodic jets (Sahai \& Trauger 1998; 
111: Soker 2002). Some combination of these mechanisms (e.g., Soker \& Rappaport 
112: 2000) may be necessary for a complete accounting of bipolarity in particular, 
113: and asphericity in general, but one of the strongest constraints on the
114: mechanism responsible is set by the timing of its observable onset. In a
115: handful of sources, bipolarity appears very early in the post-AGB stage of
116: evolution, and in some cases when it is thought that the central
117: star is still on the AGB (e.g., V~Hydrae --- Kahane et al. 1996, and OH~231.8+4.2
118: --- e.g. Kastner et al. 1998).
119: 
120: Molecular masers are ubiquitous in the circumstellar envelopes of 
121: oxygen-rich AGB stars; in general the distribution of masers
122: in the AGB stage traces more or less spherically symmetric shells
123: of gas, with SiO masers near the stellar photosphere, water masers
124: further out in the shell (a few tens to a hundred A.U.), main-line
125: OH masers next, and finally in the far reaches of the CSE (a few
126: hundred to 1000 A.U.), the 1612 MHz OH masers. The total velocity spread of
127: these masers is set by the expansion speed of the AGB ejecta, which
128: typically lies in the 5---20\kms~range. 
129: 
130: A particularly interesting sub-class of PPNs, is the group of so-called
131: ``water-fountain'' nebulae,  whose distinguishing characteristic is the
132: presence of very high velocity red- and blue-shifted \water and OH maser
133: % Morris text 5 BEGIN
134: % The Creel et al. 2007 conference paper, which, amazingly, didn't get
135: % published according to Mark C. (amazing because it was essentially
136: % finished) convinced me that 19190 is a confirmed member, and the
137: % Boboltz & Marvel work is convincing on OH12.8-0.9, so I list 5 confirmed
138: % members.  I left the Creel et al. reference in here.. how shall it be listed?
139: features, with velocity separations in the range 50 --- 150\kms.  When examined 
140: with sufficient spatial resolution (using Very Long Baseline Interferometry ---
141: VLBI,  or the VLA), the
142: red and blue-shifted water masers in these sources are typically displaced 
143: from each other and show large proper motions, so the three dimensional (3-D) separation
144: velocities of the "water fountain" sources thus exceeds 80\kms, and, as
145: we show here, can be as high as 370\kms.  The compact size and high
146: velocity imply lifetimes on the order of 50 ---- 100 years, so it is not surprising
147: that these objects are relatively rare.
148: At the moment there are five confirmed members of this
149: type of object: IRAS~16342-3814, IRAS~19134+2131, W~43A (Likkel,
150: Morris, \& Maddalena 1992, hereafter LMM92), OH12.8-0.9 (Boboltz \& Marvel 2005; 2007),
151: and IRAS~19190+1102 (Likkel 1989; Claussen et al. 2009); although a few 
152: other candidates for inclusion in this class have been identified:  
153: IRAS~16552-3050 (Suarez et al. 2007), IRAS~18043-2116
154: (Deacon et al. 2007), and IRAS~18460-0151 (Deguchi et al. 2007).
155: % Deguchi et al. ApJ 664, 1130 ;   Suarez, Gomez, \& Morata 2007, A\&A 467, 1085
156: % Note to us:  I found the objects 15445-5449 (Deacon et al 2007) and 
157: % 18286-0959 (Deguchi et al.) to be very interesting, because they 
158: % have a forest of water features spread over a very broad range.   
159: % I hesitate to include them in the listing of water fountain sources because
160: % they seem rather different, but I would love to see VLBA images of these
161: % objects.  If you or somebody else is not already doing that, we should
162: % consider a proposal.  
163: % Morris text 5 END
164: 
165: In the water-fountain nebulae, the water masers appear to have been re-born
166: in a fast outflow. The most extreme of the water fountain nebulae is
167: IRAS~16342-3814 (hereafter \i16). This interesting source has water masers
168: spread over a range of radial velocities encompassing 270\kms, the largest
169: spread of velocities known in the handful of water-fountain nebulae. In the
170: discovery single-dish spectra (Likkel \& Morris 1988) and subsequent single-dish observations
171: (LMM92), the spectra show two main groups of
172: masers: one at radial LSR velocities from approximately 150 to 180\kms,
173: and one at radial LSR velocities from $-$90 to $-60$\kms. In the LMM92 multi-epoch
174: study, no water maser emission was found in the
175: intervening range of the central velocities, i.e. approximately $-$60 to $+$150\kms.
176: LMM92 suggested that the water masers appeared to be reflection symmetric about a
177: central velocity and analyzed red/blue pairs of features (e.g. $\sim +$170, $-$85; $\sim +$
178: 145, $-$55\kms) which were thought to be 
179: symmetrically situated  about a single centroid velocity.  In their analysis, they
180: found that the central velocity was +43.2\kmss with respect to the LSR.  This central
181: velocity is comparable to that estimated from OH observations by Sahai et al. (1999).
182: LMM92 also suggested that the members of the red/blue pairs presumably arise on opposite
183: sides and at similar distances from the star.
184: 
185: \i16 has been studied in some detail in the optical and near-infrared, and
186: in OH maser emission. Hubble Space Telescope (HST) and Keck Adaptive Optics
187: (AO) images of \i16 show a small bipolar nebula, with the lobes separated
188: by a dark equatorial waist (Sahai et al. 1999: STMZL99; Sahai et al. 2005).  The
189: image morphology implies that the lobes are bubble-like reflection nebulae
190: illuminated by starlight escaping through polar holes in a dense, dusty
191: waist obscuring the central star. The AO observations reveal a remarkable
192: corkscrew-shaped structure apparently etched into the lobe walls, which is
193: inferred to be the signature of an underlying precessing jet -- this jet
194: has presumably carved out the observed bipolar cavities in a surrounding
195: AGB mass-loss envelope (Sahai et al. 2005). VLA maps of the OH maser
196: emission show features with the largest red- and blue-shifted velocities 
197: concentrated around the bright eastern and western polar lobes, respectively, 
198: whereas intermediate-velocity features generally occur at low latitudes, in the
199: dark waist region.
200: 
201: 
202: In this paper, we present a high angular resolution study of the water maser 
203: emission in \i16 with
204: the VLA and VLBA. VLBA studies of the water masers in the other two water
205: fountain PPNs, IRAS~19134+2131 and W~43A, have been carried out in the
206: past, yielding detailed information about the jets which are believed to
207: be responsible for producing the masers (Imai et al. 2002, 2004, 2007). However,
208: in \i16 and IRAS~19134+2131 we have the additional advantage of knowing the morphology
209: and orientation of the bipolar nebula. The goal of our study is to locate
210: the water maser features, which represent the fastest moving material
211: observed in this object, relative to the optical nebula, and to measure their
212: proper motions. Preliminary results from this study have been
213: presented by Sahai, Claussen, \& Morris (2003), Morris, Sahai, \&
214: Claussen, (2003), and Claussen, Sahai, \& Morris (2004).
215: 
216: \section{Observations and Data Reduction}
217: 
218: The observations that we describe here were performed with the
219: Very Long Baseline Array (VLBA) and the Very Large Array (VLA)
220: of the National Radio Astronomy Observatory.
221: The VLBA is a radio interferometer using ten 25-meter antennas (spread
222: across the continental United States, with two island locations
223: on Hawaii and the U.S. Virgin Islands (see Napier et al.  1994 for 
224: a full description of the VLBA).  The VLBA makes use of VLBI techniques; 
225: the VLA is a phase-stable, connected element interferometer, consisting 
226: of 27 25-m antennas in a Y-like configuration on the Plains of San Agustin 
227: in west-central New Mexico (reference).
228: 
229: \subsection{VLBA Observations}
230: 
231: We observed the water masers toward IRAS~16342 at a wavelength of 1.3 cm 
232: (rest frequency 22235.08 MHz) using the
233: VLBA at six different epochs:  2002 February 03, 2002 March 03, 
234: 2002 April 04, 2002 May 05, 2002 June 03, and 2002 July 03.  
235: All six observations were performed in the same manner, using
236: two sets of 8 MHz bandwidth (each set with two circular polarizations).  
237: One 8 MHz set was centered at $+$155.0\kmss with respect to the local standard
238: of rest (LSR), and the other set was centered at $-$71.5\kmss \vlsr.
239: The bandwidth of 8 MHz corresponds to a total velocity range 
240: of 108\kmss at the frequency of the water maser line.  Thus 
241: the velocity coverage was from $+$209.0\kmss to $+$101\kmss and 
242: from $-$17.5\kmss to $-$125.5\kms.  Each 8 MHz bandwidth was correlated
243: with 512 spectral points across the band.  Thus each spectral
244: point had a velocity width of 0.211\kms.
245: 
246: We observed IRAS~16342 and associated calibration sources for 
247: five hours for each epoch.  The VLBA data were 
248: correlated at the VLBA correlator in Socorro, NM.  Post-processing of 
249: the correlated data was performed using the NRAO software package
250: known as Astronomical Image Processing System (AIPS; Greisen 2003).  
251: The correlated data were amplitude-calibrated by using {\it a priori} 
252: knowledge of the gains and system temperatures of the receiving 
253: systems / antennas provided by NRAO.  Residual delays were measured 
254: using the strong continuum source J1733$-$1304, which was observed 
255: for this purpose for several minutes approximately once per hour 
256: throughout each observing run. The amplitude response across the 
257: bandpass was calibrated, also using the observations of J1733$-$1304.  
258: After delay calibration, global fitting of the fringe rates of the
259: strong maser channels at the peak of the two maser groups was performed.
260: Applying the fitted solution has the effect of placing each of the 
261: peaks at the phase center of the subsequent images, i.e. position
262: information
263: is lost at the imaging stage.  Additionally, we applied the fringe-rate
264: solution of the blue-shifted peak to the red-shifted masers before mapping,
265: and vice-versa.  This has the effect of finding the relative position
266: between
267: the peak emission in the blue-shifted maser and the red-shifted masers
268: (and vice-versa).  The {\it relative} positional accuracy obtained in
269: this way is conservatively estimated to be 100 $\mu$as (micro-arcseconds).
270: % The global fit to the fringe rate contains 
271: For the July 2002 VLBA observations, the masers were too weak for fringe-fit
272: solutions to be determined. A combination of poor weather at several stations
273: (causing the noise level per baseline be higher) and the decline in peak flux 
274: density of the strongest maser features was responsible for this deficiency.
275: This epoch of VLBA observations is excluded from the discussion in the rest 
276: of this paper. 
277: 
278: \subsection{VLA Observations}
279: At each epoch we also observed the water masers with the Very
280: Large Array, ostensibly for the purpose of determining an 
281: accurate position of the masers. The goal was to provide
282: a reference for the VLBA observations of the masers in order
283: to provide accurate absolute astrometry.  Because the VLA correlator
284: is more limited in its capabilities (e.g number of spectral channels
285: and bandwidth) as compared to the VLBA correlator, the
286: VLA  observations were made in a mode that allowed two velocity
287: settings simultaneously in different polarizations.  The bandwidth
288: for each of the velocity settings was 3.125 MHz, yielding a 
289: velocity coverage for each bandwidth of about 42\kms.
290: In order to cover all the expected maser emission, two sets of two pairs
291: of velocity settings were interleaved during each set of VLA
292: observations.  The radial velocities of the centers of the VLA
293: bands were $+$172.0\kms, $+$155.0\kms, $-$55.0\kms, and $-$88.0\kmss
294: \vlsr.  Notice that for the $-$55.0\kmss band, the range of velocities covered
295: is from $-$34.0 to $-$76.0\kms, excluding the emission seen with the VLBA
296: at $-$25 to $-33$\kms (Section 3.2).  For the first three epochs the VLA
297: was used in a ``normal'' calibrator switching mode, moving between
298: the target and a calibrator source; spending approximately 2 minutes
299: on the calibrator and 10 minutes on the target.  For the final three
300: epochs we employed the so-called ``fast-switching'' mode, spending
301: approximately 70 seconds on the calibrator for every 140 seconds on
302: the target.  For the May and June 2002 observations, an error in setting
303: the frequency for the central velocity of the $-55$\kmss band was made, which
304: resulted in a shift of the central velocity in this band to higher negative
305: velocities.
306: IRAS~16342 is at a very low southerly declination ($\sim-$38.3$\arcdeg$)
307: and thus only culminates at an elevation of 20$\arcdeg$.  Due to VLA
308: scheduling, we didn't always observe IRAS~16342 near transit.  We have
309: made no corrections to the data for atmospheric opacity.  
310: 
311: Images were made of each spectral channel for each epoch, in order to measure 
312: the positions and strengths of the water maser features.  Although the rms
313: noise per channel varied over the 6 months of observations due to differing
314: weather conditions, all the observations had at least an rms noise per channel
315: of 20 mJy/beam.
316: 
317: % The central velocities and the bandwidths of the VLA and VLBA
318: % setups were selected based on velocity coverage and spectral
319: % resolution, and the fact that no water maser emission has
320: % ever been seen toward IRAS~16342 in the velocity region (LMM92).
321: 
322: \section{Results}
323: 
324: \subsection{VLA Water Masers}
325: We have detected water masers with the VLA for each observing session
326: except for the July 2002 observation, when the weather and the phase
327: stability of the VLA was quite poor (and the masers were likely quite
328: weak).  Figures 1 and 2 show spectra, made from the VLA data,  of the red-shifted
329: and blue-shifted water masers toward IRAS~16342 for five of the six months.  The
330: systemic (stellar) velocity of IRAS~16342 is +42$\pm$2\kms, based on LMM92's analysis and
331: observations of the OH masers (Sahai et al. 1999).  Note that the 
332: peak maser flux densities vary over the period of our observations, increasing until May 
333: 2002, and then decreasing afterward.  For each epoch, the positions of the
334: red-shifted and the blue-shifted masers were measured.  
335: % Table 1 shows
336: % the results of these measurements. 
337: Three velocity complexes are apparent from
338: the VLA observations:  1) $+$150$\rightarrow$ $+$160\kms; 2) $+$166$\rightarrow$ 
339: $+$186\kms; and 3) $-$70 $\rightarrow$ $-$60\kms.  We did not detect any
340: masers in the velocity region near $+$145\kms, nor in the region near $-$55\kms.
341: These velocity ranges {\it did} have detectable masers in the study of LMM92; thus
342: significant changes have taken place in the maser spectra in the intervening 13 years.
343: 
344: \subsection{VLBA Water Maser Images}
345: We detected water masers using the VLBA for each observing session except for the
346: July 2002 observation.  Ranges of LSR velocity in which masers were detected are:
347: 1) Figure 3 shows the general distribution of the water masers  for the epoch
348: of February 03, 2002 relative to the optical bipolar nebula and the OH maser emission
349: mapped with the VLA (from STMZL99). Since the absolute astrometry of the optical
350: nebula is only accurate to about 0\farcs5 or so, accurate registration of the VLBA
351: \water maser features to the optical image cannot be done in an absolute sense. We
352: have therefore assumed that the kinematic center of symmetry of the \water maser
353: features, C(\water), coincides with the center of symmetry of the optical nebula,
354: C(opt). C(\water) is determined by taking the average of the positions of the
355: symmetric velocity-pairs\footnote{as defined by LMM92, i.e. symmetric about the systemic
356: velocity} at -65\kms and 150\kmss \vlsr. 
357: There are four velocity groups of masers found in the February 2002 data as shown in 
358: Table 2.  The flux-weighted LSR velocities of these four groups are $-$66.0, $-$29.5, 
359: $+$178.7, and $+$153.0\kms; we 
360: will refer to these groups by their flux-weighted velocities in what follows.   
361: The maps of the masers are similar for all 5 ``good" epochs (i.e.,
362: excluding the July 2002 epoch).  In only one epoch, that of May 2002, is one group
363: (the $-$29.5\kmss group) missing.  It is clear that the groups 
364: of extreme velocity masers (both red and blue-shifted) are situated  
365: just outside the optical lobes. The $-$29.5\kmss group 
366: is located in the optical lobe region, near the most
367: blue-shifted OH masers. The extreme blue-shifted water maser regions are offset slightly
368: %Morris alteration text 1 BEGIN
369: south of the lobe tip, that is, in a clockwise direction from the axis joining the central axes
370: of the two lobes, whereas the two groups of red-shifted water maser regions are 
371: offset slightly north of the lobe tip, again, clockwise from the same
372: axis.  Thus, the geometrical location of these regions fits in well
373: %Morris alteration text 1 END
374: with the overall point-symmetric geometry of the optical lobes (classified as Bcw,ps(s)
375: in the morphological scheme of Sahai et al. 2007), and is consistent
376: with the idea that a precessing jet is responsible for producing the morphology of
377: this object as well as the \water masers. 
378: 
379: While Figure 3 shows the large-scale distribution of masers relative to the 
380: optical nebula, the angular resolution of the VLBA allows us to study the 
381: fine structure of the masers in several groups or clusters of maser sources, 
382: corresponding to the velocity features in the single-dish or VLA spectra.
383: Figures 4 - 7 show the distribution of masers on scales of tens of milliarcseconds.
384: In constructing these figures, we have plotted the positions of maser spots in
385: every frequency channel, thus, in effect, ignoring the velocity width of the
386: maser lines; i.e., a given maser has some intrinsic width in velocity --- since
387: the lines are well-resolved in velocity, plotting the position from every channel
388: is somewhat redundant.  The origin of these figures is the center of the line joining
389: the extreme velocity ``pair'': the $-$65.2\kmss and $+$154.2\kmss (with respect to the
390: LSR) water maser features (see below).
391:  
392: %Sahai addition text 1 BEGIN (with some Morris alterations)
393: The $-$66 and $+$178.7\kms~features are spread along arc-like structures that are almost perpendicular
394: to the vector joining the red- and blue-shifted features.  Their 15 or 20 degree tilt with
395: respect to
396: the perpendicular is in the same sense on both sides, and is consistent with the overall point 
397: reflection symmetry of the nebula. The chords joining the end-points of the
398: largest such arcs are about 19 mas in extent. The arcs are slightly curved such that their center
399: of curvature lies towards the center of the nebula.  In contrast to these arc structures, the
400: 153\kms~features for each epoch are clumped closely together (within a $\sim$4 mas region). The
401: $-$29.5\kms~features do not show any distinct pattern.  There is no obvious systematic velocity
402: change along the arcs.
403: %do not interpret the arcs in terms of planar shocks here, that should be done later
404: %Sahai addition text 1 END
405: 
406: Although we are unable to measure absolute positions, we can 
407: estimate relative proper motions of the masers with respect to one maser
408: feature. While not as satisfying as measuring absolute proper motions, an
409: advantage to measuring relative proper motions is that motions due to
410: parallax and motions that are in common with all the masers (secular
411: motion of the system, for example) are removed. Motions of water masers
412: toward young stellar objects have been tracked by this method (Claussen et al. 1998).
413: In this paper we take a slightly different tack. We  measure the
414: length and the position angle of a line connecting two maser spots for all five of the epochs.
415: For this measurement, we have selected the maser spots at the extreme ends of 
416: the maser distributions; viz. the maser spot with LSR velocity of $+$154.2\kmss at
417: the north-east extreme, and the maser spot with LSR velocity of $-$65.2\kmss at
418: the south-west extreme.  The length of this line changes monotonically with time
419: over the course of these observations, while the position angle remains constant.
420: The accuracy to which we measure the line length is about 200 $\mu$as, depending
421: slightly on epoch.  The position angle of this line is 66.1$\arcdeg$ (measured east from 
422: north) and is measured to an accuracy of better than 0.1$\arcdeg$.  Table 1 summarizes
423: this measurement for the five epochs.  The change in the length of this line, of course,
424: is due to maser proper motions at both ends of the line, but we cannot tell with the current
425: data how much proper motion to assign to either end of the line.  Thus we simply divide
426: the proper motion into two halves and assign one-half to each group of masers, following
427: the kinematical symmetry implied by the radial velocities. Since we
428: accurately know the relative positions of all the rest of the masers with respect to
429: either of the extreme ends of the distribution, assigning the proper motions this way
430: also fixes the motions of all the rest of the masers. 
431: 
432: % 2002 February 03:  2972.6  0.2  66.1  0.05
433: % 2002 March    03:  2974.3  0.2  66.1  0.05    1.7
434: % 2002 April    04:  2976.1  0.3  66.1  0.06    1.8
435: % 2002 May      05:  2978.5  0.1  66.1  0.05    2.4
436: % 2002 June     03:  2980.0  0.3  66.1  0.07    1.5
437: %                                               7.4 / 4 = 1.85 / month
438: 
439: %({\bf probably want a table which shows the detected velocity regions for both vla
440: %and vlba })
441: 
442: % vlba:  february: -62.2 -->  -68.3  =>  -65.95 weighted
443: %                  -20.5      -33.4      -29.52
444: %                 +184.3     +170.6     +178.70 
445: %                 +155.2     +150.2     +153.00 
446: 
447: Our analysis here and below is based on the \water maser features that form a symmetric
448: velocity pair about the $+$42\kmss centroid velocity: the $-$66.0 and $+$153.0\kmss groups.  
449: The four ``separation" proper-motion vectors (i.e., the total change in the length of the 
450: line described above) which we derive from our five consecutive epochs of good data
451: for these features, with amplitudes 1.7, 1.8, 2.4, and 1.5 mas, are consistent with
452: the above interpretation. In fact, it is interesting that we do not see larger
453: fluctuations in the proper motion over time, suggesting that the distribution of
454: density and temperature in the jet head are maintained over at least several month
455: time-scales.  A linear least-squares fit to these amplitudes with time gives a formal
456: estimate of 63$\pm$2 $\mu$as day$^{-1}$ (see Figure 8).
457: 
458: We henceforth assume that $\mu=63\,\, \mu$as day$^{-1}$ (or 23.0 mas yr$^{-1}$ for the proper motion
459: of the ``separation vector'') represents the proper motion of these two groups of \water masers 
460: due to their physical 3-D motion.  Apportioning one-half of the proper 
461: motion to each group of masers at the ends of the line results in a tangential velocity for each group 
462: (i.e., the component of 3-D velocity in the sky-plane), $V_t=\mu d=109.8$\kms, where 
463: $d$ is the distance (assumed to be 2\,kpc from STMZL99).  Since the line-of-sight velocity is
464: $V_r=109.8$\,\kms, we obtain an inclination angle, $i=tan^{-1}(V_t/V_r)=45.0\arcdeg$
465: (where $i$ is the inclination of the nebular axis to the line-of-sight), which is in
466: excellent agreement with the independent estimate by STMZL99, who derived
467: $i\sim40\arcdeg$ (by assuming that the observed velocity spread in the 1612 MHz
468: emission features clustered around the base of each lobe is comparable to the
469: difference in projected radial velocities of the front and back sides of the lobe).
470: The total 3-D speed (one-sided) of material in the jet head, based on the $-$66.0\kmss and $+$153.0\kmss
471: \vlsr velocity pair, is thus 155.3\,\kms. This is very similar to the 3-D jet speeds
472: inferred for the other water fountain PPNs, W43A (150\kms: Imai et al. 2002) and
473: IRAS19134 (130\kms: Imai et al. 2004).
474: 
475: We can also perform a similar analysis relating the $+$178.7 group of masers relative
476: to the $-$66.0 group (though these two groups do {\it not} form a symmetric pair). 
477: The four ``separation'' proper-motion vectors between the $+$178.7 and
478: the $-$66.0\kmss groups have amplitudes 2.1, 1.7, 1.7, and 2.9 mas.  A linear least-square
479: fit to these amplitudes over time gives 67$\pm$5 $\mu$as day$^{-1}$, slightly higher than
480: that derived from the $-$66.0 and $153.0$ groups.  Thus the proper motion 
481: of the $+$178.7 group appears to be somewhat higher than that of the $+$153.0 velocity group.  
482: This value gives 13.0 mas yr$^{-1}$ for the proper motion of the $+$178.7 velocity group 
483: (assuming that the $-$66.0 group has a proper motion of 31.5 $\mu$as day$^{-1}$).  Thus the tangential 
484: velocity, the inclination angle, and the total 3-D speed of material in the $+$178.7\kms 
485: velocity group are 134.2\kms, 42.8$\arcdeg$, and 182.8\kms, respectively.   It is interesting
486: to note that the difference in position between the April, May, and June observations is not
487: very uniform; we suggest that this is due to a clumpy medium through which the jet is traveling.
488: 
489: We can derive a distance-independent age for the jet in \i16 by dividing the
490: projected (on the sky-plane) separation of the masers by the expansion proper motion. 
491: For the $-$66.0 and $+$153.0\kmss pair, the age estimate is $\sim$130\,yr, assuming that the average
492: proper motion has remained constant.  For the $+$178.7\kmss group the age estimate is $\sim$107\,yr.
493: However, since the speed of the underlying jet which is pushing and compressing the \water 
494: emitting material must be greater than the speed of the latter, these ages are upper limits. 
495: Making a similar assumption of constant proper motion, the ages of the jets in W43A and
496: IRAS19134, 
497: respectively, are 35 and 50 yr (Imai et al. 2004), significantly smaller than that in \i16. This
498: may explain why only \i16, amongst the three water-fountain PPNs, has a well
499: developed optical nebula.
500: 
501: Comparing the velocities of the maser features from our data set with the data sets
502: presented by LMM92, we find a systematic increase in the velocity offset of the
503: blue- and red-shifted features from the systemic velocity over the time period from
504: 1987/89 to 2002 that are in common in the two data sets. In LMM92's data, there are 
505: three sets of closely spaced velocity (see Table 2)
506: features (velocity complexes) which have counterparts in our data. In the LMM92 data,
507: these are the $-$60.5, the $+$147.7, and the $+$172.7\kmss groups, which are matched
508: in our data by the $-$66.0, $+$153.0, and $+$178.7\kmss velocity groups.
509: We have computed the peak-intensity-weighted average velocities of these
510: complexes, both from the LMM92 data and ours; in comparing these we
511: find that the outflow speed of
512: the maser features has increased roughly by about 5-6\kms~over a period of about
513: $\sim$13 --- 14 years (we use only the 1988 and 1989 data from LMM92). Thus, during this period, 
514: momentum has been added to the \water maser
515: emitting material by the jet, implying that the jet in \i16 has remained active. The
516: increase in the outflow speed may be due to an increase in the momentum of the jet,
517: or to a decrease in the resistive force of the ambient circumstellar material into which the jet is
518: presumably expanding.  An estimate of the average acceleration is $\sim$0.4\kmss yr$^{-1}$.  This
519: compares favorably with the three-dimensional acceleration of water masers in OH12.8$-$0.9
520: determined by Boboltz \& Marvel (2007).
521: 
522: We have looked for a similar effect in W43A and IRAS19134 by comparing the maser
523: velocities in LMM92 with those published by Imai et al. (2002, 2004); unlike \i16, we
524: find no evidence for a systematic change in the outflow
525: velocities for these sources.
526: 
527: 
528: \section{Discussion}
529: 
530: Proper motion of the \water masers can result from either actual physical motion of
531: % Morris mod text 2 BEGIN
532: the emitting material, or to a ``theatre marquis" phenomenon, in which dense
533: material at progressively larger distances ``lights up" sequentially in \water maser
534: emission due to, for example, the passage of a shock wave that produces an outward moving
535: compression front. 
536: % Morris mod text 2 END
537: Numerical simulations of fast collimated jets expanding in AGB
538: circumstellar envelopes (Lee \& Sahai 2003) show the presence of dense material at
539: the head of the jet, where the physical conditions for the excitation of \water
540: masers (temperature $\sim$ 400\,K, density $\sim 10^9$ cm$^{-3}$: Elitzur 1992) are
541: likely to be maintained. The forward (bow) shock along the jet axis is located within
542: this dense region at the head of the jet. 
543: %Sahai modification text 2 BEGIN (with some Morris edits)
544: The bulk of the gas in this region is material from the slowly-expanding 
545: ambient circumstellar envelope that has been swept-up, compressed, and accelerated 
546: by a fast, tenuous (i.e., underdense with respect to the ambient medium) jet-like 
547: outflow.  Hence we expect that, on the average, the proper motions of the \water 
548: masers reflects the physical motion of the dense gas in the jet head, and this motion 
549: will generally be at speeds less than the intrinsic speed of the underlying jet. 
550: The ratio of the speed of the dense region to that of the jet depends on the
551: density contrast between the jet and the ambient medium (see, e.g., Eqn. 9 of 
552: Lee \& Sahai 2003). 
553: %Sahai modification text 2 END
554: Fluctuations from the average motion may arise as the instantaneous
555: %actual elemental
556: %MM: maybe replace "actual elemental" with "instantaneous"  ?
557: masing region, whose size is small compared to the size of the dense region in the
558: jet head, may be located at different positions within the latter at different
559: times. 
560: %Morris addition text 3 BEGIN
561: If, in addition, the jet is precessing, it impacts material in progressively different
562: directions, and one might observe the projection of this progression (another 
563: potential manifestation of the theatre marquis effect), which, in general, would 
564: consist of both radial and tangential components.  We do not observe any deviations 
565: from purely radial motion (i.e., motion parallel to the jet) of the maser spots that 
566: we have followed, so any precession of the jet that might be occurring is apparently 
567: not  affecting those spots.  Rather, precession of the jet may simply produce new 
568: maser spots in the direction of the precession.  This might account for the disappearance 
569: of maser features at the south end of the $-$66\kmss arc and their appearance at the
570: north end of this arc (Fig. 4).
571: 
572: The maser spot evolution diagrams shown in Figures 4 and 6 show a 
573: potential additional phenomenon: a slight divergence of the spot motions,   
574: relative to the average proper motion vector (shown as dashed lines in each 
575: figure).  The southernmost extreme spot group that can be followed through 
576: all epochs shows a directional deviation that corresponds to such a divergence.  
577: The significance of this phenomenon is marginal, but if it is found to 
578: recur in future observations, it could provide an interesting clue to the 
579: curvature-related dynamics of the shock front.
580: %Morris addition text 3 END
581: 
582: The distribution of \water maser features which we have found in \i16 is
583: generally tangential to the inferred jet axis, in contrast to that in
584: W43A, where the features appear to lie along the curved trajectory of
585: material ejected in a precessing jet. These differences may be related to
586: differences in the intrinsic properties of the jets or the ambient
587: circumstellar environment provided by the slowly expanding mass-loss
588: envelopes formed during the progenitor AGB phase, with which the fast jet
589: must presumably interact. The IRAS far-infrared fluxes from these objects
590: are a probe of the dense, dusty circumstellar environment in these
591: objects. Both objects are believed to lie at comparable distances
592: (2.6\,kpc for W43A and 2\,kpc for \i16), and their 12 and 25 \micron~fluxes
593: are rather similar (23.7 and 103.5\,Jy in W43A; 16.2 and 199.8\,Jy in
594: \i16) hence there is no obvious reason to believe that their circumstellar
595: environments are very different.
596: 
597: %Sahai modification to text within this para on 4/18/08
598: The lobes seen in the near-infrared AO images are more extended radially than the optical
599: lobes, and, because of the smaller optical depths in the near-IR, more accurately
600: delineate the true physical extent of the bipolar cavities which which are manifested as
601: the lobes in the images. The extreme \water masers, separated by 2.978\arcsns, lie near 
602: (but not exactly) at the tips of the near-IR lobes, thus near the ends of the bipolar 
603: cavities. The point-symmetrically distributed offsets of these features from the tips 
604: of the infrared lobes is to be noted, and suggests that the current working-surfaces of 
605: the bipolar jet have moved away from the physical ends of these elongated cavities, 
606: providing further support for the idea that the cavities are being carved out
607: by a precessing bipolar jet.
608: 
609: %some Sahai modifications in this para: e.g. L band is 3.8 micron, 
610: % expected proper motion: 6 ==> 7 mas/yr
611: The $-$29.5\kms~masers do not show obvious evidence for systematic proper motion. 
612: In contrast to the extreme velocity masers, these are located near the most 
613: blue-shifted OH masers in the western optical lobe. A comparison with the 
614: $L_p$ (3.8 \micron) image in Sahai et al. (2005), shows that this
615: maser complex coincides with the bright region labeled knot W2,
616: apparently part of the corkscrew structure. The associated \water emitting
617: material clearly does not show any systematic motion like the extreme
618: velocity masers (Fig. 7). An inspection of the jet-envelope interaction
619: models of Lee \& Sahai (2003) shows that lower outflow velocities are
620: generally expected for the dense swept-up material at locations on the
621: walls of the cavities which are at lower latitudes, compared to the
622: material at the tips. Thus both the proper motion and the radial
623: velocity of this material is expected to be less than that of the
624: material at the tips of the lobes because its intrinsic, 3-dimensional
625: outflow speed is less than that of the latter. However, the expected
626: proper motion of the $-$29.5\kms, assuming the same
627: inclination angle as for the extreme velocity masers, is about 7 mas yr~$^{-1}$, 
628: so it is somewhat surprising that we do not detect this in our data. We
629: conclude that, for these masers (unlike the extreme velocity ones), the
630: location of the individual masing regions changes by distances
631: comparable to the proper motion ($\sim$1 AU) on time-scales of $\sim$1
632: month. Since maser emission requires velocity coherence, and \water maser
633: excitation is believed to be collisionally-driven, such changes imply
634: that there are much larger changes in the velocity, density,
635: temperature and structure in region W2 than at the tips of the lobes,
636: where the extreme velocity masers are produced over month-long
637: time-scales.
638: 
639: % Morris late comment BEGIN
640: % I think this point about the paired features is useful to include, and I suggest
641: % placing it before the previous point (my text 4), because that text can use the
642: % conclusion that I add below.
643: % {\bf the snapshot in time for these observations show that the ``paired''
644: % velocity features as discussed in LMM92 don't necessarily occur together
645: % all the time}
646: % include a comment that the paired velocity features imply that the jets that
647: % produce the water masers come out as a succession of perhaps stochastic,
648: % bidirectional bursts and that the directionality of the bursts may vary a little,
649: % but is very close to being constant.
650: % Morris late comment END
651: 
652: %{\bf the highest radial velocity masers (+180 \kms) are not the furthest
653: %from the center }
654: % Morris addition text 4 BEGIN; some modifications by Sahai 4/18/08
655: % this is to address this bold point -- it can be placed wherever you want:
656: The \water masers associated with the blueshifted, near lobe of the nebula, are
657: arranged spatially so that the highest velocity component is farthest from the
658: center of the system.  This could be understood simply in terms of the displacement 
659: being proportional to velocity.  However, this is not the case for the masers 
660: associated with the redshifted lobe, where the $+$153\kmss feature is located 
661: farther out than the $+$179\kmss feature.  Because all of the masers lie close 
662: to a single line joining the most distant spot groups, this is unlikely to be 
663: ascribable to purely geometrical effects in which the different velocities in 
664: either lobe result from different line-of-sight projections of jet bursts coming
665: out with a constant velocity in different directions at different times.  If the 
666: successive bidirectional jet bursts have the same initial velocities, then our 
667: data suggest that they have variable histories, with some being decelerated by 
668: direct interaction with the slow AGB wind, while others may follow the trajectories 
669: of previous spurts which may have cleared out some of the slow wind material. 
670: Of course, it is also possible that the jet bursts have variable initial velocities. 
671: 
672: % late note: I don't see any statement in the paper or in the figure captions saying
673: % what zero offset in Figs 4-7 might correspond to.  How was 0,0 chosen?  is it arbitrary?
674: % Unless it's already in there somewhere, we should state it.
675: %  Morris addition text 4 END
676: 
677: \section{Conclusions}
678: 
679: We have observed the ``water fountain'' masers from IRAS~16342$-$3814
680: with the VLBA at very high angular resolution over a total time span of 
681: four months at approximately 1 month intervals. These observations show 
682: that the detailed structure of the high radial velocity water masers
683: (making up the ``fountain'') lie on opposite sides of the optical nebula.
684: These structures appear to be bow shocks which are likely formed at the
685: impact point of a highly collimated jet with dense molecular gas found
686: along the jet axis.  Based on our VLBA and VLA observations, compared
687: with earlier single-dish observations of the water masers from \i16,
688: we find that the ``paired'' high-velocity features are not always apparent.
689: 
690: The proper motion measurements, when combined with the radial velocity
691: measurements of the water maser spots, give a three-dimensional picture
692: of the kinematics of the gas.  We calculate that the 3-D velocities of 
693: the jet at the position of the extreme velocity maser components are at
694: least 155 to 180\kms.  We also calculate that the inclination of the jet 
695: axis to the line of sight is about 44\arcdeg, in excellent agreement with
696: other estimates.  Upper limits on the dynamical age of the jet (which is
697: distance-independent) are 110 - 130 years.
698: 
699: Direct evidence for precession of the fast collimated jet is lacking in
700: our data;  the maser features do not trace a curved trajectory as they
701: do in the case of W43A.  However, there may be some indirect indication 
702: of precession from these observations, e.g. the appearance and disappearance
703: of features at the ends of the arcs, as well as the difference in position
704: between the extreme \water masers and the tips of the infrared lobes.
705: 
706: Finally, there are maser features near (but not {\it} at) the systemic velocity 
707: which show no systematic proper motion over the four months. We 
708: conclude that, for these water masers, the location of such  masing regions
709: changes in a non-systematic manner by distances similar to the proper motion 
710: (~1 AU) on one-month time scales.
711: 
712: \acknowledgments
713: 
714: RS and MM thank NASA for partially funding this work by NASA LTSA award (nos.
715: NMO710651/ 399-20-40-06); RS also received partial support for this work
716: from HST/GO awards (nos. GO-09463.01-A and GO-09801.01-A) from the Space
717: Telescope Science Institute (operated by the Association of Universities
718: for Research in Astronomy, under NASA contract NAS5-26555). RS's
719: contribution to the research described in this publication was carried
720: out at the Jet Propulsion Laboratory, California Institute of
721: Technology, under a contract with NASA.
722: 
723: \begin{thebibliography}{}
724: 
725: \bibitem[Balick \& Frank(2002)]{} Balick, B., \& Frank, 
726: A.\ 2002, \araa, 40, 439 
727:  
728: \bibitem[]{} Boboltz, D. A. \& Marvel, K. B.\ 2005, \apjl, 627, L45
729: 
730: \bibitem[]{} Boboltz, D. A. \& Marvel, K. B.\ 2007, \apj, 665, 680 
731: 
732: \bibitem[]{} Claussen, M. J, Marvel, K. B., Wootten, H. A., \& Wilking, B. A. 
733: \ 1998, \apjl, 507, L79
734: 
735: \bibitem[]{} Claussen, M., Sahai, R., \& Morris, M.\ 2004, ASP
736: Conf.~Ser.~313: Asymmetrical Planetary Nebulae III: Winds, Structure and
737: the Thunderbird, 313, 331
738: 
739: \bibitem[]{} Claussen, M. J, Pihlstrom, Y., Sahai, R., \& Morris, M. \ 2009,
740: in preparation
741: 
742: \bibitem[]{} Deacon, R. M., Chapman, J. M., Green, A. J., \& Sevenster, M. N. 
743: \ 2007, \apj, 658, 1096
744: 
745: \bibitem[]{} Deguchi, S., Nakashima, J., \& Kwok, S. \ 2007, \apj, 664, 1130
746: 
747: %\bibitem[] {gs99} Garcia-Segura, G., Langer, N., Rozyczka, M., \&
748: %Franco, J. 1999, \apj, 517, 767
749: 
750: \bibitem[]{} Elitzur, M. 1992, {\it Astronomical Masers} (Kluwer Academic, Dordrecht)
751: 
752: \bibitem[Garcia-Segura et al.(2005)] {gs05} Garcia-Segura, G., L{\'o}pez,
753: J.~A., \& Franco, J.\ 2005, \apj, 618,  919 
754: 
755: \bibitem[Greisen (2003)]{greis03} Greisen, E. W. 2003, in {Information Handling
756: in Astronomy - Historical Vistas}, Astrophysics and Space Science Library, 
757: Vol. 285, ed. A. Heck (Dordrecht: Kluwer), 109.
758:  
759: \bibitem[Imai et al.(2002)]{imai02} Imai, H., Obara, K., 
760: Diamond, P.~J., Omodaka, T., \& Sasao, T.\ 2002, \nat, 417, 829 
761:  
762: \bibitem[Imai et al.(2004)]{imai04} Imai, H., Morris, M., 
763: Sahai, R., Hachisuka, K., \& Azzollini F., J.~R.\ 2004, \aap, 420, 265 
764: 
765: \bibitem[Imai et al.(2007)]{imai07} Imai, H., Sahai, R., \& Morris, M. 
766: 2007, \apj, 669, 424
767: 
768: \bibitem[Kahane et al. (1996)]{kah96} Kahane, C., Audinos, P, Barnbaum, C,
769: \& Morris, M. \ 1996, \aap, 314, 871
770: 
771: \bibitem[Kastner et al. (1998)]{kas98} Kastner, J. H., Weintraub, D. A.,
772: Merrill, K. M. \& Gatley, I. \ 1998, \aj, 116, 1412
773: 
774: \bibitem[Likkel 1989]{lik89} Likkel, L. \ 1989, \apj, 344, 350
775: 
776: \bibitem[Likkel \& Morris 1988]{lm88} Likkel, L., \& Morris, M. 1988, ApJ,
777: 329, 914
778: 
779: \bibitem[] {lik92} Likkel, L., Morris, M., \& Maddalena, R. J. 1992, 
780: \aap, 256, 581 (LMM92)
781: 
782: \bibitem[] {mas99} Mastrodemos, N., \& Morris, M. \apj, 523, 357
783: 
784: \bibitem[] {mm81} Morris, M. 1981, \apj, 249, 572
785: 
786: \bibitem[] {mm87} Morris, M. 1987, \pasp, 99, 1115
787: 
788: \bibitem[]{} Morris, M.~R., Sahai, R., \& Claussen, M.\ 2003, Revista
789: Mexicana de Astronomia y Astrofisica Conference Series, 15, 20
790: 
791: \bibitem[]{nap94} Napier, P. J., Bagri, D. S., Clark, B. G., Rogers, 
792: A. E. E., Romney, J. D., Thompson, A. R., \& Walker, R. C. \ 1994, 
793: Proc. IEEE, 82, 658
794: 
795: \bibitem[] {rs98} Sahai, R., \& Trauger, J. T. 1998, \aj, 116, 1367
796: 
797: \bibitem[Sahai et al.(1999)]{stmzl99} Sahai, R., Te Lintel 
798: Hekkert, P., Morris, M., Zijlstra, A., \& Likkel, L.\ 1999, \apjl, 514, 
799: L115 (STMZL99) 
800:  
801: \bibitem[]{} Sahai, R., Claussen, M.~J., \& Morris, M.\ 2003, IAU
802: Symposium, 209, 519
803: 
804: \bibitem[Sahai et al.(2005)]{sah05} Sahai, R., Le Mignant, 
805: D., S{\'a}nchez Contreras, C., Campbell, R.~D., \& Chaffee, F.~H.\ 2005, 
806: \apjl, 622, L53 
807: 
808: \bibitem[Sahai et al. (2007)]{sah07} Sahai, R., Morris, M., S{\'a} Contreras, C., 
809: \& Claussen, M. J \ 2007, \aj, 134, 2200
810: 
811: \bibitem[] {sok02} Soker, N. 2002, \apj, 568, 726
812: 
813: \bibitem[] {sok00} Soker, N. \& Rappaport, S. 2000, \apj, 538, 241
814: 
815: \bibitem[]{} Suarez, O., Gomez, J. F.,  \& Morata, O. \ 2007, \aap, 467, 1085
816: 
817: \end{thebibliography}
818: 
819: \clearpage
820: \begin{figure}
821: \plotone{fig1a.eps}
822: \caption{Spectra of the red-shifted water masers observed with the
823: VLA for the five monthly epochs described in the text.  Each of these
824: spectra is comprised of two frequency settings at the VLA; the velocity
825: range of the plot is the entire coverage of the red-shifted masers observed
826: with the VLA.}
827: \label{red-spectra}
828: \end{figure}
829: \clearpage
830: 
831: \begin{figure}
832: \plotone{fig1b.eps}
833: \caption{Spectra of the blue-shifted water masers observed with the
834: VLA for the five monthly epochs described in the text.  Each of these
835: spectra is comprised of two frequency settings at the VLA; the velocity
836: range of the plot is the entire coverage of the blue-shifted masers
837: observed with the VLA.}
838: \label{blue-spectra}
839: \end{figure}
840: 
841: \begin{figure}[htb]
842: \vskip -0.9 truein
843: \plotone{i16342_hst_oh_h2o_4.eps}
844: \vskip -1.0 truein
845: \caption{
846: (a) A color image of the protoplanetary nebula IRAS16342-3814, made of exposures taken with
847: WFPC2/HST through the F814W (in red) and F555W (in green) filters, and displayed on a
848: logarithmic stretch (b) OH 1612, 1665, and 1667 MHz maser line features observed with the VLA ,
849: overlaid on the F814W image (in grey-scale). The maser features have been grouped into 6
850: color-coded velocity bins. The symbol sizes are proportional to the square-root of the
851: integrated flux, and the highest flux features have strengths of 8.83, 0.53, and 1.17 Jy
852: for the 1.612, 1.665, and 1.667 features, respectively. (c) VLBA \water 22.235 GHz features
853: observed with the VLBA, overlaid on the F814W image. Arrows representing the proper motion
854: vectors of the \water maser features are also shown; the orientation of this vector remains
855: fixed at $66\arcdeg.1$ over our monitoring period (Jan--May 2002). (d) \water maser features
856: overlaid on the $L_p$-band (3.8\micron) image, taken using Adaptive Optics on the Keck II
857: telescope (the $L_p$ image is adapted from Sahai et al. 2005).
858: }
859: \label{hst-oh-h2o}
860: \end{figure}
861: 
862: \clearpage
863: \begin{figure}
864: \plotone{fig.-65.blue.eps}
865: \caption{Zoom-in of the VLBA distribution of masers for the blue-shifted ($-$66\kmss 
866: radial velocity features) for all five epochs. This figure zooms in on
867: the masers in the extreme lower left of Figure \ref{hst-oh-h2o} (i.e. outside the
868: southwest optical lobe). The color code represents different epochs of observation:
869: blue --- 3 Feb 2002; green --- 3 Mar 2002; red --- 4 April 2002; yellow --- 5 May
870: 2002; and cyan --- 3 June 2002. All features in all spectral channels comprising
871: this range are plotted as open circles centered on their position. In the lower-left
872: hand corner is a bar showing a 10 AU extent, assuming a distance to the source of 2
873: kpc. Each of the dashed lines connects maser features at different epochs which we
874: believe arise in the same physical clump. The arrow shows a 10.8 mas yr$^{-1}$ proper
875: motion vector. The orientation of the axes here is different than that in Fig. 2.
876: }
877: \label{blue-distr}
878: \end{figure}
879: 
880: \clearpage
881: \begin{figure}
882: \plotone{fig.155.red.eps}
883: \caption{Zoom-in of the VLBA distribution of masers for the red-shifted
884: ($+$153\kmss radial velocity features) for all five epochs. This figure zooms in on the
885: masers in the extreme upper right of Figure \ref{hst-oh-h2o} (i.e. outside the
886: northeast optical lobe). The color code represents different epochs of observation:
887: blue --- 3 Feb 2002; green --- 3 Mar 2002; red --- 4 April 2002; yellow --- 5 May
888: 2002; and cyan --- 3 June 2002. All features in all spectral channels comprising
889: this range are plotted as open circles centered on their position. In the lower-left
890: hand corner is a bar showing a 10 AU extent, assuming a distance to the source of 2
891: kpc.
892: }
893: \label{red155-distr}
894: \end{figure}
895: 
896: \clearpage
897: \begin{figure}
898: \plotone{fig.180.red.eps}
899: \caption{Zoom-in of the VLBA distribution of masers for the red-shifted
900: ($+$178\kmss radial velocity features) for all five epochs. This figure zooms in on the
901: masers in the upper right of Figure \ref{hst-oh-h2o} (i.e. outside the
902: northeast optical lobe). The color code represents different epochs of observation:
903: blue --- 3 Feb 2002; green --- 3 Mar 2002; red --- 4 April 2002; yellow --- 5 May
904: 2002; and cyan --- 3 June 2002. All features in all spectral channels comprising
905: this range are plotted as open circles centered on their position. In the lower-left
906: hand corner is a bar showing a 10 AU extent, assuming a distance to the source of 2
907: kpc.  The proper motion vector here represents 14.4 mas yr${-1}$.
908: }
909: \label{red180-distr}
910: \end{figure}
911: 
912: \clearpage
913: \begin{figure}
914: \plotone{green.eps}
915: \caption{Zoom-in of the VLBA distribution of masers for the velocity range $-$21 $\rightarrow$
916: $-33$ \kmss for four epochs.  The color code represents the different epochs of observation:
917: blue --- 3 Feb 2002; green --- 3 Mar 2002; red --- 4 April 2002; 
918: and cyan --- 3 June 2002. We did not detect emission from this velocity group in May 2002.
919: All features in all spectral channels comprising this range are plotted as open circles centered 
920: on their position.   Note that there is no apparent systematic motions of these masers as
921: compared with the those at other radial velocities in Figures 4, 5, \& 6.
922: In the lower-left
923: hand corner is a bar showing a 20 AU extent, assuming a distance to the source of 2
924:  kpc.
925: }
926: \label{green-33-distr}
927: \end{figure}
928: 
929: \clearpage
930: \begin{figure}
931: \plotone{fig8.pm.eps}
932: \caption { A plot of the length of the line joining the $+$154.2\kmss feature
933: and the $-$65.2\kmss feature as a function of time for 5 VLBA epochs.  The line
934: is a linear least-squares fit to the data, giving a slope of 63$\pm$2 $\mu$as
935: day$^{-1}$. }
936: \label{pm-linelength}
937: \end{figure}
938: 
939: \begin{deluxetable}{ccc}
940: \tablecolumns{3}
941: % \tabletypesize{\scriptsize}
942: % \rotate
943: \tablewidth{0pc}
944: \tablecaption{Summary of VLBA Observations}
945: \tablehead{
946: \colhead{Date} & \colhead{Length of Line from} & \colhead{Position Angle (N$\rightarrow$E)} \\
947:  & \colhead{$-$65.2 and $154.2$\kmss Features (mas)} & \colhead{of this Line (degrees)} \\
948: }
949: \startdata
950: 2002 February 03 & 2972.6$\pm$0.2 & 66.1$\pm$0.04  \\
951: 2002 March 03    & 2974.3$\pm$0.2 & 66.1$\pm$0.04  \\
952: 2002 April 04    & 2976.1$\pm$0.3 & 66.1$\pm$0.05  \\
953: 2002 May 05      & 2978.5$\pm$0.1 & 66.1$\pm$0.04  \\
954: 2002 June 03     & 2980.0$\pm$0.3 & 66.1$\pm$0.05  \\
955: \enddata
956: \end{deluxetable}
957: 
958: \clearpage
959: \begin{deluxetable}{cccc}
960: \tablecolumns{4}
961: \tablewidth{0pc}
962: \tablecaption{Nomenclature for Flux-Density Weighted Velocity Groups }
963: \tablehead{
964: \multicolumn{2}{c}{This Paper} & \multicolumn{2}{c}{Likkel et al. (1992)} \\
965: \colhead {Name of Velocity Group} & \colhead{LSR Velocity Range} 
966:  & \colhead {Name of Velocity Group} & \colhead{LSR Velocity Range}
967: }
968: \startdata
969: $-$66.0 & $-$62.2$\,\rightarrow$$\,-$68.3 & $-$60.5 & ---  \\
970: $+$155.3 & $+$150.2$\,\rightarrow$$+\,$155.2 & $+$147.7 & $+$147.6$\,\rightarrow$$\,+$147.8 \\
971: ---     & ---                          & $-$56.5  & $-$56.2$\,\rightarrow$$\,-$55.8 \\
972: $+178.7$ & $+$170.6$\,\rightarrow$$\,+$184.3 & $+$172.7 & $+$170.2$\,\rightarrow$$\,+$175.1 \\
973: $-$29.5 & $-$20.5$\,\rightarrow$$\,-$33.4 & --- & --- \\
974: ---     & ---                          & $-$87.7  & $-$86.4$\,\rightarrow$$\,-$89.0 \\
975: \enddata
976: \end{deluxetable}
977: 
978: % vlba:  february: -62.2 -->  -68.3  =>  -65.95 weighted
979: %                  -20.5      -33.4      -29.52
980: %                 +184.3     +170.6     +178.70 
981: %                 +155.2     +150.2     +153.00 
982: 
983: 
984: \end{document}
985: 
986: Given our observed proper motion, P=11.1 mas yr$^{-1}$,
987: the distance is $D=2.1$\,kpc\,($V_t$/100\kms)/($P$/10 mas
988: yr$^{-1}$)=1.7\,kpc. 
989: 
990: