1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[preprint1]{aastex}
3: % one-column, double-spaced document:
4: % \documentclass[manuscript]{aastex}
5: % double-column, single-spaced document:
6: %\documentclass[preprint2]{aastex}
7: \usepackage{graphicx}
8: \usepackage{epsfig}
9: \usepackage{rotating}
10: %% -------------------------------------------------------------
11: \def\la{Ly$\alpha$}
12: \def\hb{H$\beta$}
13: \def\arcmin {$^{\prime}$}
14: \def\arcm {$^{\prime}$}
15: \def\arcsec {$^{\prime\prime}$}
16: \def\arcs {$^{\prime\prime}$}
17: \def\kms {km\thinspace s$^{-1}$}
18: \def\cms {cm$^{-2}$}
19: \def\etal{{\it et al. }}
20: \def\aa{{\rm A$\,$\&$\,$A}} % A \and\ A
21: \def\araa{{\rm ARA$\,$\&$\,$A}} % Annu. Rev. A \and\ A
22: \def\aar{{\rm A$\,$\&$\,$AR}} % A \and\ A Review
23: \def\aas{{\rm A$\,$\&$\,$AS}} % A \and\ A Sup.
24: \def\apj{{\rm ApJ}} % Astrophys. J.
25: \def\apjs{{\rm ApJS}} % Astrophys. J. Supl.
26: \def\apjl{{\rm ApJ Let}} % Astrophys. J. (Letters)
27: \def\aj{{\rm AJ}} % Astron. J.
28: \def\sva{{\rm SvA}} % Sov. Astron.
29: \def\pasp{{\rm PASP}} % Publ. Astr. Soc. Pacific
30: \def\pasj{{\rm PASJ}} % Publ. Astr. Soc. Japan
31: \def\mnras{{\rm MNRAS}} % Monthly Notices.
32: \def\kmsmpc {km\thinspace s$^{-1}$\thinspace Mpc$^{-1}$}
33: \def\Msol{\thinspace\hbox{$\hbox{M}_{\odot}$}}
34: \def\Zsol{\thinspace\hbox{$\hbox{Z}_{\odot}$}}
35: \def\sol{\thinspace\hbox{$_{\odot}$ }}
36: \def\deg{\hbox{$^\circ$}}
37: \def\kpc{\thinspace\hbox{kpc}}
38: \def\ojo{\fbox{\bf !`$\odot$j$\odot$!}}
39: \def\ie{{\it i.e.} } % i.e. italicized
40: \def\a4{\hsize 17.0cm \vsize 25.cm}
41:
42: %\shorttitle{...}
43: %\shortauthors{Silich et al.}
44:
45: \begin{document}
46:
47: \title{On the origin of the neutral hydrogen supershells: the
48: ionized progenitors and the limitations of the multiple
49: supernovae hypothesis}
50:
51: \author{
52: Sergiy Silich \affil{Instituto Nacional de Astrof\'\i sica Optica y
53: Electr\'onica, AP 51, 72000 Puebla, M\'exico; silich@inaoep.mx}
54: Federico Elias \affil{Instituto de Astronom\'\i a. Universidad
55: Nacional Aut\'onoma de M\'exico. AP 70 - 264, 04510 M\'exico D.F.;
56: felias@astroscu.unam.mx} \and Jos\'e Franco \affil{Instituto de
57: Astronom\'\i a. Universidad Nacional Aut\'onoma de M\'exico. AP 70 -
58: 264, 04510 M\'exico D.F.; pepe@astroscu.unam.mx}}
59:
60: \begin{abstract} Here we address the question whether the ionized shells
61: associated with giant HII regions can be progenitors of the larger
62: HI shell-like objects found in the Milky Way and other spiral and
63: dwarf irregular galaxies. We use for our analysis a sample of 12 HII
64: shells presented recently by Rela\~no et al. (2005, 2007). We
65: calculate the evolutionary tracks that these shells would have if
66: their expansion is driven by multiple supernovae explosions from the
67: parental stellar clusters. We find, contrary to Rela\~no et al.
68: (2007), that the evolutionary tracks of their sample HII shells are
69: inconsistent with the observed parameters of the largest and most
70: massive neutral hydrogen supershells. We conclude that HII shells
71: found inside giant HII regions may represent the progenitors of
72: small or intermediate HI shells, however they cannot evolve into the
73: largest HI objects unless, aside from the multiple supernovae
74: explosions, an additional energy source contributes to their
75: expansion.
76: \end{abstract}
77:
78: \keywords{ISM: bubbles --- (ISM:) HII regions --- ISM: kinematics
79: and dynamics}
80:
81: \section{Introduction}
82:
83:
84: The origin of numerous holes and shells detected in the distribution
85: of neutral hydrogen in spiral and dwarf galaxies (Heiles, 1980;
86: Brinks \& Bajaja, 1986; Puche et al., 1992) is a long standing
87: problem (Heiles, 1984; Tenorio-Tagle \& Bodenheimer, 1988). Heiles
88: (1984), Brinks \& Bajaja (1986), Puche et al. (1992), Mashchenko et
89: al. (1999), Ehlerov\'a \& Palou\v s (2005) detected in the Milky
90: Way, M31 and in the dwarf irregular galaxy Holmberg II (HoII)
91: hundreds of neutral hydrogen shells whose radii range from a few
92: tens to a thousand parsecs. On the other hand, Fabry-Perot
93: observations of the ionized gas kinematics in many galaxies revealed
94: a number of shells whose radii reach a few hundred parsecs and whose
95: velocity pattern suggests an expansion with velocities up to 60 --
96: 70~km $s^{-1}$ (see, for example, Lozinskaya 1992; Chu et al. 1990;
97: Oey \& Massey, 1995; Valdez-Guti\'errez et al. 2001; Naz\'e et al.
98: 2001; Lozinskaya et al. 2003; Rela\~no \& Beckman, 2005 and
99: references therein). The majority of these shells is associated with
100: interior stellar clusters. The standard model (McCray \& Kafatos,
101: 1987; Mac Low \& McCray 1988) was constructed with the energy
102: injection from multiple supernovae explosions and stellar winds
103: occurring in young stellar clusters which are often found inside
104: small and intermediate sized shells. In such a case the energy
105: supplied by supernovae and individual stellar winds is thermalized
106: inside the parental stellar cluster, resulting in a high central
107: overpressure which drives a high velocity outflow (the star cluster
108: wind). This outflow, when interacting with the ambient interstellar
109: medium (ISM), forms a leading and a reverse shocks which are
110: separated by a contact discontinuity (see Figure 1). The
111: interstellar gas collected by the outer shock forms an expanding
112: shell which moves because the thermal pressure in the region C,
113: between the reverse shock and the contact discontinuity, exceeds the
114: ISM pressure. The swept-up gas cools rapidly and forms an expanding
115: shell photoionized by the Lyman continuum from the embedded cluster.
116: The number of ionizing photons rapidly drops with the star cluster
117: age (Leitherer et al. 1999) and eventually (after $\sim 10$~Myrs)
118: the driving cluster (or clusters) will find itself embedded inside a
119: slowly expanding neutral hydrogen shell. The shell radius and
120: velocity depend on the amount of energy released by the cluster and
121: on the parameters of the ambient interstellar medium (see review by
122: Bisnovatyi-Kogan \& Silich, 1995 and references therein).
123:
124:
125: While the standard model is broadly consistent with the parameters
126: of many small and intermediate sized structures found around young
127: stellar clusters and OB associations (see, for example, the
128: discussion of the superbubble growth discrepancy in Oey \& Garc\'\i
129: a-Segura, 2004 and the analysis of the XMM-Newton observations of
130: the 30 Dor C in Smith \& Wang 2004), it meets a profound energy
131: problem when applied to larger structures whose radii are comparable
132: to the characteristic Z-scale of the density distribution in the
133: host galaxy. For instance, Rhode et al. (1999) found that, in the
134: case of the HoII galaxy, the stellar clusters found inside HI holes
135: are unable to remove gas from these regions. Kim et al. (1999) found
136: only a weak correlation between the Large Magellanic Cloud (LMC)
137: neutral hydrogen holes and the HII regions and concluded that the
138: hypothesis of multiple supernovae is inconsistent with their data.
139: Hatzidimitriou et al. (2005) cross-correlated the positions of 509
140: young neutral hydrogen shells detected in the Small Magellanic Cloud
141: (SMC) with the locations of known OB-associations, Wolf-Rayet stars,
142: supernova remnants and stellar clusters. They found that 59 shells
143: have no young stellar objects associated with them despite the
144: distributions of their radii and expansion velocities are consistent
145: with those predicted by the multiple SNe model. They concluded that
146: turbulence may be a promising mechanism that would allow us to
147: understand the origin of these objects, but a quantitative
148: comparison of the existing theory with observations is not possible
149: at this moment.
150:
151: Crosthwaite et al. (2000) examined the distribution of HI in the
152: nearly face-on Scd galaxy IC 342 and did not find either the
153: kinematic signatures specific to the expanding shells, or
154: indications on the distortion of the observed structures by the
155: differential galactic rotation, which is expected to be noticeable
156: if the observed structures expand because of the energy output
157: provided by the embedded stellar population (Palous et al. 1990;
158: Silich, 1992). Therefore Crosthwaite et al. (2000) have concluded
159: that the observed flocculent structure is formed via gravitational
160: instabilities in a turbulent galactic disk, as suggested by Wada \&
161: Norman (1999), and not because of the energy output provided by the
162: stellar component of the galaxy. Silich et al. (2006) have compared
163: parameters of a $\sim 500$~pc radius HI ring detected in dwarf
164: irregular galaxy IC 1613 with a combined energy output provided by a
165: number of OB associations found inside the structure. They did not
166: find a noticeable expansion velocity and concluded that the observed
167: radius and mass of the structure are inconsistent with the SNe
168: hypothesis.
169:
170: Several other mechanisms that allow to shape the ISM of the host
171: galaxy into large shell-like structures similar to those observed in
172: nearby spiral and irregular galaxies, and that do not require
173: violent stellar activity, include collisions of high velocity clouds
174: with galactic disks (Tenorio-Tagle, 1981; Comeron \& Torra, 1992),
175: non-linear instabilities in the self-gravitating turbulent galactic
176: disks (Wada et al. 2000; Dib \& Burkert, 2004) and more exotic
177: mechanisms such as the distortion of the ISM by powerful gamma-ray
178: bursts (Efremov et al. 1999).
179:
180: On the other hand, several modifications of the multiple SNe
181: hypothesis have been suggested by different authors, in an attempt
182: to reconcile the observed parameters of large HI structures with
183: those predicted by the multiple supernovae model. Elmegreen \&
184: Chiang (1982) added the effects of the radiation pressure from field
185: stars, that is able to provide additional expansion. Palous et al.
186: (1990) added the effects of galactic rotation that distorts and
187: stretches the shell along the direction of rotation.
188: McClure-Griffiths et al. (2002) found that some HI shells are
189: located between the spiral arms of the Galaxy. They suggested then
190: that density gradients that occur between the spiral arms and the
191: interarm medium could result in the enhancement of the predicted
192: sizes and migration of large shells from spiral arms into the
193: interarm medium.
194:
195: More recently, Rela\~no et al. (2007) have suggested that the
196: population of ionized H$_{\alpha}$ shells associated with large HII
197: regions (Rozas et al. 1996; Rela\~no \& Beckman, 2005) may represent
198: the precursors of the larger HI structures, and claimed that a very
199: simplified analytic model is enough to explain larger HI objects.
200: They assumed that young shells observed in H$_{\alpha}$ emission
201: evolve in an energy dominated regime until the beginning of the
202: supernova explosion phase, then make a transition to a momentum
203: dominated stage and continue to expand with the momentum supplied by
204: the embedded cluster during the initial time. The idea is not new
205: (Dyson, 1980; Bruhweiler et al. 1980) and was used by Gil de Paz et
206: al. (2002) and Silich et al. (2002), who discussed the origin of a
207: giant ($R \approx 700$~pc) ring in the low-metallicity BCD galaxy
208: Mrk 86. However, the key problem with this idea is that the momentum
209: dominated regime is inefficient in driving bubble expansion because
210: the thermal pressure in zone C (Figure 1) is approximately equal to
211: the ram pressure of the ejected gas at the location of the reverse
212: shock. Given that ram pressure drops as $r^{-2}$, the driving
213: pressure equals the external pressure at a certain radius and the
214: expansion quickly stalls after that point. Therefore we wonder how
215: the early transition to the momentum dominated stage can solve the
216: ``energy crisis'' that has been discussed by many authors during the
217: last decades.
218:
219: Here we re-analyze the hydrodynamical model presented in Rela\~no et
220: al. (2007) keeping their assumption of a fast transition to the
221: momentum dominated stage. We show that the over-simplified
222: hydrodynamic equations and the ``statistical'' initial conditions
223: adopted by them, which are not consistent with parameters for the
224: largest HI structures, lead to overestimated expansion velocities
225: and radii of the shells. Our results do not support their contention
226: that H$_{\alpha}$ shells associated with the HII regions can be
227: progenitors of the largest HI structures in the absence of any
228: additional driving mechanism (for instance, radiation pressure from
229: field stars, as suggested by Elmegreen \& Chiang 1982).
230:
231: The paper is organized as follows: in Section 2 we establish the
232: main equations of our hydrodynamical model, and compare them with
233: those used by Rela\~no et al. (2007). In Section 3 we first
234: determine the initial conditions required to perform the numerical
235: integration of the equations, and then present the resulting
236: evolutionary tracks, our main result that small shells found in
237: giant HII regions cannot be the progenitors of the largest neutral
238: hydrogen supershells detected in gaseous galaxies.
239:
240:
241:
242: \section{Main equations}
243:
244: The kinetic energy supplied by supernovae and stellar winds from
245: stellar clusters is thermalized by a shock, and this results in the
246: four-zone structure discussed in Figure 1 (Castor et al. 1975;
247: Weaver et al. 1977; Mac Low \& McCray 1988; Bisnovatyi-Kogan \&
248: Silich 1995 and references therein). If thermal pressure in zone C
249: suddenly drops, the radius of the reverse shock approaches the
250: contact discontinuity, and the expansion of the outer shell is then
251: supported directly by the momentum deposited by the stellar cluster
252: (see, for example, Koo \& McKee, 1992). Rela\~no et al. (2007)
253: suggested that interstellar shells expanding around star forming
254: regions reach this stage after a short while and then evolve in the
255: momentum-dominated mode. We follow their assumption in the next
256: sections, but note that it is in bad agreement with estimates of the
257: characteristic cooling time in zone C (e.g., Figure 1 presented in
258: Mac Low \& McCray 1988). In the 2D or 3D cases, this regime occurs
259: at the waist of the shell if a driving cluster is embedded into the
260: disk-like density distribution (Gil de Paz et al., 2002; Silich et
261: al., 2006). The expansion of the shell (or of some segments of the
262: shell) is then defined by mass and momentum conservation:
263: %---------------------------------------------------------------
264: \begin{eqnarray}
265: \label{eq1.a}
266: & & \hspace{-1.0cm}
267: M = M_0 + \frac{4 \pi}{3} (R^3 - R^3_0) \rho_{ISM} + {\dot M}_{SC} (t - t_0) ,
268: \\[0.2cm]
269: \label{eq1.b}
270: & & \hspace{-1.0cm}
271: \frac{\rm d}{{\rm d} t} (M u) = - 4 \pi R^2 (P_{ISM} - \rho_w V^2_{\infty}) ,
272: \end{eqnarray}
273: %-------------------------------------------------------------
274: where $M$, $u$ and $R$ are the mass, expansion velocity and radius
275: of the shell, respectively. $R_0$ is the initial radius of the
276: shell, $t_0$ is the initial time and $M_0$ is the initial mass of
277: the shell. The second term is the mass of the interstellar gas swept
278: up by the expanding shell, and the last term in equation
279: (\ref{eq1.a}) is the amount of ejected matter that sticks to the
280: shell from the inside. The right-hand part of equation (\ref{eq1.b})
281: represents the difference between the ambient gas pressure,
282: $P_{ISM}$, and the driving ram pressure of the ejecta. ${\dot
283: M}_{SC}$ is the mass deposition rate provided by SNe explosions and
284: stellar winds, $\rho_w(R)$ is the density of the ejected matter when
285: it reaches the shell, and $V_{\infty}$ is the terminal speed of the
286: ejected matter.
287:
288: We assume that the parameters of the cluster remain constant during
289: the evolution, and that the expansion velocity of the shell is much
290: smaller than that of the ejected matter, $u \ll V_{\infty}$. The
291: mass deposition rate, ${\dot M}_{SC}$, and the density of the
292: ejecta, $\rho_w(R)$, then are:
293: %---------------------------------------------------------------
294: \begin{eqnarray}
295: \label{eq2.a}
296: & & \hspace{-1.0cm}
297: {\dot M}_{SC} = 2 L_{SC} / V^2_{\infty}
298: \\[0.2cm]
299: \label{eq2.b}
300: & & \hspace{-1.0cm}
301: \rho_w = {\dot M}_{SC} / 4 \pi R^2 V_{\infty} ,
302: \end{eqnarray}
303: %-------------------------------------------------------------
304: where $L_{SC}$ is the rate of mechanical energy supplied by supernovae
305: and stellar winds.
306: %---------------------------------------------------------------
307: \begin{figure}[htbp]
308: \plotone{f1.eps} \caption{Schematic representation of the structure
309: formed in the ISM by multiple stellar winds and supernovae
310: explosions. The central zone (A) represents the stellar cluster
311: where stellar winds and supernovae release their energy. The
312: remaining concentric zones are the free-wind region (zone B), the
313: shocked wind region (zone C), the shell of swept up interstellar
314: matter (zone D) and the ambient ISM. When the shocked wind cools
315: rapidly, zone C vanishes and the outer shell is pushed away by the
316: momentum of the ejected matter.} \label{fig1}
317: \end{figure}
318: %---------------------------------------------------------------
319:
320: Combining equations (\ref{eq1.a}) and (\ref{eq1.b}) one obtains:
321: %---------------------------------------------------------------
322: \begin{eqnarray}
323: \label{eq3.a}
324: & & \hspace{-1.0cm}
325: \frac{{\rm d}u}{{\rm d} t} = - \frac{4 \pi R^2 V^2_{\infty}
326: (P_{ISM} + \rho_{ISM} u^2) - 2 L_{SC} (V_{\infty} - u)}
327: {[M_0 + 4 \pi \rho_{ISM} (R^3 - R^3_0)/3] V^2_{\infty} +
328: 2 L_{SC} (t - t_0)}
329: \\[0.2cm]
330: \label{eq3.b}
331: & & \hspace{-1.0cm}
332: \frac{{\rm d}R}{{\rm d} t} = u
333: \end{eqnarray}
334: %-------------------------------------------------------------
335: One can solve these equations numerically for known values of $R_0$,
336: $M_0$, $u_0$, $\rho_{ISM}$, $L_{SC}$ and $V_{\infty}$.
337:
338: For the particular case when supernovae and stellar winds deposit
339: all momentum instantaneously and the pressure in the interstellar
340: medium is zero, $P_{ISM} = 0$, equations (\ref{eq1.a}) and
341: (\ref{eq1.b}) become:
342: %---------------------------------------------------------------
343: \begin{eqnarray}
344: \label{eq4.a}
345: & & \hspace{-1.0cm}
346: M(R) = M_0 + \frac{4 \pi}{3} \rho_{ISM} (R^3 - R^3_0) ,
347: \\[0.2cm]
348: \label{eq4.b}
349: & & \hspace{-1.0cm}
350: M(R) u(t) = M_0 u_0 .
351: \end{eqnarray}
352: %-------------------------------------------------------------
353: Furthermore, neglecting the mass of the star forming cloud and
354: assuming
355: %---------------------------------------------------------------
356: \begin{equation}
357: \label{eq5}
358: M_0 = \frac{4 \pi}{3} \rho_{ISM} R^3_0 ,
359: \end{equation}
360: %---------------------------------------------------------------
361: the solutions are reduced to the set of the main equations
362: (equations 1 and 2) used by Rela\~no et al. (2007):
363: %---------------------------------------------------------------
364: \begin{eqnarray}
365: \label{eq6.a}
366: & & \hspace{-1.0cm}
367: R(t) = R_0 \left[1 + \frac{4 u_0 (t - t_0)}{R_0} \right]^{1/4} ,
368: \\[0.2cm]
369: \label{eq6.b}
370: & & \hspace{-1.0cm}
371: u(t) = \frac{3 M_0 u_0}{4 \pi \rho_{ISM} R^3} .
372: \end{eqnarray}
373: %-------------------------------------------------------------
374: Thus, their main equations represent an oversimplified 1D model that
375: neglects the effects of the ambient pressure, the mass of the star
376: forming cloud and the continuous mechanical energy injection.
377:
378: To illustrate the differences in the solutions of equations
379: \ref{eq3.a} and \ref{eq3.b} and equations \ref{eq6.a} and
380: \ref{eq6.b}, we assume an interstellar gas density
381: %---------------------------------------------------------------
382: \begin{equation}
383: \label{a1}
384: \rho_{ISM} = \frac{3 M_{HI}}{4 \pi R^3_{HI}} ,
385: \end{equation}
386: %---------------------------------------------------------------
387: where $M_{HI}$ and $R_{HI}$ are the mass and radius of an evolved HI
388: shell (for the comparison we use the particular case of GSH
389: 285-02+86, whose radius and mass are $R_{HI} = 385$~pc and $M_{HI} =
390: 44\times10^5$ M$_{\odot}$, respectively). The number density of the
391: interstellar gas then is $n_{ISM} = 0.75$~cm$^{-3}$. We further
392: assume that the initial radius of the shell is $R_0 = 104$~pc and
393: obtain the initial mass of the shell, $M_0$, from equation
394: (\ref{eq5}). Then we use the observed velocity of the progenitor
395: shell (64.7 km~s$^{-1}$) as the initial value for the solution of
396: equations \ref{eq3.a} and \ref{eq3.b}. For the oversimplified case
397: of equations \ref{eq6.a} and \ref{eq6.b} we use energy conservation,
398: and the initial velocity in this case is
399: %---------------------------------------------------------------
400: \begin{equation}
401: \label{a2}
402: u_0 = \frac{2 E_{kin}}{M_0} ,
403: \end{equation}
404: %---------------------------------------------------------------
405: where the kinetic energy of the progenitor shell, $E_{kin} =
406: 36.1\times10^{52}$~erg s$^{-1}$, has been derived from the
407: Starburst99 synthetic model (Leitherer et al. 1999). Note that one
408: cannot use identical initial conditions (the same initial velocity)
409: in both approaches because in one case the energy and momentum are
410: deposited instantaneously and, in the other, they are supplied
411: continuously and grow with time during the shell evolution.
412:
413: Figure 2 presents the expansion velocities predicted by both sets of
414: equations. Certainly, the initial velocity in the oversimplified
415: model is much larger than that of equations \ref{eq3.a} and
416: \ref{eq3.b}, but it drops faster. When the radius of the shell
417: becomes about three or four times that of the initial value, the
418: difference between the two calculations becomes small. Later on, the
419: effect of the ambient pressure becomes important, but the shell
420: continues to expand in the oversimplified model.
421:
422: In the next section we drop all simplifications associated with the
423: analytic solutions and solve equations (\ref{eq3.a}) and
424: (\ref{eq3.b}) numerically.
425:
426:
427: \section{The expansion of the momentum-dominated shell into the
428: homogeneous ISM}
429:
430: \subsection{Initial conditions}
431:
432: We start the integrations at the initial time, $t_0 = 10^6$ yr, and
433: with the initial mass, $M_0$. We adopt initial radii, $R_0$, and
434: expansion velocities, $u_0$, derived from the H$_{\alpha}$
435: observations of the ionized shells. $R_0$ is approximately 0.3 times
436: the radius of the HII region, and $u_0$ is the observed velocity of
437: the H$_{\alpha}$ shell. For example, in the case of NGC 1530-8, $R_0
438: = 104$~pc and $u_0 = 64.7$~km~s$^{-1}$, respectively. In order to
439: calculate the initial mass of the shell, we substitute into equation
440: (\ref{eq5}) the value of the initial radius and the average density
441: 0.1~cm$^{-3}$ assumed by Rela\~no et al. We also
442: assume that the embedded cluster continuously expels the gas
443: released by SNe and stellar winds, whose momentum supports the
444: expansion of the outer shell, during $\sim 40$~Myr, the
445: characteristic life-time of a $~8$\Msol \, star - the lowest mass
446: star which will eventually explode as a supernova. Also, the
447: pressure in the surrounding medium is explicitly included here.
448:
449: We calculate the density of the ISM using the masses and radii of
450: the HI shells and their progenitors:
451: %---------------------------------------------------------------
452: \begin{equation}
453: \label{eq7}
454: \rho_{ISM} = \frac{M_{HI} - M_0}{\frac{4 \pi}{3} (R^3_{HI} - R^3_0)}
455: \end{equation}
456: %---------------------------------------------------------------
457: Thus, we derive the density of the ISM for each couple of HI and HII
458: shells from their observed parameters. For instance, in the case of
459: GSH 285-02+86, GSH 304--00-12 and GSH 305+01-24 from the list of
460: McClure-Griffiths et al. (2002), the interstellar gas number density
461: would be 0.75, 1.38 and 0.62 cm$^{-3}$, respectively, if one uses as
462: the progenitor the HII shell N8 found by Rela\~no et al. (2007) in
463: the spiral galaxy NGC~1530.
464:
465: The average mechanical luminosity of the embedded cluster has been
466: calculated from the Starburst99 synthetic model with an
467: instantaneous burst of star formation (Leitherer et al. 1999). The
468: average mechanical luminosity of the cluster then is:
469: %---------------------------------------------------------------
470: \begin{equation}
471: \label{eq8}
472: L_{SC} = E_{SW+SN} / \tau ,
473: \end{equation}
474: %---------------------------------------------------------------
475: where we use $\tau = 10$ Myr. The mechanical luminosities of the
476: embedded clusters, as well as the initial masses, radii and
477: velocities for all shells are listed in Table~1.
478:
479: The temperature of the interstellar medium is assumed to be $T_{ISM}
480: = 6000$K, which is a typical value in the warm neutral component of
481: the ISM (Brinks, 1990). The thermal pressure in the ambient medium
482: then is $P_{ISM} = k~n_{ISM}~T_{ISM}$, where $k$ is the Boltzmann's
483: constant and $n_{ISM} = \rho_{ISM} / m_{H}$ is the interstellar gas
484: number density obtained from equation (\ref{eq7}).
485:
486: The last input parameter for our model, the terminal speed of the
487: star cluster wind $V_{\infty}$, is determined by the energy and mass
488: deposition rates, and is close to the terminal velocity of
489: individual stellar winds (Raga et al. 2001, Stevens \& Hartwell
490: 2003). We assume for our calculations that the star cluster wind
491: terminal speed is constant and falls in the range $1500-3000$ km
492: s$^{-1}$ (e.g., Leitherer at al. 1999). Note that this parameter
493: defines the amount of momentum deposited by the cluster and
494: therefore affects the dynamics of the momentum-dominated shell.
495:
496:
497: \subsection{Results and discussion}
498:
499:
500: We calculate the evolutionary tracks for the HII shells associated
501: with the list of large HII regions studied by Rela\~no et al. (2007), and
502: compare them with the observed parameters of the HI shells. For the
503: comparison we have chosen three HI shells from the list of
504: McClure-Griffiths et al. (2002) whose morphologies are close to the
505: spherical shape: GSH 285-02+86, GSH 304-00-12 and GSH 305+01-24.
506: They are representative of high, intermediate and low mass objects,
507: respectively.
508:
509: The results of the calculations are presented in Figure 3 and in
510: Table 2. The top left panel in Figure 2 compares the results of the
511: calculations with the parameters of low mass neutral shells. It
512: shows that most of the HII shells detected in the giant HII regions
513: certainly can evolve into objects whose parameters are identical to
514: those of low mass HI shells. For example, in the case of GSH
515: 305+01-24, nine out of the twelve HII shells (see Table 2) can
516: easily reach the observed size of the HI shell having a proper mass
517: ($3.9\times10^5$ M$_{\odot}$) and an expansion velocity which is similar
518: or even higher than the observed velocity of GSH 305+01-24. This implies
519: that the HII shells found by Rela\~no et al. (2007) in giant HII regions
520: can be progenitors of such low mass HI shells.
521:
522: The top right panel in Figure 3 compares the evolutionary tracks of
523: the progenitor shells with the observed parameters of the
524: intermediate mass ($1.9\times10^6$ M$_{\odot}$) HI object GSC
525: 304-00-12. In this case only the four most energetic HII shells (NGC
526: 1530-8, NGC 1530-22, NGC 3359-6 and NGC 6951-2) can eventually reach
527: the required radius sweeping enough interstellar mass (see Table 2).
528: However only two of them (NGC 1530-8 and NGC 6951-2) present the
529: expansion velocities that are similar to the observed one. The
530: expansion velocities of the rest fall well below the observed value.
531: Thus GSC 304-00-12 represents a limit case that separates the low
532: mass HI objects driven by multiple SNe explosions from the largest
533: ones whose parameters are not consistent with the multiple SNe
534: hypothesis. The latter case is illustrated by the bottom panel in
535: Figure 3.
536:
537: The bottom panel compares the evolutionary tracks of Rela\~no's
538: shells with parameters of the very massive shell GSH 285-02+86,
539: whose mass is $4.4 \times 10^6$ M$_{\odot}$ (McClure-Griffiths et
540: al. 2002). Here we also present (dashed line) the expansion velocity
541: predicted by the analytic model for NGC 1530-8, the most energetic
542: shell from the list of Rela\~no et al. (2005). The plot clearly
543: demonstrates that the analytic model leads to overestimated
544: expansion velocities. This is because the analytic calculations
545: adopt a low value for the interstellar gas density which is not
546: consistent with the masses and radii of the most massive neutral
547: shells (in the analytic approach the mass of the shell is only $0.6
548: \times 10^6$ \Msol \, when its radius reaches the observed value of
549: 385~pc, whereas in our calculations it is $4.4 \times 10^6$ \Msol).
550:
551:
552: The bottom panel in Figure 3 shows that HII shells found by Rela\~no
553: et al. (2005) inside giant HII regions can never evolve into the
554: largest and most massive neutral hydrogen supershells. In this case
555: the energy and momentum supplied by the central cluster are not
556: sufficient to make the evolution from the original shell finally fit
557: all the observed parameters of the HI supershell: its mass, radius
558: and expansion velocity. In this case only the three most energetic
559: progenitor shells (NGC 1530-8, NGC 3359-6 and NGC 6951-2) can reach
560: a size and accumulate a mass that are comparable to those of GSH
561: 285-02+86. However, in all these cases the model predicted expansion
562: velocities are much smaller than that measured by McClure-Griffiths
563: et al. (2002) for GSH 285-02+86. This implies that, contrary to the
564: statement made by Rela\~no et al. (2007), the young ionized shells
565: found in giant HII regions cannot evolve into the largest HI
566: supershells (i.e., eventually fit simultaneously the mass, radius
567: and expansion velocity of the HI object) unless some additional
568: physical mechanism, other than the multiple supernovae explosions,
569: contributes to the formation of the largest shells at the later
570: stages of their evolution.
571:
572: Figure 3 presents the results of the calculations under the
573: assumption that the star cluster terminal speed is $V_{\infty} =
574: 1500$~km s$^{-1}$. In order to learn how this parameter affects our
575: results we have provided a set of numerical calculations with
576: initial conditions which are identical to those used in the case of
577: GSH 285-02+86 (Figure 3, bottom panel) but with a star cluster wind
578: terminal speed twice as large as the previous one: $V_{\infty} =
579: 3000$~km~s$^{-1}$. The results of these calculations are presented
580: in Figure 4 which demonstrates that the model predicted expansion
581: velocity becomes smaller when the star cluster wind terminal speed
582: grows up. This implies that one cannot avoid the discrepancy between
583: the predicted and the observed velocities by varying the star
584: cluster wind terminal speed within a reasonable velocity interval.
585:
586: \section{Conclusions}
587:
588: Here we have critically examined the evolution of the ionized shells
589: found inside giant HII regions, as predicted by the multiple SNe
590: model in one dimension. We have used the observed parameters of the
591: HII shells as initial conditions for our numerical model and
592: compared the results of the calculations with three representative
593: cases of low, intermediate and high mass HI objects from the list of
594: McClure-Griffiths et al. (2002).
595:
596: We have found that the ionized shells observed within giant HII
597: regions cannot evolve into the largest neutral hydrogen supershells
598: if the multiple supernovae explosions of massive stars is the only
599: driving mechanism. Some additional physical mechanism must
600: contribute to the formation of the largest shells for the model to
601: be in agreement with the observed parameters of the most massive
602: neutral hydrogen supershells detected in our and other galaxies.
603:
604: Note that spherically-symmetric models must be taken with care when
605: compared with objects whose radii are comparable with several
606: characteristic scale heights in the ISM density distribution. If
607: that is the case, the expansion velocities at the top and at the
608: waist of the shell are different and the shell acquires a distorted
609: hour-glass form. The majority of the swept-up interstellar matter is
610: concentrated in a thin layer nearby the plane of the host galaxy. 3D
611: calculations are then required in order to investigate the shell's
612: morphology and kinematics in the differentially rotating galactic
613: disk (see, for example, Silich et al. 1996).
614:
615: \begin{acknowledgements}
616: We thank an anonymous referee for central comments and suggestions
617: which helped us to clarify our formulations and improve the quality
618: of the paper. We also appreciate usuful comments from our Korean and
619: Mexican colleagues during the IV Korea-Mexico workshop in Daejon.
620: This study has been supported by Conacyt (M\'exico) grant 47534-F.
621: \end{acknowledgements}
622:
623: \begin{thebibliography}
624:
625: \bibitem{1} Bisnovatyi-Kogan, G. S. \& Silich, S. A. 1995, Rev. Mod.
626: Phys. 67, 661
627:
628: \bibitem{2} Brinks, E., 1990, ASSL, 161, 39
629:
630: \bibitem{3} Brinks, E. \& Bajaja, E. 1986, A\&A, 169, 14
631:
632: \bibitem{4} Bruhweiler, F.C., Gull, T.R., Kafatos, M. \& Sofia, S.
633: 1980, ApJL, 238, 27
634:
635: \bibitem{5} Castor, J., McCray, R. \& Weaver, R., 1975, ApJ, 200, 107
636:
637: \bibitem{6} Chu, Y.-H. \& Mac Low, M.-M. 1990, ApJ, 365, 510
638:
639: \bibitem{7} Comeron, F. \& Torra, J. 1992, A\&A, 349, 41
640:
641: \bibitem{8} Crosthwaite L.P., Turner, J.L. \& Ho, P.T.P. 2000, AJ, 119, 1720
642:
643: \bibitem{9} Dib, S. \& Burkert, A. 2004, Ap\&SS, 292, 135
644:
645: \bibitem{10} Dyson, J.E. 1980, Physics of the Interstellar Medium,
646: NY, John Wiley \& Sons, 145
647:
648: \bibitem{11} Efremov, Y. N., Ehlerov\'a, S. \& Palou\v s, J. 1999,
649: ApJ, 350, 457
650:
651: \bibitem{12} Ehlerov\'a, S. \& Palou\v s, J. 2005, A\&A, 437, 101
652:
653: \bibitem{13} Elmegreen, B. G. \& Chiang, W.-H. 1982, ApJ, 253, 666
654:
655: \bibitem{14} Gil de Paz, A., Silich, S.A., Madore, B.,F., S\'anchez Contreras,
656: C., Zamorano, J. \& Gallego, J. 2002, ApJ, 573, 101
657:
658: \bibitem{15} Hatzidimitriou, D., Stanimirovic, S., Maragoudaki, F.,
659: Stavely-Smith, L., Dapergolas, A. \& Bratsolis, E.
660: 2005, MNRAS, 360, 1171
661:
662: \bibitem{16} Heiles, C. 1980, ApJ, 235, 833
663:
664: \bibitem{17} Heiles, C. 1984, ApJS, 55, 585
665:
666: \bibitem{18} Kim, S., Dopita, M. A., Stavelet-Smith, L. \&
667: Bessel, M. 1999, A\&A, 350, 230
668:
669: \bibitem{19} Koo, B.-C. \& McKee, C.F., 1992, ApJ, 388, 93
670:
671: \bibitem{20} Leitherer, C., et al. 1999, ApJS, 123, 3
672:
673: \bibitem{21} Lozinskaya, T.A. 1992, Supernova and Stellar Wind in the
674: Interstellar Medium, AIP, NY, 223p.
675:
676: \bibitem{22} Lozinskaya, T., Moiseev, A., Podorvanyuk, N., 2003,
677: Astronomy Letters, 29, 77, (Astro-ph/0301214)
678:
679: \bibitem{23} Mac Low, M.-M. \& McCray, R. 1988, ApJ, 324, 776
680:
681: \bibitem{24} Mashchenko, S.Y., Thilker, D.A. \& Braun, R. 1999, A\&A, 343, 352
682:
683: %\bibitem{25} Mashchenko, S.Y. \& Silich, S.A., 1997, ARep, 41, 19
684:
685: \bibitem{26} McClure-Griffiths, N.M., Dickey, J.M., Gaensler, B.M. \&
686: Green, A.J. 2002, AJ, 578, 176
687:
688: \bibitem{27} McCray, R. \& Kafatos, M. 1987, ApJ, 317, 190
689:
690: \bibitem{28} Naz\'e, Y., Chu, Y.-H., Points, S.D., Danforth, C.W.;
691: Rosado, M. \& Chen, C.-H.R.
692:
693: %\bibitem{29} Oey, M.S., 1996, ApJ, 467, 666
694:
695: \bibitem{30} Oey, M.S. \& Garc\'\i a-Segura, G., 2004, ApJ, 613, 302
696:
697: \bibitem{31} Oey, M.S. \& Massey, P. 1995, ApJ, 452, 210
698:
699: \bibitem{32} Palous J., Franco, J. \& Tenorio-Tagle, G. 1990,
700: Astron. Astrophys. 227, 175
701:
702: \bibitem{33} Puche, D., Westpfahl, D., Brinks, E. \& Roy, J-R. 1992, AJ, 103,
703: 1841
704:
705: \bibitem{34} Raga, A.C., Vel\'azquez, P.F., Cant\'o, J., Masciadri, E.,
706: Rodr\'\i guez, L.F., 2001, ApJ, 559, 33
707:
708: \bibitem{35} Rela\~no, M. \& Beckman, J.E. 2005 A\&A, 430, 911
709:
710: \bibitem{36} Rela\~no, M., Beckman, J.E., Daigle, O., Carignan, C. 2007,
711: A\&A, 467, 1117
712:
713: \bibitem{37} Rhode, K. L., Salzer, J. J., Westpfahl, D. \& Radice, L. A.
714: 1999, AJ, 118, 323
715:
716: \bibitem{38} Rozas, M., Beckman, J.E. \& Knapen, J-H. 1996, A\&A, 307, 735
717:
718: \bibitem{39} Silich, S.A. 1992, Ap\&SS 195, 317.
719:
720: \bibitem{40} Silich, S.A., Franco, J., Palou\v{s} \& Tenorio-Tagle, G.
721: 1996, ApJ, 468, 722
722:
723:
724: %\bibitem{40} Silich, S.A., Mashchenko, S.Y., Tenorio-Tagle, G., Franco, J.,
725: % 1996, MNRAS, 280, 711
726:
727: \bibitem{41} Silich, S.A., Tenorio-Tagle G.,
728: Mu\~noz-Tu\~non, C., Cairos L.-M. \&
729: Gil de Paz A., 2002, in ASP Conf. Ser. 282 ``Galaxies: The Third
730: Dimension'' Edts. M. Rosado, L. Binette \& L. Arias,
731: p58
732:
733: \bibitem{42} Silich, S., Lozinskaya, T., Moisseev, A., Podorvanuk, N.,
734: Rosado, M., Borissova, J. \& Valdez-Guti\'errez, M.
735: 2006, Astron. Astrophys., 448, 123
736:
737: \bibitem{43} Smith, D.A. \& Wang, Q.D., 2004, ApJ, 611, 881
738:
739: \bibitem{44} Stevens, I.R. \& Hartwell, J.M., 2003, MNRAS, 339, 280
740:
741: \bibitem{45} Tenorio-Tagle, G. 1981, Astron. Astrophys. 94, 338
742:
743: \bibitem{46} Tenorio-Tagle, G. \& Bodenheimer, P., 1988, ARA\&A, 26, 145
744:
745: \bibitem{47} Valdez-Guti\'errez M., Rosado M., Georgiev L., Borissova J.,
746: Kurtev R. 2001, Astron. Astrophys., 366, 35.
747:
748: \bibitem{48} Wada, K. \& Norman, C.A. 1999, ApJ, 516, L13
749:
750: \bibitem{49} Wada, K., Spaans, M. \& Kim, S., 2000, ApJ, 540, 797
751:
752: \bibitem{50} Weaver, R., McCray, R., Castor, J., Shapiro, P., Moore,
753: R., 1977, ApJ, 218, 377
754:
755: \end{thebibliography}
756:
757: %------------------------------------------------------------------------
758: \clearpage
759:
760: \begin{figure}
761: \includegraphics[angle=0,scale=0.60]{f2.eps}
762: \caption{The comparison of analytic results with numerical models (the
763: example corresponds to the HI shell GSH 285-02+86 and the HII progenitor
764: shell NGC 1530-8). The solid line represents the numeric
765: solution of equations \ref{eq3.a} and \ref{eq3.b}. The dashed line
766: is the analytic solution of equations \ref{eq6.a} and \ref{eq6.b}.
767: The initial radii and masses of the shells are identical in both
768: calculations, however the initial velocities are different. This is
769: because the analytic formulation is based on the assumption that all
770: energy and momentum are deposited instantaneously at the beginning
771: of the momentum dominated stage, whereas the numerical model assumes
772: that the energy and momentum are supplied continuously until the
773: last supernova explosion.} \label{fig2}
774: \end{figure}
775:
776: \begin{figure}
777: \vspace{14.0cm} \special{psfile=f3A.eps angle=270 hoffset=-50
778: voffset=500
779: vscale=50 hscale=40}
780: \special{psfile=f3B.eps angle=270 hoffset=+210 voffset=500
781: vscale=50 hscale=40}
782: \special{psfile=f3C.eps angle=0 hoffset=100 voffset=-90
783: vscale=50 hscale=40}
784: %\special{psfile=f2C.eps angle=270 hoffset=75 voffset=265
785: % vscale=50 hscale=40}
786: \caption{The comparison of the predicted expansion velocities with
787: the observed values. The results of the calculations are compared
788: with parameters of low, intermediate and high mass objects from the
789: list of McClure-Griffiths et al. (2002). Different lines correspond
790: to different initial conditions (associated with 12 HII progenitor
791: shells from the list of Rela\~no et al. 2007). The lines are
792: labeled with the numbers that identify the progenitor shells in
793: Table 1. The observed velocities and radii of the HI shells are
794: marked by the horizontal and vertical lines, respectively. The top
795: left panel compares the model predicted radii and velocities with
796: that of the HI supershell GSH 305+01-24 from the list of
797: McClure-Griffiths et al. (2002). Similarly, the top right and bottom
798: panels present the results for the more massive supershells GSH
799: 304-00-12 and GSH 285-02+86 (the dashed line in the latter is
800: Rela\~no's et al. analytic solution for NGC 1530-8). A
801: 1500~km~s$^{-1}$ star cluster wind terminal speed was adopted for
802: all calculations.} \label{fig3}
803: %\end{center}
804: \end{figure}
805: %\clearpage
806:
807: %\clearpage
808:
809: \begin{figure}
810: \includegraphics[angle=270,scale=0.60]{f4.eps}
811: \caption{The impact of the star cluster wind terminal speed on the
812: shell evolution. The calculations were performed for the HI shell
813: GSH 285-02+86, with a star cluster wind terminal speed $V_{\infty} =
814: 3000$ km s$^{-1}$. This can be compared with that presented in
815: Figure 3 (bottom panel) where a value of $V_{\infty} = 1500$ km
816: s$^{-1}$ for the wind terminal speed has been adopted.} \label{fig4}
817: \end{figure}
818: %------------------------------------------------------------------------
819:
820:
821: \clearpage
822:
823: \begin{deluxetable}{rlcccc}
824: \tablecolumns{6} \tablewidth{0pt} \tabletypesize{\scriptsize}
825: \tablecaption{Initial conditions\label{tbl-1}}
826:
827: \tablehead{\colhead{ID} & \colhead{Name} & \colhead{$M_0$} &
828: \colhead{$R_0$} & \colhead{$u_0$} & \colhead{$L$}\\
829: \colhead{} & \colhead{} & \colhead{$10^4 M_{\odot}$} &
830: \colhead{$pc$} & \colhead{$km~s^{-1}$} & \colhead{$10^{38}
831: erg~s^{-1}$}}
832:
833: \startdata
834:
835: $1$ & NGC 1530--8 & $8.9$ & $104$ & $64.72$ & $11.4$ \\
836: $2$ & NGC 1530--22 & $3.7$ & $91$ & $49.38$ & $5.3$ \\
837: $3$ & NGC 1530--92 & $0.3$ & $41$ & $27.24$ & $0.9$ \\
838: $4$ & NGC 3359--6 & $5.5$ & $104$ & $63.97$ & $5.7$ \\
839: $5$ & NGC 3359--42 & $1.9$ & $73$ & $40.91$ & $1.5$ \\
840: $6$ & NGC 3359--92 & $0.8$ & $55$ & $51.98$ & $0.6$ \\
841: $7$ & NGC 6951--2 & $4.8$ & $100$ & $60.93$ & $7.6$ \\
842: $8$ & NGC 6951--18 & $1.9$ & $73$ & $47.62$ & $1.9$ \\
843: $9$ & NGC 6951--41 & $0.8$ & $56$ & $50.82$ & $0.9$ \\
844: $10$ & NGC 5194--312 & $0.7$ & $53$ & $56.75$ & $1.7$ \\
845: $11$ & NGC 5194--403 & $4.2$ & $96$ & $46.75$ & $3.3$ \\
846: $12$ & NGC 5194--416 & $2.3$ & $79$ & $50.80$ & $1.2$ \\
847:
848: \enddata
849: \end{deluxetable}
850:
851:
852: \clearpage
853: \begin{sidewaystable}
854: {\scriptsize \caption{Model predictions\label{tbl-2}}
855: \begin{tabular}{rl|ccc|ccc|ccc}
856:
857: \multicolumn{2}{c}{} & \multicolumn{3}{|c}{GSH 285--02+86} &
858: \multicolumn{3}{|c}{GSH 304--00-12} & \multicolumn{3}{|c}{GSH 305+01-24}\\
859: \tableline \tableline \\
860: ID & Name & $M$ & $R$ & $V$ & $M$ & $R$ & $V$ & $M$ & $R$ & $V$\\
861: & & $10^5 M_{\odot}$ & $pc$ & $km s^{-1}$ & $10^5 M_{\odot}$ & $pc$ & $km s^{-1}$ & $10^5 M_{\odot}$ & $pc$ & $km
862: s^{-1}$\\
863:
864: \cutinhead{Observed parameters}
865:
866: & & $44$ & $375-395$ & $21$ & $19$ & $200-280$ & $9$ & $3.9$ & $140-220$ & $6$ \\
867:
868: \cutinhead{Predicted parameters}
869:
870: $1$ & NGC 1530--8 & $44$ & $385$ & $7.4$ & $19$ & $240$ & $9.7$ & $3.9$ & $179$ & $26.1$ \\
871: $2$ & NGC 1530--22 & $42$ & $378$ & $0.1$ & $19$ & $240$ & $3.9$ & $3.9$ & $179$ & $14.2$ \\
872: $3$ & NGC 1530--92 & $27$ & $152$ & $0.0$ & $19$ & $112$ & $0.2$ & $2.9$ & $163$ & $0.2$ \\
873: $4$ & NGC 3359--6 & $44$ & $385$ & $1.4$ & $19$ & $240$ & $4.5$ & $3.9$ & $180$ & $16.1$ \\
874: $5$ & NGC 3359--42 & $6.2$ & $201$ & $0.1$ & $4.3$ & $148$ & $0.0$ & $3.9$ & $180$ & $4.6$ \\
875: $6$ & NGC 3359--92 & $1.9$ & $136$ & $0.0$ & $14$ & $103$ & $0.1$ & $2.0$ & $145$ & $0.1$ \\
876: $7$ & NGC 6951--2 & $44$ & $385$ & $4.3$ & $19$ & $240$ & $6.5$ & $3.9$ & $180$ & $17.8$ \\
877: $8$ & NGC 6951--18 & $9.0$ & $227$ & $0.0$ & $6.2$ & $167$ & $0.1$ & $3.9$ & $180$ & $6.5$ \\
878: $9$ & NGC 6951--41 & $2.8$ & $154$ & $0.0$ & $1.9$ & $115$ & $0.1$ & $3.0$ & $165$ & $0.2$ \\
879: $10$ & NGC 5194--312 & $7.9$ & $217$ & $0.1$ & $5.6$ & $160$ & $0.1$ & $3.9$ & $180$ & $5.7$ \\
880: $11$ & NGC 5194--403 & $21$ & $299$ & $0.1$ & $14$ & $218$ & $0.0$ & $3.9$ & $180$ & $10.7$ \\
881: $12$ & NGC 5194--416 & $4.8$ & $185$ & $0.1$ & $3.5$ & $142$ & $0.0$ & $3.9$ & $180$ & $3.4$ \\
882:
883: \end{tabular} }
884: \end{sidewaystable}
885:
886:
887: \end{document}
888:
889:
890: