1: \documentclass{aa}
2: %\documentclass[referee]{aa}
3: \usepackage{graphicx}
4: \usepackage{natbib}
5: \usepackage{latexsym}
6:
7: \newcommand {\ha}{H$\alpha$}
8: \newcommand {\hb}{H$\beta$}
9: \newcommand {\nii}{[\ion{N}{2}]}
10: \newcommand {\hii}{\ion{H}{2}}
11:
12: \begin{document}
13:
14: \title{Diffuse X-ray emission from the planetary nebula NGC\,7009}
15:
16: \author{Mart\'{\i}n A.\ Guerrero
17: \and Robert A. Gruendl
18: \and You-Hua Chu}
19:
20: \institute{Astronomy Department, University of Illinois,
21: 1002 W. Green Street, Urbana, IL 61801, USA}
22:
23: \thanks{
24: Based on observations obtained with {\sl XMM-Newton}, an ESA science
25: mission with instruments and contributions directly funded by ESA
26: Member States and NASA. }
27:
28: \offprints{Mart\'{\i}n A.\ Guerrero, \email{mar@astro.uiuc.edu}}
29:
30: \date{Received / Accepted }
31:
32:
33: \abstract{
34:
35: {\sl XMM-Newton} EPIC observations of the planetary nebula (PN)
36: NGC\,7009, the Saturn Nebula, have detected extended X-ray emission
37: in its central cavity.
38: The diffuse X-ray emission must originate in the shocked fast
39: stellar wind.
40: The spectra show that the temperature of the hot gas is
41: 1.8~10$^6$~K.
42: The rms density derived from the volume emission measure is a few tens
43: H-atom~cm$^{-3}$.
44: The hot gas does not appear over-pressurized with respect to the
45: nebular shell.
46: NGC\,7009 represents an evolutionary stage at which the
47: influence of the hot gas in the PN interior starts to decline
48: due to the diminishing strength of the fast stellar wind and the
49: expansion of the central cavity.
50:
51: \keywords{
52: (ISM:) planetary nebulae: general --
53: (ISM:) planetary nebulae: individual: NGC\,7009 --
54: stars: winds, outflows --
55: X-rays: ISM
56: }
57: }
58:
59: \authorrunning{Guerrero et al.}
60:
61: \maketitle
62:
63: \section{Introduction}
64:
65: It has been commonly accepted that the shell structure of a
66: planetary nebula (PN) is formed by the current fast stellar wind
67: sweeping up the previous asymptotic giant branch (AGB) wind
68: \citep[e.g.][]{Kwok83}.
69: In this interacting-stellar-winds model, a PN will comprise a central
70: cavity filled with shocked fast wind at temperatures of 10$^6-10^8$ K,
71: a dense shell of swept-up AGB wind at 10$^4$ K, and an outer envelope
72: of unperturbed expanding AGB wind.
73: The shocked fast wind has a high temperature but does not emit
74: appreciably in X-rays because its density is low.
75: At the interface between the shocked fast wind and the dense nebular
76: shell, heat conduction lowers the temperature and mass evaporation
77: raises the density of the shocked fast wind, producing optimal
78: conditions for X-ray emission \citep{Weav77}. Thus, soft X-ray
79: emission from the shocked fast wind in a PN interior is expected to
80: peak near the inner wall of the nebular shell.
81:
82: The $ROSAT$ X-ray satellite had been used to search
83: for diffuse X-ray emission from PNe with limited success
84: \citep{GCG00}.
85: It is only recently that the {\it Chandra X-ray Observatory} has
86: unambiguously detected diffuse X-ray emission in the central
87: cavities of three young PNe: BD\,+30$^\circ$3639 \citep{Kast00},
88: NGC\,6543 \citep{CGGWK01}, and NGC\,7027 \citep{Kast01}.
89: In the case of NGC\,6543 where the X-ray emission is well-resolved,
90: a clear limb-brightening is observed.
91: The plasma temperatures implied by their X-ray spectra are a few
92: $10^6$~K.
93: {\it Chandra} observations of the evolved PN NGC\,7293 (the Helix
94: Nebula), on the other hand, does not show any diffuse X-ray emission.
95: These results suggest that the hot gas content of a PN diminishes as
96: the PN ages.
97: To confirm this hypothesis, PNe at intermediate evolutionary stages
98: need to be observed.
99:
100: We have examined all pointed and serendipitous observations of
101: PNe in the $ROSAT$ archive and found that NGC\,7009, the Saturn
102: Nebula, hosts an X-ray source with spectral characteristics of
103: plasma emission.
104: NGC\,7009 is more evolved than the three PNe with diffuse X-ray
105: emission because of its larger nebular size and dynamical age.
106: Consequently, NGC\,7009 is an excellent candidate not only for diffuse
107: X-ray emission, but also as a PN at intermediate evolutionary stages.
108: We have obtained X-ray observations of NGC\,7009 with the
109: {\it XMM-Newton Observatory}, and found that the X-ray emission
110: is indeed extended.
111: This paper reports our analysis of the diffuse X-ray emission from
112: NGC\,7009.
113:
114: \section{Observations}
115:
116: NGC\,7009 was observed with the {\it XMM-Newton Observatory}
117: in Revolution 255 on 2001 April 30 using the EPIC/MOS1, EPIC/MOS2,
118: and EPIC/pn CCD cameras.
119: The two EPIC/MOS cameras were operated in the Full-Frame Mode for a
120: total exposure time of 39.4 ks, while the EPIC/pn camera was operated
121: in the Extended Full-Frame Mode for a total exposure time of 31.6 ks.
122: The EPIC/pn observations started about 4 ks later than the EPIC/MOS
123: observations because of a larger overhead required by the
124: EPIC/pn, but both observations concluded at the same time.
125: For all observations, the Medium filter was used.
126: At energies below 1 keV, the point spread function (PSF) of the EPIC
127: instruments has a small core with FWHM $\sim 6''$ but extended
128: wings with a half energy width of $\sim 15''$.
129: For the EPIC/MOS and EPIC/pn CCDs, the pixel sizes are 1\farcs1 and
130: 4\farcs1, and the energy resolutions are $\sim60$ eV and $\sim100$ eV
131: at 1 keV, respectively (Dahlem 1999; Str\"uder et al.\ 2001).
132:
133: We received the {\it XMM} pipeline products which included the raw
134: and processed data of all instruments.
135: These data were further processed using the {\it XMM-Newton} Science
136: Analysis Software (SAS version 5.2.0) and the calibration files from
137: the Calibration Access Layer available on 2002 January 24.
138:
139: The event files were further screened to eliminate bad events, such as
140: those due to charged particles, and periods of high background.
141: For the EPIC/MOS observations, only events with CCD patterns 0--12
142: (similar to $ASCA$ grades 0--4) were selected;
143: for the EPIC/pn observation, only events with CCD pattern 0 (single
144: pixel events) were selected.
145: To assess the background rate, we binned the counts over 25 s time
146: intervals for each instrument in the 10--12 keV energy range, where
147: the counts were dominated by the background.
148: The time intervals with high background, i.e.\ count rates $\ge 0.4$
149: cnts~s$^{-1}$ for the EPIC/MOS or $\ge 1.5$ cnts~s$^{-1}$ for the
150: EPIC/pn, were discarded.
151: The resulting exposure times are 35.0 ks, 35.4 ks, and 26.2 ks for
152: the EPIC/MOS1, EPIC/MOS2, and EPIC/pn observations, respectively.
153: A total of $480\pm25$ cnts, $470\pm25$ cnts, and $1,610\pm45$ cnts
154: from NGC\,7009 were detected in the EPIC/MOS1, EPIC/MOS2, and EPIC/pn
155: observations, respectively.
156:
157:
158: \section{Spatial Analysis}
159:
160: We have analyzed the spatial distribution of the X-ray emission
161: from NGC\,7009 and compared it to the optical nebula.
162: As the data analysis is fairly complex, we present images at
163: various stages of the processing in Figure 1.
164: First, the event files from the three EPIC instruments are merged
165: together to construct a raw EPIC image with a pixel size of 1\farcs5
166: in the 0.15--1.25 keV energy band (Fig.~1a).
167: As shown in \S 4, this energy band includes all photons detected
168: from NGC\,7009.
169: The raw EPIC image is then adaptively smoothed using Gaussian
170: profiles with FWHM ranging from 5$''$ to 17$''$.
171: A grey-scale presentation of this adaptively smoothed EPIC image
172: is shown in Fig.~1b.
173: The contours from this smoothed EPIC image are overplotted on an
174: H$\alpha$ image of NGC\,7009 taken with the {\it Hubble Space
175: Telescope (HST)} Wide Field Planetary Camera 2 (WFPC2) and presented
176: in Fig.~1c.
177: Finally, to improve the spatial resolution, this smoothed EPIC image
178: is deconvolved using the Lucy-Richardson algorithm and a theoretical
179: EPIC PSF provided in the {\it XMM-Newton} Calibration Access Layer.
180: The contours extracted from the PSF-deconvolved EPIC image are
181: overplotted on the {\it HST} WFPC2 H$\alpha$ image in Fig.~1d.
182:
183: The X-ray emission from NGC\,7009 is clearly extended.
184: Its spatial distribution in the adaptively smoothed EPIC image
185: (Fig.~1b and 1c) can be described by an elliptical Gaussian with
186: FWHM $\sim 23'' \times 14''$ and a major axis along the position
187: angle $\sim70^\circ$.
188: There is an excellent correspondence between the X-ray emission
189: and the nebular shell of NGC\,7009.
190: The {\it HST} WFPC2 H$\alpha$ image of NGC\,7009 shows a bright
191: $25'' \times 10''$ inner shell surrounded by a $25'' \times
192: 20''$ envelope, with two ansae extending from the tips of the
193: major axes of the inner elliptical shell out to $25''$ from the
194: central star.
195: The position and orientation of the 50\% contour of the smoothed EPIC
196: image matches almost perfectly the inner shell of the nebula
197: (Fig.~1c), indicating that the X-ray emission peaks within the central
198: cavity.
199: The PSF-deconvolved EPIC image further shows that the majority of the
200: X-ray emission originates from the central cavity.
201: As the EPIC PSF has extended wings and our PSF-deconvolution is
202: limited by the number of counts and the validity of the theoretical
203: PSF adopted, it is most likely that all diffuse X-ray emission from
204: NGC\,7009 is confined within the inner shell, similar to what is
205: observed in BD\,+30$^\circ$3639 and NGC\,6543.
206: Furthermore, NGC\,7009 has a sharp-edged inner shell similar to that of
207: NGC\,6543, which implies an expansion driven by the pressure of the hot
208: gas in the central cavity \citep{CGGWK01}.
209: Therefore, we suggest that the hot, X-ray-emitting gas in NGC\,7009 is
210: also confined within its central cavity, but the detailed distribution
211: of the hot gas cannot be determined from the currently available
212: X-ray observations.
213: In particular, it is not possible to address whether X-ray emission from
214: its central star is present or not.
215:
216: \section{Spectral Analysis}
217:
218: Spectra of NGC\,7009 have been extracted from the event files
219: of the three {\sl XMM-Newton} EPIC cameras separately because
220: they have different instrumental response functions.
221: A $50\arcsec \times 40\arcsec$ elliptical source aperture is used
222: to include all X-ray emission detected from NGC\,7009.
223: An elliptical annular region exterior to and $\sim10$ times as
224: large as the source aperture is used to assess the background
225: level.
226: The background-subtracted spectra of NGC\,7009 are shown in
227: Fig.~2-{\it left}.
228:
229: The EPIC/pn spectrum peaks at $\sim0.55$ keV and drops abruptly to a
230: faint plateau between 0.7 and 1.0 keV, but stays essentially flat
231: from 0.5 keV down to 0.3 keV.
232: The EPIC/MOS spectra also peak at $\sim0.55$ keV, but have
233: smaller numbers of counts because of lower sensitivities.
234: The detailed differences between the EPIC/pn and EPIC/MOS spectra
235: reflect the different instrumental responses among these cameras
236: and the low S/N ratios in the latter spectra.
237: Neither the EPIC/MOS spectra nor the EPIC/pn spectrum shows
238: any detectable emission at energies greater than 1.25 keV.
239:
240: The EPIC spectra of NGC\,7009 are clearly inconsistent with a
241: power-law or blackbody spectrum.
242: Thus, the X-ray emission from NGC\,7009 must be thermal plasma
243: emission and the spectral peak at $\sim$ 0.55 keV corresponds
244: to the He-like triplet of O~{\sc vii} at $\sim0.57$ keV.
245: We can model the observed spectra in order to determine the
246: physical conditions of the X-ray-emitting plasma and the amount
247: of foreground absorption.
248: We have adopted the thin plasma emission model of \citet{RS77}
249: and the nebular abundances determined from optical and UV spectra
250: of NGC\,7009.
251: The abundances of He, C, N, O, Ne, and S relative to hydrogen
252: by number have been measured to be 0.13, 2.9~10$^{-4}$,
253: 2.2~10$^{-4}$, 4.3~10$^{-4}$, 7.2~10$^{-5}$, and 1.9~10$^{-5}$,
254: respectively (Kwitter \& Henry 1998).
255: Accordingly, we have adopted nebular abundances, relative to the solar
256: values \citep{AG89}, of 1.33, 0.8, 1.9, 0.7, 0.6, and 1.2 for He, C,
257: N, O, Ne, and S, respectively, and 1.0 for the other elements.
258: We have assumed solar abundances for the foreground interstellar
259: absorption, and adopted absorption cross-sections from \citet{BM92}.
260:
261: The spectral fits are carried out by folding the Raymond-Smith model
262: spectrum through the corresponding EPIC/MOS1, EPIC/MOS2, and EPIC/pn
263: response matrices, and fitting the modeled spectra to the observed
264: EPIC spectra in the 0.3--1.5 keV energy range using the $\chi^2$
265: statistics.
266: The best-fit model, overplotted on the EPIC/pn and EPIC/MOS spectra in
267: Fig.~2-{\it left}, has a plasma temperature of $T$ =
268: 1.8~10$^6$ K (or $kT$ = 0.152 keV), an absorption column density
269: of $N_{\rm H}$ = 8.1~10$^{20}$~cm$^{-2}$, and a volume emission
270: measure of 1.9~10$^{54} d^2$ cm$^{-3}$ where $d$
271: is the distance to NGC\,7009 in kpc\footnote{
272: The distance estimates to NGC\,7009 range from 0.5 to 2.1 kpc.
273: In the following we will adopt a distance of 1.2 kpc.}.
274:
275:
276: The goodness of the spectral fits is illustrated by the plot
277: of reduced $\chi^2$ of the fits as a function of $kT$ and
278: $N_{\rm H}$ shown in Fig.~2-{\it right}.
279: The 99\% confidence contour spans
280: $T$ = 1.6--2.0~10$^6$ K (or $kT$ = 0.14--0.17 keV),
281: $N_{\rm H}$ = 2--17~10$^{20}$~cm$^{-2}$, and volume emission
282: measure 1.5--2.4~10$^{54} d^2$ cm$^{-3}$.
283: Assuming that the H\,{\sc I} column density is similar to the
284: absorption column density and adopting a typical gas-to-dust ratio
285: (Bohlin, Savage, \& Drake 1978), the corresponding visual extinction
286: is $A_V$ = $0.44^{+0.5}_{-0.3}$ mag, and the logarithmic extinction
287: at the H$\beta$ line is $c_{\rm H\beta}$ = $0.15^{+0.22}_{-0.15}$.
288: This extinction is consistent with other estimates of extinction
289: ($c_{\rm H\beta} = 0.24\pm0.06$, Lame \& Pogge 1996;
290: $c_{\rm H\beta} = 0.10^{+0.06}_{-0.08}$, Rubin et al.\ 2001).
291: The observed (absorbed) X-ray flux of NGC\,7009 in the 0.3--1.5 keV
292: energy range is 7.2~10$^{-14}$ ergs~cm$^{-2}$~s$^{-1}$, and the
293: unabsorbed (intrinsic) X-ray flux is 1.7~10$^{-13}$
294: ergs~cm$^{-2}$~s$^{-1}$.
295: The total X-ray luminosity in the 0.3--1.5 keV energy range is
296: $L_{\rm X} =$ 2.0~10$^{31} d^2$ ergs~s$^{-1}$ = 3~10$^{31}$
297: ergs~s$^{-1}$ for a distance of 1.2 kpc.
298:
299:
300: \section{Discussion}
301:
302: \noindent
303: {\sl XMM-Newton} EPIC observations of NGC\,7009 have shown that
304: the diffuse X-ray emission is confined within the central cavity.
305: The temperature of the X-ray-emitting gas, 1.8~10$^6$~K, is below
306: that expected for the shocked fast wind from its central star which
307: has a terminal velocity $\sim 2,800$ km~s$^{-1}$ \citep{CP89,IS01}.
308: Similar situation is observed in the other PNe with diffuse X-ray
309: emission: the plasma temperatures implied by the X-ray spectra of
310: BD\,+30$^\circ$3639 and NGC\,6543 are several times lower than the
311: expected post-shock temperature for their fast stellar wind
312: \citep{Kast00, CGGWK01}.
313: This temperature difference is expected if heat conduction is prevalent
314: at the interface between the shocked fast wind and the dense nebular
315: shell \citep{S94,ZP96}.
316:
317: In order to determine the role played by the hot gas in the nebular
318: dynamics of NGC\,7009, it is possible to derive the rms density and
319: pressure of the X-ray-emitting gas and compare them to those of the
320: surrounding cool nebular shell.
321: Assuming a prolate ellipsoidal central cavity, the volume occupied
322: by the X-ray-emitting gas is $\sim$6.7~10$^{51} (d/1.2)^3 \epsilon$
323: cm$^3$, where $\epsilon$ is the filling factor.
324: From this volume and the aforementioned volume emission measure
325: we derive an rms density of 20~$(d/1.2)^{-1/2}~\epsilon^{-1/2}$
326: cm$^{-3}$, and a total mass for the X-ray-emitting gas of
327: 8~10$^{-5}~(d/1.2)^{5/2}~\epsilon^{1/2} M_\odot$.
328: This rms density and the plasma temperature of 1.8~10$^6$~K imply that
329: the thermal pressure of the X-ray-emitting gas is
330: $\sim$4~10$^{-9}~(d/1.2)^{-1/2}~\epsilon^{-1/2}$ dynes~cm$^{-2}$.
331: The nebular shell surrounding the hot gas has a density of 5,000
332: cm$^{-3}$, a temperature of 10,500 K (Hyung \& Aller 1995; Rubin
333: et al.\ 2001), and thus a thermal pressure of
334: $\sim$10$^{-8}$ dynes~cm$^{-2}$.
335: The thermal pressure of the nebular shell is comparable or higher
336: than that of the X-ray-emitting gas unless the filling factor
337: $\epsilon$ is $\ll$0.2.
338:
339: The X-ray luminosity of NGC\,7009 and the density and pressure of the
340: hot gas in its interior are the lowest among the four PNe for wich
341: diffuse X-ray emission has been detected thus far.
342: These differences may be associated with the nebular evolution.
343: NGC\,7009 has a large nebular shell, 0.145$\times$0.06 pc in size,
344: and dynamical age, $\sim1,700$ yrs for an expansion velocity of
345: 17~km~s$^{-1}$ \citep{BLA94}.
346: Therefore, it appears to be in the most advanced evolutionary stage
347: among these four PNe.
348: At a more advanced evolutionary stage, NGC\,7293 does not have
349: any detectable diffuse X-ray emission (Guerrero et al.\ 2001).
350: The {\it Chandra} and {\it XMM-Newton} observations of these five
351: PNe suggest that PNe are brightest in X-rays when they are young
352: and NGC\,7009 may be at an evolutionary stage when the diffuse
353: X-ray emission starts to decline.
354: This evolutionary effect is expected as the stellar wind
355: diminishes and the central cavity grows.
356: More observations of PNe at different evolutionary stages are
357: needed to evaluate the duration of the pressure-driven phase in
358: PNe and its effects on the nebular dynamics.
359: As the evolved PNe are expected to have low X-ray surface
360: brightness, the high sensitivity of {\it XMM-Newton} is essential
361: in detecting such PNe.
362:
363:
364: \begin{acknowledgements}
365: We thank Steve Snowden for advice on the {\it XMM-Newton} data
366: reduction, Guillermo Garc\'{\i}a-Segura for enlightening
367: discussion on the X-ray emission from PNe, and Rosina Iping,
368: George Sonneborn, and Robert H.\ Rubin for useful discussion on
369: the physical properties of NGC\,7009.
370: We also thank the referee, Joel Kastner, for his constructive
371: comments.
372: This research was supported by the NASA grant NAG~5-10042.
373: \end{acknowledgements}
374:
375:
376:
377: \begin{thebibliography}{}
378:
379: \bibitem[Anders \& Grevesse(1989)]{AG89}
380: Anders, E., \& Grevesse, N.\ 1989, Geochimica et Cosmochimica
381: Acta 53, 197
382:
383: \bibitem[Balucinska-Church \& McCammon (1992)]{BM92}
384: Balucinska-Church, M., \& McCammon, D.\ 1992, ApJ, 400, 699
385:
386: \bibitem[Bohigas, L\'opez, \& Aguilar(1994)]{BLA94}
387: Bohigas, J., L\'opez, J.\ A., \& Aguilar, L.\ 1994, \aap, 291,
388: 595
389:
390: \bibitem[Bohlin, Savage, \& Drake 1978]{BSD78}
391: Bohlin, R.\ C., Savage, B.\ D., \& Drake, J.\ F.\ 1978, ApJ,
392: 224, 132
393:
394: \bibitem[Cerruti-Sola \& Perinotto(1989)]{CP89}
395: Cerruti-Sola, M., \& Perinotto, M.\ 1989, \apj, 345, 339
396:
397: \bibitem[Chu et al.(2001)]{CGGWK01}
398: Chu, Y.-H., Guerrero, M.\ A., Gruendl, R.\ A., Williams, R.\ M.,
399: \& Kaler, J.\ B.\ 2001, \apjl, 553, L69
400:
401: \bibitem[Dahlem (1999)]{D99}
402: Dahlem, M.\ 1999, XMM Users' Handbook
403:
404: \bibitem[Guerrero et al.(2000)]{GCG00}
405: Guerrero, M., Chu, Y.-H., \& Gruendl, R. A.\ 2000, ApJS, 129, 295
406:
407: \bibitem[Guerrero et al.(2001)]{GCGWK01}
408: Guerrero, M.\ A., Chu, Y.-H., Gruendl, R.\ A., Williams, R.\ M.,
409: \& Kaler, J.\ B.\ 2001, \apjl, 553, L55
410:
411: \bibitem[Hyung \& Aller(1995)]{HA95}
412: Hyung, S., \& Aller, L.\ H.\ 1995, MNRAS, 273, 973
413:
414: \bibitem[Iping \& Sonneborn(2001)]{IS01}
415: Iping, R.\ C., \& Sonneborn, G.\ 2001, American Astronomical
416: Society Meeting, 198, 1306
417:
418:
419: \bibitem[Kastner et al.(2000)]{Kast00}
420: Kastner, J.\ H., Soker, N., Vrtilek, S.\ D., \& Dgani, R.\ 2000,
421: \apjl, 545, L57
422:
423: \bibitem[Kastner et al.(2001)]{Kast01}
424: Kastner, J.\ H., Vrtilek, S.\ D., \& Soker, N.\ 2001, \apjl, 550,
425: L189
426:
427: \bibitem[Kwitter \& Henry(1998)]{KH98}
428: Kwitter, K.\ B., \& Henry, R.\ B.\ C. 1998, ApJ, 493, 247
429:
430: \bibitem[Kwok(1983)]{Kwok83}
431: Kwok, S.\ 1983, in IAU Symposium 103, Planetary Nebulae, ed. D.\
432: R.\ Flower (Dordrecht: Reidel), 293
433:
434: \bibitem[Lame \& Pogge(1996)]{LP96}
435: Lame, N.\ J., \& Pogge, R.\ W.\ 1996, AJ, 111, 2320
436:
437: \bibitem[Raymond \& Smith (1977)]{RS77}
438: Raymond, J.\ C., \& Smith, B.\ W.\ 1977, ApJS, 35, 419
439:
440: \bibitem[Rubin et al.(2001)]{Rubin01}
441: Rubin, R.\ et al.\ 2001, submitted to MNRAS
442:
443: \bibitem[Soker(1994)]{S94}
444: Soker, N.\ 1994, \aj, 107, 276
445:
446: \bibitem[Str\"uder et al.(2001)]{St01}
447: Str\"uder, L.\ et al.\ 2001, A\&A, 365, 18
448:
449: \bibitem[Weaver et al.(1977)]{Weav77}
450: Weaver, R., McCray, R., Castor, J., Shapiro, P., \& Moore, R.\ 1977,
451: ApJ, 218, 377
452:
453: \bibitem[Zhekov \& Perinotto(1996)]{ZP96}
454: Zhekov, S.\ A., \& Perinotto, M.\ 1996, \aap, 309, 648
455:
456: \end{thebibliography}
457:
458: \begin{figure*}
459: \centering
460: \includegraphics[width=15.5cm]{fig1.eps}
461: \caption{
462: Raw {\it (a)} and smoothed {\it (b)} {\it XMM-Newton} EPIC images of
463: NGC\,7009 in the 0.15--1.25 keV energy band.
464: X-ray contours from the smoothed EPIC image {\it (c)} and from the
465: PSF-deconvolved EPIC image {\it (d)} overplotted on an $HST$ WFPC2
466: H$\alpha$ image of NGC\,7009.
467: The contour levels are chosen at 10, 25, 50, 75, and 90\% of the emission
468: peak. }
469: \end{figure*}
470:
471: \begin{figure*}
472: \centering
473: \includegraphics[width=15.5cm]{fig2.eps}
474: \caption{
475: {\it (left)}
476: {\it XMM-Newton} EPIC/pn {\it (top)}, and EPIC/MOS1 and EPIC/MOS2 {\it
477: (bottom)} background-substracted spectra of NGC\,7009.
478: The spectra are overplotted with the best-fit Raymond \& Smith model
479: (1977) to the EPIC/pn and the combined EPIC/MOS spectra.
480: Both spectra and models are plotted in 30 eV wide energy bins to
481: preserve the EPIC spectral resolution ($\sim$ 60--100 eV).
482: {\it (right)}
483: $\chi^2$ grid plot of the spectral fit of the EPIC/pn and the combined
484: EPIC/MOS spectra.
485: The contours represent 68\%, 90\%, and 99\% confidence levels, and the
486: `+' sign corresponds to the parameters of the best-fit model.
487: }
488: \end{figure*}
489:
490: \end{document}
491:
492:
493:
494: