astro-ph0104431/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: 
3: \def\gs{\mathrel{\raise0.35ex\hbox{$\scriptstyle >$}\kern-0.6em
4: \lower0.40ex\hbox{{$\scriptstyle \sim$}}}}
5: \def\ls{\mathrel{\raise0.35ex\hbox{$\scriptstyle <$}\kern-0.6em
6: \lower0.40ex\hbox{{$\scriptstyle \sim$}}}}
7: 
8: \shorttitle{Molecular  gas  in  HR\,10}
9: \shortauthors{Papadopoulos \& Ivison}
10: 
11: \begin{document}
12: 
13: \title{Low-excitation gas in HR\,10: possible 
14:       implications for estimates of metal-rich H$_2$ mass at high
15:       redshifts}
16: 
17: \author{P.\ P.\ Papadopoulos}
18: \affil{Astrophysics Division, Space Science Department of ESA,
19:              ESTEC, Postbus 299, NL-2200 AG, Noordwijk,
20:              The Netherlands; ppapadop@rssd.esa.int}
21: 
22: \and
23: 
24: \author{R.\ J.\ Ivison}
25: \affil{Astronomy Technology Centre, Royal Observatory, Blackford Hill,
26: Edinburgh EH9 3HJ, UK; rji@roe.ac.uk}
27: 
28: \begin{abstract}
29: We examine the physical conditions of the molecular gas in the
30: extremely red object (ERO), HR\,10 (J164502+4626.4, or [HR94]\,10) at
31: $z=1.44$, as constrained by the recent detection of CO $J+1
32: \rightarrow J, J=1, 4$ lines.  The line ratios of such widely spaced
33: CO transitions are very sensitive probes of gas excitation; in the
34: case of HR\,10, they provide a rare opportunity to study the state of
35: H$_2$ gas in a gas-rich starburst at a cosmologically significant
36: distance.  Contrary to earlier claims, we find the CO $J=5-4$
37: transition to be of low excitation with a brightness temperature ratio
38: of ($5-4$)/($2-1$) $\sim$ 0.16.  Depending upon how often such
39: conditions prevail in intense starbursts, this may have serious
40: consequences for molecular gas mass estimates and the detectability of
41: high-$J$ CO lines from similar objects at high redshifts ($z\ga 3$).
42: \end{abstract}
43: 
44: \keywords{galaxies: evolution --- galaxies: individual (HR\,10) ---
45: galaxies: ISM --- galaxies: starburst --- infrared: galaxies}
46: 
47: \section{Introduction}
48: 
49: A defining characteristic of highly obscured active galactic nuclei
50: and star-forming galaxies is their red optical/infrared (IR) colours.
51: The first few EROs (usually defined such that $R-K>6$ or $I-K>5$) were
52: found soon after the advent of near-IR array detectors (Elston, Rieke
53: \& Rieke 1988) but were viewed as little more than curiosities.  It
54: has become  clear,  however, that  while   EROs provide  a  negligible
55: contribution  to the  optical extragalactic background,  they may host
56: power  sources which dominate the  background at other wavelengths, in
57: particular in the submillimetre (submm) and hard X-ray wavebands.
58: 
59: Recent  work suggests that  EROs  form a  bi-modal population: a large
60: fraction, possibly around  half at  $K>20$,   can be identified   with
61: obscured starbursts (e.g., Dey  et al.\ 1999); the remainder represent
62: the end products of these starbursts --- massive ellipticals at $z
63: \sim$ 1--2 (e.g., Dunlop et al.\ 1996) whose colours are due to their
64: evolved stellar populations.
65: 
66: Submm surveys with SCUBA (Holland et al.\ 1999) have demonstrated that
67: a large fraction of the extragalactic submm background originates in
68: ultra-luminous IR galaxies (ULIRGs, $L_{\rm FIR}\ga
69: 10^{12}$\,L$_{\odot}$) at high redshifts (Smail, Ivison \& Blain
70: 1997), a population which provides well below 1 per cent of the local
71: luminosity density and which must have evolved dramatically, having
72: provided closer to half of the total background at earlier times. Like
73: the local examples of ULIRGs, many submm sources are distant, gas-rich
74: mergers, the probable progenitors of luminous present-day ellipticals
75: (Ivison et al.\ 1998, 2000; Smail et al.\ 1998, 2000; Frayer et al.\
76: 1998, 1999; Eales et al.\ 1999).  Unsurprisingly, given the bi-modal
77: population discussed earlier, several have been identified with EROs
78: (Smail et al.\ 1999). Submm data, in fact, have proved to be an ideal
79: discriminant between the obscured starburst and old elliptical EROs
80: (Dey et al.\ 1999).
81: 
82: HR\,10, an optical/IR-selected galaxy at $z=1.44$ (Hu \& Ridgway 1994;
83: Graham \& Dey 1996), was the first ERO to be detected at submm
84: wavelengths (Cimatti et al.\ 1998; Dey et al.\ 1999) and to be thus
85: revealed unequivocally as an extremely luminous, dust-obscured
86: starburst ($\rm L_{\rm FIR}\sim 10^{12}\ L_{\odot }$), a fact that
87: would have been difficult to surmise from its non-descript
88: optical/near-IR appearance (Dey et al.\ 1999).
89: 
90: The  proper estimation  of  molecular  gas   mass in such    gas-rich,
91: IR-luminous objects is of central importance since the gas fuels their
92: often   enormous  ($\sim 10^3$ M$_{\odot  }$~yr$^{-1}$) star-formation
93: rates (e.g., Solomon et al.\ 1992; Dunlop et al.\ 1994; Rowan-Robinson
94: 2000 and references therein).  Comparing $M$(H$_2$) with the dynamical
95: mass allows the determination of  the evolutionary status of a galaxy,
96: while a comparison  of its dynamical  mass with that  of a present-day
97: spiral or elliptical can point towards its  possible descendent at the
98: current   epoch.   Finally,   since  a  dust-mass   estimate is  often
99: available, the resulting  gas-to-dust ratio is a significant indicator
100: of the   metal enrichment and  chemical evolution   of galaxies in the
101: early Universe (e.g., Frayer \& Brown 1997).
102: 
103: In this paper we examine the  physical conditions of the molecular gas
104: in  the ERO,  HR\,10  at  $z=1.44$, and  their  effect on  H$_2$  mass
105: estimates.  Finally, to the extent that such conditions are frequently
106: encountered in  starburst  environments,  there   may be  far-reaching
107: consequences  for the detection and interpretation  of high-J CO lines
108: emitted from  starbursts at high redshifts  --- for the prioritisation
109: of  observing bands and  receivers  for the  Atacama  Large Millimetre
110: Array (ALMA),  for example. Throughout  this work, we  adopt $q_{\circ
111: }=0.5$ and $H_{\circ }=75\rm \ km\ s^{-1}\ Mpc^{-1}$.
112: 
113: 
114: \section{Physical properties of H$_2$ in HR\,10}
115: 
116: The detection  of the  two  widely  spaced CO transitions  in   HR\,10
117: ($J=2-1, J=5-4$, Andreani et al.\ 2000) offers a unique opportunity to
118: examine  the  physical  conditions  of the   molecular  gas using  the
119: excitation-sensitive ratio  of their brightness temperatures.  This is
120: a   rare occurrence  for high-redshift objects   ($z\ga   1$) since it
121: requires a fortuitous combination of redshift and available receivers.
122: For low-redshift objects ($z\la 0.1$), systematic studies of this sort
123: are hindered    by  large atmospheric  attenuation   and  low  antenna
124: efficiencies for frequencies $\nu > 400$\,GHz, where the CO $J+1
125: \rightarrow J, J+1 > 3$ transitions lie and, until the commissioning
126: of a new generation of wide-band heterodyne receivers, by the
127: bandwidth and sensitivity limitations of available instrumentation.
128: 
129: 
130: The detection of the  CO $J=2-1$ transition is  particularly important
131: since, together with CO $J=1-0$, is  the most widely studied molecular
132: line in the local Universe.  The critical  densities of CO $J=1-0$ and
133: $J=2-1$ (i.e.  the densities at which the collisional  rate out of the
134: upper level and its  radiative de-excitation become equal) are $n_{\rm
135: crit}\sim 2.4\times  10^2-2.4\times 10^3$\,cm$^{-3}$ (for  $\rm T_{\rm
136: k}=100$ K;  see e.g. Kamp \& van  Zadelhoff 2001), and $\Delta E/k\sim
137: 5-10$\,{\sc  k}.  These  minimal  excitation requirements allows these
138: transitions to trace the bulk of the  metal-enriched molecular gas and
139: there  are  several surveys measuring their  relative  strengths for a
140: wide variety of environments (e.g.,  Braine  \& Combes 1992;  Hasegawa
141: 1997; Papadopoulos \& Seaquist 1998).
142: 
143: The CO $J+1 \rightarrow J$ line luminosity in units of {\sc
144: k}\,km\,s$^{-1}$\,pc$^{2}$ is given by
145: 
146: \noindent
147: \begin{equation} 
148: L_{\rm CO}(J+1\rightarrow J) = \frac{c^2}{2\ k\ \nu_{J+1,J}^2}
149: \left[\frac{D^2_{\rm L}}{\ 1+z\ }\right] \int _{\Delta v} S_{\nu _{\rm
150: obs}} d v,
151: \end{equation}
152: 
153: \noindent
154: where $D_{\rm L}=2\ c\ H_{\circ}^{-1} (1+z -\sqrt{1+z})$ is the
155: luminosity distance, and $\nu_{J+1, J}$, $S_{\nu_{\rm obs}}$, are the
156: rest-frame frequency and the observed flux density of the CO $J+1
157: \rightarrow J$ transition (see, e.g., Ivison et al.\ 1996).
158: 
159: Andreani et al.\ (2000) have reported almost equal velocity-integrated
160: flux  densities for  the  $J=5-4$ and  $J=2-1$ transitions, hence  the
161: area/velocity-integrated brightness temperature ratio will simply be
162: 
163: \noindent
164: \begin{equation}
165: \frac{L_{\rm CO}(5-4)}{L_{\rm CO}(2-1)}\sim \left( \frac{\nu
166: _{21}}{\nu _{54}}\right)^2=\left( \frac{2}{5}\right)^2= 0.16
167: \end{equation}
168: 
169: \noindent
170: which is $\sim 4$ times lower than  their reported value and indicates
171: the low excitation of the CO $J=5-4$ line.  The new value is bracketed
172: by the global value of $\sim 0.07$ for  the Milky Way, as derived from
173: {\em Cosmic Background  Explorer} measurements (Wright et al.\  1991),
174: and  $\sim  0.35-1$ expected  for the  conditions  in  the inner $\sim
175: $500\,pc of  the archetypal starburst  M\,82 (using data from Seaquist
176: \& Frayer 2000; Mao et al.\ 2000 and references therein).
177:  
178: The  surprising fact  is not that   the aforementioned  line  ratio is
179: higher for the starburst environment of  HR\,10 than for the quiescent
180: one of the Milky  Way but that it  is at least  a  factor of $\sim  2$
181: lower  than in  the prototype  starburst galaxy,  M\,82.  The  case of
182: HR\,10 shows  that  very different  excitation  conditions may  indeed
183: prevail in  such   systems.    It is worth    noting  that   such  low
184: (5--4)/(2--1) ratios are  associated with the low-excitation gas phase
185: present in  M\,82  (e.g., Mao  et al.\  2000  and references therein),
186: suggesting that this phase is dominant in HR\,10.
187: 
188: The estimate of H$_2$ mass  in the Galaxy  and other galaxies utilises
189: the bright  CO $J=1-0$ emission  and the so-called Galactic conversion
190: factor,  $X_{\rm  CO}$ (e.g., Young  \&  Scoville  1982, 1991; Bloemen
191: 1985; Dickman, Snell \& Schloerb 1986; Bryant \&  Scoville 1996).  For
192: an observed CO $J+1 \rightarrow J$ line, the  H$_2$ mass is then given
193: by
194: 
195: \noindent
196: \begin{equation}
197: M({\rm H}_2)=X_{\rm CO}\ R_{J+1, J}^{-1} \ L_{\rm CO}(J+1, J)
198: \end{equation}
199: 
200: \noindent
201: where $R_{J+1, J}$ is ratio of the CO $J+1 \rightarrow J$ and $J=1-0$
202: line luminosities.
203: 
204: The  range  of  $R_{21}$  observed   in  a variety    of  Galactic and
205: extragalactic environments  is $\sim  0.5-1$ (e.g., Braine   \& Combes
206: 1992; Hasegawa 1997; Papadopoulos \& Seaquist 1998) hence the expected
207: range of  $R_{54}$ for  HR\,10  is $\sim 0.08-0.16$  (see Equation~2),
208: significantly lower than $\sim 1$,   the value for an optically  thick
209: and thermalized line.
210: 
211: Unlike strongly lensed objects, where   differential lensing may  bias
212: the dust continuum and line ratios towards a warmer dust/gas component
213: (Blain 1999; Papadopoulos  et   al.\    2000), the  lack  of    strong
214: gravitational lensing  in   HR\,10 (Dey  et al.\ 1999)   allows a more
215: straightforward interpretation  of  the observed line   ratio.  In the
216: absence  of more data,  we  examine the physical conditions compatible
217: with the observed (5--4)/(2--1) line ratio assuming a single gas phase
218: and  using  a  large-velocity gradient   (LVG)  code (e.g., Richardson
219: 1985).  The code was run for the typical range of temperatures $T_{\rm
220: kin}=30-60$\,{\sc k} that characterise  the warm-dust component (where
221: the  bright   CO    emission  also arises)   in    IR-bright  galaxies
222: (e.g.  Devereux  \& Young 1990   and references  therein).   A typical
223: solution    is      $T_{\rm     kin}=45$\,{\sc k}    and       $n({\rm
224: H}_2)=10^3$\,cm$^{-3}$  ($T_{\rm dust}\sim 40$\,{\sc  k}, Dey  et al.\
225: 1999).      Such a  diffuse,    low-excitation gas    phase with  such
226: temperatures is  not unusual in  extreme starburst systems  (Downes \&
227: Solomon 1998); it is present (but not dominant)  in much less luminous
228: starbursts like M\,82 (Hughes, Gear, \& Robson 1994; Mao et al.\ 2000)
229: and is   probably the product  of their  UV-intense  and kinematically
230: violent environments (Aalto et al.\ 1995).
231: 
232: For these conditions $R_{54}\sim 0.14$,  while higher transitions have
233: $R_{65}\sim   0.035$  and $R_{\rm J+1,   J}   \la 5\times 10^{-3}$ for
234: $J+1>6$.  The low excitation of the $J=5-4$  line in particular is due
235: to the low density rather than  temperature since, for optically thick
236: and thermalized gas, such a low ratio implies $T_{\rm kin} < 10$\,{\sc
237: k},  much  lower than  the observed  dust temperature.    The claim by
238: Andreani et al.\ that $T_{\rm kin}$ is usually significantly less than
239: $T_{\rm dust}$ is not valid; usually they are  similar, except for the
240: photodissociation  regions  (PDRs) in  the surfaces  of UV-illuminated
241: molecular clouds where $T_{\rm kin} > T_{\rm dust}$ (e.g., Hollenbach
242: \& Tielens 1999). Here we must stress that all other LVG solutions
243: compatible with the low (5--4)/(2--1) ratio result to $\rm R_{\rm J+1,
244: J} \la 0.05$ for $\rm J+1\geq 6$.
245: 
246: \section{Consequences for high-redshift galaxies}
247: 
248: At high  redshifts  ($z\ga 3$), the  usual  instruments of  choice for
249: detecting CO lines are millimetre-wave interferometers. These are only
250: capable  of  accessing high-$J$  (i.e.\  $J>2$) CO $J+1 \rightarrow J$
251: lines.  In  such cases, the  obvious consequence of the low excitation
252: of such lines  is that they become  difficult to detect, even if there
253: is a  large quantity of molecular  gas.  Equivalently, any  H$_2$ mass
254: estimates  based on the assumption  of optically thick and thermalized
255: high-$J$ CO lines  (e.g., Evans et al.\  1996; van Ojik et al.\  1997)
256: may underestimate the mass  present, or  its  upper limit, by  a large
257: factor ($\ga 10$).
258: 
259: In the case of HR\,10, scaling its CO $J=5-4$ flux to $z=4$
260: (reasonably typical for the large samples of quasars, radio galaxies,
261: Lyman-break and submm-selected galaxies that are now being assembled
262: -- e.g., Steidel et al.\ 1999; Smail et al.\ 2000) yields a
263: velocity-integrated flux density of $\sim 0.8$\,Jy\,km\,s$^{-1}$ (from
264: Equation~1), received at $\nu_{\rm obs} \sim 115$\,GHz.  This is a
265: rather weak signal. While still detectable in a dedicated observing
266: run of a few tens of hours with a sensitive present-day
267: millimetre-wave interferometer, it would evade detection in a typical
268: survey for such lines in full samples of high-redshift galaxies (e.g.,
269: van Ojik et al.\ 1997). Moreover, an estimate of $M$(H$_2$) based on
270: the CO $J=5-4$ line only, and the assumption that it is optically
271: thick and thermalized, would yield a value $\sim 6-12$ times too low
272: (Equation~3, $R_{54}\sim 0.08-0.16$).  The previous simple excitation
273: analysis suggests that CO $J+1 \rightarrow J$ ( with $J+1 \geq 5$)
274: lines can be poor global mass tracers in such environments,
275: underestimating the H$_2$ mass by factors of $\ga 10$, depending on
276: the particular line used and the prevailing physical conditions.
277: 
278: 
279: For the least excitation-biased mass estimates of metal-enriched
280: molecular gas it is clearly important to observe the lowest two
281: transitions of CO.  At high redshifts this is currently possible only
282: in cases where there is a fortuitous combination of redshift and
283: available receivers.  Sadly, this precludes a systematic inventory of
284: molecular gas mass at $z\ga 3$ with present instruments, though the
285: Green Bank Telescope, the upgraded Very Large Array (VLA) and ALMA
286: will provide the necessary instantaneous bandwidth (velocity
287: coverage), tuning range (redshift coverage) and sensitivity to detect
288: many CO lines at such redshifts.  Fig.~1 demonstrates that several CO
289: line ratios will be available for galaxies at virtually all redshifts,
290: with a small gap at $z\sim 2.7-3.4$ ($z\sim 2.7-4.1$ if one considers
291: only ALMA).  For the CO(1--0) and CO(2--1) lines, however, ALMA
292: provides the only coverage at $z\sim 2.0-2.7$, a potentially important
293: range for a number of known galaxy populations (Steidel et al.\ 1999;
294: Ivison et al.\ 1998, 2000; Smail et al.\ 2000), and to do this ALMA
295: requires the so-called `Band~1' receivers (31.3--45\,GHz).
296: 
297: Currently, the commissioning of a new generation of sensitive
298: receivers with wider tuning ranges on mm/sub-mm telescopes around the
299: world will enable the extension of multi-line studies of nearby
300: starbursts ($\rm L_{\rm FIR}\sim 10^{9}-10^{10} L_{\odot }$) towards
301: more luminous systems ($\rm L_{\rm FIR}\sim 10^{11}-10^{12} L_{\odot
302: }$) at larger distances.  This will reveal whether a low-excitation
303: gas phase dominating the emission of the $ ^{12}$CO rotational lines
304: is indeed more frequently encountered at higher FIR luminosities and
305: will yield a firm assessment of the expected luminosities of
306: higher-$J$ lines in luminous starbursts found at high redshift.
307: 
308: \section{Conclusions}
309: 
310: We  have  presented a  brief  analysis  of the  physical conditions of
311: molecular gas in the $z=1.44$  extremely red object, HR\,10, using the
312: CO $J=2-1,  J=5-4$   lines.  HR\,10 provides   a  rare opportunity  to
313: investigate   low-$J$ {\it  and} high-$J$   CO  lines  in an   intense
314: starburst at a cosmologically significant distance.  Its case can thus
315: be instructive  for similar  environments in  objects at still  higher
316: redshifts.  Our conclusions are as follows:
317: 
318: \begin{itemize}
319: \item
320: Contrary to earlier claims, we find the CO $J=5-4$ line to be of low
321: excitation, with a brightness temperature ratio of $(5-4)/(2-1) \sim
322: 0.16$, likely due to low densities rather than low temperatures.
323: \item
324: The CO $J+1\rightarrow J, J+1 \geq 5$ emission lines from similar
325: galaxies at $z\ga 3$ would be hard to detect using current
326: millimeter-wave arrays.  If detected, the usual assumption made for
327: their excitation, namely that they are thermalized and optically
328: thick, may underestimate H$_2$ gas masses by a factor $\ga 10$.
329: \end{itemize}
330: 
331: This analysis may have some bearing  on the relative priority given to
332: the so-called `Band~1' frequencies/receivers (31.3--45\,GHz) for ALMA.
333: It is at these frequencies that one would detect the lowest-excitation
334: CO $J=1-0$     transition   from  a substantial     fraction    of the
335: submm-selected  galaxy population   (`SCUBA galaxies'),  systems which
336: appear to have a median redshift of 2.5--3  (Smail et al.\ 2000).  The
337: upgraded VLA may also have a significant role to play;  it will have a
338: similar {\em effective} aperture to ALMA at those frequencies, but the
339: expected phase stability at Atacama  and the $2\times$ larger field of
340: view means that ALMA is likely to be the interferometer of choice.
341: 
342: \acknowledgments
343: 
344: We thank the referee Christian Henkel for useful comments that
345: improved the present work and acknowledge support by a Joint Project
346: grant from the Royal Society.
347: 
348: \begin{thebibliography}{}
349: 
350: \bibitem[]{} Aalto, S., Booth, R.\ S., Black, J.\ H., \& Johansson, L.\
351:              E.\ B.\ 1995, A\&A, 300, 369{}{}
352: \bibitem[]{} Andreani, P., Cimatti, A., Loinard, L., \& R\"ottgering, H.\ 2000,
353:              A\&A, 354, L1{}{}
354: \bibitem[]{} Bloemen, J.\ B.\ G.\ M.\ 1985, Ph.\ D.\ thesis, Univ.\ Leiden{}{}
355: \bibitem[]{} Blain, A.\ W.\ 1999, MNRAS, 304, 669{}{}
356: \bibitem[]{} Braine, J., \& Combes, F.\ 1992, A\&A, 264, 433{}{}
357: \bibitem[]{} Bryant, P.\ M., \& Scoville, N.\ Z.\ 1996, ApJ, 457, 678{}{}
358: \bibitem[]{} Cimatti, A., Andreani, P., R\"ottgering, H., \& Tilanus, R.\ 1998,
359:              Nature, 371, 586{}{}
360: \bibitem[]{} Devereux, N., \& Young, J.\ S.\ 1990, ApJ, 359, 42{}{}
361: \bibitem[]{} Dey, A., Graham, J.\ R., Ivison, R.\ J., Smail, I.,
362:              Wright, G.\ S., \& Liu, M.\ 1999, ApJ, 519, 610{}{}
363: \bibitem[]{} Dickman, R.\ L., Snell, R.\ L., \& Schloerb, F.\ P.\
364:              1986, ApJ, 309, 326{}{}
365: \bibitem[]{} Downes, D., \& Solomon, P.\ M.\ 1998, ApJ, 507, 615{}{}
366: \bibitem[]{} Dunlop, J.\ S., Hughes, D.\ H., Rawlings, S., Eales, S.\
367:              A., \& Ward, M.\ J.\ 1994, Nature, 370, 347{}{}
368: \bibitem[]{} Dunlop, J., Peacock, J., Spinrad, H., Dey, A., Jimenez,
369:              R., Stern, D., \& Windhorst, R.\ 1996, Nature, 381, 581{}{}
370: \bibitem[]{} Eales, S.\ A, Lilly, S.\ J., Gear, W.\ K., Dunne, L.,
371:              Bond, J.\ R., Hammer, F., Le F{\`e}vre, O., \& Crampton,
372:              D.\ 1999, ApJ, 515, 518{}{}
373: \bibitem[]{} Elston, R., Rieke, G.\ H., \& Rieke, M.\ J.\
374:              1988, ApJ, 331, L77{}{}
375: \bibitem[]{} Evans, A.\ S., Sanders, D.\ B., Mazzarella, J.\ M.,
376:              Solomon, P.\ M., Downes, D.,  Kramer, C., \& Radford, S.\
377:              J.\ E.\ 1996, ApJ, 457, 658{}{}
378: \bibitem[]{} Frayer, D.\ T., \& Brown, R.\ L.\ 1997, ApJS, 113, 221{}{}
379: \bibitem[]{} Frayer, D.\ T., Ivison, R.\ J., Scoville, N.\ Z., Yun,
380:              M., Evans, A.\ S., Smail, I., Blain, A.\ W., \& Kneib,
381:              J.-P.\ 1998, ApJ, 506, L7{}{}
382: \bibitem[]{} Frayer, D.\ T., Ivison, R.\ J., Scoville, N.\ Z., Evans,
383:              A.\ S., Yun, M., Smail, I., Barger, A.\ J., Blain, A.\
384:              W., \& Kneib, J.-P.\ 1999, ApJ, 514, L13{}{}
385: \bibitem[]{} Graham, J.\ R., \& Dey, A.\ 1996, ApJ, 471, 720{}{}
386: \bibitem[]{} Hasegawa, T.\ 1997, in IAU Sym.\ 170: CO: Twenty-five years of
387:              millimeter-wave spectroscopy, Latter W.\ B., Radford, S.\
388:              J.\ E., Jewell, P.\ R., Mangum, J.\ G., \& Bally, J.\
389:              eds, Kluwer, Dordrecht, p.\ 39{}{}
390: \bibitem[]{} Holland, W.\ S.\ et al.\ 1999, MNRAS, 303, 659{}{}
391: \bibitem[]{} Hollenbach, D.\ J.\ Tielens, A.\ G.\ G.\ M.\ 1999,
392:              Rev.\ Modern Physics, 71, 174{}{}
393: \bibitem[]{} Hu, E.\ M., \& Ridgway, S.\ E.\ 1994, AJ, 107, 1303{}{}
394: \bibitem[]{} Hughes, D.\ H., Gear, W.\ K., \& Robson, E.\ I.\ 1994,
395:              270, 641{}{}
396: \bibitem[]{} Kamp, I, \& van Zadelhoff, G.-J.\ 2001, A\&A 373, 641{}{}
397: \bibitem[]{} Ivison, R.\ J., Papadopoulos, P.\ P., Seaquist, E.\ R.,
398:              \& Eales, S.\ A.\ 1996, MNRAS, 278, 669{}{}
399: \bibitem[]{} Ivison, R.\ J., Smail, I., Le Borgne, J.-F., Blain, A.\
400:              W., Kneib, J.-P., B\'ezecourt, J., Kerr, T.\ H., \& Davies,
401:              J.\ K.\ 1998, MNRAS, 298, 583{}{}
402: \bibitem[]{} Ivison, R.\ J., Smail, I., Barger, A., Kneib, J.-P.,
403:              Blain, A.\ W., Owen, F.\ N., Kerr, T.\ H., \& Cowie, L.\ L.\
404:              2000, MNRAS, 315, 209{}{}
405: \bibitem[]{} Mao, R.\ Q., Henkel, C., Schulz, A., Zielinsky, M.,
406:              Mauersberger, R., Stoerzer, H., Wilson, T.\ L., \&
407:              Gensheimer, P.\ 2000, A\&A, 358, 433{}{}
408: \bibitem[]{} Papadopoulos, P.\ P., \& Seaquist, E.\ R.\ 1998, ApJ, 492, 521{}{}
409: \bibitem[]{} Papadopoulos, P.\ P., R\"ottgering, H.\ J.\ A., van der
410:              Werf, P.\ P., Guilloteau, S., Omont, A., van Breugel,
411:              W.\ J.\ M., \& Tilanus, R.\ P.\ J.\ 2000, ApJ, 528, 626{}{}
412: \bibitem[]{} Perley, R.\ A.\ 2000. `The Very Large Array Observational Status
413:              Summary' (http://www.aoc.nrao.edu/vla/obstatus/vlas/)
414: \bibitem[]{} Richardson, K.\ J.\ 1985, Ph.\ D.\ thesis,
415:              Queen Mary College, Univ.\ of London{}{}
416: \bibitem[]{} Rowan-Robinson, M.\ 2000, MNRAS, 316, 885{}{}
417: \bibitem[]{} Seaquist, E.\ R., \& Frayer, D.\ T.\ 2000, ApJ, 540, 765{}{}
418: \bibitem[]{} Smail, I., Ivison, R.\ J., \& Blain, A.\ W.\ 1997, ApJ,
419:              490, L5{}{}
420: \bibitem[]{} Smail, I., Ivison, R.\ J., Kneib, J.-P., Cowie, L.\ L.,
421:              Blain, A.\ W., Barger, A.\ J., Owen, F.\ N., \& Morrison,
422:              G.\ 1999, MNRAS, 308, 1061{}{}
423: \bibitem[]{} Smail, I., Ivison, R.\ J., Owen, F.\ N., Blain, A.\ W., \&
424:              Kneib, J.-P.\ 2000, ApJ, 528, 612{}{}
425: \bibitem[]{} Solomon, P.\ M., Downes, D., \& Radford, S.\ J.\ E.\
426:              1992, ApJ, 398, L29{}{}
427: \bibitem[]{} Steidel, C.\ C., Adelberger, K.\ L., Dickinson, M.,
428:              Giavalisco, M., \& Pettini, M.\ 1999,  ApJ, 519, 1{}{}
429: \bibitem[]{} van Ojik, R.,  R\"ottgering, H.\ J.\ A., van der Werf,
430:              P.\ P., Miley, G.\ K., Carilli, C.\ L., Visser, A.,
431:              Isaak, K.\ G., Lacy, M., Jenness, T., Sleath, J., \&
432:              Wink, J.\ 1997, A\&A, 321, 389{}{}
433: \bibitem[]{} Wright, E.\ L.\ et al.\ 1991, ApJ, 381, 200{}{}
434: \bibitem[]{} Young, J.\ S., \& Scoville, N.\ Z.\ 1982, ApJ, 258, 467{}{}
435: \bibitem[]{} Young, J.\ S., \& Scoville, N.\ Z.\ 1991, ARA\&A, 29, 581{}{}
436: \end{thebibliography}
437: 
438: \clearpage
439: 
440: %
441: % Figure 1
442: %
443: \begin{figure}
444: \plotone{f1.eps}
445: \caption{The expected velocity-integrated flux density, $\int S_{\nu
446: _{\rm obs}} (\rm J+1, J)\ dV$, ($J$ is labeled, $J=0-6$) emitted by
447: $10^{11}$\,M$_{\odot}$ of metal-enriched H$_2$ gas ($\rm Z/Z_{\odot
448: }=1$) with properties identical to those of HR\,10.  Each line
449: represents the velocity-integrated flux density appropriate for the
450: redshifts where each $J+1 \rightarrow J$ transition is available to
451: the receivers at a particular telescope.  For the VLA, the bands are
452: taken to be the regions where current receivers have at least 50 per
453: cent of their optimum sensitivity (Perley 2000).  Plans for the
454: Extended VLA include continuous frequency coverage from 1.1 to
455: 50\,GHz, which will provide coverage of low-$J$ CO transitions at
456: $z\ge 1.3$.  For ALMA, the definitions of `Bands 1--10' were taken
457: from the ALMA Construction Project Book
458: (www.tuc.nrao.edu/$\sim$demerson/almapbk/construc/cons\_toc.htm), and
459: Band 1 coverage is shown by dashed lines; for the GBT, the bands were
460: taken from the Short Guide to the Green Bank Telescope
461: (www.nrao.edu/GBT/proposals/short\_guide.html).}
462: \end{figure}
463: 
464: \end{document}
465: