1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[preprint2]{aastex}
3: %\documentclass[twocolumn]{aa}
4: %\documentclass[referee]{aa} % for a referee version
5: %\usepackage{graphicx}
6: %\usepackage{txfonts}
7: %\usepackage{natbib}
8:
9: %\makeatother
10:
11: \slugcomment{}
12:
13: \shorttitle{3-D modeling and distance of NGC 6369 }
14: \shortauthors{Monteiro et al.}
15:
16: \received{2003 November 11}
17: \begin{document}
18:
19: \title{3-D Photoionization Structure and Distances of Planetary Nebulae I.
20: NGC 6369}
21:
22:
23: \author{Hektor Monteiro$^{1,2}$, Hugo E. Schwarz$^1$, Ruth Gruenwald$^2$}
24: \affil{1 Cerro Tololo Inter-American Observatory\altaffilmark{1}, Casilla 603,
25: Colina El Pino S/N, La Serena, Chile}
26: \affil{2 Instituto de Astronomia, Geof\'isica e Ciencias Atmosf\'{e}ricas,
27: S\~{a}o Paulo, Brasil}
28:
29: \author{Steve Heathcote}
30: \affil{SOAR Telescope, Casilla 603, Colina El Pino S/N,
31: La Serena, Chile}
32:
33: \altaffiltext{1}{Cerro Tololo Inter-American Observatory, National Optical
34: Astronomy Observatory, operated by the Association of Universities for
35: Research in Astronomy, Inc., under a cooperative agreement with the
36: National Science Foundation.}
37:
38: \begin{abstract}
39: We present the results of mapping the planetary nebula NGC\,6369 using
40: multiple long slit spectra taken with the CTIO 1.5m telescope. We
41: create two dimensional emission line images from our spectra, and use
42: these to derive fluxes for 17 lines, the H$\alpha$/H$\beta$ extinction
43: map, the [SII] line ratio density map, and the [NII] temperature map
44: of the nebula. We use our photoionization code constrained by these
45: data to determine the distance, the ionizing star characteristics, and
46: show that a clumpy hour-glass shape is the most likely
47: three-dimensional structure for NGC\,6369. Note that our knowledge of
48: the nebular structure eliminates all uncertainties associated with
49: classical distance determinations, and our method can be applied to
50: {\it any spatially resolved emission line nebula}. We use the central
51: star, nebular emission line, and optical+IR luminosities to show that
52: NGC\,6369 is matter bound, as about 70\% of the Lyman continuum flux
53: escapes. Using evolutionary tracks from
54: \cite{B95} we derive a central star mass of about 0.65\,M$_{\odot}$.
55:
56: \end{abstract}
57:
58: \keywords{planetary nebulae: individual (NGC 6369) --- techniques:
59: spectroscopic}
60:
61: \section{Introduction}
62:
63: The planetary nebula (PN) NGC\,6369 -- shown in Fig.\,\ref{n2ima} in
64: the light of H$\alpha$\,+\,N[II]$\lambda$658.4nm -- is an object with
65: a complex morphology, consisting of a main bright annulus of diameter
66: 40\arcsec, and fainter, curved outer structures on two sides in the
67: E-W direction. Deep narrow band H$\alpha$\, and
68: \,N[II]$\lambda$6584 images of NGC\,6369 were obtained by \cite{CSSP03}
69: in a survey to look for faint outer halos around PNe. Apart from
70: being deeper, their image is not significantly different from the one
71: shown in Fig.\,1, and no large faint halo was found.
72:
73: The HST image of NGC\,6369 available at
74: http://heritage.stsci.edu/2002/25/index.html also shows the same main
75: features as our image but has additional resolved details, and tags
76: the H$\alpha$, [OIII]500.7nm, and [NII]658.4nm lines with the colors
77: red, blue and green respectively. In the section on observational
78: results we discuss this composite HST image in more detail.
79:
80: \clearpage
81: \begin{figure}
82: %\includegraphics[scale=0.65]{schwarz_ima_n2.eps}
83: \includegraphics[scale=0.65]{f1.eps}
84: \caption{Narrow band filter image of NGC\,6369 in
85: H$\alpha $+{[}NII{]} from \cite{SCM92}. Note the bright ring and outer
86: faint ansae to the E and W of the object. N is up, E to the left, and
87: the plate scale is 0.26 ''/pix. \label{n2ima}}
88: \end{figure}
89: \clearpage
90:
91: The equatorial and galactic coordinates for NGC\,6369 are $\alpha$ =
92: 17$^h$\,29$^m$\,20$^s$.40 and $\delta$ = 23$^o$\,45\arcmin\,37\arcsec.9
93: (ICRS2000), and l = 2.43$^o$ b = +5.85$^o$ respectively. The extreme distances
94: determined for NGC\,6369 are 0.33\,kpc obtained by
95: \cite{AGNR84}, and 2.00\,kpc obtained by \cite{GPG86}, out of a total of
96: 10 listed by \cite{AO92}, from which we obtain a mean value for the
97: distance to NGC\,6369 of 1039$\pm$ 538 pc. Note that these ten
98: distances were determined using several statistical methods, each with
99: their own systematic biases and assumptions, so that the average
100: probably is a reasonably unbiased estimator, within the (large) error.
101:
102: The temperature of the central star obtained from the Zanstra method
103: by \cite{GP89} is $T(H)_\mathrm{Z}=67\,600~\mathrm{K}$. The star has
104: been classified as a WC4 star by \cite{TAS93}.
105:
106: Spectra of NGC\,6369 were obtained by \cite{AKSJ1989}, using an IDS
107: (Image Dissector Scanner) as detector with double
108: 4\,\arcsec~apertures, by \cite{AK87} using an ITS (Image Tube Scanner)
109: with a $2\arcsec\,\times\,10\arcsec$ slit, and by \cite{PPPP} using an
110: IDS and a combination of different apertures not bigger than
111: $13\arcsec\,\times\,13\arcsec$. An expansion velocity of $41.5$ km
112: s$^{-1}$ has been determined by \cite{MWF88}. The system radial
113: velocity was reported to be -106\,km.s$^{-1}$ by \cite{W53}, and
114: -101\,km.s$^{-1}$ by \cite{MWF88}.
115:
116: The spectral energy distribution (SED) of NGC\,6369 is shown in
117: Fig.\,\ref{sed} and has been computed using flux values from
118: \cite{AO92} and \cite{SK02} over a range of wavelengths from
119: 0.44 to 1100\,$\mu$m. Note that we plot $\lambda$\,F$_{\lambda}$ against
120: wavelength and below we will use the F$_{\lambda}$ values to determine the
121: observed optical+IR luminosity of NGC\,6369.
122:
123: \clearpage
124: \begin{figure}
125: %\includegraphics[width=\columnwidth]{sed.ps}
126: \includegraphics[width=\columnwidth]{f2.eps}
127: \caption{The SED for NGC\,6369 between 0.44 and 1100\,$\mu$m using flux
128: values from the literature (see text).\label{sed}}
129: \end{figure}
130: \clearpage
131:
132: Planetary nebulae are generally classified according to their
133: appearance, and such classification is then used for studies of the
134: formation and evolution of these objects (for example
135: \citep{B87}). Such a classification, based on the observed two
136: dimensional (2-D) brightness distribution in a given line or lines can
137: be quite misleading. This could be the case for NGC 6369, for which a
138: round morphology is evident from the observed images. Different 3-D
139: geometries can produce the same observed 2-D morphology as has been
140: shown by \cite{M00}. Furthermore, images produced by ions of different
141: ionization degree can be very different, due to the radiation transfer
142: within the nebulae. Only detailed modeling, which reproduces the
143: brightness intensity distribution of different lines, as well other
144: observed parameters, can provide more reliable information about the
145: 3-D geometry of the gas distribution. The determination of the actual
146: 3-D gas distribution in planetary nebulae is essential for
147: understanding their formation and evolution, as well as that of the
148: ionizing star.
149:
150: By determining the 3-D structure of nebulae, we eliminate the large
151: uncertainties that have plagued classical statistical distance
152: determination methods for over 5 decades. Assumptions about the
153: filling factor, constancy of ionized mass or diameter, mass-radius
154: relationships etc. are not needed here: we {\it know} what the
155: structure and ionized mass are, and can therefore determine distances
156: to much greater accuracy than before. In this paper we explain in detail how
157: we can determine this 3-D structure from long slit spectra and our
158: 3-D photo-ionization model, and apply it to the case of NGC\,6369
159:
160: In summary, we obtain the spatial structure of the object along with
161: its chemical composition, ionizing source temperature and luminosity,
162: mass, as well as an {\it independent} distance, in a self-consistent
163: manner. In \S2 we discuss the observations and basic reduction
164: procedures, as well as the details of the image reconstruction
165: technique used to obtain the line intensity maps. In
166: \S3 the results obtained from these maps are discussed: the reddening
167: correction of the images, total line fluxes and the computed
168: temperature and density maps. In \S4 we present the model results of
169: the 3-D photoionization code, and we discuss the derived quantities
170: Finally, in \S5 we give our conclusions.
171:
172: \section{Observations and Mapping}
173:
174: \subsection{Observations and general reduction}
175:
176: Observations were made using the CTIO 1.5\,m Ritchey-Chr\'etien
177: telescope with the RC Spectrograph on the nights of 12 and 13th June,
178: 2002. We used a grating with 600\,l/mm blazed at 600\,nm giving a
179: spectral resolution of 0.65\,nm per pixel and a plate scale of
180: 1.3\arcsec/pixel with a slit width of 4\arcsec. The spectral coverage
181: obtained with this configuration was approximately 450\,nm to
182: 700\,nm. We took three 1200s exposures at each slit position for ease
183: of cosmic hit removal etc.. For details of the instrument and
184: telescope see http://www.ctio.noao.edu and click on ``Optical
185: Spectrographs'', then on ``1.5M RC spectrograph''.
186:
187:
188: By taking sets of exposures at several parallel long-slit positions
189: across the nebula, we obtained line intensity profiles for each
190: slit. These profiles were then combined to create emission line images
191: of the nebula with a spatial resolution of about
192: $4\arcsec\,\times\,4\arcsec$, in a way similar to radio mapping. This
193: method has been applied before for the PN NGC\,3132 as discussed in
194: \cite{M02}. A total of 10 positions were observed, moving the slit in 4\arcsec
195: steps in the N-S direction between exposures. The third position
196: observed from south to north (east is to the left, and north is up in
197: all the figures) was not observed due to pointing problems. For this
198: position an average of the two adjacent exposures was adopted.
199:
200: The seeing conditions during the observing run were 1.5\arcsec or
201: better, obtained from the seeing monitor at CTIO. In both nights the
202: atmospheric conditions were photometric.
203:
204: The individual slit spectra were reduced using standard procedures for
205: long-slit spectroscopy, using IRAF reduction packages. The wavelength
206: calibration was applied to the two dimensional images using the tasks
207: FITCOORDS and TRANSFORM after re-identifying the arc spectra in many
208: spatial positions. This procedure helps to remove distortions in the
209: image that may be present due to slit curvature and misalignment with
210: respect to the CCD. The final corrected images show displacements of
211: the order of one pixel, which is the precision limit of this method.
212: An additional fine correction for slit misalignment was made using the
213: H$\alpha $ and H$\beta $ profiles for each exposure. Using IDL the
214: images were re-dimensioned to 100 times their original size. The
215: normalized H$\alpha $ and H$\beta $ profiles were then matched and the
216: final result re-dimensioned to original values. This procedure yields
217: the precise alignment necessary for calculation of diagnostic line
218: intensity ratios. Minor shifts of the order of one pixel can introduce
219: considerable errors in line ratios, if this method is not applied. The
220: error introduced in line ratios by misalignments can reach 4\% on sharp
221: edges of the brightness profile.
222:
223: For the flux calibrations two standard stars, LTT\,7379 and LTT\,9239
224: were observed on both nights. The final sensitivity function obtained
225: from these stars, using the IRAF task SENSFUNC, has an RMS of 0.02,
226: indicating the good photometric quality of the nights. Based on
227: previous similar spectrophotometric observations made with the INT on
228: La Palma, we compute the {\it random} error on individual line flux
229: calibrations to be 3\%, while the overall possible {\it systematic}
230: uncertainty in the stellar flux is about 5\%.
231:
232: \subsection{Image reconstruction}
233:
234: The final emission line images were obtained from the spatially
235: integrated flux profile in each slit. To obtain the slit profile of
236: integrated fluxes in a given transition, a Gaussian function was
237: fitted to each point along the spatial direction using IDL
238: routines. The errors adopted for the fits were the standard deviations
239: calculated by the fitting procedure. The profiles for each slit
240: position were then combined and interpolated using a cubic convolution
241: algorithm \citep{PS83,RM74}, in order to reconstruct a 2-D image of the
242: nebula for that transition. Integration of this final image provides
243: the total observed flux for the nebula, for each line. Note that this
244: procedure gives better estimates of the total flux and relative line
245: intensities than observing just one slit position or aperture, since
246: it takes into account the entire nebula.
247:
248: Signal to noise images were also obtained using the fitted profiles by
249: subtracting the fitted profile from the original section and analyzing
250: the remaining noise. This is a combination of all noise sources in the
251: image (photon statistics, noise introduced by the data reduction
252: procedures, etc.). These errors were then combined with the fitting
253: errors mentioned above, using the usual error propagation expressions
254: to make the final signal to noise image. We then quadratically added
255: the 3\% random calibration error and the de-reddening error computed
256: from the H$\alpha$/H$\beta$ reddening map (see \S3.1), to arrive at
257: the overall error for each line flux. This procedure provides precise
258: error estimates for the total fluxes and diagnostic line ratio
259: calculations.
260:
261:
262: \section{Observational results}
263:
264: Images were created for all 17 lines detected with sufficient
265: signal-to-noise ratio. In Fig.\,\ref{maps_full} the images for the
266: most important lines are shown. The important [OIII]436.3nm line fell
267: outside our available spectral range and could therefore not be
268: included in our analysis. The images were corrected for reddening as
269: described in the following section. The extracted total line
270: intensities with their fractional total errors are given in Table~1.
271:
272: Note that all main features are reproduced in these reconstructed line
273: images. The outer ansae only show up in the [NII]685.4nm line, and
274: are extremely faint or invisible in all other lines, as they are in
275: the HST original images shown at
276: http://heritage.stsci.edu/2002/25/original.html The [OIII]500.7nm
277: emission comes mainly from the central part of the bright ring-like
278: structure, while H$\alpha$ comes from further out, and the outermost
279: part is dominated by [NII]685.4nm. Clearly, we do not reproduce the
280: fine detail visible in the HST image because of our effective spatial
281: resolution of 4'' in the observational data. However, the expected
282: ionization stratification for the different ions and the overall
283: nebular structure are clearly seen in our images.
284:
285: \begin{figure*}
286: %\includegraphics[]{line_maps_full.ps}
287: \includegraphics[]{f3.eps}
288: \caption{Reddening corrected, reconstructed emission line contour plots
289: for NGC6369. The peak intensity in each frame is normalized to unity,
290: and there are 7 contour steps between the maximum (black) and zero
291: (white). N is up, E to the left.\label{maps_full} }
292: \end{figure*}
293:
294: \subsection{Reddening correction and total fluxes}
295:
296: The reconstructed images for each line were corrected for reddening
297: using the H$\alpha $/H$\beta $ ratio map shown in Fig.\,\ref{ha/hb}.
298: The logarithmic correction constant was calculated pixel by pixel
299: using the theoretical value of H$\alpha $/H$\beta $=2.87 (\citep{O89})
300: and the reddening curve of \cite{S79}. We investigated the effect of
301: differential atmospheric refraction on this ratio map. From the
302: airmasses of our observed positions and the values given by \cite{F82}
303: we computed a correction which we applied to our data. Since we used
304: wide slits (4\arcsec) and the object is extended, the effect was
305: small, but not negligible in the steep gradients near the bright ring
306: structure. The average error due to this effect is about 2\% in the
307: high signal to noise areas and about 20\% in the low signal to noise
308: areas. The nett effect on the final calculated relative total fluxes
309: is about 0.2\% for strong lines and of 5\% for weak ones, well within
310: the other observed uncertainties.
311:
312: In Fig.\,\ref{n2corrected} we show the {[}NII{]}658.4nm image as an
313: example before and after the reddening and differential refraction
314: corrections.
315:
316: \clearpage
317: \begin{figure}
318: %\includegraphics[width=\columnwidth]{ha_ratio_cont.ps}
319: \includegraphics[width=\columnwidth]{f4.eps}
320: \caption{H$\alpha $/H$\beta $ ratio map with contour overlay of H$\alpha $
321: (Image is cut for S/N lower than 10). N is up, E to the left. \label{ha/hb}}
322: \end{figure}
323:
324: \clearpage
325: \begin{figure}
326: %\includegraphics[width=\columnwidth]{n2_im_comp.ps}
327: \includegraphics[width=\columnwidth]{f5.eps}
328: \caption{Comparison of [NII] image before and after reddening correction.
329: N is up, E to the left.
330: \label{n2corrected}}
331: \end{figure}
332: \clearpage
333:
334: The reddening correction was also obtained from the spatially
335: integrated fluxes calculated from the images. The H$\alpha
336: $ flux was divided by the H$\beta $ flux and a single value
337: for the logarithmic extinction constant was obtained for the whole
338: nebula. This value was then used to correct the total line
339: intensities. The results for the line intensities were compared with
340: those obtained from the first method and showed no differences to
341: within the calculated uncertainties.
342:
343: The resulting emission line fluxes relative to H$\beta$ and their
344: corresponding total errors are shown in Table\,\ref{line fluxes}. All
345: total fluxes were obtained by integrating the reddening corrected
346: (pixel by pixel) images. The value we obtained for the total
347: de-reddened H$\beta$ flux was $F_{H\beta
348: }\,=\,7.2\,\times\,10^{-13}W\,m^{-2}$ and the uncorrected flux
349: 5.9$\,\times\,10^{-15}W\,m^{-2}$ corresponding to a value of
350: E(B-V)\,=\,1.4 magnitudes.
351:
352: \clearpage
353: \begin{table}
354:
355: \centering
356:
357: \caption{Line fluxes relative to H$\beta$ \label{line fluxes}}
358:
359:
360: \begin{tabular}{lccc}
361: \hline
362: { Line}& { Flux}& {Dered. Flux}& {Error\,(\%)}\\
363: \hline\hline
364:
365: {HeII$\lambda$468.6} & {0.011} & {0.014} & {19}\\
366: {[OIII]$\lambda$495.9} & {4.59} & {4.11} & {7}\\
367: {[OIII]$\lambda$500.7} & {14.60} & {12.41} & {7}\\
368: {[CIII]$\lambda$551.8} & {0.01} & {0.01} & {25}\\
369: {[CIII]$\lambda$553.8} & {0.01} & {0.01} & {22}\\
370: {[NII]$\lambda$575.5} & {0.04} & {0.02} & {17}\\
371: {HeI$\lambda$587.6} & {0.47} & {0.16} & {8}\\
372: {[OI]$\lambda$630.0} & {0.22} & {0.05} & {8}\\
373: {[SIII]$\lambda$631.1}& {0.05} & {0.014} & {13}\\
374: {[OI]$\lambda$636.3} & {0.07} & {0.02} & {13}\\
375: {[NII]$\lambda$654.9} & {1.20} & {0.24} & {7}\\
376: {$H\alpha\lambda$656.3} & {13.80} & {2.87} & {7}\\
377: {[NII]$\lambda$658.4} & {3.70} & {0.73} & {7}\\
378: {HeI$\lambda$667.8} & {0.22} & {0.04} & {9}\\
379: {[SII]$\lambda$671.7} & {0.25} & {0.04} & {10}\\
380: {[SII]$\lambda$673.1} & {0.37} & {0.07} & {9}\\
381:
382: \hline
383:
384: \end{tabular}
385: \end{table}
386: \clearpage
387:
388: \subsection{Gas density and temperature}
389:
390: We calculated the density and temperature maps from the reddening
391: corrected maps of the {[}SII{]} and {[}NII{]} lines respectively. The
392: expression relating the sulfur doublet line intensity ratios to the
393: gas density is from \cite{D87} which is also used in the IRAF package {\it
394: temden}. The density map is shown in Fig.\,\ref{dens_map}. For the
395: temperature map we used the expression given by \cite{M84} as this
396: allowed us to take the density variations over the nebula into
397: account, which we did by using our density map. The difference between
398: using this method and one that assumes constant density is small
399: --since the temperature is only weakly dependent on the density-- but
400: systematic.
401:
402: For the temperature maps, correction for slit misalignment was carried
403: out as done for the H$\alpha $/H$\beta $ maps discussed in section
404: 2.1. This correction was not necessary for the density maps since the
405: small wavelength separation of the two lines generates very little
406: deviation. The temperature map obtained after de-reddening is shown in
407: Fig.\,\ref{temp_comp}(A). We also show the difference between the
408: temperature map obtained using constant density and our map in
409: Fig.\,\ref{temp_comp}(B). Note that the differences are less than 500\,K.
410: The images are clipped for data values with S/N lower than 10 for
411: visual clarity.
412:
413: \clearpage
414: \begin{figure}
415: \includegraphics[width=\columnwidth]{f6.eps}
416: \caption{Density map obtained from the de-reddened {[}SII{]}671.7,673.1
417: line ratio using the method of \cite{D87}. N is up, E to the left. \label{dens_map}}
418: \end{figure}
419:
420: \clearpage
421: \begin{figure}
422: \includegraphics[width=\columnwidth]{f7.eps}
423: \caption{Temperature map obtained from the de-reddened [NII] line ratio (A),
424: and the difference between this map and the one obtained using constant
425: density (B). Note that the maximum difference is less than 500\,K. N is up,
426: E to the left.
427: \label{temp_comp}}
428: \end{figure}
429: \clearpage
430:
431: \subsection{Temperature of the central star}
432:
433: With the total fluxes corrected for interstellar reddening, the central
434: star temperature can be calculated. For this, we apply the method described by
435: \cite{HS66}, using the observed fluxes presented in \S3.1 and the
436: central star flux $F_{4793\AA}= 1.20\,\times\,10^{-15}\,erg\,cm^{-2}\,
437: s^{-1}$ obtained from \cite{GP88}. It is important to note that this
438: method assumes spherical symmetry and a black-body spectral energy
439: distribution for the ionizing star. This is not necessarily true as
440: has been shown by many authors (for recent work see \cite{RDDW00})
441: Here we use a blackbody spectrum modified by the major H and He
442: absorption edges, and below we show by comparison with stellar
443: evolution models from \cite{B95} that this is a reasonable method.
444:
445: From the H$\beta $ and $HeII \lambda 468.6$ line fluxes, we obtain,
446: respectively, $T_\mathrm{Z}(H)= 70000~\mathrm{K}$ and
447: $T_\mathrm{Z}(HeII)= 106000~\mathrm{K}$. The value obtained for
448: $T_\mathrm{Z}(H)$ with our fluxes differs slightly from the one
449: obtained by \cite{GP89} but is within the expected (large)
450: uncertainties of the method. These values are more reliable than
451: previously calculated ones since we have precise integrated fluxes for
452: the whole nebula.
453:
454: The Zanstra temperature for H and HeII are not necessarily equal, and
455: do not correspond to the effective temperature of the central star if
456: the nebula is not optically thick in all directions
457: \citep{GV00}. NGC\,6369 is optically thin for radiation short-ward of
458: 91.2\,nm (Lyman continuum) since a rough estimate of the total
459: luminosity in the optical lines is $\approx$\,14\% of that shortward
460: of 91.2\,nm. Most of the UV escapes the nebula, and therefore we
461: expect the T(H) do be lower than the T(He), as indeed found above.
462:
463:
464: \section{Photoionization Models}
465:
466: The photoionization code applied here was described in detail by
467: \cite{GVB97}. It uses a cube divided into cells, each
468: having a given density. This allows arbitrary density distributions to
469: be studied. The input parameters are the ionization source
470: characteristics (luminosity, spectrum, and temperature), element
471: abundances, density distribution, and the distance to
472: the object. The conditions are assumed uniform within each cell where
473: the code calculates the temperature and ionic fractions. These values
474: are used to obtain emission line emissivities for each cell.
475:
476: The final data cube can be oriented in order to reproduce the observed
477: morphology. The orientation of the structure is thus also
478: determined. The line intensities and other relevant quantities are
479: then obtained after projection onto the plane of the sky. Note that
480: because we lack velocity information, we cannot specify if an ansa is
481: in front of or behind the main nebula.
482:
483: The structure adopted for NGC 6369 is shown in Fig.\,\ref{strut3D} as
484: it is oriented relative to the observer. The choice of the structure
485: was based on the density map obtained from the observations. As shown
486: in \cite{M00}, closed shells were not able to reproduce the decrease
487: in the density profile in the central regions of NGC\,3132. For the
488: same reason, we adopt an hour-glass shape for the main structure of
489: NGC\,6369, as its density map shows the same decrease in the central
490: regions. We include random fluctuations in the density grid to
491: simulate the clumpiness observed in images, as well as a density gradient
492: along the main axis of the bipolar structure. The fluctuations were
493: set to vary between the maximum and the minimum densities measured
494: from the observational data. The rotation angles relative to the x, y
495: and z axis are 70$^o$, 10$^o$, and 0$^o$ respectively, with the
496: symmetry axis of the main structure being x. The model resolution of
497: $100^3$ cells was limited by computer memory.
498:
499: The ionizing spectrum adopted for the central source is a modified
500: blackbody with a break at 54.4eV. This was necessary to fit both the
501: [OIII] and HeII line intensities, the last one being particularly
502: sensitive to the depth of the break. This break is also present in
503: the more precise theoretical atmospheric models presented by Rauch et
504: al. (2003). The cutoff at 54.4\,eV is one of the most prominent
505: features that differentiates this spectrum from that of a blackbody,
506: followed by the 13.6\,eV hydrogen cutoff and absorption lines. The effective
507: temperature and luminosity of this modified blackbody are listed in
508: Table\,\ref{mod-res}.
509:
510: \clearpage
511: \begin{figure}
512: %\includegraphics[width=\columnwidth]{n6369_strut.ps}
513: \includegraphics[width=\columnwidth]{f8.eps}
514: \caption{Iso-density plot of the 3-D structure used for the NGC\,6369
515: model calculations. In the main image N is up, E to the left. The
516: small insets view the structure along the other two axes, allowing the
517: 3-D structure and orientation of the central object and the ansae to be
518: clearly seen.\label{strut3D}}
519: \end{figure}
520: \clearpage
521:
522: \subsection{Model Results}
523:
524: We present here the main results obtained with the photoionization
525: model. The total line intensities are given in Table\,\ref{mod-res},
526: as well as the fitted abundances and ionizing star
527: characteristics. Projected line images for the most important
528: transitions are shown in Fig.\,\ref{ims-mod}.
529:
530: The model image size is fitted to the observed size for the
531: [NII]$\lambda658.4\,nm$ line, as well as constrained by the absolute
532: $H\beta$ flux. The distance of 1550\,pc we determine is also
533: constrained by fitting all the line fluxes and the shape of the nebula
534: simultaneously, and for this reason is fundamentally different from
535: classical distance determination methods, as it needs no assumptions
536: to be made. One cannot increase the distance, and simply make the
537: nebula larger to fit the optical image, as the line ratios would
538: change significantly. 3-D ionization models such as this one cannot
539: simply be scaled, thus placing much more stringent constraints on all
540: derived parameters, including the distance. Note that this model
541: distance is within one adjusted standard deviation of the average of
542: all the observationally determined distances given in
543: Sec.\,1. Fig.\,\ref{mod-comp} shows the observed [NII] image with the
544: corresponding model image contours overlaid for the obtained distance.
545:
546: \begin{figure*}[!ht]
547: %\includegraphics[]{figs_mod_6369-1.ps}
548: \includegraphics[]{f9.eps}
549: \caption{Projected line images obtained with the photoionization
550: model. N is up, E to the left.\label{ims-mod}}
551: \end{figure*}
552:
553: \clearpage
554: \begin{figure}
555: %\includegraphics[width=\columnwidth]{cont_over_mod_6369-1.ps}
556: \includegraphics[width=\columnwidth]{f10.eps}
557: \caption{Contours of the model [NII] image over the observed image for d=1550
558: pc. N is up, E to the left. \label{mod-comp}}
559: \end{figure}
560: \clearpage
561:
562: \begin{table}
563: \begin{center}
564: \caption{Observed and model line fluxes and model central star
565: properties.\label{mod-res}}
566: \begin{tabular}{lll}\hline
567:
568: & {Observed}
569: & {Model}
570: \\\hline\hline
571:
572: &
573: &
574: \\
575: {$T_*$ (K)} & {70kK-106kK} & {91kK}\\
576: {$L_*/L_{\odot}$} & {-} & {8100}\\
577: {Density} & {1000-5400} & {300-5400}\\
578: {He/H} & {-} & {$1.14\times 10^{-1}$}\\
579: {C/H} & {-} & {$3.3\times 10^{-4}$}\\
580: {N/H} & {-} & {$1.0\times 10^{-4}$}\\
581: {O/H} & {-} & {$6.1\times 10^{-4}$}\\
582: {Ne/H} & {-} & {$6.5\times 10^{-5}$}\\
583: {S/H} & {-} & {$7.5\times 10^{-6}$}\\
584: {log($H\beta$)} & {-12.14} & {-12.16}\\
585: {[NeIII]386.8$^{a}$}& {0.9}& {0.9}\\
586: {[NII]658.4} & {0.73} & {0.78}\\
587: {[OIII]500.7}& {12.41} & {12.33}\\
588: {HeII468.6} & {0.014} & {0.013}\\
589: {HeI587.6} & {0.163} & {0.169}\\
590: {[OI]630.0} & {0.052} & {0.048}\\
591: {[SII]671.7}& {0.044} & {0.047}\\
592: {[SII]673.1}& {0.066} & {0.063}\\
593: &
594: &
595:
596: \\\hline
597:
598: \end{tabular}
599:
600: $^{a}$ Value obtained from \cite{PSM01}
601: \end{center}
602: \end{table}
603: \clearpage
604: The [SII] ratio map obtained with the model is shown in Fig.\,\ref{s2ratio}.
605:
606: \clearpage
607: \begin{figure}
608: %\includegraphics[width=\columnwidth]{s2_ratio_6369.ps}
609: \includegraphics[width=\columnwidth]{f11.eps}
610: \caption{The [SII] line ratio obtained with the model. N is up, E to the left.\label{s2ratio}}
611: \end{figure}
612: \clearpage
613:
614: \section{Discussion and conclusions}
615:
616: We have presented spectrophotometric maps of NGC\,6369. These maps
617: provided spatially resolved information for many emission lines and
618: precise total fluxes for the whole nebula. The images produced with
619: this technique were used to study the nebula with the usual diagnostic
620: ratios. Each image was corrected for reddening pixel by pixel using
621: the H$\alpha $/H$\beta$ image. This correction lead to some
622: significant differences between the corrected and uncorrected images,
623: as was shown in Figs.\,5 \& 7.
624:
625: The H$\alpha $/H$\beta$ map shows some interesting features. In
626: Fig.\,4 we show that the extinction is not uniform across the nebula,
627: as it varies between 6 and 18. The structure follows the main nebular
628: morphology, which could indicate that dust and neutral material are
629: present. This is to be expected since the nebula shows a very clumpy
630: structure, which may cause shadowing within the gas, allowing for the
631: survival of grains and neutral material. The prominent features on
632: either side of the nebula do not show significant differences in
633: extinction when compared to the main region.
634:
635: The temperature map derived from the observed integrated line of sight
636: intensities showed some structure but due to the low signal to noise
637: ratio, this should be treated with caution. Within the errors and
638: resolution of our observations the temperature can be considered to be
639: constant across the nebula.
640:
641: We also show that the density map indicates a decrease in density
642: for the central regions. This decrease, as has been discussed by
643: \cite{M00}, is not compatible with a closed shell structure. Based on
644: this map we propose an hour-glass structure for the main nebula, which
645: has reproduced all the observational features of NGC\,6369. This
646: indicates that the one needs to use the [SII] ratio in addition to images
647: to distiguish between open and closed structures.
648:
649: The position of the outer condensations or ansae --which are off-set
650: by 30$^o$ from the main nebular symmetry axis-- was obtained by
651: matching model images with the observed line images
652: (Fig.\,\ref{maps_full}) as well as matching the model [SII] density
653: map with the observed map (Fig.\,\ref{dens_map}). The presence of
654: these condensations can be accounted for by earlier ejection of matter
655: and precession can account for their deviation from the symmetry
656: axis. Many authors have dealt with these issues; for more complete
657: discussions and models see for example
658: \cite{CFJ96}, \cite{CFLJ95}, \cite{G97}, and \cite{LP97}.
659:
660:
661: Some of the parameters we have derived (and use as a sanity check)
662: are:
663:
664: The total luminosity of the observed lines,
665: L$_l$\,=\,1150\,L$_{\odot}$; that derived from the SED by integrating
666: the F$_{\lambda}$ curve, L$_o$\,=\,7.1$\,\times$\,10$^{-4}$\,d$^2$
667: where d is the distance to NGC\,6369 in pc, resulting in
668: L$_o$\,=\,1700\,L$_{\odot}$ . Correcting this value using the method
669: of Myers et al. (1987) we obtain 2550\,L$_{\odot}$. The luminosity of
670: the central star is L$_{tot}$\,=\,8100\,L$_{\odot}$ so that the ratios
671: of the line and optical+IR luminosities to the total luminosity are
672: respectively: L$_l$/L$_{tot}$\,=\,0.14 and
673: L$_o$/L$_{tot}$\,=\,0.3. Assuming that the absorbed flux is
674: re-radiated in the IR, mainly the IRAS bands, the integrated
675: optical+IR luminosity indicates that about 70\% of the UV flux (Lyman
676: continuum) escapes from the nebula. This is not implausible as the
677: nebula is open at it's ``poles'' and is clumpy, allowing the radiation
678: to pass through the many ``holes''.
679:
680: From the input matter distribution and abundances used in the model,
681: we calculate the mass of the nebular ionized gas to be
682: $M_{neb}\,=\,1.8\,M_{\odot}$. If we now use our values of luminosity
683: and temperature for the central star and compare them with the
684: evolutionary models of
685: \cite{B95} we can obtain the mass of the central core. Using this
686: procedure we can see from Fig.\,\ref{evol_track} that the best fitting
687: track for our data corresponds to a mass somewhat higher than
688: $M_{core}\,=\,0.625 M_{\odot}$, say, $\approx\,0.65\,M_{\odot}$, and
689: to an initial mass of $M_{0}=3 M_{\odot}$. If we sum our nebular and
690: core masses, we obtain $M_{core}\,+\,M_{neb}\,=\,2.4\,M_{\odot}$
691: approximately. If we use the line intensity errors as a measure of the
692: goodness of the model results and combine the luminosity and
693: temperature errors we obtain approximately 20-25\% for the uncertainty
694: in $M_{core}\,+\,M_{neb}$, placing the obtained initial stellar mass
695: within the uncertainty of our determination. The value for $M_{neb}$
696: calculated from the model input density is a lower limit estimate of
697: the total nebular mass as at least some material is present in the
698: form of dust, as indicated qualitatively by our extinction map
699: structure. Additionally, there may be neutral gas for wich we have no
700: constraints.
701:
702: Note that typical parameters for a WC4 PN central star from \cite{A64}
703: are T\,=\,90kK, and R\,=\,0.4\,R$_{\odot}$, very close to our derived
704: values of 91000\,K and 0.4\,R$_{\odot}$. Using
705: g\,=\,G\,.\,M\,/\,R$^2$ in CGS units, we obtain log(g)\,=\,5.1, which
706: is much lower than the value of 5.8 derived from the depth of our He
707: break. This is due to the fact that the central star has an extended
708: atmosphere, again showing the consistency of the model results.
709:
710: \clearpage
711: \begin{figure}
712: \includegraphics[width=\columnwidth]{f12.eps}
713: \caption{Comparison of model temperature and luminosity obtained for the
714: central star with model tracks calculated by \cite{B95}.\label{evol_track}}
715: \end{figure}
716: \clearpage
717:
718: Using our photoionization code and the proposed structure we obtained
719: a complete 3-D model for the NGC\,6369. The fitted model line
720: intensities show excellent agreement with the observed values. The
721: obtained distance of d=1550 pc, is well within the range present in
722: the literature obtained from different methods. The model temperature
723: for the ionizing star is similar to the Zanstra (He) value discussed
724: in \S3.3. The temperature and luminosity values obtained from the
725: model as well as the total nebular plus central star mass show good
726: agreement with the stellar evolution models of \cite{B95}.
727:
728: Using multiple long slit spectroscopy, we can determine accurate
729: distances, 3-D structures, abundances, ionized masses, central star
730: masses, luminosities, and temperatures to any spatially resolved
731: emission line nebula, assuming there are no strong shocks or extreme
732: morphologies involved.
733:
734: \begin{acknowledgements}
735: HM acknowledges support from Fapesp grant 00/03126-5, and
736: NOAO's Science Fund.
737: \end{acknowledgements}
738:
739:
740:
741: \begin{thebibliography}
742:
743: \bibitem[Acker, Koppen, Stenholm, \& Jasniewicz\,(1989)]{AKSJ1989}Acker, A.,
744: Koppen, J., Stenholm, B., \& Jasniewicz, G. 1989, A\&AS,
745: 80, 201
746: \bibitem[Acker et al.\,(1992)]{AO92}Acker, A., Ochsenbein, F., Stenholm, B.,
747: Tylenda, R., Marcout, J., \& Schohn, C. 1992, Strabourg-ESO Catalogue of
748: Galactic PNe.
749: \bibitem[Allen, 1964]{A64}Allen, C. W., 1964 AQ, p216
750:
751: \bibitem[Aller \& Keyes\,(1987)]{AK87}Aller, L. H., \& Keyes, C. D. 1987,
752: ApJS, 65, 405
753: \bibitem[Amnuel, Guseinov, Novruzova, \& Rustamov\,(1984)]{AGNR84}Amnuel, P. R.,
754: Guseinov, O. K., Novruzova, K. I. \& Rustamov, I. S.1984,
755: ApSS, 107, 19
756:
757: \bibitem[Balick\,(1987)]{B87} Balick, B.\ 1987, \aj, 94, 671
758:
759: \bibitem[Bl\"{o}cker(1995)]{B95} Bl\"{o}cker, T.\ 1995, \aap,
760: 299, 755
761:
762: \bibitem[Cliffe, Frank, \& Jones\,(1996)]{CFJ96} Cliffe, J.~A.,
763: Frank, A., \& Jones, T.~W.\ 1996, MNRAS, 282, 1114
764:
765: \bibitem[Cliffe, Frank, Livio, \& Jones\,(1995)]{CFLJ95} Cliffe,
766: J.~A., Frank, A., Livio, M., \& Jones, T.~W.\ 1995, ApJ, 447, 49
767:
768: \bibitem[Corradi, Sch{\" o}nberner, Steffen, \&
769: Perinotto\,(2003)]{CSSP03} Corradi, R. L. M., Sch{\" o}nberner,
770: D., Steffen, M., \& Perinotto, M. 2003, MNRAS, 340, 417
771:
772:
773: \bibitem[de Robertis, Dufour, \& Hunt(1987)]{D87} de
774: Robertis, M.~M., Dufour, R.~J., \& Hunt, R.~W.\ 1987, \jrasc, 81, 195
775:
776: \bibitem[Filippenko(1982)]{F82} Filippenko, A.~V.\ 1982,
777: \pasp, 94, 715
778:
779:
780: \bibitem[Garc{\'i}a-Segura\,(1997)]{G97} Garc{\'i}a-Segura, G.\ 1997,
781: \apjl, 489, L189
782:
783: \bibitem[Gathier, Pottasch, \& Goss\,(1986)]{GPG86}Gathier, R., Pottasch, S. R.,
784: \& Goss, W. M. 1986, AAp, 157, 191
785: \bibitem[Gathier, \& Pottasch\,(1988)]{GP88} Gathier, R., \& Pottasch, S.~R.\ 1988
786: , A\&A, 197, 266
787: \bibitem[Gathier, \& Pottasch\,(1989)]{GP89}Gathier, R., \& Pottasch, S. R. 1989,
788: A\&A, 209, 369
789:
790: \bibitem[Gruenwald, Viegas, \& Broguiere\,(1997)]{GVB97}
791: Gruenwald, R., Viegas, S.~M., \& Broguiere, D.\ 1997, \apj, 480, 283
792:
793: \bibitem[Gruenwald, \& Viegas\,(2000)]{GV00} Gruenwald, R., \& Viegas, S.~M.\
794: 2000, ApJ, 543, 889
795:
796: \bibitem[Harman, \& Seaton(1966)]{HS66} Harman, R.F., \& Seaton, M.~J. 1966,
797: MNRAS, 132, 15
798:
799: \bibitem[Livio, \& Pringle\,(1997)]{LP97} Livio, M., \& Pringle,
800: J.~E.\ 1997, ApJ, 486, 835
801:
802: \bibitem[McCall\,(1984)]{M84}McCall, M. L. 1984, MNRAS, 208, 253
803:
804: \bibitem[Meatheringham, Wood, \& Faulkner\,(1988)]{MWF88}Meatheringham, S. J.,
805: Wood, P. R., \& Faulkner, D. J. 1988, ApJ, 334,
806: 862
807: \bibitem[Monteiro, Gruenwald, Morisset, \& Viegas\,(2002)]{M02}Monteiro, H.,
808: Gruenwald, R., Morisset,
809: C., \& Viegas, S.~M.\ 2002, Revista Mexicana de Astronom{\'i}a y Astrof{\'i}sica
810: Conference Series, 12, 170
811:
812: \bibitem[Monteiro, Morisset, Gruenwald, \& Viegas\,(2000)]{M00} Monteiro, H.,
813: Morisset, C., Gruenwald,
814: R., \& Viegas, S.~M.\ 2000, \apj, 537, 853
815:
816: \bibitem[Myers, et al.\,1987)]{M87} Myers, P.C. et al. 1987, ApJ, 319, 340
817:
818: \bibitem[Osterbrock(1989)]{O89} Osterbrock, D.~E.\ 1989,
819: Research supported by the University of California, John Simon Guggenheim
820: Memorial Foundation, University of Minnesota, et al.~Mill Valley, CA,
821: University Science Books, 1989, 422 p.,
822:
823:
824: \bibitem[Pe{\~ n}a, Stasi{\' n}ska, \& Medina(2001)]{PSM01} Pe{\~ n}a, M., Stasi{\' n}ska, G., \& Medina, S.\ 2001, \aap, 367, 983
825:
826: \bibitem[Perinotto, Purgathofer, Pasquali, \& Patriarchi\,(1994)]{PPPP}
827: Perinotto, M., Purgathofer, A., Pasquali, A., \& Patriarchi, P. 1994,
828: A\&AS, 107, 481
829:
830: \bibitem[Rifman, \& McKinnon\,(1974)]{RM74}Rifman, S.S., \& McKinnon, D.M.,
831: \char`\"{}Evaluation of Digital Correction Techniques for ERTS Images; Final
832: Report\char`\"{}, Report 20634-6003-TU-00,
833: TRW Systems, Redondo Beach, CA, July 1974
834:
835: \bibitem[Rauch, Deetjen, Dreizler, \& Werner\,(2000)]{RDDW00}Rauch, T.,
836: Deetjen, J. L., Dreizler, S., \& Werner, K. 2000, Asymmetrical Planetary Nebulae
837: II: From origins to microstructures, ASP Conference Series 199, 337
838:
839: \bibitem[Park, \& Schowengerdt\,(1983)]{PS83}Park, S., \& Schowengerdt, R. 1983
840: \char`\"{}Image Reconstruction by
841: Parametric Cubic Convolution\char`\"{}, Computer Vision, Graphics
842: \& Image Processing 23, 256
843:
844: \bibitem[Schmeja, \& Kimeswenger\,(2002)]{SK02} Schmeja, S., \& Kimeswenger, S.\
845: 2002, Revista Mexicana de Astronom{\'i}a y Astrof{\'i}sica
846: Conference Series, 12, 176
847:
848: \bibitem[Schwarz, Corradi, \& Melnick\,(1992)]{SCM92}Schwarz, H. E.,
849: Corradi, R. L. M., \& Melnick, J. 1992, AApS, 96, 23
850:
851: \bibitem[Seaton\,(1979)]{S79}Seaton, M. J. 1979, MNRAS, 187, 73
852:
853: \bibitem[Tylenda, Acker, \& Stenholm\,(1993)]{TAS93}Tylenda, R., Acker, A.,
854: \& Stenholm, B. 1993, AApS, 102, 595
855:
856: \bibitem[Wilson\,(1953)]{W53}Wilson, R. E.,
857: 1953, GCRV, Carnegie Inst. Publ. 601
858:
859: \end{thebibliography}
860:
861: \end{document}
862:
863:
864:
865:
866:
867: