1: \documentstyle[12pt,aasms4,psfig]{article}
2: \begin{document}
3: \title{On the large escape of ionizing radiation from
4: GEHRs} \author{Marcelo Castellanos\altaffilmark{1}, {\'A}ngeles I. D{\'\i}az\altaffilmark{2}
5: \affil{Dpt. F\'{\i}sica Te{\'o}rica C-XI,
6: Universidad Aut{\'o}noma de Madrid, \\
7: E-28049 Cantoblanco,
8: Madrid, Spain\altaffiltext{1}{e-mail: marcelo@pollux.ft.uam.es}
9: \altaffiltext{2}{e-mail: angeles@pollux.ft.uam.es}}
10: Guillermo Tenorio-Tagle\altaffilmark{3,4}\affil{INAOE,\\
11: Apartado de Correos 51 y 216, 72000 Puebla, M\'exico}
12: \altaffiltext{3}{e-mail: gtt@inaoep.mx}
13: \altaffiltext{4}{IBERDROLA Visiting Professor for Science and Technology, U.A.M.}}
14: \begin{abstract}
15: A thorough analysis of well studied giant HII regions on galactic discs
16: for which we know the
17: ionizing stellar population, the gas metallicity
18: and the Wolf-Rayet population, leads to photoionization
19: models which can only match all observed line intensity ratios ([OIII] ,
20: [OII] , [NII], [SII] and [SIII] with respect to the intensity of H$\beta$), as
21: well as the H$\beta$ luminosity and equivalent width if one allows for an important escape of energetic ionizing
22: radiation. For the three regions presented here, the fractions of escaping Lyman continuum
23: photons amount to 10 to 73 \% and, in all cases, the larger
24: fraction of escaping photons has energies between 13.6 and 24.4 eV. These escaping photons clearly must have an important impact as a
25: source of ionization of the diffuse ionized gas (DIG) found surrounding many
26: galaxies, as well as of the intergalactic medium (IGM).
27: \end{abstract}
28: \keywords{galaxies: individual (NGC 628, NGC 1232, NGC 4258) --- galaxies: ISM --- HII regions --- intergalactic medium --- stars: Wolf-Rayet} \section{INTRODUCTION}
29: The issue of whether or not ionizing star clusters are the sources of
30: radiation responsible for the diffuse ionized gas (DIG) and, moreover, for
31: the background radiation at low redshift (Heckman et al 2001), is still
32: controversial. UV observations of four nearby starburst
33: galaxies (Leitherer et al. 1995; Hurwitz et al. 1997)
34: showed that more than 10\% of the ionizing photons could escape from these
35: galaxies. On the other hand, Dove et al. (2000) obtained from theoretical
36: arguments that 7\% of photons produced by OB
37: associations would have to escape into the DIG to sustain it and could also
38: be consistent with the estimated flux required to photoionize the Magellanic
39: Stream (Weiner \& Williams 1996; Bland-Hawthorn \& Maloney 1999). All these
40: aspects at low redshift are undoubtedly important to establish whether massive
41: stars in starburst galaxies can rival QSO's as sources of the
42: metagalactic background radiation at high redshifts (z $>$ 3; see Ricotti \&
43: Shull 2000). Furthermore, an estimate of the escape fraction at low redshift
44: is to provide an upper limit to the expected escape fraction at high redshift,
45: when the universe was much denser. Beckman et al. (2000) presented evidence,
46: based on the analysis of the H$\alpha$ luminosity function of HII region
47: populations in nearby spiral galaxies, that luminous HII regions may be matter
48: bounded. This would imply that these regions are an important source
49: of photons to the DIG.
50: Here we report strong evidence that points at Giant
51: Extragalactic HII Regions (GEHRs) to be matter bounded, and thus as an
52: important source of ionization at large distances from their exciting stars.
53: Our conclusions are based on evolutionary
54: synthesis models (Leitherer et al. 1999; hereafter ST99) applied to previously
55: analyzed GEHRs with observed Wolf-Rayet (WR) features (D{\'\i}az et al. 2000;
56: Castellanos et al. 2002; hereafter referred to as Paper I). The observed
57: parameters inside these regions, i.e. the fluxes and equivalent widths of the
58: WR features, the emission line intensities of [OII]($\lambda$3727),
59: [OIII]($\lambda$5007), [NII]($\lambda$6584), [SII]($\lambda$6717), and
60: [SIII]($\lambda$9069) relative to H$\beta$, the observed H$\beta$ luminosity and equivalent width, cannot be fitted
61: simultaneously unless an important fraction of the ionizing radiation escapes
62: from these regions.
63:
64: \section{MODELS AND RESULTS}
65: The analyzed GEHRs are regions H13 in NGC 628, CDT3
66: in NGC 1232 and 74C in NGC 4258. The three of them, particularly 74C
67: and H13, show spectral signatures of prominent Wolf-Rayet features.
68: The
69: presence of these stars allows to constrain the age of the ionizing
70: clusters to between 3Myr and 6Myr. These values depend on
71: the stellar mass loss rates and, consequently, on the stellar metallicity that
72: ultimately controls when massive O stars enter the Wolf-Rayet phase (Meynet
73: 1995). This metallicity has been assumed to be similar to that
74: found through the analysis of the
75: emission line spectrum (see Paper I). Once the age and metallicity
76: are known, a unique Spectral Energy Distribution (SED) can be
77: provided by evolutionary synthesis models.
78:
79: To estimate the age of the ionizing stellar clusters from the
80: measured strength and equivalent widths
81: of the observed WR features we used the models of Schaerer \& Vacca (1998).
82: Ages between 4 and 4.5 Myr, 3 and 3.5 Myr and 3.5 and 4 Myr are found for
83: regions H13, CDT3 and 74C with oxygen abundances of 0.2, 0.4 and 0.5 solar,
84: respectively.
85: We have then used the ST99 models to predict the SEDs of the ionizing
86: clusters.
87: These models assume the same stellar tracks as those used by Schaerer \&
88: Vacca (1998) and, at the metallicities here considered,
89: represent a good approach to the ionizing spectra (see Stasi{\'n}ska et
90: al. 2001).
91:
92: % although differ slightly in the treatment of the
93:
94: %stellar atmospheres. This difference only seems to affect the emergent
95:
96: %continuum flux at energies higher than 54 ev (He$^+$ continuum) and
97:
98: %therefore provide a nearly selfconsistent set of models.
99:
100:
101: The predicted SEDs were then used as input to the latest version of CLOUDY
102: (Ferland 1999). The models also require of an electron density (assumed to be constant
103: throughout the nebula),
104: the derived gas metallicity (see paper I) and an ionization parameter, U =
105: Q(H)/(4$\pi$R$^2$n$_e$c); where Q(H) is the total flux of ionizing photons
106: emitted by the cluster per second, R the distance from the ionizing
107: source to the illuminated face of the cloud, n$_e$ the electron density and c the speed of light. In spherical
108: geometry, the average U is proportional to
109: (Q(H)n$_e\epsilon^2$)$^{1/3}$ where $\epsilon$ is the gas filling factor.
110: In the case of ionization bounded regions, Q(H) must equal, in the
111: absence of dust, the number of ionizing photons derived from H$\alpha$
112: recombinations. Our models have considered values of the ionization parameter
113: between -4.0 $<$ log U $<$ -2.0, through the span of the WR
114: phase which, combined with the observed numbers of ionizing
115: photons (Q(H)) imply R
116: values of the order of 10$^{\rm 20}$ cm - 10$^{\rm 21}$ cm ($\approx$ 30 - 300 pc). The thickness of
117: the ionized gas shell is of the order of 10$^{\rm 19}$, less than 10 \% of the total
118: dimensions of the region, which results in a plane-paralel geometry.
119:
120: Figure 1 shows the run
121: of the emission line intensities relative to H$\beta$, and the
122: H$\beta$ equivalent width and luminosity as a function of the ionized gas
123: shell thickness. The figure also compares the model predictions in each case
124: with the observed values (horizontal bars).
125:
126:
127: Figure 1 clearly shows that a consistent fit to all observed
128: quantities can only be found if the regions are considered to be matter
129: bounded, i.e. only if there is an important escape of ionizing radiation from
130: the nebulae. Models in which all photons are absorbed (most right-hand
131: points in Figure 1) lead to large departures from all observed values.
132: However, if one assumes the nebulae to be matter
133: bounded, a satisfactory agreement between all predicted quantities and the
134: observations is found at a common distance to the ionizing
135: cluster for each case. Clearly, for these models we have also demanded for
136: consistency with the observed H$\alpha$ luminosity which has been found
137: through an iterative process. Successful models have hydrogen column
138: densities of the order of 10$^{\rm 19}$ cm$^{\rm -2}$ while ionization bounded models
139: reach column densities close to 10$^{\rm 20}$ cm$^{\rm -2}$.
140:
141: The fraction of escaping photons depends both
142: on their energy and the gas column density. Table 1 gives the main physical
143: conditions of the observed regions. Columns 1, 2 and 3 identify the GEHRs and
144: list the observed metallicity and resultant age of the best fitting models
145: for the ionizing clusters. Columns 4 indicates the range in ionization
146: parameter used, and columns 5-7 indicate both, the absolute and relative values
147: of the photon flux escaping in three different energy bins (13.6 - 24.5 eV,
148: 24.5 - 54.4 eV, and larger than 54.4 eV). The log of the total values of
149: photons escaping the nebulae range from 49.80 $<$ log Q$_{esc}$(H) $<$ 51.51,
150: which imply an escape fraction between 10\% (for region H13) and 73\% (for
151: region 74C).
152:
153:
154:
155: \section{DISCUSSION AND CONCLUSIONS}
156:
157:
158: The observed emission lines of GEHR have been widely used to derive the
159: properties of the ionizing radiation and link them to those of the
160: stellar populations generating it, but always under the assumption that
161: the ionized region is ionization bounded. In the absence of an
162: independent constraint on the age of this stellar population, such as
163: the presence of WR stars, successful fits to the emission line spectra
164: of most GEHR have been found within a narrow range of ages, between 2 and
165: 2.5 Myr (Bresolin et al. 1999; Dopita et al. 2000). In fact, if
166: similar methods of analysis were applied to the regions modelled here,
167: we would obtain a similar result. For these regions, however,
168: we have accurate determinations of the gas metallicity and their age,
169: provided by the analysis of the WR spectral features, and therefore we
170: can synthesize the corresponding SEDs.
171: However, the derived
172: SEDs are unable to reproduce the observed emission line spectra
173: %As mentioned above, at the metallicities involved, there are no strong
174: %reasons why the stellar evolution and atmosphere models be fundamentaly
175: %wrong (they actually reproduce rather
176: %accurately the observed WR feature properties).
177: unless the GEHRs are density bounded. In such a case, we have shown here that
178: a satisfactory fit can be found in the three analyzed cases.
179:
180: Our results suggest that the leaking of ionizing photons do depend on
181: the evolutionary stage of the regions. Regions 74C and CDT3 show
182: the highest escape fractions, consistent with the prominent observed
183: WR features in the former spectrum, and the high excitation
184: lines ([FeII],, [FeIII]) in the latter one. However, region H13, which also
185: shows a prominent WR feature, shows a far lower escape
186: fraction which can be understood from its lower metallicity (0.2 solar
187: {\it versus} 0.5 solar in 74C and CDT3), affecting the WR phase
188: strength. It is plausible that large photon escaping fractions might be
189: expected in those regions that experience a hard WR phase, this one
190: depending on both the age and metallicity.
191:
192: An important implication from the escape of ionizing photons concerns
193: the EW(H$\beta$) - age relation, that does not hold for matter bounded
194: regions. It is a well known fact that few HII regions show EW(H$\beta$) as
195: large as those predicted from evolutionary synthesis models (e.g. Melnick,
196: Terlevich \& Eggleton, 1985; Mas-Hesse \& Kunth 1998). The leak of Lyman
197: continuum radiation provides a natural explanation in the correct direction.
198:
199: A close analysis of the energy distribution of the escaping radiation
200: (see Table 1) shows that the largest fraction
201: has energies between 13.6 and 24.4 eV (up to 80\% in the case of 74C).
202: However, if escaping to incident fractions are considered, helium ionizing
203: photons are clearly dominant (e.g. up to 84\% of the incident photons
204: with energies between 24.4 and 54.4 eV escape from region
205: 74C). This would explain why the observed emission line spectrum is so
206: similar to that produced by a star cluster without WR stars.
207:
208:
209:
210: Furthermore, the derived large amounts of escaping ionizing
211: photons indicate that GEHRs and stellar OB associations may significantly
212: contribute to the ionization of the diffuse gas layers observed above the disks of spiral
213: galaxies. Though, to our
214: knowledge, no measurements of the diffuse interstellar gas (DIG) are
215: available for NGC 628, NGC 1232 and NGC 4258, recent
216: studies by Ferguson et al. (1996), Oey \& Kennicutt (1997) and Zurita et
217: al. (2000) stress the
218: fact that the integrated escaping flux from disk galaxies can attain up
219: to 10$^{41}$ erg s$^{-1}$,
220: enough to account for the diffuse H$\alpha$ flux and fully compatible with our derived values for
221: single GEHRs.
222:
223: In our models the lower energy photons emitted by the central star
224: cluster are more effectively absorbed by the HII region gas while a large
225: fraction of the photons with energy between 1.8 to 4 Ry escape the
226: nebula.
227: Therefore, the resulting spectral energy distribution of the escaping
228: photons is slightly harder than that of the ionizing star cluster. For
229: the three studied regions the Q(He)/Q(H) ratio of the escaping photons
230: is: 0.22, 0.20 and 0.19 which should be compared to 0.10, 0.14 and 0.15
231: for the respective ionizing clusters.
232:
233: One still unsolved problem for the interpretation of
234: observations of the DIG above the galactic disk is the apparently low
235: value of the HeI$\lambda$5876 ${\AA}$ /H$\beta$ (Reynolds \& Tufte 1995) which would
236: be in contradiction with our derived Q(He)/Q(H) ratios. Our studied
237: regions, however,
238: are not representative of galactic HII regions ionized by a single star,
239: but of high luminosity Giant
240: Extragalactic HII Regions on the disks of spiral galaxies like, for
241: example, NGC891. For the DIG in this galaxy the average value of
242: HeI$\lambda$5876 ${\AA}$ /H$\beta$ is 0.1 (Rand 1997). According to Bresolin, Kennicutt
243: \& Garnett (1999) the value of this ratio for stellar effective
244: temperatures between 40000 and 50000 K is highly dependent of
245: metallicity and reaches the value of 0.1 for a metallicity of about 1/4
246: solar. In fact, most HII galaxies show values of HeI$\lambda$5876 ${\AA}$ /H$\beta$ around
247: 0.10 (Izotov \& Thuan 1998 and references therein).
248:
249: In summary, by applying evolutionary synthesis models to well-studied GEHRs with observed
250: WR features, we find that these regions need to be matter bounded in order to
251: reconcile model predictions with observations. The amount of escaping ionizing
252: photons per unit time from these regions range from
253: 49.80 $<$ log Q$_{esc}$(H) $<$ 51.51, which implies an escape
254: fraction between 10\% and 73\% of the available incident photons. These
255: fractions seem to increase with both the strength of the WR phase and the
256: metallicity of the ISM. The implication is thus that the mass of the
257: ionizing clusters is larger than the value derived in a straight manner from the observed
258: H$\alpha$ luminosity. It is also clear that the ISM is highly non-uniform and that the
259: matter swept up in the shells produced by the mechanical energy deposited by
260: the WR sources and other massive stars is not sufficient to trap the ionization
261: fronts. This is also the
262: result
263: obtained from numerical calculations that consider both the mechanical
264: energy
265: of massive stellar clusters as well as their ionizing luminosity.
266: Tenorio-Tagle
267: et al. (1999) have shown how once the shell of swept up matter becomes
268: Rayleigh - Taylor unstable and fragments, as it evolves out of a galaxy
269: disk into the halo
270: (the blowout phenomenon), it then allows not only for the venting of the
271: hot (wind and SN) matter into the halo, but also for the leakage of a
272: large
273: fraction of the ionizing radiation. The latter soon establishes a giant
274: conical HII
275: region in the low density halo. The low densities in the halo lead to a
276: long recombination time and thus to a large leakage of photons into the
277: IGM. The
278: situation changes later, once the shock is able to sweep enough halo
279: matter,
280: enhancing locally the number of recombinations in the expanding shell.
281: This leads eventually to the trapping of the ionization front.
282: Consecuently,
283: depending on the stage of the evolution, a large fraction of the
284: ionizing flux escapes the
285: nebulae and is freely available to impact upon the gas at large distances from
286: the host galaxy plane (Collins \& Rand 2001) and it is also likely to escape the
287: galaxy and cause an important ionization of the IGM. Clearly, further analysis of this
288: kind must be done to infer whether or not our derived escape fractions are typical of other GEHRs.
289:
290: \section*{Acknowledgements}
291: We thank the anonymous referee for many comments and suggestions which significantly
292: improved the content of the paper.
293:
294: G.T.-T. is grateful to an IBERDROLA Visiting Professorship
295: to UAM during which part of this work was completed.
296: This work has been partially supported by DGICYT project AYA-2000-0973.
297: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
298:
299:
300: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
301:
302:
303: \begin{thebibliography}{99}
304: \bibitem[\protect\citename{}]{} Beckman, J.E., Rozas, M., Zurita, A., Watson, R.A., \& Knapen, J.H. 2000, AJ , 119, 2728
305: \bibitem[\protect\citename{}]{} Bland-Hawthorn, J., \& Maloney, P.R. 1999, ApJ, 510, L33
306: \bibitem[\protect\citename{}]{} Bresolin, F., Kennicutt, R.C., \& Garnett, D.R. 1999, ApJ, 510, 104
307: \bibitem[\protect\citename{}]{} Castellanos, M., D{\'\i}az, A.I., \& Terlevich, E.
308: 2002, MNRAS, 329, 315 Paper I
309: \bibitem[\protect\citename{}]{} Collins, J.A., \& Rand, R.J. 2001, ApJ, 551, 57
310: \bibitem[\protect\citename{}]{} D{\'\i}az, A.I.,
311: Castellanos, M., Terlevich, E., \& Garc{\'\i}a-Vargas, M.L. 2000, MNRAS, 318, 462
312: \bibitem[\protect\citename{}]{} Dopita, M.A., Kewley, L.J., Heisler, C.A., \&
313: Sutherland, R.S. 2000, ApJ, 542, 224
314: \bibitem[\protect\citename{}]{} Dove, J.B., Shull, J.M., \& Ferrara, A. 2000, ApJ, 531, 846
315: \bibitem[\protect\citename{}]{} Ferguson, A.M., Wyse, R.G., Gallagher, J.S.,
316: \& Hunter, D.A. 1996, AJ, 111, 2265
317: \bibitem[\protect\citename{}]{} Ferland, G.J.
318: 1999, HAZY: A Brief Introduction to CLOUDY. Univ. Kentucky internal report
319: \bibitem[\protect\citename{}]{} Heckman, T.M., Sembach, K.R., Meurer, G.R.,
320: Leitherer, C., Calzetti, D., \& Martin, C.L. 2001, ApJ, 558, 56
321: \bibitem[\protect\citename{}]{} Hurwitz, M., Jelinsky, P., \& Van Dyke Dixon, W.
322: 1997, ApJ, 481, L31
323: \bibitem[\protect\citename{}]{} Izotov, Y.I., \& Thuan, T.X. 1998, ApJ, 497, 227
324: \bibitem[\protect\citename{}]{} Leitherer, C., {\it et al.}
325: 1999, ApJS, 123, 3 (ST99)
326: \bibitem[\protect\citename{}]{} Leitherer, C.,
327: Ferguson, H.C., Heckman, T.M., \& Lowenthal, J.D. 1995, ApJ, 454, L19
328: \bibitem[\protect\citename{}]{} Mas-Hesse, J.M., \& Kunth, D. 1999, A\&A, 349, 765
329: \bibitem[\protect\citename{}]{} Melnick, J., Terlevich, R., \& Eggleton, P.P.
330: 1985, MNRAS, 216, 255
331: \bibitem[\protect\citename{}]{} Meynet, G. 1995, A\&A,
332: 298, 767
333: \bibitem[\protect\citename{}]{} Oey, M.S., \& Kennicutt, R.C. 1997,
334: MNRAS, 289, 827
335: \bibitem[\protect\citename{}]{} Rand, R.J. 1997, ApJ, 474, 129
336: \bibitem[\protect\citename{}]{} Reynolds, R.J., \& Tufte, S.L. 1995, ApJ, 439, L17
337: \bibitem[\protect\citename{}]{} Riccoti, M., \& Shull, J.M. 2000,
338: ApJ, 542, 548
339: \bibitem[\protect\citename{}]{} Schaerer, D., \& Vacca, W.D. 1998,
340: ApJ, 497, 618
341: \bibitem[\protect\citename{}]{} Stasi{\'n}ska, G., Schaerer, D.,
342: \& Leitherer, C. 2001, A\&A, 370, 1
343: \bibitem[\protect\citename{}]{} Tenorio-Tagle, G., Silich, S.A., Kunth, D., Terlevich, E.,
344: \& Terlevich, R. 1999, MNRAS, 309, 332
345: \bibitem[\protect\citename{}]{} Weiner, B.J., \& Williams, T.B. 1996, AJ, 111, 1156
346: \bibitem[\protect\citename{}]{} Zurita, A., Rozas, M., \& Beckman, J.E. 2000, A\&A, 363, 9
347: \end{thebibliography}
348: \clearpage
349: %%%%%%%%%%%%%%%%%%%%%%%%%%%%TABLAS%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
350:
351: %
352:
353: % TABLA 1 : RESULTS
354:
355: %
356:
357: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
358:
359: \begin{table*}
360: \setcounter{table}{0}
361: \begin{minipage}{140mm}
362: \caption{Model results}
363: \hspace{0.5cm}
364: \scriptsize{
365: \begin{tabular}{@{}lccccccc@{}}
366: \hline
367: \hline
368: Region & {Z/Z$_{\odot}$\footnote {Gas metallicity relative to solar}} & {t(Myr)\footnote {Derived age from WR population models}} & log U & {log Q$_{esc}$\footnote {Columns 5 to 8: Lower and upper values for the number of escaping ionizing photons in the three energy bins and the total one.\\ The brackets in each column provide the ratios of escaping to incident radiation in each energy bin and for the total one.}} & log Q$_{esc}$ & log Q$_{esc}$ & log Q$_{esc}$ \\
369: & & & & (13.6 - 24.6 eV) & (24.6 - 54.4 eV) & ($\geq$ 54.4 eV) & total \\
370: \hline
371: H13 & 0.2 & 4 - 4.5 & -3.00 $\pm$ 0.05 & 49.68 (8.5\%) & 49.15 (25\%) & 43.51 (0.1\%) & 49.80 (10\%) \\
372: (NGC 628)& & & & 50.17 (26\%) & 49.41 (45\%) & 43.70 (0.3\%) & 50.24 (28\%) \\
373: \hline
374: CDT3 & 0.4 & 3 - 3.5 & -3.35 $\pm$ 0.05 & 51.30 (41\%) & 50.70 (63\%) & 45.72 (4\%) & 51.40 (43\%) \\
375: (NGC 1232)& & & & 51.38 (48\%) & 50.74 (68\%) & 45.76 (5\%) & 51.47 (51\%) \\
376: \hline
377: 74C & 0.5 & 3.5 - 4 & -3.25 $\pm$ 0.05 & 51.13 (51\%) & 50.52 (72\%) & 47.08 (11\%) & 51.23 (54\%) \\
378: (NGC 4258)& & & & 51.42 (71\%) & 50.74 (84\%) & 47.60 (26\%) & 51.51 (73\%) \\
379: \hline
380: \end{tabular}
381: }
382: \end{minipage}
383: \end{table*}
384:
385: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
386: \clearpage
387: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
388:
389: %
390:
391: % FIGURA 1: matter bounded models
392:
393: %
394:
395: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
396:
397:
398: \begin{figure}
399: \setcounter{figure}{0}
400: \centering
401: \begin{minipage}[c]{20mm}
402: \centering\psfig{figure=mbh13.epsi,width=7.4cm,height=7.4cm,clip=}
403: \end{minipage}%
404: \hspace{1cm}
405: \begin{minipage}[c]{10mm}
406: \centering\psfig{figure=mb74c.epsi,width=7.4cm,height=7.4cm,clip=}
407: \end{minipage}
408: \\
409: \vspace{1cm}
410: \begin{minipage}[c]{10mm}
411: \centering
412: \psfig{figure=mbcdt3.epsi,width=7.4cm,height=7.4cm,clip=}
413: \end{minipage}
414: \caption{The panels show the run of different line intensity ratios ([OII]($\lambda$3727),
415: [OIII]($\lambda$5007), [NII]($\lambda$6584), [SII]($\lambda$6717), and
416: [SIII]($\lambda$9069) relative to H$\beta$), as well as the H$\beta$ equivalent
417: width and H$\beta$ luminosity (both of them scaled by a constant factor), as
418: a function of ionized gas shell thickness in the three nebulae here
419: considered. The models have been derived from the ionization parameter range
420: given in Table 1 and extend until all photons are used up. The horizontal bars
421: across every predicted quantity trend indicate the observed values and their
422: errors (see Paper I).}
423: \end{figure}
424: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
425:
426: \end{document}
427:
428:
429: