1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2005 December 5
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8:
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12:
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: %%\documentclass[12pt,preprint]{aastex}
19:
20: %% manuscript produces a one-column, double-spaced document:
21:
22: %%\documentclass[manuscript]{aastex}
23:
24: %% preprint2 produces a double-column, single-spaced document:
25:
26: \documentclass[preprint2]{aastex}
27:
28: \newcommand{\bs}{\textbackslash}
29: \newcommand{\CS}[1]{\texttt{\textbackslash #1}}
30: \newcommand{\shape}{{\it Shape\ }}
31: \newcommand{\Mdot}{\dot{M}~}
32: \newcommand{\kms}{\mbox{ km s$^{-1}$}~}
33: \newcommand{\Moy}{\mbox{M$_{\odot}$ yr$^{-1}$}~}
34:
35:
36: \shorttitle{Hydrodynamical Velocity Fields in Planetary Nebulae}
37: \shortauthors{Steffen and Garc\'{\i}a-Segura}
38:
39:
40: \begin{document}
41:
42:
43: \title{Hydrodynamical Velocity Fields in Planetary Nebulae}
44:
45: \author{ Wolfgang Steffen, }%\altaffilmark{1}
46: \affil{Instituto de Astronom\'{\i}a,
47: Universidad Nacional Aut\'onoma de M\'exico,
48: Ensenada, B.C., M\'exico}
49: \email{wsteffen@astrosen.unam.mx}
50:
51:
52: \author{
53: Guillermo Garc\'{\i}a-Segura,}%\altaffilmark{1}
54: \affil{Instituto de Astronom\'{\i}a,
55: Universidad Nacional Aut\'onoma de M\'exico,
56: Ensenada, B.C., M\'exico}
57: \email{ggs@astrosen.unam.mx}
58:
59:
60: \and
61:
62: \author{
63: N. Koning}%\altaffilmark{2}
64: \affil{Space Astronomy Laboratory, University of Calgary, Canada.}
65: \email{nkoning@iras.ucalgary.ca}
66:
67:
68:
69: \begin{abstract}
70: Based on axi-symmetric hydrodynamical simulations and 3D reconstructions with {\it Shape},
71: we investigate the kinematic
72: signatures of deviations from homologous (``Hubble-type'') outflows in some typical
73: shapes of planetary nebulae. We find that, in most situations considered
74: in our simulations,
75: the deviations from a Hubble-type flow are significant and
76: observable. The deviations are systematic and a simple
77: parameterization of them considerably improves morpho-kinematical models of the
78: simulations.
79: We describe such extensions to a homologous expansion law that capture the
80: global velocity structure of hydrodynamical axi-symmetric nebulae during their
81: wind-blown phase. It is the size of the poloidal velocity component that strongly
82: influences the shape of the position velocity diagrams that are obtained,
83: not so much the variation of the radial component.
84: The deviations increase with the degree of collimation of the nebula and
85: they are stronger at intermediate latitudes.
86: We describe potential deformations
87: which these deviations might produce in 3D reconstructions
88: that assume ''Hubble-type'' outflows. The general conclusion
89: is that detailed morpho-kinematic observations and modeling of planetary nebulae
90: can reveal whether a nebula is still in a hydrodynamically active stage
91: (windy phase) or whether it has reached ballistic expansion.
92: \end{abstract}
93:
94:
95: \keywords{Planetary Nebulae, ISM: Outflows, Stars: Post-AGB, Stars: Mass loss, Hydrodynamics}
96:
97:
98: \section{Introduction}
99:
100: \label{sec:intro}
101:
102: Most methods used for the reconstruction of three dimensional
103: structures of planetary nebulae are based on imaging and kinematic
104: data. Usually, they assume homologous expansion, i.e., purely radial velocity
105: with magnitude proportional to the distance from the
106: central star (e.g. Sabbadin et al. 2006; Santander-Garc\'{\i}a et al. 2004;
107: Hajian et al. 2007; Steffen \& L\'opez 2006;
108: Morisset et al. 2008; Dobrin{\v c}i{\'c} et al. 2008).
109: The general justification for adopting this so called
110: "Hubble-type expansion" is that observed expansion speeds directly
111: correlate with distance from the center of the nebula
112: (Wilson 1950, O'Connor et al. 2000, Meaburn et al. 2008). Furthermore, the structures
113: seen in position-velocity (P-V) diagrams are often very similar to the direct
114: image
115: (e.g. Mz-3: Santander-Garc\'{\i}a 2004; or NGC6302: Meaburn et al. 2005).
116: A technical reason for adopting this type of expansion law
117: is that, it is the simplest direct mapping between
118: the velocity of a gas
119: parcel and its position along the line of sight.
120: Still, this method requires an additional constraint
121: to determine the constant of
122: proportionality that links the Doppler-velocity to the position along
123: the line of sight. This constraint may be an estimate of the extent
124: of the object along the line of sight as deduced from partial symmetries
125: or other arguments.
126:
127: The homologous expansion
128: is a reasonable approximation if the motion of the gas has become
129: ballistic only a short time after the leaving the center
130: of the nebula as compared to the current age of the nebula (for a discussion
131: of this issue see e.g. Steffen \& L\'opez 2004). Observational and theoretical evidence is mounting that
132: the hydrodynamical events that determine the structure formation in many planetary nebulae are rather short-lived (e.g. Alcolea, Neri \& Bujarrabal 2007; Meaburn et al. 2005; Dennis et al. 2008; Akashi \& Soker 2008). In order to decide whether the extended interacting wind scenario can be held up as it has been proposed almost two decades ago, it becomes important to look for kinematic evidence.
133: During the action of the interacting winds, important deviations from a Hubble-type expansion law are to be expected in the velocity magnitude and direction.
134:
135: We study the kinematics of hydrodynamical simulations as a base to look
136: for such deviations in observed nebulae. This study also aims to improve
137: three-dimensional modeling that is being
138: carried out with photo-ionization codes (e.g. Morisset 2006; Ercolano, Barlow \& Storey 2005) and morpho-kinematic reconstruction methods (e.g. Sabbadin et al. 2006; Steffen \& L\'opez 2006).
139:
140: The morpho-kinematical modeling code \shape has been designed to allow the 3D modeling
141: of very complex emissivity structures and velocity fields (Steffen \& L\'opez 2006).
142: {\it Shape}, and most current photo-ionization codes, do not actually calculate the structure
143: and velocities from physical initial conditions. It is therefore important to have a physical guidance when deviations from a homologous expansion are to be included in morpho-kinematical modeling.
144: To provide improved velocity fields we study the kinematics of some typical morphologies of planetary nebulae as created by two different mechanisms, purely hydrodynamics and magnetic confinement.
145: %We characterize the velocity fields in such a way that
146: %serve as a guidance to models of actual planetary nebulae with similar morphology
147: %as those of our simulations.
148:
149: As discussed by Li, Harrington \& Borkowski (2002) detailed kinematic models are required to improve the accuracy of distance measurements to planetary nebulae from kinematic data. As these authors stress, deviations from homologous expansion should be taken into account to improve accuracy. Our paper aims at validating the \shape code for later application to the type of data presented by Li, Harrington \& Borkowski (2002) for BD+30$^{o}$3639.
150:
151: First, we apply purely hydrodynamical numerical simulations to create elliptical and bipolar nebulae from an equatorial density enhancement in the environment. Highly elongated bipolar nebulae are simulated with magneto-hydrodynamic calculations.
152: The kinematics of the simulated data is analyzed and then reproduced as a morpho-kinematical model using {\it Shape}.
153: We assess the distortions that might occur in 3D reconstructions of such objects.
154: In addition to providing template velocity fields, the direct comparison of the simulations
155: with the reconstructed models using \shape allows us to validate the modeling
156: capabilities of \shape itself.
157:
158: In Section 2 we describe the hydrodynamical model setup.
159: Section 3 contains a description of the morpho-kinematical modeling technique and in Section 4 we discuss and summarize our results.
160:
161:
162: \section{Numerical simulations}
163: \label{sec:models}
164:
165:
166:
167: The four reference simulations have been performed using the
168: magneto-hydrodynamic code ZEUS-3D (version 3.4), developed by M. L. Norman
169: and the Laboratory for Computational Astrophysics.
170: This is a finite-difference, fully explicit,
171: Eulerian code descended from the code described in Stone \& Norman (1992).
172: A method of characteristics is used to compute magnetic fields as described
173: in Clarke (1996).
174: The models include the Raymond \& Smith (1977) cooling curve above $10^4$ K.
175: For temperatures below $10^4$ K, {\bf the radiation cooling term is set to
176: zero}.
177: Photoionization is not included for simplicity.
178: The initial and minimum temperature allowed in all models is set to $10^2$ K.
179:
180: The computations are done in spherical coordinates (r, $\theta$),
181: with outflowing outer boundary conditions and reflective conditions at the
182: pole and equator.
183: The models have grid resolutions of $200 \times 180 $ in r and $\theta$,
184: with a radial extent of 0.1 pc and 90$^\circ$, respectively.
185: The inner radial boundary is located at $2.5 \times 10^{-3}$ pc, i.e.,
186: we skip the inner $2.5 \%$ of the computational mesh.
187:
188: The winds are set based on the rotating wind solutions given by
189: Bjorkman \& Cassinelli (1993), Ignace et al. (1996)
190: and the approach to those equation given by Garc\'{\i}a-Segura et al. (1999).
191: These equations permit the introduction of the winds self-consistently.
192: The magnetic field of an outflowing wind from a rotating star
193: can be described by two components, $B_{\phi}$ and $B_{\rm r}$
194: (Chevalier \& Luo 1994; R\'o\.zyczka \& Franco 1996; Garc\'{\i}a-Segura
195: et al. 1999).
196: The radial field component can be neglected since $B_{\rm r}
197: \sim r^{-2}$ while the toroidal component $B_{\phi} \sim r^{-1}$, and the
198: field configuration obeys $ \nabla \cdot B = 0 $ .
199:
200: Two useful dimensionless parameters
201: containing the basic information of the models are used.
202: The first one, $\Omega$, is the velocity ratio of the stellar rotation to the
203: critical rotation, and the second parameter, $\sigma$, is the ratio of
204: the magnetic field energy density to the kinetic energy density in the
205: wind.
206: As a initial condition, the AGB wind is set to fill up the whole
207: computational mesh, while the fast wind is set only at the inner
208: radial boundary, using equation (2) and (3) in Garc\'{\i}a-Segura et al.(1999)
209: for both winds with their respective parameters. Note that equation (2)
210: only sets the radial component of the wind velocity, since $v_{\phi}$
211: is negligible (at the inner boundary distance) due to angular momentum
212: conservation.
213:
214:
215: The four runs (Figure \ref{fig:simulations})
216: include, a bipolar (B) and an elliptical (E) case without
217: magnetic fields, and a bipolar (Bm) and an elliptical (Em) case with magnetic
218: fields on. The AGB slow winds have velocities of 10 \kms and
219: $\Mdot = 10^{-5} $ \Moy in all cases, while the fast winds have
220: velocities of 1000 \kms and $\Mdot = 10^{-8} $ \Moy.
221: Bipolar runs (B, Bm) have $\Omega=0.98$ in the AGB phases, while
222: elliptical ones $\Omega=0.6$.
223: Magnetic models have $\sigma = 0.01$ only in the fast wind.
224: AGB winds are set unmagnetized (see Garc\'{\i}a-Segura et al. 1999 for
225: further details).
226:
227:
228: \begin{figure}[!t]\centering
229: \vspace{0pt}
230: \includegraphics[angle=00,width=0.9\columnwidth]{f1.eps}
231: % \includegraphics[angle=00,width=0.95\columnwidth]{f1.eps}
232: \caption{Density distributions of model B at 2500 yr (top-left),
233: model E at 3500 yr (bottom-left),
234: model Bm at 1500 yr (top-right), and
235: model Em at 2000 yr (bottom-right).}
236: \label{fig:simulations}
237: \vspace{0pt}
238:
239: \end{figure}
240:
241:
242: \section{Morpho-kinematical modeling with SHAPE}
243: \label{sec:shape}
244:
245: In this study, the hydrodynamical simulations serve two purposes.
246: First, we use them as observable objects, of which we know
247: their structure, dynamics and kinematics. For a given time in the
248: evolution, we analyze and visualize the projected images and
249: position velocity (P-V) diagrams as if they where true observations.
250: These images and P-V diagrams are generated with \shape after the
251: data from the simulations have been imported in a suitable format (see below).
252:
253: The second purpose of using hydrodynamical simulations is to validate the
254: functionality of \shape itself. For this purpose we reproduce the velocity
255: field of the simulation and its structure in detail using the structure
256: and velocity modeling tools of \shape. We then check
257: whether the images and P-V diagrams are consistent with those of the
258: simulation.
259:
260: We consider only the dense expanding
261: shell that has been swept up by the fast stellar wind. It is this
262: shell, that is expected to be visible as the bright main shell in
263: a planetary nebula. We assume that the nebula is fully ionized and that
264: the emission measure, proportional to the square of the density,
265: provides the relative brightness.
266:
267: In order to extract the region of the expanding shell we have
268: selected cells from the hydrodynamic simulation according to
269: their radial velocity and density contrast with neighboring cells.
270: The selection is based on the fact that the shell is faster
271: than the external medium (AGB-wind), but slower and denser than
272: the active fast stellar wind.
273:
274: In {\it Shape}, each cell corresponds to a particle with properties
275: of position, velocity and brightness.
276: In order to convert the axi-symmetric two-dimensional hydro-data
277: into a 3-D spatial particle structure, 18 particle copies for each
278: extracted cell have been distributed randomly along the corresponding
279: circle around the cylindrical symmetry axis. The velocity components
280: of the particles have been adjusted according to the rotated position.
281: To visualize, analyze and further process the hydrodynamic simulations
282: we import the simulated particle data into \shape
283: (Version $2.7$).
284:
285: Since Version 1.0, \shape has an integrated tool for the creation of
286: 3-D structures and their velocity field. The simulated planetary nebula
287: will be used to validate our modeling software, its method and accuracy.
288: The hydrodynamic simulations
289: that we are processing have been performed in spherical coordinates.
290: Because of the cylindrical symmetry, only the radial and the poloidal
291: velocity components are non-cero. \shape has two different modeling
292: modes for the velocity components. One is providing an analytic formula
293: and the other is an editable curve, in which any number of
294: linear segments can be used to describe the magnitude of each velocity
295: component ($v_r, v_{\theta}$) as a function of the corresponding
296: space coordinate ($r$, $\theta$).
297:
298: The main shell of
299: the nebula is modeled as a closed surface with cylindrical symmetry
300: using the surface modeling tools in \shape. As an example,
301: Figure \ref{fig:model_mesh} we show a visualization of the surface adjusted
302: to run B of the simulations. The surface density of particles is constant within the
303: deviations from the random particle statistics. About $6\times 10^{4}$ particles have
304: been used for each object.
305: In order to model the total emission, the relative particle emissivity
306: is given as a function of position with a similar tool described
307: for the velocity modeling (see above).
308:
309: For the rendering of the images,
310: the emission is directly integrated along the line of sight.
311: In this work we assume optically thin emission. The resulting raw
312: image is then convolved with a gaussian point-spread-function that models the
313: seeing conditions of an optical observation or the beam of a radiotelescope.
314:
315: Similarly, a P-V diagram is generated by distributing the emission
316: contained in a simulated spectrograph slit in a position-velocity image
317: (velocity is horizontal and position along the slit is vertical).
318: The raw image is then convolved with an ellipsoidal gaussian that
319: has width and height corresponding to the velocity and spatial resolution
320: of the simulated instrument.
321:
322:
323: \begin{figure}[!t]\centering
324: \vspace{0pt}
325: \includegraphics[angle=00,width=0.9\columnwidth]{f2.eps}
326: % \includegraphics[angle=00,width=0.95\columnwidth]{f2.eps}
327: \caption{The surface model mesh that has been adjusted to the extracted shell from run B.
328: The dots that are plotted represent only 5 \% of the total particles that have been
329: stochastically applied to sample the object.}
330: \label{fig:model_mesh}
331: \vspace{0pt}
332:
333: \end{figure}
334:
335:
336:
337: \begin{figure}[!t]\centering
338: \vspace{0pt}
339: \includegraphics[angle=-90,width=0.9\columnwidth]{f3.eps}
340: %% \includegraphics[angle=-90,width=0.95\columnwidth]{f3.eps}
341: \caption{The radial velocity component of the ellipsoidal and bipolar lobed hydrodynamical simulations (E \& B)is plotted against the distance from the center. The lines represent the piecewise linear representation of the velocity field in {\it Shape}. The dotted line is a simplified approximation with a single linear curve. The simulated data marked by (+) show those of model B and those marked as (x) are from model E.}
342: \label{fig:run1_2_r_vr}
343: \vspace{0pt}
344:
345: \end{figure}
346:
347:
348:
349: \begin{figure}[!t]\centering
350: \vspace{0pt}
351:
352: \includegraphics[angle=-90,width=0.9\columnwidth]{f4.eps}
353: %% \includegraphics[angle=-90,width=0.95\columnwidth]{f4.eps}
354: \caption{The radial velocity component of the MHD-simulations is plotted against the distance from the center. For a description of the lines, see Figure (\ref{fig:run1_2_r_vr}). Here, two linear curve segments have been adopted for the simplified velocity field. The simulated data marked by (+) show those of model Bm and those marked as (x) are from model Em.}
355: \label{fig:run3_4_r_vr}
356: \vspace{0pt}
357:
358: \end{figure}
359:
360: \begin{figure}[!t]\centering
361:
362: \includegraphics[angle=-90,width=0.9\columnwidth]{f5.eps}
363: %% \includegraphics[angle=-90,width=0.95\columnwidth]{f5.eps}
364: \caption{The poloidal velocity component of the ellipsoidal and bipolar lobed hydrodynamical simulations is plotted against the angle in degrees from the symmetry axis (equator = $90^o$). For a description of the lines and symbols, see Figure (\ref{fig:run1_2_r_vr}). Here, two linear curve segments have been adopted for the simplified velocity field.}
365: \label{fig:run1_2_dt}
366: \vspace{0pt}
367:
368: \end{figure}
369:
370:
371:
372: \begin{figure}[!t]\centering
373: \vspace{0pt}
374:
375: \includegraphics[angle=-90,width=0.9\columnwidth]{f6.eps}
376: %% \includegraphics[angle=-90,width=0.95\columnwidth]{f6.eps}
377: \caption{The poloidal velocity component of the MHD collimated simulations is plotted against the angle in degrees from the symmetry axis (equator = $90^o$). For a description of the lines and symbols, see Figure (\ref{fig:run3_4_r_vr}). Here, three linear curve segments have been adopted for the simplified velocity field.}
378: \label{fig:run3_4_dt}
379: \vspace{0pt}
380:
381: \end{figure}
382:
383:
384: \section{Results and discussion}
385:
386: In order to search for systematic tell-tale signatures which
387: might reveal information about the physics of their initial formation
388: process, we have investigated the kinematics of typical
389: expanding planetary nebulae based on 4 hydrodynamic simulations (Figure \ref{fig:simulations}).
390:
391:
392: Figures \ref{fig:run1_2_r_vr} and \ref{fig:run3_4_r_vr} show the distribution of the radial velocities ($v_r$) of the four simulations
393: as a function of distance from the center. Figures \ref{fig:run1_2_dt} and \ref{fig:run3_4_dt} display the poloidal velocity ($v_\theta$) component as a function of the angle from the polar axis of the simulations. For software validation the graphs have been approximated in \shape using piece-wise linear segments as shown by the continuous lines with dots superimposed on the data from the simulations. The resulting P-V diagrams follow those of the simulation to within the scatter of the points from the simulation, validating the our \shape code.
394:
395: Figures (\ref{fig:renders1}) to (\ref{fig:renders4}) show the rendered hydrodynamical simulations (columns ''a''),
396: their slit-less spectra or Hubble-like reconstructions (columns ''b'') and the reconstructed models with \shape (columns ''c'').
397: For the slit-less spectra, the simulated spectrograph slit was opened to include the complete object.
398: Note that a slit-less spectrum is equivalent to a 3D reconstruction that makes the assumption of a ''Hubble-type'' velocity distribution,
399: projected perpendicular to the line of sight.
400: A small gaussian blur has been applied to simulate observational seeing and reduce the stochastic particle-noise.
401: We discuss the results for each simulation separately.
402:
403:
404: \begin{figure}[!t]\centering
405:
406: \includegraphics[angle=00,width=0.95\columnwidth]{f7.eps}
407: %% \includegraphics[angle=00,width=0.95\columnwidth]{f7.eps}
408:
409: \caption{Column ''a''shows the rendered images (emission measure) of the hydrodynamical simulation E at different inclination angles ($0^o, 45^o, 90^o$) as seen from the observer's point of view. Column ''b'' contains the corresponding slit-less spectra of the simulation. It represents the view of the object as seen perpendicular to the line of sight in a reconstruction that is based on homologous expansion. This image reveals the deformations expected from such reconstructions. Despite of the cylindrical symmetry of the object, the shape of the ''reconstruction'' is quite different to the direct image of the simulation in column ''a''. Column ''c'' shows the rendered image of our model as reconstructed with \shape (column ''c'') with corrections to the homologous expansion velocity field as seen from the same direction as column ''b''. }
410: \label{fig:renders1}
411: \vspace{0pt}
412:
413: \end{figure}
414:
415:
416:
417:
418: \begin{figure}[!t]\centering
419:
420: \includegraphics[angle=00,width=0.95\columnwidth]{f8.eps}
421: %% \includegraphics[angle=00,width=0.95\columnwidth]{f8.eps}
422:
423: \caption{The rendered images of the hydrodynamical simulation B (column ''a''), slit-less spectra or homologous reconstruction (column ''b'') and the rendered
424: image of the reconstructed model with \shape (column ''c''). For further descriptions see Figure (\ref{fig:renders1}).}
425: \label{fig:renders2}
426: \vspace{0pt}
427:
428: \end{figure}
429:
430:
431: \begin{figure}[!t]\centering
432:
433: \includegraphics[angle=00,width=0.95\columnwidth]{f9.eps}
434: %% \includegraphics[angle=00,width=0.95\columnwidth]{f9.eps}
435:
436: \caption{The rendered images of the hydrodynamical simulation Em (column ''a''), slit-less spectra or homologous reconstruction (column ''b'') and the rendered
437: image of the reconstructed model with \shape (column ''c''). For further descriptions see Figure (\ref{fig:renders1}).}
438: \label{fig:renders3}
439: \vspace{0pt}
440:
441: \end{figure}
442:
443:
444: \begin{figure}[!t]\centering
445:
446: \includegraphics[angle=00,width=0.95\columnwidth]{f10.eps}
447: %% \includegraphics[angle=00,width=0.95\columnwidth]{f10.eps}
448:
449: \caption{The rendered images of the hydrodynamical simulation Bm (column ''a''), slit-less spectra or homologous reconstruction (column ''b'') and the rendered
450: image of the reconstructed model with \shape (column ''c''). For further descriptions see Figure (\ref{fig:renders1}).}
451: \label{fig:renders4}
452: \vspace{0pt}
453:
454: \end{figure}
455:
456:
457:
458: \subsection{Non-MHD elliptical nebula}
459:
460: Figures (\ref{fig:run1_2_r_vr}) and (\ref{fig:run1_2_dt}) show the statistics of the velocity component distribution
461: in the hydrodynamical simulation of run E which produced an ellipsoidal nebula (Figure \ref{fig:renders1}). The radial velocity component
462: increases approximately linearly with distance (Figure \ref{fig:run1_2_r_vr}). This might suggest that a homologous expansion is a good description of the velocity field. However,
463: the plot of the poloidal velocity component versus position angle
464: reveals that there are deviations of the velocity
465: vector from the radial direction (Figure \ref{fig:run1_2_dt}). The deviations increase from the
466: equator to poles until they reach approximately 7 km s$^{-1}$ at 20 $\deg$ from the pole. Close to the poles the poloidal velocity decreases again
467: to zero. Where the poloidal component is largest, noticeable deformations
468: will occur along the line of sight in a 3D reconstruction that assumes a homologous expansion law.
469:
470: The image, the slit-less spectra and the reconstructed models are compared in Figure (\ref{fig:renders1}). The shape of
471: the slit-less spectrum at $90^o$ (b) is more rectangular than the direct image (a). However, the reconstructed model (c) fits
472: quite well. Although the deviations in (b) are small they are significant at intermediate to high latitudes.
473: For the $0^o$ and $90^o$ viewing angles, the qualitative difference is small and therefore might not influence significantly the physical interpretation of the
474: object. However, at intermediate viewing angles the P-V diagram is more distorted and becomes point-symmetric.
475: We consider it cautiously acceptable to directly reconstruct this type of objects from their spectral data assuming a homologous expansion law.
476:
477: Both, the radial and the poloidal velocity show small-scale fluctuations which are associated with local instabilities. In a reconstruction, unresolved fluctuations will result in an increased thickness of the reconstructed surface structure.
478:
479:
480: \begin{figure}[!t]\centering
481:
482: \plotone{f11.eps}
483:
484: \caption{Central slit P-V diagrams from the simulations are represented in grey-scale. The superimposed contours
485: are from the same simulations, but with a simple piece-wise linear velocity distributions applied (dashed lines in Figures (\ref{fig:run1_2_r_vr}) to (\ref{fig:run3_4_dt}).}
486: \label{fig:simple_model}
487: \vspace{0pt}
488:
489: \end{figure}
490:
491:
492: \subsection{Non-MHD bipolar lobes}
493:
494: The radial velocity distribution for the bipolar lobes from the purely hydrodynamic model B is shown in Figure (\ref{fig:run1_2_r_vr}) and the distribution of the poloidal velocity is given in Figure (\ref{fig:run1_2_dt}). As in the ellipsoidal case of run E, we find a nearly linear increase of the radial velocity component with distance. Deviations from linearity are stronger far away and very close to the center. Also, the small-scale fluctuations increase at a certain distance. This distance marks the transition between two different dynamical regions in this bipolar nebula.
495:
496: The general trend of the poloidal velocity is similar to the one in the ellipsoidal simulation, a linear increase with angle up to a maximum value and then a decrease towards the pole. The vanishing poloidal velocity at the pole and the equator are enforced in these simulations due to the boundary conditions of reflection at the equator and polar axis. However, it is expected that this behavior is very closely sustained in real objects with approximate axi-symmetry.
497:
498: Column ''b'' in Figure (\ref{fig:renders2}) shows the slitless spectra of simulation B for 3 different viewing angles.
499: This corresponds to a 3D reconstruction that assumes homologous expansion as viewed along a direction perpendicular to the observer's line of sight. We note very significant differences to the observer's view in column ''a''. Especially, the ''Hubble-type reconstructions'' at $45^o$ and $0^o$ inclination angles show strong deviations. As in the case of run E, the intermediate angles introduce a point-symmetry which is now even stronger. Column ''c'' of Figure (\ref{fig:renders2}) shows the corresponding views of the axi-symmetric reconstructed
500: model using \shape which applies the simple piece-wise linear velocity law (see below).
501:
502: In order to not only reproduce the images, but also the P-V diagrams with a
503: model that improves the homologous expansion law with only a few additional parameters,
504: we propose a simple piece-wise linear approximation to the velocity components as a
505: function of position.
506:
507: For runs E and B, a two-parameter curve of two linear segments that start at zero at
508: the pole and equator and meet at a point ($\theta$, v$_{\theta}$) are sufficient.
509: The point ($\theta$, v$_{\theta}$) is a free parameter of the model. In addition,
510: a linear relation of the radial velocity with distance could be used. The result of
511: such a fit is shown in the P-V diagrams of Figure (\ref{fig:simple_model}, column E \& B).
512: The grey-scale P-V diagrams are from the hydrodynamical simulations with its original
513: velocity information. The contours are from the same structural data, but with the new
514: simple piece-wise linear velocity field. As an example, the corresponding velocity components
515: of run B are
516: plotted as dotted lines in Figures (\ref{fig:run1_2_r_vr}) \& (\ref{fig:run1_2_dt}).
517:
518: We find that the agreement of this simple piece-wise linear velocity field with the
519: simulation is very satisfactory. At any viewing angle deviations in the P-V diagram
520: are only significant at the small scales of instabilities. Hence, for an elliptical
521: or bipolar hydrodynamical nebula a very significant improvement in its morpho-kinematical
522: modeling may be achieved by including only one and two linear segments in the distribution
523: of the radial and poloidal velocity components, respectively. This procedure retains the
524: axi-symmetric structure of the object and, simultaneously, provides a good fit to the
525: point-symmetric P-V diagrams.
526:
527:
528:
529: \subsection{MHD collimated nebulae}
530: \label{sec:non-mhd}
531:
532: We now perform a similar analysis to the MHD-simulations.
533: Figure (\ref{fig:run3_4_r_vr}) shows the distribution of the radial velocity
534: components of simulations Bm and Em as a function of distance from the center.
535: The poloidal velocity component is plotted in Figure (\ref{fig:run3_4_dt}). Both
536: simulations show similar velocity fields, but with different positions and amplitudes
537: of the dynamical regions.
538: The radial velocity component has a nearly linear increase up to about half of the
539: extent of the nebula. Then an abrupt change is noticeable with a small drop in the
540: velocity and further out a significant gradual increase in the slope of the velocity graph.
541:
542: The poloidal velocity component is slowly growing from the equator towards the pole and
543: shows an accelerated increase from mid-latitudes on. It reaches
544: a maximum near the tip of the outflow, after which it declines rapidly
545: (as expected from the symmetry of the flow, see above). There are no strong
546: small-scale variations in either velocity component as opposed to those found in the
547: purely hydrodynamic simulations.
548:
549: Figures (\ref{fig:renders3}) and (\ref{fig:renders4}) compare the projected images rendered from the MHD-simulations (column ``a'')
550: with the ``3D Hubble-type reconstructions'', i.e., the slit-less spectra (column ``b''),
551: and the reconstructed model using \shape (column ``c'').
552: The distortions observed in columns ''b'' are even stronger than in the non-MHD
553: simulations.
554: Instead of being closed bipolar lobes, they appear to be open. In all cases of
555: intermediate inclination angles, the axi-symmetry of the simulation is lost and replaced
556: by point-symmetry. A physical interpretation of such artificial symmetry features, in
557: 3D reconstructions that assume homologous expansion, may lead to seriously erroneous
558: results.
559:
560: In order to capture the main features of the velocity field in this type of objects in
561: morpho-kinematical modeling, we propose to use at least two linear segments for the
562: radial component and three segments for the poloidal velocity component. Such
563: approximations for model Bm have been plotted in Figures (\ref{fig:run3_4_r_vr}) and
564: (\ref{fig:run3_4_dt}) as dotted lines. See Figure (\ref{fig:simple_model}) to
565: compare the P-V diagrams from the simplified \shape model with those of the simulation.
566: The deviations from the simulation are very small, hence, also for the MHD cases a
567: considerable improvement of the velocity field can be obtained by the simple piece-wise
568: linear velocity field proposed here.
569:
570:
571: \begin{figure}[!t]\centering
572:
573: \includegraphics[angle=00,width=0.45\columnwidth]{f12a.eps}
574: \includegraphics[angle=00,width=0.45\columnwidth]{f12b.eps}
575: %% \includegraphics[angle=00,width=0.45\columnwidth]{f12a.eps}
576: %% \includegraphics[angle=00,width=0.45\columnwidth]{f12b.eps}
577: \caption{Proper motion vectors are shown for model B at 90 and 45 degrees viewing angle on the left and right, respectively. The top half has the velocity field as obtained from the simulation, while the bottom has a homologous expansion law.}
578: \label{fig:proper_motion_1}
579: \vspace{0pt}
580:
581: \end{figure}
582:
583: \begin{figure}[!t]\centering
584:
585: \includegraphics[angle=00,width=0.45\columnwidth]{f13a.eps}
586: \includegraphics[angle=00,width=0.45\columnwidth]{f13b.eps}
587: %% \includegraphics[angle=00,width=0.45\columnwidth]{f13a.eps}
588: %% \includegraphics[angle=00,width=0.45\columnwidth]{f13b.eps}
589: \caption{Proper motion vectors are shown for model Bm at 90 and 45 degrees viewing angle on the left and right, respectively. The top half has the velocity field as obtained from the simulation, while the bottom has a homologous expansion law. The top half includes part of the central collimated outflow with very large velocity vectors.}
590: \label{fig:proper_motion_3}
591: \vspace{0pt}
592:
593: \end{figure}
594:
595:
596: \section{Proper motion patterns}
597: \label{sec:proper_motion}
598:
599: Additional kinematic information from an expanding object can be obtained from internal
600: proper motion measurements. Such measurements have been obtained for a few nearby
601: planetary nebulae (e.g. NGC6302: Meaburn et al. 2008; NGC7009: Rodr\'{\i}guez \& G\'omez
602: 2007). Measurements of Doppler-shift and proper motion of a large number of features in a
603: nebula (e.g. Li, Harrington \& Borkowski 2002) might reveal whether an object is
604: hydrodynamically active or expanding homologously due to ballistic motion.
605: In Figures (\ref{fig:proper_motion_1}) and (\ref{fig:proper_motion_3}) we compare the
606: proper motion vectors of our simulations B and Bm with the corresponding homologous
607: motion patterns for two different viewing angles. The top halves of the images represent
608: the velocity pattern from the hydrodynamical simulations, while in the bottom halves we
609: replaced the original velocity field by a homologous velocity field. Clearly, such
610: measurements might reveal a lot of information about the dynamical state of a planetary
611: nebula. When the expansion is homologous, the proper motion pattern reflects this in
612: purely radial vectors with magnitudes that increase linearly with distance.
613: This pattern is basically independent of the structure and orientation of the object.
614:
615: However, in a dynamically active nebula or in a nebula with several kinematic sub-systems,
616: the motion pattern will depend on the orientation and structure. Some of the proper
617: motion vectors in Figures (\ref{fig:proper_motion_1}) and (\ref{fig:proper_motion_3})
618: almost reverse direction in the inclined objects (right side). Such a behavior is found
619: in BD+30$^{o}$3639. As seen from the proper motion pattern in Li, Harrington \&
620: Borkowski (2002), BD+30$^{o}$3639 is likely to still be a dynamically active nebula.
621: The spectral and proper motion data from Meaburn et al. (2008) indicate that
622: NGC~6302 is expanding radially and, hence, presumably ballistically.
623:
624: In general, because of the evolution of the fast stellar wind, it is expected that
625: young planetary nebulae will be more likely to be active than older nebulae. It is
626: therefore surprising that very young non-spherical nebulae, like, for instance,
627: Minkowski's footprint (M1-92, Alcolea, Neri \& Bujarrabal 2007) show Doppler-shift
628: patterns that are consistent with axi-symmetric homologous expansion.
629: It is important to systematically survey the kinematic properties of planetary nebulae
630: for evidence of deviations from homologous expansion. This will reveal information
631: about the evolution of the fast stellar wind and the evolution of central stars
632: in general. Hopefully, the analysis of large kinematical catalogs like those of Hajian et al. (2007) and L\'opez et al. (2006) might be helpful in studying the transition phase between hydrodynamically active and ballistic expansion.
633:
634:
635: \section{Conclusions}
636:
637: Based on axi-symmetric hydrodynamical simulations and 3D reconstructions with
638: the software package {\it Shape}, we have investigated the kinematic
639: signatures of deviations from ``Hubble-type'' outflows in some typical shapes of planetary nebulae.
640: The deviations are systematic and their description as
641: given in this paper may
642: help to improve morpho-kinematic and dynamic models of planetary nebulae.
643: Excluding the small-scale variations from local instabilities,
644: we find that the radial velocity as a function of distance can be well described by a
645: linear increase for the cases of purely hydrodynamical ellipsoidal and bipolar-lobed
646: nebulae. The poloidal component is, however, not negligable. Its parameterized
647: approximation requires two linear segments, one increasing and the other decreasing
648: with zero velocity at the pole and equator.
649: We also find that the magneto-hydrodynamically generated cases require two linear
650: segments for the radial velocity component and three segments for the poloidal component.
651: These simple extensions to a homologous expansion law capture the global velocity
652: structure of hydrodynamically active axi-symmetric nebulae except for small-scale
653: variations.
654:
655: We find that, in most situations considered in our simulations,
656: the deviations from a homologous flow, both in magnitude and
657: direction of the velocity vector, are significant and
658: observable. The only case where only small deviations were
659: found was the case of an ellipsoidal nebula.
660:
661: If significant deviations from a homologous expansion model can
662: be inferred, then it is very likely that it is still
663: dynamically active in the sense that the fast wind is still significant or
664: the inner region of the nebula is still over-pressurized with respect to
665: the exterior. In a future paper we will explore more precisely how
666: the transition from the hydrodynamically to the ballistic regimes occurs.
667:
668: It is the relative
669: size of the poloidal velocity component that strongly influences the
670: shape of the position velocity diagrams that are obtained,
671: not so much the variation of the radial component with distance from the center.
672: The deviations increase with the degree of collimation of the nebula.
673: They are stronger at intermediate latitudes.
674: There is a slow increase in deviations as we move from the equator to
675: the poles, usually with a sudden increase at mid-latitudes.
676:
677: A good description of the velocity field in the MHD-simulations requires one more
678: linear segment than that of the non-MHD simulations. This would imply that
679: this difference might provide observational evidence for the presence or absence
680: of dynamical active magnetic fields in the formation process of planetary nebulae.
681: However, this aspect might require more detailed and specific studies.
682:
683: Summing up, detailed morpho-kinematic observations and modeling of planetary
684: nebulae can reveal whether a nebula is still in a hydrodynamically active stage
685: or whether it has reached ballistic expansion.
686:
687: \acknowledgments
688:
689: We thank Michael L. Norman and the Laboratory for Computational
690: Astrophysics for the use of ZEUS-3D. We also thank J.A. L\'opez for his very
691: useful discussions on the observed properties of planetary nebulae and their
692: interpretation. We acknowledge financial support from CONACYT grant 49447 and
693: UNAM DGAPA-PAPIIT IN108506-2. We thank the anonymous referee for useful suggestions.
694:
695:
696: \begin{thebibliography}{}
697: \bibitem[]{686} Akashi, M., Soker, N., 2008, arXiv:0805.2332v1 [astro-ph]
698: \bibitem[]{687} Alcolea, J.; Neri, R.; Bujarrabal, V., 2007, A\&A, 468. L41-L44
699: \bibitem[]{688} Bjorkman, J. E. \& Cassinelli, J. P. 1993, ApJ, 409, 429
700: \bibitem[]{689} Chevalier, R. A. \& Luo, D. 1994, ApJ, 421, 225
701: \bibitem[]{690} Clarke, D. A. 1996, ApJ, 457, 291
702: \bibitem[]{691} Dennis, T.~J., Cunningham, A.~J., Frank, A., Balick, B., Blackman, E.~G. , Mitran, S., 2008, ApJ, 679, 1327-1337
703: \bibitem[]{692} Dobrin{\v c}i{\'c}, M. and Villaver, E. and Guerrero, M.~A. and
704: Manchado, A., 2008, \aj, 135, 2199-2211
705: \bibitem[]{694} Ercolano, B., Barlow, M. J., Storey, P. J., 2005, MNRAS, 362, 1038-1046
706: \bibitem[]{695} Garc\'{\i}a-Segura, G., Langer, N., R\'o\.zyczka, M., \&
707: Franco, J. 1999, ApJ, 517, 767
708: \bibitem[]{697} {Hajian}, A.~R., and {Movit}, S.~M., {Trofimov}, D., {Balick}, B.,
709: {Terzian}, Y., {Knuth}, K.~H., {Granquist-Fraser}, D.,
710: {Huyser}, K.~A., {Jalobeanu}, A., {McIntosh}, D., {Jaskot}, A.~E.,
711: {Palen}, S., {Panagia}, N., 2007, ApJS, 169, 289-327
712: \bibitem[]{701} Ignace, R., Cassinelli, J. P. \& Bjorkman, J. E. 1996, ApJ, 459, 671
713: \bibitem[]{702} Li, J., Harrington, J. P., Borkowski, K. J., 2002, AJ, 123, 2676-2688
714: \bibitem[]{703} L\'opez, J. A., Richer, M., Riesgo, H., Steffen, W., Meaburn, J., García-Segura, G., Escalante, K., 2006, XI IAU Regional Latin American Meeting of Astronomy (Eds. L. Infante \& M. Rubio), RMxAC, 26, 161
715: \bibitem[]{704} Meaburn, J., L\'opez, J.A., Steffen, W., Graham, M.F., Holloway, A.J., 2005, , AJ, 130, 2303-2311
716: \bibitem[]{705} Meaburn, J., LLoyd, M., Vaytet, N.M.H., L\'opez, J.A., 2008, MNRAS, 385, 269-273
717: \bibitem[]{706} Morisset, C., Stasinska, G., 2008, RMxAA, 44, 171-180
718: \bibitem[]{707} Morisset, C., 2006, IAU proceedings on "Planetary Nebulae in our Galaxy and Beyond", 2: 467-468 Cambridge University Press
719: \bibitem[]{708} O'Connor, J. A., Redman, M. P., Holloway, A. J., Bryce, M., L\'opez, J. A., Meaburn, J., 2000, ApJ, 531, 336-344
720: \bibitem[]{709} Raymond, J. C. \& Smith, B. W. 1977, ApJS, 35, 419
721: \bibitem[]{710} Rodr{\'{\i}}guez, L.F., G{\'o}mez, Y., 2007, RMxAA, 453, 173-177
722: \bibitem[]{711} R\'o\.zyczka, M. \& Franco, J. 1996, ApJL, 469, L127
723: \bibitem[]{712} Sabbadin, F. and Turatto, M. and Ragazzoni, R. and Cappellaro, E. and
724: Benetti, S., 2006, A\&A, 451, 937-949
725: \bibitem[]{714} Santander-Garc{\'{\i}}a, M. Corradi, R.~L.~M. Balick, B.
726: Mampaso, A., 2004, A\&A, 426, 185-194
727: \bibitem[]{716} Steffen, W., L\'opez, J.A., 2004, ApJ, 612, 319-331
728: \bibitem[]{717} Steffen, W., L\'opez, J.A., 2006, RMxAA, 42, 99-105
729: \bibitem[]{718} Stone, J. M. \& Norman, M. L. 1992, ApJS, 80, 753
730: \bibitem[]{719} Wilson, O., 1950, ApJ, 111, 279
731:
732: \end{thebibliography}
733:
734: \end{document}
735:
736: