astro-ph0405275/ms.tex
1: %                                                                 aa.dem
2: % AA vers. 5.2, LaTeX class for Astronomy & Astrophysics
3: % demonstration file
4: %                                                 (c) Springer-Verlag HD
5: %                                                revised by EDP Sciences
6: %-----------------------------------------------------------------------
7: %
8: %\documentclass[referee]{aa} % for a referee version
9: %
10: \documentclass[twocolumn]{aa}
11: \usepackage{graphicx}
12: \usepackage{natbib}
13: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
14: %%%\usepackage{txfonts}  aws
15: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
16: %
17: \def\gray {$\gamma$-ray\ }
18: \def\grays{$\gamma$-rays\ }
19: \def\grayse{$\gamma$-rays}
20: 
21: \def\piodecay{$\pi^0$-decay\ }
22: \def\EB{EGRB\ }
23: \def\Xco{$X_{\rm CO}$}
24: \def\Xcounits{$10^{20}$~molecules~cm$^{-2}$/~(K km s$^{-1}$)\ }
25: \def\Xcounitsa{molecules~cm$^{-2}$/~(K km s$^{-1}$)\ }
26: 
27: \def\deg{^\circ}
28: \def\intensityunits{10$^{-6}$~cm$^{-2}$~sr$^{-1}$~s$^{-1}$\ }
29: \def\regionA{$330\deg<l<30\deg,    |b|<5\deg$\  }
30: \def\regionB{$30\deg<l<330\deg,    |b|<5\deg$\ }
31: \def\regionC{$90\deg<l<270\deg,   |b|<10\deg$\  }
32: \def\regionD{$360\deg<l<0\deg,10\deg<|b|<20\deg$\ }
33: \def\regionE{$360\deg<l<0\deg,20\deg<|b|<60\deg$\ }
34: \def\regionF{$360\deg<l<0\deg,60\deg<|b|<90\deg$\ }
35: \def\regionG{$360\deg<l<0\deg,10\deg<|b|<90\deg$\ }
36: \def\regionH{$300\deg<l<60\deg,    |b|<10\deg$\ } 
37: 
38: \def\spidiffit{{\it spidiffit\ }}
39: \def\fha{48mm}%48
40: \def\fhb{56mm}%61
41: 
42: %
43: \begin{document}
44: %
45: \title{The distribution of cosmic-ray sources in the Galaxy, 
46: \grays and the gradient in the CO-to-H$_2$ relation  }
47: 
48:    \subtitle{ }
49: 
50:    \author{A. W. Strong\inst{1},
51:            I. V. Moskalenko  \inst{2,3},
52:            O. Reimer   \inst{4},
53:            S. Digel   \inst{5},
54:            \and
55:            R. Diehl   \inst{1}
56:           }
57: 
58: \offprints{A. W. Strong, aws@mpe.mpg.de}
59: 
60: \institute{ Max-Planck-Institut f\"ur extraterrestrische Physik,
61: Postfach 1312, D-85741 Garching, Germany           
62:           %    \email{aws@mpe.mpg.de,rod@mpe.mpg.de}
63: \and
64: NASA/Goddard Space Flight Center, Code 661,  Greenbelt, MD 20771, USA 
65: \and
66: Joint Center for Astrophysics, University of Maryland, Baltimore County, 
67: Baltimore, MD 21250, USA
68:  \and
69: Ruhr-Universit\"at  Bochum, D-44780 Bochum, Germany
70: \and
71: W.W. Hansen Experimental Physics Laboratory, Stanford University, 
72: Stanford, CA 94305, USA
73:              }
74: 
75: \date{Received / Accepted  }
76: 
77: 
78: \abstract{
79: We present a solution to the apparent discrepancy between the radial
80: gradient in the diffuse Galactic \gray emissivity and the
81: distribution of  supernova remnants, believed to be the sources of
82: cosmic rays. Recent determinations of the pulsar distribution have
83: made the discrepancy even more apparent. The problem is shown to be
84: plausibly solved  by a variation in the $W_{\rm CO}$-to-$N$(H$_2$) scaling factor.
85: If this factor increases by a factor of 5--10 from the inner to the
86: outer Galaxy, as expected from the Galactic metallicity gradient, 
87: we show that the source distribution 
88: required to match the radial gradient of \grays can be reconciled with
89: the distribution of supernova remnants as traced by current studies of
90: pulsars.  The resulting model fits the EGRET \gray profiles
91: extremely well in longitude, and reproduces the mid-latitude inner
92: Galaxy intensities better than previous models.
93: 
94: \keywords{gamma rays -- Galactic structure -- interstellar medium -- 
95: cosmic rays -- supernova remnants -- pulsars
96:    }
97:    }
98: \titlerunning{Distribution of cosmic ray sources in the Galaxy}
99: \authorrunning{Strong A.W. et al.}
100: \maketitle
101: %
102: %________________________________________________________________
103: 
104: \section{Introduction}
105: 
106: The puzzle of the Galactic \gray gradient goes back to the time of the COS-B
107: satellite \citep{bloemen86,strong88}; using HI and CO surveys to trace the atomic and
108: molecular gas,  the Galactic distribution of emissivity per H atom is a
109: measure of the cosmic-ray (CR) flux, for the gas-related bremsstrahlung and
110: pion-decay components. However the gradient  determined in this way is much
111: smaller than expected if supernova remnants (SNR) are the sources of cosmic
112: rays, as is generally believed. This discrepancy was confirmed with the
113: much more precise data from EGRET on the COMPTON Gamma Ray Observatory,
114: even allowing for the fact that inverse-Compton emission (unrelated to
115: the gas) is more important than originally supposed \citep{SMR00}.
116: A possible explanation of the small gradient in terms of CR propagation, 
117: involving radial variations of a Galactic wind,  was recently put forward by 
118: \citet{breitschwerdt02}.
119: 
120: However the derivation of the Galactic distribution of SNR, 
121: commonly based on radio surveys, is subject to large
122: observational selection effects, so that it can be argued that the
123: discrepancy is not so serious. But other tracers of the distribution
124: of SNR are available, in particular pulsars; the new sensitive Parkes
125: Multibeam survey with 914 pulsars has been used by \citet{lorimer04}
126: to derive the Galactic distribution, and this confirms the
127: concentration to the inner Galaxy.  Fig.~\ref{fig1} compares the
128: pulsar distribution from \citet{lorimer04} with a CR source
129: distribution  which fits the EGRET \gray data \citep{SMR00}.  If the
130: pulsar distribution indeed traces the SNR, then  there is a serious
131: discrepancy with \grayse.  The distribution of SNR given by
132: \citet{case98} is not so peaked, but the number of known SNR is much
133: less than the number of pulsars and the systematic effects very
134: difficult to account for \citep{green96}. But even this flatter
135: distribution is hard to reconcile with that required for \grayse.
136: Another, quite independent, tracer of the SNR  distribution is the
137: 1809 keV line of $^{26}$Al; whether this originates mainly in type II
138: supernovae or masssive stars is not important in this context, since
139: both trace star-formation/SNR. The COMPTEL  $^{26}$Al maps
140: \citep{knodlseder99,pluschke01} show that the emission is very
141: concentrated to the inner radian of the Galaxy.  The density of free
142: electrons shows a similar distribution \citep{cordeslazio03}.  The
143: $^{26}$Al measurements are not subject to the selection effects of
144: other methods; although they have their own uncertainties, they
145: support  the type of distribution which we adopt in this paper.
146: 
147: %-----------------------------------------------------------
148:    \begin{figure}
149: %   \centering     \includegraphics[width=65mm,height=60mm]{fig1.ps}
150:    \centering     \includegraphics[width=0.99\linewidth,height=\fha]{fig1.ps}
151:    \caption{CR source density as function of Galactocentric
152: radius $R$.  Dotted: as used in \citet{SMR00}, solid line:based on
153: pulsars \citep{lorimer04}  as used in this work, vertical bars: SNR
154: data points from \citet{case98}.  Distributions are normalized at $R =
155: 8.5$ kpc.}  \label{fig1}
156:    \end{figure}
157: %______________________________________________________________
158: 
159: A major uncertainty in the models of diffuse Galactic \gray emission
160: is the  distribution of molecular hydrogen, as traced by the
161: integrated intensity of the $J$ = 1--0  transition of $^{12}$CO,
162: $W_{\rm CO}$.  Gamma-ray analyses have in fact provided one of the
163: standard values for the scaling factor\footnote{units: \Xcounitsa}
164: \Xco = $N(H_2)/W_{\rm CO}$; with only the assumption that cosmic rays
165: penetrate molecular clouds freely, the \gray values are free of the
166: uncertainties of other methods (e.g. those based on the assumption of
167: molecular cloud
168:  virialization). However previous analyses, e.g.
169: \citet{strongmattox96}, \citet{hunter97}, \citet{SMR00}, have usually assumed that
170: \Xco\ is independent of Galactocentric radius $R$, since otherwise the
171: model has too many free parameters. But  there is now good reason to
172: believe that \Xco\ increases with $R$, both from COBE/DIRBE studies
173: \citep{sodroski95,sodroski97} and from the measurement of a Galactic
174: metallicity gradient combined with the strong inverse dependence of
175: \Xco\ on metallicity in external galaxies  \citep{israel97,israel00}.
176: A rather rapid radial variation of \Xco\ is expected,  based on a
177: gradient in [O/H] of 0.04--0.07 dex/kpc
178: \citep{hou00,deharveng00,rolleston00,smartt01,andrievsky02} and the
179: dependence of \Xco\ on metallicity in external galaxies: $\log$
180: \Xco$\propto -2.5$   [O/H] \citep{israel97,israel00}, giving \Xco
181: $\propto 10^{(-0.14\pm0.04)R}$, amounting to a factor 1.3--1.5 per
182: kpc, or an order of magnitude between the inner and outer Galaxy.
183: \footnote{The values given by \citet{israel97,israel00} include the
184: effects of the radiation field, implicitly containing the radiation
185: field/metallicity correlation of his galaxy sample. \Xco\ is
186: positively and almost linearly correlated with radiation field, so the
187: dependence of \Xco\ for constant radiation field is even larger: $\log$
188: \Xco$\propto -4$  [O/H] \citep{israel00}. By adopting the coefficient
189: --2.5 we implicitly assume the same radiation/metallicity correlation
190: within the Galaxy as over his galaxy sample.}  A less rapid
191: dependence, $\log$ \Xco$\propto$--1.0  [O/H], was found by
192: \citet{boselli02}, which however still implies a significant \Xco($R$)
193: variation.  \citet{boissier03} also combine the metallicity gradient
194: with \Xco$(Z)$ within individual galaxies, to obtain radial profiles
195: of H$_2$, and give arguments for the validity of this procedure.
196: \citet{digel90} found that molecular clouds in the outer Galaxy 
197: ($R$$\sim$12 kpc) are underluminous in CO, with \Xco\ a factor 4$\pm$2
198: times the  inner Galaxy value.  \citet{sodroski95,sodroski97} derived a
199: similar variation ($\log$\Xco/10$^{20} = 0.12R - 0.34$) when  modelling
200: dust emission for COBE data. \citet{pak98} predicted the physical origin  for a variation of \Xco\ with Z. \citet{papadopoulos02}
201: and \citet{papadopoulos04} discuss  the physical state of this metal-poor gas phase
202: in the outer parts of spiral galaxies (relatively warm and diffuse).  Observations of H$_2$ line emission from NGC~891
203: with ISO \citep{valentijn99}  indicate a massive cool molecular
204: component in the outer regions of this galaxy, supporting the trend
205: found in our Galaxy.
206:  
207: %----------------------------------------------------------- 
208:    \begin{figure}
209:    \centering    
210:    \includegraphics[width=0.99\linewidth,height=\fha]{fig2.ps}
211:    \caption{\Xco\ as function of  $R$.
212:     Dotted horizontal line, black: as used in \citet{strongmattox96,SMR00}; 
213:     solid  line, black: as used for \grays in this work; 
214:     dashed, dark blue: from \citet{sodroski95};
215:     dash-dot, red: using metallicity gradient as described in the text, 
216:     \Xco$\propto Z^{-2.5}$ \citep{israel00}, two lines
217:     for [O/H] = 0.04 and 0.07 dex/kpc;
218:     dash-dot-dot,light blue: using \Xco$\propto Z^{-1.0}$ \citep{boselli02}
219:     and [O/H] =  0.07 dex/kpc.
220:     The values using metallicity are normalized approximately to those 
221:     from the \gray analysis.} \label{fig2}
222:    \end{figure}
223: %______________________________________________________________
224: 
225: Fig.~\ref{fig2} illustrates some of the possible \Xco\ variations
226: implied by these studies.  For the cases where metallicity is used to
227: estimate  \Xco, the values are normalized approximately to the
228: values used in the present \gray analysis, since we are only
229: interested in comparing the variations of \Xco.  From the viewpoint of
230: \grayse, the effect of a steeper CR source distribution is
231: compensated by the increase of \Xco. Thus we might expect to resolve
232: the  apparent discrepancy in the source distribution, and improve our
233: understanding of the Galactic \gray emission. In this paper we
234: investigate quantitatively this possibility.  Note that the \grays
235: include major  contributions from interactions with atomic hydrogen
236: and from inverse Compton scattering, both of which are independent of
237: \Xco; this means that the \Xco\ variation has to be quite large to
238: have a significant effect.
239: 
240: %-----------------------------------------------------------------
241: \section{Data}
242: 
243: The EGRET and COMPTEL data are the same as described in
244: \citet{SMR00,SMR04a}.  The EGRET data consist of the standard product
245: counts and exposure for 30 MeV -- 10 GeV, augmented with data for 10--120
246: GeV. The \gray point sources in the 3EG catalogue have been
247: removed as described in \citet{SMR00}.  The HI and CO data are as
248: described in \citet{moskalenko02} and \citet{SMR04a}; they consist of combined
249: surveys divided into 8 Galactocentric rings on the basis of kinematic
250: information.  Full details of the procedures for comparing models with
251: data are given in \citet{SMR04a} to which the reader is referred.
252: 
253: %----------------------------------------------------------- 
254:    \begin{figure}
255:    \centering    
256:    \includegraphics[width=0.99\linewidth,height=\fhb]{fig3.ps}
257:    \caption{Longitude profile of \grays for 1000--2000 MeV, averaged over 
258: $|b|<5.5\deg$. 
259: Vertical bars: EGRET data; lines are model components convolved with the EGRET 
260: point-spread function:
261: green: inverse Compton emission, red: $\pi^0$-decay, light blue: bremsstrahlung,
262: dark blue: total.} \label{fig3}
263:    \end{figure}
264: %______________________________________________________________
265: 
266: %----------------------------------------------------------------------
267: \section{Model and Method}
268: We use the GALPROP program \citep{SMR00,SMR04a} to compute the models.
269: GALPROP was extended to allow a variable \Xco($R$) to be input.  The
270: distribution of CR sources is assumed to follow that of
271: pulsars in the form given by \citet{lorimer04}, as shown in
272: Fig.~\ref{fig1}. The other parameters, in particular the CR
273: nucleon and electron injection spectral shape and propagation
274: parameters, are taken from the ``optimized model'' of
275: \citet{SMR04a}.  As before the halo height is taken as $z_h$ = 4 kpc,
276: and the maximum radius $R$ = 20 kpc.  The isotropic background is as
277: derived in \citet{SMR04b}.  Since in this work we simply wish to
278: demonstrate the possibility to obtain a plausible solution, we adopt a
279: heuristic approach, adjusting \Xco($R$)  to obtain a satisfactory
280: solution as shown in Fig.~\ref{fig2}.  The electron flux has been
281: scaled down by a factor 0.7 relative to \citet{SMR04a} to obtain an
282: optimal fit.
283: 
284: %-----------------------------------------------------------------------------
285: \section{Results}
286: Figs.~\ref{fig3} and \ref{fig4} show the longitude and latitude
287: distributions  for 1--2 GeV, compared to EGRET data.  A rather rapid
288: variation of \Xco\ is required to compensate the CR source gradient,
289: but it is fully compatible with the expected variation based on
290: metallicity gradients and the COBE result, as described in the
291: Introduction.  The longitude and latitude fits are good except in the
292: outer Galaxy where the prediction is rather low. One possible reason
293: for this is that the CR source density does not fall off so fast
294: beyond the Solar circle as given by the adopted pulsar distribution,
295: which has an exponential decay.  Another possibility could be even
296: larger amounts of H$_2$ in the outer Galaxy than we have assumed (see
297: discussion in Introduction).  We have chosen the range 1--2 GeV for
298: the profiles since this is where the gas contribution and hence the
299: effect of \Xco\ is maximal.  An exhaustive comparison of profiles in
300: all energy ranges is beyond the scope of this Letter, but in fact the
301: agreement is good at all energies.  The larger CR gradient in this
302: model has another consequence:  the predicted inverse-Compton emission
303: in the inner Galaxy  is more intense  at intermediate latitudes where
304: the interstellar radiation field is still high; this is precisely the
305: region where previous models \citep{hunter97,SMR00,SMR04a} have had
306: problems to reproduce the EGRET data.  Fig.~\ref{fig5} shows the model
307: spectrum of the inner Galaxy compared with EGRET data; the fit is
308: similar to that of  models \citep{SMR04a} with {\it ad hoc} source
309: gradient and constant \Xco. The prediction is rather high above 20
310: GeV, however the EGRET data are least certain in this range
311: \citep{SMR04a}.
312:   
313: %----------------------------------------------------------- 
314:    \begin{figure}
315:    \centering    
316:    \includegraphics[width=0.99\linewidth,height=\fhb]{fig4.ps}
317:    \caption{Latitude profile of \grays for 1000--2000 MeV, averaged over 
318: $330\deg<l<30\deg$. Data and curves as in Fig.~\ref{fig3}.  
319: The extragalactic background is shown  as a black horizontal line. } \label{fig4}
320:    \end{figure}
321: %______________________________________________________________
322: 
323: 
324: %------------------------------------------------------------------------------
325: \section{Discussion}
326: We have shown that a good fit to the EGRET data is obtained with the
327: particular  combination of parameters chosen.  We can however ask
328: whether the pulsar source distribution combined with a constant  \Xco\
329: could also give a good fit if we reduce the CR electron intensity, to
330: supress  the inner Galaxy peak from inverse Compton emission. This can
331: indeed reproduce the  longitude profile in the inner Galaxy, but fails
332: badly to account for the latitude  distribution, since it has a large
333: deficit at intermediate latitudes.  Some variation of \Xco\ is
334: therefore required.  The suggested variation of \Xco\ would have
335: significant impact on the  Galactic H$_2$  mass and distribution.
336: Warm molecular hydrogen in the outer parts of spiral  galaxies that is
337: not traced by CO emission may be detectable by the Spitzer
338: observatory in 28 $\mu$m vibrational emission.  These issues will be
339: addressed in future work.
340:  
341: %----------------------------------------------------------- 
342:    \begin{figure}
343:    \centering    
344:    \includegraphics[width=0.99\linewidth,height=\fhb]{fig5.ps}
345:    \caption{Spectrum of inner Galaxy, $330\deg<l<30\deg, |b|<5.5\deg$.
346: Vertical bars: EGRET data  (red), COMPTEL data (green). 
347: Curves: predicted intensities; inverse Compton (green), 
348: $\pi^0$-decay (red), bremsstrahlung (light blue), extragalactic background 
349: (black), total (dark blue). } \label{fig5}
350:    \end{figure}
351: %______________________________________________________________
352: 
353: %------------------------------------------------------------------------------
354: \section{Conclusions}
355: Two {\it a priori} motivated developments allow us to obtain a more
356: physically plausible model for Galactic \grayse, simultaneously
357: allowing a CR source distribution similar to SNR as traced by pulsars
358: and an expected variation in the $W_{\rm CO}$-to-$N$(H$_2$) conversion
359: factor.  Obviously the uncertainty in both the source distribution and
360: \Xco\ are large so our solution is far from unique, but it
361: demonstrates the possibility to obtain a physically-motivated model
362: without resorting to an {\it ad hoc} source distribution. This result
363: supports the SNR origin of CR.  The resulting model also gives
364: improved predictions for \grays in the inner Galaxy at mid-latitudes.
365: We have therefore  achieved a step towards a better understanding of
366: the diffuse Galactic \gray emission.  This result is important input
367: to the development of models for the upcoming GLAST mission.   This
368: Letter is intended only to point out the potential importance of the
369: effect.  The next step will be a more quantitative analysis to derive
370: \Xco($R$) from the \gray data themselves.
371: 
372: %______________________________________________________________
373: \begin{acknowledgements}
374: We  thank F.~Israel and D.~Lorimer and the referee for useful discussions.
375: I.V.M.\ acknowledges partial support from a NASA Astrophysics Theory Program
376: grant, O.R.\ acknowledges 
377: support from the BMBF through DLR grant QV0002.
378:  
379: %of his \Xco - metallicity relation to the Galaxy.
380: % and D. Lorimer for
381: %discussion of the pulsar distribution. 
382: %We thank the referee for pointing out additional relevant  material
383: %and  implications of this work, in particular the relevance for the 
384: %Spitzer observatory.
385: 
386: \end{acknowledgements}
387: 
388: 
389: %%%%%%%
390: 
391: \begin{thebibliography}{} 
392: 
393: \bibitem[Andrievsky et al.(2002)]{andrievsky02}
394: Andrievsky, S. M., Bersier, D., Kovtyukh, V. V., et al. 2002, A\&A, 384, 140 
395: 
396: \bibitem[Boissier et al.(2003)]{boissier03}
397: Boissier, S., Prantzos, N., Boselli, A., \& Gavazzi, G. 2003,
398: MNRAS, 346, 1215
399: 
400: \bibitem[Bloemen et al.(1986)]{bloemen86}
401: Bloemen, J. B. G. M., Strong, A. W., Mayer-Hasselwander, H. A., et al. 1986, 
402: A\&A, 154,  25
403: 
404: \bibitem[Breitschwerdt et al.(2002)]{breitschwerdt02}
405: Breitschwerdt, D., Dogiel, V. A., \& V\"olk, H. J. 2002, A\&A, 385,  216
406: 
407: \bibitem[Boselli et al.(2002)]{boselli02}
408: Boselli, A., Lequeux, J., \& Gavazzi, G. 2002, A\&A, 384,  33
409: 
410: \bibitem[Case \& Bhattacharya(1998)]{case98}
411: Case, G. L., \& Bhattacharya, D. 1998, ApJ, 504, 761
412: 
413: \bibitem[Cordes \& Lazio(2003)]{cordeslazio03}
414: Cordes, J. M., \& Lazio, T. J. W. 2003, astro-ph/0207156
415: 
416: \bibitem[Deharveng et al.(2000)]{deharveng00}
417: Deharveng, L., Pena, M., Caplan, J., \& Costero, R. 2000, MNRAS 311, 329
418: 
419: \bibitem[Digel et al.(1990)]{digel90}
420: Digel, S., Bally, J., \& Thaddeus, P. 1990, ApJ, 357, L29
421: 
422: \bibitem[Green(1996)]{green96}
423: Green, D. A. 1996, in Supernovae and Supernova Remnants, 
424: IAU Coll.\ 145, R.~McCray and Z.~Wang (eds),
425: Cambridge University Press, p.341
426: 
427: \bibitem[Hou et al.(2000)]{hou00}
428: Hou, J. L., Prantzos, N., \& Boissier, S. 2000, A\&A 362, 921
429: 
430: \bibitem[Hunter et al.(1997)]{hunter97}
431: Hunter, S. D., Bertsch, D. L., Dingus, B.L., et al. 1997, ApJ, 481, 205
432: 
433: \bibitem[Kn\"odlseder et al.(1999)]{knodlseder99}
434: Kn\"odlseder, J., Dixon, D. D., Bennett, K., et al. 1999, A\&A 344, 68
435: 
436: \bibitem[Israel(1997)]{israel97}
437: Israel, F. P. 1997, A\&A, 328, 471 
438: 
439: \bibitem[Israel(2000)]{israel00}
440: Israel, F. P. 2000, in Molecular Hydrogen in Space, F.\ Combes and G.\ 
441: Pineau des For\^ets (eds), p.293
442: 
443: \bibitem[Lorimer(2004)]{lorimer04}
444: Lorimer, D. R. 2004, in Young Neutron Stars and Their Environments, IAU Symp.\ 218, 
445: F.\ Camilo and B.\ M.\ Gaensler (eds), astro-ph/0308501
446: 
447: \bibitem[Moskalenko et al.(2002)]{moskalenko02}   
448: Moskalenko, I. V., Strong, A. W., Ormes, J. F., \& Potgieter, M. S. 2002, ApJ, 565, 280
449: 
450: \bibitem[Papadopoulos et al.(2002)]{papadopoulos02} 
451: Papadopoulos, P. P., Thi, W.-F., \& Viti, S. 2002, ApJ, 579, 270 %(erratum 583, 524)
452: 
453: \bibitem[Papadopoulos(2004)]{papadopoulos04} 
454: Papadopoulos, P. P. 2004, in The Neutral ISM in Starburst Galaxies, 
455: PASP Conf.\ Series, in press, astro-ph/0403087
456: 
457: \bibitem[Pak et al.(1998)]{pak98}
458: Pak, S., Jaffe, D. T., van Dishoeck, E. F., Johansson, L. E. B., 
459: Ewine, F. \& Booth, R. S. 1998, ApJ, 498, 735
460: 
461: \bibitem[Pl\"uschke et al.(2001)]{pluschke01}
462: Pl\"uschke, S., Diehl, R., Sch\"onfelder, V., et al. 2001, ESA SP--459, 55
463: 
464: \bibitem[Rolleston et al.(2000)]{rolleston00}
465: Rolleston, W. R. J., Smartt, S. J., Dufton, P. L., \& Ryans, R. S. I. 2000, 
466: A\&A, 363, 537
467: 
468: \bibitem[Sodroski et al.(1995)]{sodroski95}
469: Sodroski, T. J., Odegard, T. J., Dwek, E., et al. 1995, ApJ, 452, 262
470: 
471: \bibitem[Sodroski et al.(1997)]{sodroski97}
472: Sodroski, T. J., Odegard, N., Arendt, R., et al. 1997, ApJ, 480, 173
473: 
474: \bibitem[Smartt(2001)]{smartt01}
475: Smartt, S. J., Venn, K. A., Dufton, P. L., Lennon, D. J., 
476: Rolleston, W. R. J., \& Keenan, F. P. 2001, A\&A, 367, 86
477: 
478: \bibitem[Strong et al.(1988)]{strong88}
479: Strong, A. W., Bloemen, J. B. G. M.,  Dame, T. M., et al. 1988, A\&A, 207, 1
480: 
481: \bibitem[Strong \& Mattox(1996)]{strongmattox96}
482: Strong, A. W., \& Mattox, J. R. 1996, A\&A, 308, L21
483: 
484: \bibitem[Strong et al.(2000)]{SMR00}
485: Strong, A. W., Moskalenko, I. V., \& Reimer, O. 2000, ApJ, 537, 763
486: 
487: \bibitem[Strong et al.(2004a)]{SMR04a}
488: Strong, A. W., Moskalenko, I. V., \& Reimer, O. 2004a, submitted to ApJ 
489: 
490: \bibitem[Strong et al.(2004b)]{SMR04b}
491: Strong, A. W., Moskalenko, I. V., \& Reimer, O. 2004b, submitted to ApJ 
492: 
493: \bibitem[Valentijn \& van der Werf(1999)]{valentijn99}
494: Valentijn, E. A. \& van der Werf, P. P. 1999, ApJ, 522, L29
495: 
496: \end{thebibliography}
497: \end{document}
498: