1: \documentstyle[11pt,newpasp,twoside,epsf]{article}
2: \markboth{Yao}{APS Conf. Ser. Style}
3: \pagestyle{myheadings}
4: \nofiles
5:
6: % Some definitions I use in these instructions.
7:
8: \def\emphasize#1{{\sl#1\/}}
9: \def\arg#1{{\it#1\/}}
10: \let\prog=\arg
11:
12: \def\edcomment#1{\iffalse\marginpar{\raggedright\sl#1\/}\else\relax\fi}
13: \marginparwidth 1.25in
14: \marginparsep .125in
15: \marginparpush .25in
16: \reversemarginpar
17:
18: \begin{document}
19: \title{Properties of CO Molecular Gas in IR Luminous Galaxies}
20: \author{Lihong Yao}
21: \affil{Department of Astronomy \& Astrophysics, University of Toronto,
22: Toronto, ON M5S 3H8, Canada}
23: %\author{Ima Co-Author}
24: %\affil{The Name of My Institution, The Full Address of My Institution}
25:
26: \begin{abstract}
27: We present the properties of the $^{12}$CO(1-0) and (3-2) line
28: emission from the nuclei of 60 IR luminous SLUGS galaxies.
29: This subsample is flux limited at $S_{60 \mu m}$ $\ge$ 5.24 Jy
30: with FIR luminosities mostly at $L_{FIR}$ $>$ 10$^{10}$ L$_{\odot}$.
31: The emission line strengths of $^{12}$CO(1-0) and (3-2) transitions
32: were compared at a common resolution of $\sim$ 15$^{\prime\prime}$,
33: and the line ratios $r_{31}$ vary from 0.22 to 1.72 with a mean value
34: of 0.66 for the sources observed, indicating a large spread of the
35: degree of excitation of CO in the sample. Our analysis shows that
36: (1) there is a non-linear relation between CO and FIR luminosities, such
37: that their ratio $L_{CO}$/$L_{FIR}$ decreases linearly with increasing
38: $L_{FIR}$, (2) we find a possible dependence of the degree of
39: CO gas excitation on the efficiency of star forming activity,
40: (3) using the large velocity gradient (LVG) approximation to model
41: the observed data, the results show that the mean value of the CO-to-H$_2$
42: conversion factor $X$ for the SLUGS sample is lower by a factor of 10
43: compared to the conventional value derived for the Galaxy, assuming
44: that the abundance of CO relative to H$_2$ is 10$^{-4}$,
45: (4) due to a contribution to the SCUBA brightness measurement by
46: $^{12}$CO(3-2) emission, the average dust mass is reduced by 25-38\% ,
47: and the mean global gas-to-dust mass ratio is reduced from 430
48: to 360, but is further reduced to 100 when applied to
49: the nuclear regions of the SLUGS galaxies, (5) for a subset of
50: 12 galaxies with H I maps, we derive a mean total face-on surface
51: density of H$_2$+H I of about 42 M$_{\odot}$ pc$^{-2}$ within $\sim$
52: 2 kpc of the nucleus, which is intermediate between that in
53: galaxies like our own and those with strong star formation.
54: \end{abstract}
55:
56: \section{Introduction}
57:
58: Understanding the properties and evolution of the gas and dust content in
59: nearby IR luminous galaxies is essential for understanding the cause
60: and temporal evolution of starburst activity and its role in the
61: cosmic evolution of galaxies. Therefore, there is a need to investigate
62: large statistical samples of IR luminous galaxies using a multitude
63: of different types of data, including CO, H I, and continuum in the
64: submillimeter (sub-mm), FIR and radio, in order to constrain theories
65: of how the interstellar medium (ISM) evolves. Millimeter and submillimeter
66: CO rotational lines are often used as tracers of molecular hydrogen.
67: The ratio of $^{12}$CO(3-2) to (1-0) line emission provides more sensitive
68: measure of gas temperature and density than the ratio of $^{12}$CO(2-1) to
69: (1-0) lines. Recently, we have presented the largest statistical CO
70: survey (Yao et al. 2003) for the nearby universe by investigating the
71: properties of CO molecular gas in a nearly complete subsample of 60
72: IR luminous galaxies selected from the SLUGS Survey (Dunne et al. 2000).
73: The SLUGS survey containing 104 galaxies is based on the Revised Bright
74: Galaxy Sample of {\it IRAS} galaxies (Soifer et al. 1987) within
75: -10$^{\circ}$ $\le$ $\delta$ $\le$ +50$^{\circ}$ and with $cz$ $\ge$
76: 1900 km s$^{-1}$, a flux limit of $S_{60\mu m}$ $\ge$ 5.24 Jy, and
77: the FIR luminosity $L_{FIR}$ $\ge$ 10$^{10}$ L$_{\odot}$.
78:
79: \section{Properties of Molecular Gas and Dust}
80:
81: A study of molecular gas of a complete flux-limited subsample is important
82: to complement the study of the dust. The angular resolution of the
83: $^{12}$CO(1-0) and (3-2) point observations presented in our CO survey are
84: nearly identical ($\sim$ 15$^\prime$$^\prime$). A complete discussion
85: of the observations, the CO line spectra, and the data analysis is contained
86: in Yao et al. (2003) and references therein. Together with the sub-mm
87: fluxes (Dunne et al. 2000), plus existing data on H I (Thomas et al. 2002),
88: radio continuum (Condon et al. 1990), and FIR (Dunne et al. 2000;
89: NED\footnote{The NASA/IPAC Extragalactic Database operated by the Jet
90: propulsion Laboratory.}), we are able to search for a relationship
91: between the degree of excitation of the CO in this sample and the star
92: forming properties.
93:
94: \begin{figure}
95: \plotone{figure1.eps}
96: \caption{(a) CO luminosity vs. FIR luminosity applicable to the
97: 15$^{\prime\prime}$ beam, (b) the $r_{31}$ ratio vs. star formation
98: efficiency, the line represents a linear regression fit to the data
99: with SFE $\le$ 200 L$_{\odot}$/M$_{\odot}$, the $M$(H$_2$) is computed
100: using our derived $X$-factor, i.e. 2.7$\times$10$^{19}$
101: cm$^{-2}$(K km s$^{-1}$)$^{-1}$, (c) $r_{31}$ vs. $T_{dust}$ for
102: different H$_2$ gas densities log$_{10}$ $n$(H$_2$) based on LVG model
103: for $\Lambda$ = 10$^{-5}$ (km s$^{-1}$ pc$^{-1}$), assuming
104: $T_{kin}$ = $T_{dust}$, the $^{12}$CO abundance is 10$^{-4}$,
105: and the isotope abundance ratio $^{12}$CO/$^{13}$CO = 40, (d) a
106: comparison between the distributions of computed (dotted line)
107: and observed isotope $^{12}$CO(1-0)/$^{13}$CO(1-0) ratios (solid line)
108: for the best agreement regime of $\Lambda$ = 10$^{-5}$, (e) SCUBA equivalent
109: flux $S$($V$) produced by $^{12}$CO(3-2) line vs. SCUBA flux at the central
110: pixel for SLUGS sources for $v$ $<$ 10,000 km s$^{-1}$ (filled circles) and
111: $v$ $>$ 10,0000 km s$^{-1}$ (open circles), the slope for the linear regression
112: fit is 0.26$\pm$0.01, and (f) histogram of $S$($V$), the dash curve represents
113: the expected distribution for a true ratio of $r$ = 0.24 and mean fractional
114: uncertainties of respectively 32\% and 25\% for $S$($V$) (numerator)
115: and $S_{850\mu m}$ (denominator).}
116: \end{figure}
117:
118: \subsection{CO and FIR Luminosities}
119:
120: Figure 1(a) shows a comparison between the CO and FIR luminosities
121: for two different FIR color regimes. A linear fit to the 1-0 data in the
122: log-log plane yields log$_{10} L_{^{12}CO(1-0)}$ = (0.2 $\pm$ 0.7) +
123: (0.57 $\pm$ 0.07) log$_{10} L_{FIR}$. The nonlinear variation of $L_{CO}$
124: with $L_{FIR}$ implies a decrease in the ratio $L_{CO}$/$L_{FIR}$ with
125: increasing $L_{FIR}$, so that more IR luminous galaxies have higher
126: dust temperatures and star formation efficiency. No correlation with
127: the projected beam size on the galaxies is evident. The ratio depends
128: intrinsically on the total FIR luminosity. This implies that for high-$z$
129: objects where $L_{FIR}$ is exceedingly high, the corresponding $L_{CO}$
130: will be seriously overestimated if a linear relation between $L_{CO}$
131: and $L_{FIR}$ is assumed.
132:
133: \subsection{Star Formation Efficiency}
134:
135: We examined the excitation of CO and its relation with the properties
136: of gas/dust and star formation in the the central starburst regions
137: in the SLUGS sample. There are no significant correlations between
138: the $^{12}$CO(3-2) to (1-0) line ratio $r_{31}$ and the distance of
139: galaxies, star formation rate, $T_{dust}$ and $M_{dust}$, H$_2$ gas mass,
140: and the color indices. This might reflect a range of localized conditions
141: in the molecular clouds. But the degree of CO excitation measured by
142: $r_{31}$ varies greatly from galaxy to galaxy. There is a trend for
143: the $r_{31}$ to increase with increasing concentration and efficiency
144: of star-forming activity (see Figure 1 (b)). The saturation of $r_{31}$
145: seen at higher SFE implies that the gas is denser and warmer in
146: regions of higher SFE, which is consistent with the nonlinearity
147: between $L_{CO}$ and $L_{FIR}$.
148:
149: \subsection{The CO-to-H$_2$ Conversion Factor $X$}
150:
151: Our CO line measurements, together with dust and CO isotope data
152: taken from the literature (Dunne et al. 2000; Aalto et al. 1995;
153: Taniguchi et al. 1999), are modeled using the LVG approximation
154: yielding $X$ = $\frac{n(H_2) \Lambda}{X_{CO} T_{rad}}$
155: (see Figure 1 (c) and (d)) to estimate that the controversial
156: $X$ lies between 1.3$\times$10$^{19}$ and 6.7$\times$10$^{19}$
157: cm$^{-2}$(K km s$^{-1}$)$^{-1}$ for SLUGS galaxies, which is
158: about 4-20 times lower than the conventional $X$ factor derived
159: from the Galaxy. The mean value 2.7$\times$10$^{19}$ is comparable
160: to that estimated from diffuse clouds in the Galaxy and with that
161: found for extreme starbursts in nearby galaxies.
162:
163: \subsection{Virial Stability of the Molecular Clouds}
164:
165: Using virial analysis, we show that the molecular clouds in starburst
166: regions are not gravitationally bound, confirming the suggestion
167: recently made by Zhu et al. (2003) for the Antennae galaxies,
168: unless one is willing to accept a 9-90 times lower [CO/H$_2$]
169: abundance ratio. The cause could be the destruction by the stellar
170: winds and expanding shells of newborn massive stars. Most of the
171: CO line emission originates from the nonvirialized warm and
172: diffuse gas clouds.
173:
174: \subsection{Revised Masses of Dust and Gas of SLUGS Galaxies}
175:
176: Our $^{12}$CO(3-2) observations provide an important database for
177: correcting the SCUBA 850 $\mu$m data in the SLUGS sample (see Figure 1 (e)
178: and (f)), thus permitting a revised characterization of the masses of
179: dust and gas (see Seaquist et al. 2003). The average downward correction
180: of dust mass is 25-38\%, which has no bearing on earlier conclusions
181: concerning the shapes of the dust mass-luminosity function derived
182: from SLUGS survey. Using the $^{12}$CO(3-2)/(1-0) ratios in Yao et al.
183: (2003), we estimate that it is unlikely that any correction is required
184: for the contribution by $^{12}$CO(6-5) to 450 $\mu$m fluxes of the
185: SLUGS galaxies measured with SCUBA. The revised mean gas-to-dust ratio
186: (g/d) is reduced from 310 to 100, which is significantly lower than
187: those reported for the global g/d ratios for IR luminous galaxies.
188:
189: \acknowledgements I thank my advisor Prof. Ernie Seaquist and
190: Reinhardt Travel Fellowship at Univ. of Toronto for their support. I also
191: thank Kavli Institute for Theoretical Physics for their generosity.
192:
193:
194: \begin{references}
195: \reference Aalto, S., et al. 1995, \aap, 300, 369
196: \reference Condon, J. J., Helou, G., Sanders, D. B., \& Soifer, B. T. 1990, \apjs, 73, 359
197: \reference Dunne, L., et al. 2000, \mnras, 315, 115
198: \reference Seaquist, E. R., Yao, Lihong, Dunne, L., \& Cameron, H. 2003, \mnras, in press
199: \reference Soifer, B. T., et al. 1987, \apj, 320, 238
200: \reference Taniguchi, Y., Ohyama, Y., \& Sanders, D. B. 1999, \apj, 522, 214
201: \reference Thomas, H. C., et al. 2002, \mnras, 329, 747
202: \reference Yao, Lihong, Seaquist, E. R., Kuno, N., \& Dunne, L. 2003, \apj, 588, 771
203: \reference Zhu, M., Seaquist, E. R., \& Kuno, N. 2003, ApJ, 588, 243
204: \end{references}
205:
206: \end{document}
207: