Detection of the 158 micron [CII] Transition at z=1.3: Evidence for a Galaxy-Wide Starburst

Abstract

We report the detection of 158 micron [CII] fine-structure line emission from MIPS J142824.0+352619, a hyperluminous (L_IR ~ 10^13 L_sun) starburst galaxy at z=1.3. The line is bright, and corresponds to a fraction L_[CII]/L_FIR = 2 x 10^-3 of the far-IR (FIR) continuum. The [CII], CO, and FIR continuum emission may be modeled as arising from photodissociation regions (PDRs) that have a characteristic gas density of n ~ 10^4.2 cm^-3, and that are illuminated by a far-UV radiation field ~10^3.2 times more intense than the local interstellar radiation field. The mass in these PDRs accounts for approximately half of the molecular gas mass in this galaxy. The L_[CII]/L_FIR ratio is higher than observed in local ULIRGs or in the few high-redshift QSOs detected in [CII], but the L_[CII]/L_FIR and L_CO/L_FIR ratios are similar to the values seen in nearby starburst galaxies. This suggests that MIPS J142824.0+352619 is a scaled-up version of a starburst nucleus, with the burst extended over several kiloparsecs. Comment: 6 pages, 3 figures, accepted by ApJ Letters

Full text

Page 1

arXiv:1003.2174v1 [astro-ph.CO] 10 Mar 2010
Accepted for Publication in ApJ Letters
Preprint typeset using LATEX style emulateapj v. 08/13/06
DETECTION OF THE 158 µm [CII] TRANSITION AT z = 1.3: EVIDENCE FOR A GALAXY-WIDE
STARBURST
S. Hailey-Dunsheath1,2, T. Nikola1, G. J. Stacey1, T. E. Oberst1,3, S. C. Parshley1, D. J. Benford4, J. G.
Staguhn4,5,6, and C. E. Tucker7
Accepted for Publication in ApJ Letters
ABSTRACT
We report the detection of 158 µm [CII] fine-structure line emission from MIPS J142824.0+352619,
a hyperluminous (LIR∼ 1013L⊙) starburst galaxy at z = 1.3. The line is bright, and corresponds to
a fraction L[CII]/LFIR≈ 2×10−3of the far-IR (FIR) continuum. The [CII], CO, and FIR continuum
emission may be modeled as arising from photodissociation regions (PDRs) that have a characteristic
gas density of n ∼ 104.2cm−3, and that are illuminated by a far-UV radiation field ∼103.2times more
intense than the local interstellar radiation field. The mass in these PDRs accounts for approximately
half of the molecular gas mass in this galaxy. The L[CII]/LFIRratio is higher than observed in local
ULIRGs or in the few high-redshift QSOs detected in [CII], but the L[CII]/LFIRand LCO/LFIRratios
are similar to the values seen in nearby starburst galaxies. This suggests that MIPS J142824.0+352619
is a scaled-up version of a starburst nucleus, with the burst extended over several kiloparsecs.
Subject headings: galaxies: individual(MIPS J142824.0+352619)— galaxies: high-redshift — galaxies:
ISM — galaxies: starburst — infrared: galaxies — submillimeter
1. INTRODUCTION
The2P3/2→2P1/2transition of C+(λ = 157.74 µm) is
one of the brightest emission lines in star-forming galax-
ies, typically accounting for 0.1−1% of the far-IR (FIR)
continuum (Stacey et al. 1991; Malhotra et al. 2001).
Much of this emission arises from the warm and dense
photodissociation regions (PDRs) that form on the UV-
illuminated surfaces of molecular clouds. The [CII] tran-
sition is a primary PDR coolant, and is a sensitive probe
of both the physical conditions of the photodissociated
gas, and the intensity of the ambient stellar radiation
field (Hollenbach & Tielens 1999). In ultraluminous in-
frared galaxies (ULIRGs), the L[CII]/LFIRratio is a fac-
tor of ∼7 times lower than in less luminous systems (Luh-
man et al. 2003), for reasons that are not well under-
stood.
Large aperture (sub)millimeter telescopes have been
used to search for [CII] emission from sources at z > 3,
and 3 FIR-luminous QSOs at z = 4.4 − 6.4 have been
detected thus far. SDSS J1148 (Maiolino et al. 2005) and
BR 1202N (Iono et al. 2006c) have low L[CII]/LFIRratios
similar to or smaller than the mean local ULIRG value,
while BRI 0952 (Maiolino et al. 2009) has a somewhat
larger ratio falling at the lower end of the range seen
in normal galaxies. We have initiated a search for [CII]
1Department of Astronomy, Cornell University, Ithaca, NY
14853.
2CurrentAddress: Max-Planck-Institut
restrische Physik, Postfach 1312, D-85741 Garching, Germany;
steve@mpe.mpg.de.
3Current Address: Department of Physics, Westminster Col-
lege, 319 S. Market St., New Wilmington, PA 16172.
4Observational Cosmology Laboratory (Code 665),
Goddard Space Flight Center, Greenbelt, MD 20771.
5Department of Astronomy, University of Maryland, College
Park, MD 20742.
6Current Address: Department of Physics & Astronomy, Johns
Hopkins University, Baltimore, MD 21218.
7School of Physics and Astronomy, Cardiff University, Cardiff
CF24 3AA, UK.
f¨ ur extrater-
NASA
emission from galaxies at z = 1−2, concentrating on star-
formation–dominated systems selected by their FIR or
submillimeter continuum brightness. Here we report our
first detection, from the z = 1.3 hyperluminous starburst
galaxy MIPS J142824.0+352619(hereafter MIPS J1428).
MIPS J1428 was identified in a Spitzer/MIPS blank
field survey as a bright 160 µm source with red
optical/near-IR colors, and subsequent spectroscopic and
photometric measurements established it as a hyperlu-
minous (LIR[8 − 1000 µm] = 3.2 × 1013L⊙) galaxy at
z = 1.325 (Borys et al. 2006). There are several indica-
tions that this IR emission is powered by star formation,
including the brightness of the PAH features and the non-
detection in hard X-ray emission, and with a starburst
origin the estimated LIR corresponds to a star forma-
tion rate of ∼5500 M⊙ yr−1(Borys et al. 2006; Desai
et al. 2006). MIPS J1428 was detected in CO(2→1) and
CO(3→2), and the line luminosities correspond to a large
gas mass of MH2∼ 1011M⊙(Iono et al. 2006b). How-
ever, a gravitationallens may amplify the flux by as much
as a factor of 10 (Borys et al. 2006; Iono et al. 2006a,b),
making the actual star formation rate and gas mass cor-
respondingly lower. Independent of the unknown lensing
amplification, the large value of LIR/L′CO≈ 320 L⊙(K
km s−1)−1(Iono et al. 2006b) indicates that MIPS J1428
is forming stars with high efficiency, similar to that seen
in local ULIRGs and high-redshift submillimeter galaxies
(SMGs) (Solomon & Vanden Bout 2005).
This is the first detection of [CII] in the z = 1−3 epoch
of peak star formation activity in the Universe, and the
first detection from a high-redshift galaxy not associated
with a QSO.
2. OBSERVATIONS
We observed the [CII] line toward MIPS J1428 in
March 2008 with ZEUS (Stacey et al. 2007; Hailey-
Dunsheath 2009) at the Caltech Submillimeter Obser-

Page 2

2 Hailey-Dunsheath et al.
Fig. 1.— ZEUS/CSO spectrum of the 158 µm [CII] transition in
MIPS J1428. The velocity scale is referenced to z = 1.325, and the
pixel centered at v = 1250 km s−1is removed due to the presence of
a telluric absorption feature. The integrated intensity is obtained
by summing the 2 central spectral bins, which cover ∆v = 416 km
s−1. The line is detected at 6σ significance.
vatory (CSO)8on Mauna Kea, Hawaii. ZEUS is a dual-
band (350 and 450 µm) grating spectrometer equipped
with a 1 × 32 semiconductor bolometer array. For this
observing run the instrument was configured as described
in Hailey-Dunsheath et al. (2008).
wavelength (λ = 366.75 µm) of the [CII] transition the
spectrometer provides a slit-limited resolving power of
λ/∆λ ≈ 900, and the 14 detectors operating in the 350
µm window combine to provide an instantaneous band-
width corresponding to δv = 2825 km s−1.
Orion (BN-KL) was used as spectral calibrator, and
Mars (TB = 213 K; Hildebrand et al. 1985) as flux
calibrator. We integrated on MIPS J1428 for a total
of 135 minutes (including chopping) in good submil-
limeter weather, with τ350 µm = 1.0 − 1.5.
ZEUS/CSO beam was centered at R.A. = 14h28m24.s1,
decl. = +35◦26′19′′(J2000.0), and the data were ob-
tained by chopping and nodding the telescope with a
30′′throw. A linear baseline is removed from the final
spectrum, shown in Figure 1. The integrated line flux is
Fν∆v = 713 Jy km s−1, and we compare this with other
observations in Table 1. With LFIR= 2.6×1013L⊙(Bo-
rys et al. 2006), we estimate L[CII]/LFIR≈ 2.1 × 10−3.
At the redshifted
The 13′′
3. RESULTS
3.1. Origins of the [CII] Line
Carbon has a lower ionization potential (11.26 eV)
than hydrogen, and consequently the [CII] transition is
an important coolant of both ionized gas and diffuse
atomic gas. However, most of the [CII] emission from
nearby IR-bright galaxies arises from dense PDRs il-
luminated by the far-UV (FUV) radiation from young
stars (Crawford et al. 1985; Stacey et al. 1991). Com-
bined HII region and PDR modeling of the starburst tem-
plates NGC 253 (Carral et al. 1994) and M82 (Lord et al.
1996; Colbert et al. 1999) has demonstrated that PDRs
account for at least 70% of the [CII] emission in these
sources. The available data suggests that such starburst
8The Caltech Submillimeter Observatory is supported by NSF
contract AST-0229008
TABLE 1
Spectral Line and Continuum Observations of MIPS J1428
TracerFluxLσref
[Jy km s−1]
713
5.3
13.9
...
...
[L⊙] [%]
30
23
35
22
22
[CII]5.4 × 1010
4.9 × 107
1.9 × 108
3.2 × 1013
2.6 × 1013
1
2
2
3
3
CO(2→1)
CO(3→2)
IR (8-1000 µm)
FIRa
Note. — Luminosities are uncorrected for lensing, and are cal-
culated using ΩM = 0.27, ΩΛ= 0.73, H0 = 71 km s−1Mpc−1.
References: (1) this work, (2) Iono et al. (2006b), (3) Borys et al.
(2006).
aWe estimate the rest-frame 60 µm and 100 µm luminosity den-
sities from the Borys et al. (2006) model SED, and calculate LFIR
following FFIR= 1.26 × 10−14(2.58f60+ f100) W m−2(Sanders
& Mirabel 1996).
galaxies are the best local analogs of MIPS J1428 (see
section 4), so here we assume that 70% of the [CII] emis-
sion from MIPS J1428 arises in PDRs. We further as-
sume that PDRs produce all of the CO(2→1), CO(3→2),
and FIR continuum emission in this source, and we model
these tracers as arising from a single representative PDR.
3.2. PDR Analysis
We use the PDR models of Kaufman et al. (1999),
in which the emission is determined by the gas density,
n, and the incident FUV (6 eV < hν < 13.6 eV) flux,
G0 (expressed in units of the Habing Field, 1.6 × 10−3
ergs cm−2s−1). These models calculate the line emission
generated by a cloud illuminated on only one side, and
must be modified to model clouds in the active regions
of galaxies that are illuminated on all sides. For this ge-
ometry, Kaufman et al. (1999) note that an observer will
detect optically thin radiation emitted by both the near
and far sides of the cloud, while optically thick emission
will only be visible from the near side. The [CII] line is
generally moderately optically thin in these calculations,
and the FIR continuum is also assumed to be optically
thin, but the low-J CO transitions are optically thick for
the nominal AV = 10 cloud depth. We therefore increase
the measured CO fluxes by a factor of 2 with respect to
these other tracers before comparing with the models.
In Figure 2 we show the values of n and G0 allowed
by the observations listed in Table 1, after correcting the
[CII] and CO measurements as described above. In the
n ≈ 103−105cm−3range the L[CII]/LFIRratio provides
a strong constraint on G0, while the L[CII]/LCO(2→1)
ratio is primarily a function of density.
solution obtained from these two ratios is consistent
with the n ? 103.5cm−3lower limit imposed by the
LCO(3→2)/LCO(2→1)ratio, and we conclude that the rel-
ative strengths of the [CII], CO(2→1), CO(3→2), and
FIR continuum emission may be reproduced by a PDR
with n ∼ 104.2cm−3and G0∼ 103.2. Reducing the PDR
contribution to the total [CII] emission by a factor of 2
(3) would increase n by a factor of 7.2 (22), but G0by
only a factor of 1.6 (1.8). We also note that without the
factor of 2 correction to the observed CO fluxes, n would
be 4.2 times lower, but G0would remain unchanged.
An estimate of the atomic gas mass associated with
the PDRs in MIPS J1428 is obtained from the inferred
[CII] luminosity and this PDR solution. Assuming the
The best fit

Page 3

[CII] in MIPS J14283
Fig. 2.—
L[CII]/LFIR,L[CII]/LCO(2→1), and
LCO(3→2)/LCO(2→1) as a function of density (n) and FUV
flux (G0), from the PDR models of Kaufman et al. (1999).
Contours are drawn for the measured values (solid) and ±1σ
uncertainties (dot-dashed) of MIPS J1428, after scaling the [CII]
flux by 70% to account for a non-PDR contribution, and increasing
the CO fluxes by a factor of 2 to account for geometrical effects
(see section 3).
[CII] emission is optically thin, this mass is given as
Ma
M⊙
= 0.77
?0.7L[CII]
L⊙
?1 + 2exp(−91K/T)+ ncrit/n
2exp(−91K/T)
??1.4 × 10−4
XC+
?
×
?
, (1)
where we use a C+abundance per hydrogen atom of
XC+ = 1.4 × 10−4(Savage & Sembach 1996). The sur-
face temperature of a PDR with n = 104.2cm−3and
G0 = 103.2is TS ≈ 230 K (Kaufman et al. 1999). As-
suming this temperature is representative of the full C+
region, and adopting a critical density of ncrit= 2.7×103
cm−3(Launay & Roueff 1977), this expression yields
Ma = 5.5 × 1010µ−1M⊙ (where µ is the lensing am-
plification). This is 55% of the molecular gas mass esti-
mated from the CO luminosity (Iono et al. 2006b), which
is similar to the PDR mass fractions in starburst galax-
ies (Stacey et al. 1991, corrected for the smaller XC+
used here). Much like in nearby starburst nuclei, the
PDR mass in MIPS J1428 is a significant fraction of the
total molecular gas mass.
4. DISCUSSION
4.1. [CII], CO, and the FIR Continuum: A
Comparison to Other Galaxies
To place our [CII] measurement of MIPS J1428 in con-
text, we compare the [CII], CO, and FIR continuum
emission from this source with that from other galaxies
and Galactic sources. In Figure 3 we plot the L[CII]/LFIR
and LCO(1→0)/LFIRratios for samples of Galactic star-
forming regions and nearby gas-rich galaxies (Stacey
et al. 1991), normal galaxies (Malhotra et al. 2001), local
ULIRGs (Luhman et al. 2003), and MIPS J1428 and the
3 other high-redshift sources detected in [CII] (Maiolino
et al. 2005; Iono et al. 2006c; Maiolino et al. 2009). We
estimate the CO(1→0) luminosity of MIPS J1428 assum-
ing an L′CO(2→1)/L′CO(1→0)= 0.9 ratio intermediate be-
tween the ratios observed in local ULIRGs (≈0.7; Downes
& Solomon 1998) and in starburst nuclei (≈1.1; Harrison
et al. 1999; Weiß et al. 2005). For reference, we overplot
the PDR model curves from Kaufman et al. (1999). A
vector attached to the MIPS J1428 data point indicates
how the measured ratios were adjusted for the PDR anal-
ysis in section 3.2 (to account for non-PDR contributions
to the [CII] emission, and optical depth effects in the CO
transititions), and we further note that MIPS J1428 is
placed in Figure 3 assuming an L′CO(2→1)/L′CO(1→0)ra-
tio ≈20% lower than our best fit PDR solution. We com-
pare MIPS J1428 with each of the populations shown in
Figure 3 in turn.
Stacey et al. (1991) study the [CII] emission from
the central regions of a sample of nearby galaxies, and
find that the L[CII]/LCO(1→0)and L[CII]/LFIRratios are
sensitive tracers of the star formation activity in these
sources. Starburst galaxies are well characterized by the
same L[CII]/LCO(1→0)= 4100 ratio seen in Galactic star-
forming regions, which is a factor of ∼3 times larger than
in non-starburst galaxies and Galactic molecular clouds
(Fig. 3). However, the L[CII]/LFIR ratios in starburst
galaxies are higher than those observed in the most in-
tense Galactic OB star formation regions. These results
imply that while much of the molecular gas in a starburst
nucleus is photodissociated by the FUV radiation from
young stars, the very intense fields found in the most
extreme Galactic star-forming regions are not providing
the bulk of the luminosity. This picture is supported
by PDR modeling, which indicates that the FUV fields
producing the emission in the centers of starburst galax-
ies are intermediate in intensity between those producing
the emission in non-starburst galaxies, and those found
in Galactic OB star formation regions (Fig. 3). As the
data point for MIPS J1428 falls near the high-G0end of
the distribution of starburst points in Figure 3, we con-
clude that in MIPS J1428, like in nearby starburst nuclei,
the bulk of the molecular gas is exposed to moderately-
intense FUV radiation produced by young stars.
Malhotra et al. (2001) observe [CII] and other FIR fine-
structure line emission from a sample of 60 normal galax-
ies with varying levels of star formation activity, which
are all sufficiently distant that the FIR emission is con-
tained within the ≈70′′ISO-LWS beam. These authors
find that the FIR continuum and much of the [CII] emis-
sion arises from PDRs, which in many of their galaxies
are characterized by similarly elevated values of G0as de-
duced here for MIPS J1428. However, the Malhotra et al.
(2001) sources have lower L[CII]/LCOratios than MIPS
J1428 or its starburst analogs, indicating that most of the
CO emission is produced by molecular gas residing in less
active star-forming regions than those that produce the
fine-structure line emission. This difference helps rein-
force our conclusion that in MIPS J1428 it is the entire
galaxy, rather than just an active central region, that is
host to vigorous star formation.
The median L[CII]/LFIRratio in local ULIRGs is a fac-
tor of ∼7 times lower than in normal star-forming galax-
ies (Luhman et al. 2003), and a factor of ∼4 times lower
than in MIPS J1428. One possible explanation for this
low ratio is that the [CII] emission is produced in dense
PDRs illuminated by intense FUV radiation (e.g., Pa-
padopoulos et al. 2007), as indicated by the PDR model

Page 4

4Hailey-Dunsheath et al.
Fig. 3.— L[CII]/LFIRvs. LCO(1→0)/LFIRfor Galactic star-forming regions (crosses), starburst nuclei (filled squares), non-starburst
nuclei (open squares), normal galaxies (triangles), local ULIRGs (circles), and high-redshift sources (asterisks). CO(1→0) luminosities of
the normal galaxies and local ULIRGs are taken from the literature and from J. Graci´ a-Carpio et al. (2010, in preparation). CO(1→0)
luminosities for MIPS J1428 and 2 of the other high-redshift sources are estimated from the measurements of higher-J lines as described
in the text. Starburst galaxies and Galactic star-forming regions are characterized by L[CII]/LCO(1→0)≈ 4100 (dashed). Overplotted are
the PDR model calculations for gas density (n) and FUV field strength (G0) from Kaufman et al. (1999), and a vector indicates how the
MIPS J1428 data point was shifted for the PDR analysis in section 3.2.
overlays in Figure 3. Alternatively, the [CII] emission
may originate in PDRs with more modest values of n and
G0, but an additional source of FIR continuum emission
may lower the global L[CII]/LFIR ratio (Luhman et al.
2003; Abel et al. 2009). Regardless of the origins of the
weak [CII] emission in local ULIRGs, it is clear that these
galaxies provide poor templates for MIPS J1428.
We also include the 3 high-redshift QSOs detected
in [CII] in Figure 3.SDSS J1148 has been detected
in CO(3→2) and BRI 0952 in CO(5→4), and we use
the measured L′CO(3→2)/L′CO(1→0)= 0.72 and modeled
L′CO(5→4)/L′CO(1→0) = 0.58 ratios for Mrk 231 (Pa-
padopoulos et al. 2007) to estimate the CO(1→0) lumi-
nosities for these sources. To account for the uncertain-
ties in the CO excitation of high-redshift systems (e.g.,
Hainline et al. 2006), we show these data points with
errors bars corresponding to 0.2 dex and 0.3 dex, re-
spectively. The low L[CII]/LFIRand LCO/LFIRratios in
SDSS J1148 and BR 1202N indicate similar physical con-
ditions as in local ULIRGs. Indeed, Maiolino et al. (2005)
found that the [CII], CO, and FIR continuum emission
of SDSS J1148 is consistent with a high-density, high-
G0PDR, and Walter et al. (2009) showed that the star
formation in this system is confined to a compact re-
gion with a high surface brightness comparable to that
at the center of Arp 220. However, BRI 0952 has a larger
L[CII]/LFIRratio and a L[CII]/LCOratio similar to that
in MIPS J1428. We suggest that while SDSS J1148 and
BR 1202N are likely dissimilar to MIPS J1428, the con-
nection drawn here between MIPS J1428 and local star-
burst nuclei may also provide insight into the nature of
the host galaxy of BRI 0952.
Finally, it is important to note that while we assume
that the L[CII]/LFIRand LCO/LFIRratios in MIPS J1428
are unaffected by the potential gravitational lensing, we
cannot rule out the possibility that these ratios are al-
tered by differential magnification.
could construct a model in which the intrinsic emis-
sion ratios of MIPS J1428 are similar to those in local
ULIRGs, but where the L[CII]/LFIRratio is enhanced by
the relatively large magnification of a localized HII re-
gion that falls near a caustic. However, in the absence
of other evidence supporting such a scenario, we assume
that the measured flux ratios are equal to the intrinsic
luminosity ratios.
For example, one
4.2. MIPS J1428 as an Extended Starburst
The molecular gas in MIPS J1428 is exposed to similar
UV radiation fields, and has similar excitation, as the gas
in the central region of a typical starburst galaxy, and we

Page 5

[CII] in MIPS J14285
suggest that MIPS J1428 may be modeled as a scaled-
up version of a starburst nucleus. As an example, we
consider scaling up the central region of the prototypical
starburst galaxy M82 in such a manner as to match the
much larger luminosity of MIPS J1428, while at the same
time conserving the strength of the characteristic FUV
radiation field that controls the PDR emission. We as-
sume that in a starburst environment this characteristic
field is determined by the global star formation density,
rather than by the purely local properties of the inter-
actions between individual star clusters and their natal
molecular clouds. This interpretation was adopted for
M82 and NGC 253 based on multiple-line PDR analy-
sis (Wolfire et al. 1990; Carral et al. 1994; Lord et al.
1996), and is also motivated by the non-linear relation
between I[CII] and IFIR observed in larger samples of
galactic nuclei (Crawford et al. 1985; Stacey et al. 1991).
For a starburst region in which young stellar clusters
and molecular clouds are randomly distributed, Wolfire
et al. (1990) modeled the relationship between the aver-
age FUV flux incident on the molecular gas (G0), and
the size (D) and total luminosity (LIR) of the region.
This relationship depends on the mean free path of a
FUV photon (λ), but in the limits of λ ≪ D and λ ≫ D
these models give G0∝ LIR/D3and G0∝ LIR/D2, re-
spectively. Adopting LIR∼ 3 × 1010L⊙for the central
D ∼ 300 pc starburst region of M82 (Telesco & Harper
1980; Joy et al. 1987), and assuming that MIPS J1428
and M82 have the same average G0(Fig. 3), the factor of
∼1000µ−1greater luminosity of MIPS J1428 then trans-
lates into a size scale of D ≈ 3µ−1/3− 10µ−1/2kpc. In
short, if the large luminosity of MIPS J1428 is produced
by only moderate-intensity radiation, the star formation
generating this radiation must be extended over a large
area.
While MIPS J1428 remains unresolved in dust con-
tinuum and molecular gas emission, integral field spec-
troscopy shows Hα emission extended over ≈0.7′′(≈6
kpc) (Swinbank et al. 2006). This result is consistent
with our interpretation of MIPS J1428 as a source pow-
ered by extended star formation, although the degree to
which the Hα image is distorted by the foreground lens,
or by the presence of large and spatially-varying extinc-
tion, is unknown. Further interpretation of the connec-
tion between the bright [CII] emission in MIPS J1428 and
the potentially large extent of the star formation will ben-
efit from a size measurement in an extinction-free tracer,
and better constraints on the lensing magnification.
If MIPS J1428 is indeed powered by a starburst ex-
tending over several kiloparsecs this would be in sharp
contrast to local ULIRGs, in which most of the emission
traced by CO interferometry is contained within the cen-
tral ∼1 kpc (Downes & Solomon 1998). However, there
is ample evidence that the most luminous star-forming
galaxies at higher redshifts are more extended. The best
studied sample of such sources are the SMGs, which like
MIPS J1428 are submillimeter-bright galaxies with lumi-
nosities of LIR ∼ 1013L⊙ that are generated predomi-
nantly by star formation (Alexander et al. 2005). Inter-
ferometric imaging of the radio continuum of z = 1 − 3
SMGs has shown that many of these sources have de-
tectable emission extended over ∼10 kpc (Chapman et al.
2004), and that the population has a median FWHM size
of ∼5 kpc (Biggs & Ivison 2008). Similarly, many of the
SMGs studied by Tacconi et al. (2006, 2008) have CO
FWHM sizes of ∼4 kpc. The model of extended star
formation we suggest here for MIPS J1428 is therefore
consistent with the large sizes of SMGs, which indicate
that the most luminous star-forming galaxies at z = 1−3
are powered by starbursts extending over much larger re-
gions than in local systems.
We thank Javier Graci´ a-Carpio for kindly providing
unpublished CO data, and the CSO staff for their sup-
port of ZEUS operations. We also thank an anonymous
referee for many helpful comments on an earlier draft
of this manuscript. This work was supported by NSF
grants AST-0096881, AST-0352855, AST-0705256, and
AST-0722220, and by NASA grants NGT5-50470 and
NNG05GK70H.
REFERENCES
Abel, N. P., Dudley, C., Fischer, J., Satyapal, S., & van Hoof,
P. A. M. 2009, ApJ, 701, 1147
Alexander, D. M., Bauer, F. E., Chapman, S. C., Smail, I., Blain,
A. W., Brandt, W. N., & Ivison, R. J. 2005, ApJ, 632, 736
Biggs, A. D., & Ivison, R. J. 2008, MNRAS, 385, 893
Borys, C., Blain, A. W., Dey, A., Le Floc’h, E., Jannuzi, B. T.,
Barnard, V., Bian, C., Brodwin, M., Men´ endez-Delmestre, K.,
Thompson, D., Brand, K., Brown, M. J. I., Dowell, C. D.,
Eisenhardt, P., Farrah, D., Frayer, D. T., Higdon, J., Higdon,
S., Phillips, T., Soifer, B. T., Stern, D., & Weedman, D. 2006,
ApJ, 636, 134
Carral, P., Hollenbach, D. J., Lord, S. D., Colgan, S. W. J., Haas,
M. R., Rubin, R. H., & Erickson, E. F. 1994, ApJ, 423, 223
Chapman, S. C., Smail, I., Windhorst, R., Muxlow, T., & Ivison,
R. J. 2004, ApJ, 611, 732
Colbert, J. W., Malkan, M. A., Clegg, P. E., Cox, P., Fischer, J.,
Lord, S. D., Luhman, M., Satyapal, S., Smith, H. A., Spinoglio,
L., Stacey, G., & Unger, S. J. 1999, ApJ, 511, 721
Crawford, M. K., Genzel, R., Townes, C. H., & Watson, D. M.
1985, ApJ, 291, 755
Desai, V., Armus, L., Soifer, B. T., Weedman, D. W., Higdon, S.,
Bian, C., Borys, C., Spoon, H. W. W., Charmandaris, V., Brand,
K., Brown, M. J. I., Dey, A., Higdon, J., Houck, J., Jannuzi,
B. T., Le Floc’h, E., Ashby, M. L. N., & Smith, H. A. 2006,
ApJ, 641, 133
Downes, D., & Solomon, P. M. 1998, ApJ, 507, 615
Hailey-Dunsheath, S. 2009, PhD thesis, Cornell University
Hailey-Dunsheath, S., Nikola, T., Stacey, G. J., Oberst, T. E.,
Parshley, S. C., Bradford, C. M., Ade, P. A. R., & Tucker, C. E.
2008, ApJ, 689, L109
Hainline, L. J., Blain, A. W., Greve, T. R., Chapman, S. C., Smail,
I., & Ivison, R. J. 2006, ApJ, 650, 614
Harrison, A., Henkel, C., & Russell, A. 1999, MNRAS, 303, 157
Hildebrand, R. H., Loewenstein, R. F., Harper, D. A., Orton, G. S.,
Keene, J., & Whitcomb, S. E. 1985, Icarus, 64, 64
Hollenbach, D. J., & Tielens, A. G. G. M. 1999, Reviews of Modern
Physics, 71, 173
Iono, D., Peck, A. B., Pope, A., Borys, C., Scott, D., Wilner, D. J.,
Gurwell, M., Ho, P. T. P., Yun, M. S., Matsushita, S., Petitpas,
G. R., Dunlop, J. S., Elvis, M., Blain, A., & Le Floc’h, E. 2006a,
ApJ, 640, L1
Iono, D., Tamura, Y., Nakanishi, K., Kawabe, R., Kohno, K.,
Okuda, T., Yamada, K., Hatsukade, B., & Sameshima, M. 2006b,
PASJ, 58, 957
Iono, D., Yun, M. S., Elvis, M., Peck, A. B., Ho, P. T. P., Wilner,
D. J., Hunter, T. R., Matsushita, S., & Muller, S. 2006c, ApJ,
645, L97
Joy, M., Lester, D. F., & Harvey, P. M. 1987, ApJ, 319, 314
Kaufman, M. J., Wolfire, M. G., Hollenbach, D. J., & Luhman,
M. L. 1999, ApJ, 527, 795

Page 6

6Hailey-Dunsheath et al.
Launay, J., & Roueff, E. 1977, Journal of Physics B Atomic
Molecular Physics, 10, 879
Lord, S. D., Hollenbach, D. J., Haas, M. R., Rubin, R. H., Colgan,
S. W. J., & Erickson, E. F. 1996, ApJ, 465, 703
Luhman, M. L., Satyapal, S., Fischer, J., Wolfire, M. G., Sturm,
E., Dudley, C. C., Lutz, D., & Genzel, R. 2003, ApJ, 594, 758
Maiolino, R., Caselli, P., Nagao, T., Walmsley, M., De Breuck, C.,
& Meneghetti, M. 2009, A&A, 500, L1
Maiolino, R., Cox, P., Caselli, P., Beelen, A., Bertoldi, F., Carilli,
C. L., Kaufman, M. J., Menten, K. M., Nagao, T., Omont, A.,
Weiß, A., Walmsley, C. M., & Walter, F. 2005, A&A, 440, L51
Malhotra, S., Kaufman, M. J., Hollenbach, D., Helou, G., Rubin,
R. H., Brauher, J., Dale, D., Lu, N. Y., Lord, S., Stacey, G.,
Contursi, A., Hunter, D. A., & Dinerstein, H. 2001, ApJ, 561,
766
Papadopoulos, P. P., Isaak, K. G., & van der Werf, P. P. 2007, ApJ,
668, 815
Sanders, D. B., & Mirabel, I. F. 1996, ARA&A, 34, 749
Savage, B. D., & Sembach, K. R. 1996, ARA&A, 34, 279
Solomon, P. M., & Vanden Bout, P. A. 2005, ARA&A, 43, 677
Stacey, G. J., Geis, N., Genzel, R., Lugten, J. B., Poglitsch, A.,
Sternberg, A., & Townes, C. H. 1991, ApJ, 373, 423
Stacey, G. J., Hailey-Dunsheath, S., Nikola, T., Oberst, T. E.,
Parshley, S. C., Benford, D. J., Staguhn, J. G., Moseley,
S. H., & Tucker, C. 2007, in Astronomical Society of the
Pacific Conference Series, Vol. 375, From Z-Machines to ALMA:
(Sub)Millimeter Spectroscopy of Galaxies, ed. A. J. Baker,
J. Glenn, A. I. Harris, J. G. Mangum, & M. S. Yun , 52–+
Swinbank, A. M., Chapman, S. C., Smail, I., Lindner, C., Borys,
C., Blain, A. W., Ivison, R. J., & Lewis, G. F. 2006, MNRAS,
371, 465
Tacconi, L. J., Genzel, R., Smail, I., Neri, R., Chapman, S. C.,
Ivison, R. J., Blain, A., Cox, P., Omont, A., Bertoldi, F., Greve,
T., F¨ orster Schreiber, N. M., Genel, S., Lutz, D., Swinbank,
A. M., Shapley, A. E., Erb, D. K., Cimatti, A., Daddi, E., &
Baker, A. J. 2008, ApJ, 680, 246
Tacconi, L. J., Neri, R., Chapman, S. C., Genzel, R., Smail, I.,
Ivison, R. J., Bertoldi, F., Blain, A., Cox, P., Greve, T., &
Omont, A. 2006, ApJ, 640, 228
Telesco, C. M., & Harper, D. A. 1980, ApJ, 235, 392
Walter, F., Riechers, D., Cox, P., Neri, R., Carilli, C., Bertoldi, F.,
Weiss, A., & Maiolino, R. 2009, Nature, 457, 699
Weiß, A., Walter, F., & Scoville, N. Z. 2005, A&A, 438, 533
Wolfire, M. G., Tielens, A. G. G. M., & Hollenbach, D. 1990, ApJ,
358, 116