() - 1306.3572v2.pdf

arXiv:1306.3572v2 [astro-ph.CO] 24 Aug 2013

Accepted for Publication in The Astrophysical Journal

Preprint typeset using L

X style emulateapj v. 03/07/07

AN INTENSELY STAR-FORMING GALAXY AT

∼

7 WITH LOW DUST AND METAL CONTENT

REVEALED BY DEEP ALMA AND

HST

OBSERVATIONS

Masami Ouchi

1,2

, Richard Ellis

, Yoshiaki Ono

, Kouichiro Nakanishi

4,5

, Kotaro Kohno

6,7

Rieko Momose

, Yasutaka Kurono

, M. L. N. Ashby

, Kazuhiro Shimasaku

7,9

S. P. Willner

, G. G. Fazio

, Yoichi Tamura

, and Daisuke Iono

Accepted for Publication in The Astrophysical Journal

ABSTRACT

We report deep ALMA observations complemented with associated

HST

imaging for a luminous

(

= 25) galaxy, ‘Himiko’, at a redshift z=6.595. The galaxy is remarkable f

or its high star

formation rate, 100

⊙

−

, securely estimated from our deep

HST

and

Spitzer

photometry, and

the absence of any evidence for strong AGN activity or gravitation

al lensing magnification. Our

ALMA observations probe an order of magnitude deeper than prev

ious IRAM observations, yet fail

to detect a 1.2mm dust continuum, indicating a flux

Jy comparable with or weaker than that

of local dwarf irregulars with much lower star formation rates. We lik

ewise provide a strong upper

limit for the flux of

[Cii]

158

[CII]

⊙

, a diagnostic of the hot interstellar gas often

described as a valuable probe for early galaxies. In fact, our obser

vations indicate Himiko lies off the

local

[CII]

- star formation rate scaling relation by a factor of more than 30. B

oth aspects of our

ALMA observations suggest Himiko is an unique object with a very low d

ust content and perhaps

nearly primordial interstellar gas. Our

HST

images provide unique insight into the morphology of this

remarkable source, highlighting an extremely blue core of activity an

d two less extreme associated

clumps. Himiko is undergoing a triple major merger event whose exten

sive ionized nebula of Lyman

alpha emitting gas, discovered in our earlier work with Subaru, is powe

red by star formation and the

dense circum-galactic gas. We are likely witnessing an early massive ga

laxy during a key period of its

mass assembly close to the end of the reionization era.

Subject headings:

galaxies: formation — galaxies: high-redshift — cosmology: observa

tions

INTRODUCTION

Much progress has been achieved in recent years in

charting the abundance and integrated properties of

the earliest galaxies beyond a redshift of

≃

6 se-

lected via optical and near-infrared (NIR) photome-

try (e.g. Bouwens et al. 2010a; McLure et al. 2010;

Castellano et al. 2010; Ouchi et al. 2010; Ellis et al.

2013; McLure et al. 2012; Schenker et al. 2012). The

emerging picture indicates that the redshift period

10 was a formative one in the assem-

bly history of normal galaxies. Sources at

≃

−

8 show moderately blue ultraviolet continua possibly

Institute for Cosmic Ray Research, The University of

Tokyo, 5-1-5 Kashiwanoha, Kashiwa, Chiba 277-8582, Japan;

ouchims@icrr.u-tokyo.ac.jp

Kavli Institute for the Physics and Mathematics of the Unive

rse

(WPI), The University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa

Chiba 277-8583, Japan

Department of Astrophysics, California Institute of Techn

ol-

ogy, MS 249-17, Pasadena, CA 91125, USA

The Graduate University for Advanced Studies (SOKENDAI),

2-21-1 Osawa, Mitaka, Tokyo 181-8588 Japan

Joint ALMA Observatory, Alonso de Cordova 3107, Vitacura,

Santiago 763-0355, Chile

Institute of Astronomy, University of Tokyo, 2-21-1 Osawa,

Mi-

taka, Tokyo 181-0015, Japan

Research Center for the Early Universe (WPI), University of

Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-0033, Japan

Harvard-Smithsonian Center for Astrophysics, 60 Garden St

Cambridge, MA 02138, USA

Department of Astronomy, Graduate School of Science, The

University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-003

Japan

National Astronomical Observatory of Japan, 2-21-1 Osawa,

Mitaka, Tokyo 181-8588, Japan

consistent with young, metal-poor stellar populations

with a star-formation rate (SFR) of 1

−

⊙

−

(e.g. Bouwens et al. 2010b; Finkelstein et al. 2010;

Schaerer & de Barros 2010; Dunlop et al. 2012). Their

small physical sizes (

≃

7 kpc; Oesch et al. 2010;

Ono et al. 2013) and modest stellar masses (10

−

⊙

; Labb ́e et al. 2010) suggest they quickly merge

into larger, more luminous systems. The abundance of

sub-luminous, small galaxies at high redshift also indi-

cates significant merging occurred at early times, given

the faint-end slope of the UV luminosity function changes

from a steep

≃ −

9 at

= 7

−

8 (Schenker et al.

2012; McLure et al. 2012) to

≃ −

7 at

= 2

−

3 (e.g.

Reddy & Steidel 2009).

In practice it is hard to decipher the physical pro-

cesses that govern the early assembly of galaxies from

integrated properties alone. We therefore seek to com-

plement statistical measures such as star formation

rates and stellar masses by detailed evidence from well-

studied individual examples. Likewise, our understand-

ing of early cosmic history may be incomplete given

so much is currently deduced from optical and near-

infrared data alone(Robertson et al. 2013). Although op-

tical and near-infrared selected sources at high redshift

suggest they contain little or no dust (Bouwens et al.

2012; Dunlop et al. 2013), this may be a selection bias.

Star formation obscured by dust cannot be quantified

without identifying cold dust emission. Furthermore, the

gas phase metallicity remains a key measurement for un-

derstanding early systems, most notably in locating the

highly-prized pristine ‘first generation’ systems unpol-

luted by supernova enrichment. Neither optical nor near-

Ouchi et al.

infrared facilities can currently address this important

quest given the diagnostic metal lines used at lower red-

shift, such as

[Oii]

λλ

3726,3729

A and

[Oiii]

λλ

5007,4959

A , cannot be measured beyond

≃

5 until the launch of

the James Web Space Telescope.

It is for this reason that state of the art sub-millimeter

facilities such as the Atacama Large Millimeter Array

(ALMA) offer enormous promise. First, they can quan-

tify the possible bias in our current ”optical” view of

early galaxy formation by detecting the hidden cold dust

in high redshift galaxies. Secondly, the CO/

[Cii]

158

features prominent in star forming regions in the local

Universe offer a valuable tracer of metallicity at early

times. Thus far, neither cold dust continuum nor these

low-ionization tracers of metallicity have been observed

beyond

∼

6 (Vieira et al. 2013, Capak et al. 2011,

Riechers et al. 2010, Coppin et al. 2010). Although a

few QSOs have been observed at sub-mm wavelengths

= 6

−

1 (Maiolino et al. 2005; Iono et al. 2006;

Walter et al. 2009; Venemans et al. 2012; Willott et al.

2013; Wang et al. 2013), the presence of a powerful AGN

undoubtedly complicates any understanding of the phys-

ical conditions in their host galaxies.

Detecting these important diagnostic signals of dust

and metallicity from typical

≃

7 galaxies is clearly

a major observational challenge. Only upper limits

[Cii]

and submm continuum fluxes have been pre-

sented so far for the abundant population of Lyman

break galaxies (LBGs) and Ly

emitters (LAEs) at

∼

7. These limits have come from deep exposures

with the Submillimetre Common-User Bolometer Array

(SCUBA;Holland et al. 1999) facility on the James Clerk

Maxwell Telescope and Plateau de Bure interferometric

observations (e.g. Ouchi et al. 2009a; Walter et al. 2012;

Kanekar et al. 2013). Very recently, one

= 6

34 source

has been studied in this way following a comprehensive

search for red objects in the Herschel HerMES blank field

survey at 50

−

500

m Riechers et al. (2013). This source,

HFLS3, has a very strong far-infrared continuum emis-

sion and prominent molecular/low-ionization lines. Its

star formation rate, inferred from its far-infrared lumi-

nosity, is extremely high, 2900

⊙

−

. Clearly we need

to understand the context of this remarkable object by

observing other sources at a similar redshift.

The present work is concerned with undertaking such

a study for an extraordinarily luminous star-forming

galaxy which will hopefully complement the study of

HFLS3 by Riechers et al. (2013). Ouchi et al. (2009a)

reported the discovery of the star-forming galaxy at

= 6

595, ’Himiko’

, with a

Spitzer

/IRAC counter-

part. This source was identified from an extensive 1 deg

optical survey for

= 6

6 galaxies in the UKIDSS/UDS

field conducted with the Subaru telescope. The red-

shift was subsequently confirmed spectroscopically using

Keck/DEIMOS. The unique features of this remarkable

source are evident in comparison to the total sample of

207 galaxies at

= 6

6 found in the panoramic Subaru

survey. Not only is Himiko by far the most luminous ex-

ample (

= 25;

(Ly

) = 4

erg s

−

), but it is

spatially extended in Ly

emission whose largest isopho-

tal area is 5.22 arcsec

, corresponding to a linear extent

of over 17 kpc. The lower limit,

SF R >

⊙

−

, is

See Ouchi et al. (2009a) for the meaning of this name.

placed on the SFR of Himiko by the spectral energy dis-

tribution (SED) fitting analysis with the early photomet-

ric measurements and the stellar-synthesis and nebular-

emission models (Ouchi et al. 2009a). Due to the large

uncertainties of photometric measurements, Ouchi et al.

(2009a) cannot constrain

(

−

), and provide only the

lower limit of SFR with

(

−

)

≥

The present paper is concerned with the analysis of

uniquely deep ALMA and

HST

observations of this re-

markable source. Given its intense luminosity and high

star formation rate, the presumption is that it is being

observed at a special time in its assembly history. We

seek to use the cold dust continuum and

[Cii]

measures

from ALMA to understand its dust content and gas phase

metallicity. Likewise the matched resolution of

HST

will

allow us to address its morphologic nature. By good for-

tune, one of the

HST

intermediate band filters closely

matches the intense Lyman

emission observed for this

source with Subaru. Ultimately, we then seek to un-

derstand the physical source of energy that powers the

extensive Lyman

nebula.

A plan of the paper follows. We describe our ALMA

and

HST

observations in

2, and present the detailed

properties such as dust-continuum and metal-line emis-

sion, morphology, and stellar population in

3. We dis-

cuss the nature of this object in

4, and summarize

our findings in

5. Throughout this paper, magnitudes

are in the AB system. We adopt (

Ω

, n

, σ

) =

8).

OBSERVATIONS AND MEASUREMENTS

2.1.

ALMA

To understand whether obscured star-formation is an

important issue as well as the metallicity of Himiko, a

key source at high redshift, we carried out deep ALMA

Band 6 observations in 2012 July 15, 18, 28, and 31 with

16 12m-antenna array under the the extended configura-

tion of 36-400m baseline. The precipitable water vapor

(PWV) ranged from 0.7 to 1.6 mm during the obser-

vations. We targeted Himiko’s

[Cii]

line of rest-frame

1900.54 GHz (157.74

m) which is redshifted to 250.24

GHz (1.198mm) at a redshift of

= 6

595. Be-

cause a brighter dust continuum is expected at a higher

frequency in the 1.2mm regime, we extended our up-

per sideband (USB) to the high-frequency side. Thus,

we targeted the

[Cii]

line with the lowest spectral win-

dow (among 4 spectral windows) in the lower sideband

(LSB) and set the central frequency of the 4 spectral

windows are 250.24 and 252.11 GHz in LSB, and 265.90

and 267.78 GHz in USB with a bandwidth of 1875 MHz.

The two spectral windows and each sideband cover the

frequency ranges contiguously. The total on-source in-

tegration time was 3.17 hours. We used 3c454.3 and

J0423-013 for bandpass calibrators and J0217+017 for a

phase calibrator. The absolute flux scale was established

by observations of Neptune and Callisto. Our data were

reduced with Common Astronomy Software Applications

(CASA) package. We rebin our data to a resolution of

166 MHz (200 km

−

). The FWHM beam size of the final

image is 0

′′

58 with a position angle of 79

◦

5. The

noise of continuum image is

cont

= 17

Jy beam

−

over the the total bandwidth of 19

417 GHz whose 7

GHz is sampled. The 1

noise of

[Cii]

line image is

line

= 83

Jy beam

−

at 250.239 GHz over a channel

Star-Forming Galaxy Unveiled with ALMA and HST

TABLE 1

ALMA Observations and Sensitivities

cont

line

cont

line

cont

line

FIR

[CII]

(GHz) (GHz) (

Jy beam

−

) (

Jy beam

−

) (

Jy) (

Jy) (10

⊙

) (10

⊙

)

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

259.007 250.239

17.4

83.3

250

Note

. — (1)-(2): Central frequencies of continuum and

[Cii]

line observations that correspond to

16 and 1

20 mm, respectively. (3)-(4): 1

sensitivities for continuum and

[Cii]

line in a unit of

beam

−

. The continuum sensitivity is given in the total bandwidth f

or the continuum measurement

is 19

417 GHz or 86

894

m that is a sum of 4 spectral windows (see text). The line sensi

tivity is

defined with a channel width of 200 km

−

. (5)-(6): 3

upper limits of continuum and

[Cii]

line in

a unit of

Jy. (7)-(8): 3

upper limits of far-infrared continuum luminosity (8

−

1000

m) and

[Cii]

line luminosity in a unit of 10

and 10

solar luminosities, respectively. We estimate 3

upper limits

of far-infrared continuum luminosities at 40

−

500

m and 42

−

122

m to be

and

⊙

, respectively. These far-infrared luminosities are estim

ated with the assumptions of

the graybody,

= 1

5, and the dust temperature of

= 40K.

width of 200 km

−

Further details of the ALMA observations and sensi-

tivities are summarized in Table 1.

We averaged fluxes over the two spectral windows of

LSB (249.30-253.05 GHz or 1.203-1.185mm) and USB

(264.96-268.71 GHz or 1.131-1.116mm) in the range of

frequency free from the

[Cii]

line. Figure 1 presents

the resulting ALMA continuum data at 259

01 GHz in

frequency (or 1.167mm in wavelength) with a 1

sensi-

tivity of 17

Jy beam

−

. There is a

∼

flux peak

in the beam size on the position of Himiko. However,

there are a series of negative pixels nearby that corre-

spond to the 2

−

level per beam. We conclude there-

fore that Himiko remains undetected in the 1.2mm con-

tinuum with a 3

upper limit is

Jy beam

−

We note that this sensitivity is two and one order(s)

of magnitudes better than those previously obtained by

deep SCUBA/SHADES and IRAM/PdBI observations

(Ouchi et al. 2009a; Walter et al. 2012). This clearly in-

dicates Himiko has very weak millimeter emission. Table

1 summarizes the flux upper limits for the continuum and

[Cii]

line derived from our ALMA data.

Figure 2 shows the ALMA

[Cii]

velocity channel maps

for Himiko. We searched for a signal of

[Cii]

over 600 km

−

around the frequency corresponding to

= 6

595.

In Figure 2, there are noise peaks barely reaching at the

level in 0 km s

−

and

−

200 km s

−

that are slightly

north and south of the Himiko’s optical center position,

respectively. However, neither of these sources is a reli-

able counterpart of Himiko. Although the ALMA data

reach a 1

noise of 83

Jy beam

−

, no

[Cii]

line is de-

tected. The corresponding 3

upper limit for

[Cii]

250

Jy in a channel width of 200 km s

−

. The associ-

ated luminosity limit is

⊙

2.2.

HST

The primary goal of the associated

HST

observations

of Himiko relate to the morphological nature of this re-

markable source. We carried out deep

HST

/WFC3-IR

broad-band (

125

and

160

)

thus maximizing the in-

formation content on its stellar content for SED fitting

(

3.3). The intermediate-band filter of

098

fortu-

itously includes the spectroscopically-confirmed Ly

line

125

and

160

are referred to as

125

and

160

, respec-

tively. and medium-band (

098

) observations for Himiko. The

two broad-band filters of

125

and

160

measure the rest-frame

UV continuum fluxes, and are free from contamination from Ly

emission,

Fig. 1.—

ALMA continuum data for Himiko at 259 GHz

(1.16mm). The gray scale indicates the intensity at each pos

ition

where darker regions imply higher intensities. The black co

ntours

denote

−

2, and

−

levels, while yellow contours show +1,

+2, and +3

significance levels, where the 1

flux corresponds to

Jy beam

−

. The white cross indicates the position of Himiko.

The ellipse in the lower corner denotes the beam size.

Fig. 2.—

As Figure 1, but for

[Cii]

velocity channel maps of

Himiko whose 1

intensity is 83

Jy beam

−

. The six panels

present maps of 200 km

−

width at central velocities of

−

600,

−

400,

−

200, 0, +200, and +400 km s

−

from the top left to the

bottom right. 0 km s

−

corresponds to

[Cii]

emission at the redshift

= 6

595, i.e. 250.24 GHz (1.198mm).

Ouchi et al.

Fig. 3.—

Color composite image of Himiko. Blue and green rep-

resent

HST

/WFC3 continua of

125

and

160

, respectively. Red

indicates Ly

emission resolved with sub-arcsec seeing Subaru ob-

servations. The Ly

emission image comprises the Subaru

921

narrowband data with a subtraction of the continuum estimat

from the seeing-matched

HST

/WFC3 data. The three continuum

clumps are labeled A, B and C.

of Himiko at 9233

A (Ouchi et al. 2009a) with a system

throughput of 40%, close to the peak throughput of this

filter (

∼

45%). Thus, the

098

image is ideal for for

mapping the distribution of Ly

emitting gas.

Our observations were conducted in 2010 September

9, 12, 15-16, 18, and 26 with an ORIENT of 275 de-

grees. Some observations were partially lost because

HST

went into ‘safe mode’ on 2010 September 9, 22:30

during the execution of one visit. The total integra-

tion times for usable imaging data are 15670.5, 13245.5,

18064.6 seconds for

098

125

, and

160

, respectively.

The various WFC3 images were reduced with WFC3

and MULTIDRIZZLE packages on PyRAF. To optimize

our analyses, in the multidrizzle processing we chose a

final

pixfrac= 0

5 and pixel scale of 0

′′

05132. We de-

graded images of

098

and

125

to match the PSFs

of these images with the one of

160

that has the largest

size among the

HST

images. We ensured the final WFC3

images have a matched PSF size of 0

′′

19 FWHM.

Figure 3 presents a color composite

HST

UV-

continuum image of Himiko, together with a large ion-

ized Ly

cloud identified by the Subaru observations

(Ouchi et al. 2009a). This image reveals that the system

comprises 3 bright clumps of starlight surrounded by a

vast Ly

nebula

17 kpc across. We denote the three

clumps as A, B, and C. Figure 4 shows the

HST

, Sub-

aru, and

Spitzer

images separately. The

098

image

in Figure 4 detects only marginal extended Ly

emission,

because of the shallower surface brightness limit of the

2.4m

HST

compared to the 8m Subaru telescope. Never-

theless, we have found a possible bright extended compo-

nent at position D in Figure 4. We perform 0

′′

4-diameter

aperture photometry for the clumps A-C and location D

as well as 2

′′

-diameter aperture photometry which we

adopt as the total magnitude of the system. Tables 2

and 3 summarize the photometric properties. It should

be noted that Himiko is not only identified as an LAE,

but also would be regarded as a LBG or ‘dropout’ galaxy.

Fig. 4.—

HST

, Subaru, and

Spitzer

images of Himiko; north

is up and east is to the left. Each panel presents 5

′′

im-

ages at

098

125

, and

160

bands from

HST

/WFC3, 3

and 4

m bands from

Spitzer

SEDS. The Ly

image is a Subaru

921 image continuum subtracted using

125

and includes in-

tensity contours. The Subaru image has a PSF size of 0”

8. The

red-solid circles indicate the positions of 0

′′

4-diameter apertures

for Clumps A, B, C, and D photometry in the

HST

images (see

Section 2.2 for details), while the red-dashed circles deno

te 2

′′

diameter apertures used for the defining the total magnitude

Using the optical photometry of Ouchi et al. (2009a) (see

also Table 3), we find no blue continuum fluxes for the fil-

ters from

through

′

to the relevant detection limits of

−

29 mag. The very red color of

′

−

′

1 meets typ-

ical dropout selection criteria (e.g. Bouwens et al. 2011).

Because the

′

-band photometry includes the Ly

emis-

sion line and a Ly

-continuum break, we can also esti-

mate the continuum-break color using our

HST

photom-

etry of

125

and

160

and the optical

-band photometry.

Assuming the continuum spectrum is flat (

=const.),

we obtain a continuum break color

′

−

125

0 or

′

−

160

0, further supporting that Himiko as a

LBG. Importantly, these classifications apply also to the

clumps A-C ruling out that some could be foreground

sources.

The UV continuum magnitudes of clumps A-C range

from 26.4 to 27.0 magnitudes in

125

and

160

. Each

clump has a UV luminosity corresponding to the char-

acteristic luminosity

∗

of a

∼

7 galaxy,

=26

mag (Ouchi et al. 2009b; Bouwens et al. 2011)). More-

over, the variation in luminosity across the components is

small; there is no single dominant point source in this sys-

tem, confirming earlier deductions that the system does

not contain an active nucleus.

The

098

image shows that Ly

emission is not uni-

formly distributed across the 3 clumps. Clump A shows

intense Ly

emission with a rest-frame equivalent width

(

) of 68

+14

−

A placing it in the category of a Lyman

alpha Emitter (LAE), whereas clumps B and C are have

emission more typical of Lyman break galaxies (LBGs)

with a rest-frame Ly

equivalent width (

) less than

Star-Forming Galaxy Unveiled with ALMA and HST

TABLE 2

Properties of the Subcomponents

Component

(pix)

125

160

(Ly

)

SFR(UV) SFR(Ly

)

(mag)

(10

erg s

−

)

(

⊙

−

) (

⊙

−

)

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

Total

2815.4 2790.4 23

05 24

08 24

−

(F098M)

· · ·

(24

08)

· · ·

(30

6) (61

+28

−

)

· · ·

(28

14)

A(clump)

2803.0 2789.0 26

04 26

−

+14

−

B(clump)

2816.5 2790.5 27

09 27

07 27

−

C(clump)

2826.3 2791.9 26

05 26

04 26

−

+12

−

2821 2801 28

· · ·

123

Note

. — (1): Name of the component. (2)-(3): Positions in pixels.

(4)-(6): Aperture magnitudes in

125

, and

160

indicates a nar-

row/intermediate band magnitude determined with

098

for all lines except in Total(NB921). The upper limit corres

ponds to a 3

limit. (7): UV-continuum

slope. (8): Ly

luminosity. (9): Rest-frame apparent equivalent width of L

emission line in

A. (10): Star-formation rate estimated from the UV magnitud

(11): Star-formation rate estimated from the UV magnitude.

The

magnitude corresponds to

921. The quantities of (8)-(9) and (11) are estimated from

921 photometry.

A given the measurement uncertainties.

In summary, the

HST

and Subaru data indicates

Himiko is a triple

∗

galaxy system comprising one LAE

and two LBGs surrounded by an extensive 17 kpc dif-

fuse Ly

halo. Importantly, from the above morpholog-

ical studies, the possibility that Himiko is gravitational

lensed by a foreground concentration can be readily elim-

inated. Already, Ouchi et al. (2009b) made a strong case

against lensing given the Keck spectroscopy revealed a

velocity gradient of 60km s

−

across the system. We

can further reject this supposition given there are clear

asymmetries in the outermost images (one has strong

emission and the other does not).

2.3.

Spitzer

Although

Spitzer

cannot match the resolution of

the above morphological data, we use the very deep

Spitzer

/IRAC SEDS data reaching 26 mag at the 3

level (Ashby et al. 2013) to investigate the counterpart

of the overall Himiko system at 3

m and 4

m bands.

To improve the relative astrometric accuracy, we have

re-aligned the SEDS images to the

HST

images, refer-

ring bright stellar objects commonly detected in the Sp-

tizer and

HST

images. The relative astrometric errors

are estimated to be

≃

′′

1 rms. We obtain total magni-

tudes of

Spitzer

/IRAC images from a 3

′′

-diameter aper-

ture and use an aperture correction given in Yan et al.

(2005). The total magnitudes are 23

09 mag and

19 mag at 3

m and 4

m bands, respec-

tively. Because the

Spitzer

/IRAC 5

m and 8

and

Spitzer

/MIPS 24

m band images are not available

in the SEDS data set, we use the relatively shallow

Spitzer

/SpUDS (Dunlop et al.) photometry measure-

ments presented in Ouchi et al. (2009a). Table 3 sum-

marizes these total magnitudes and fluxes.

RESULTS

Ouchi et al. (2009a) found that Himiko has a high SFR

(

⊙

−

) and derived a moderately high stellar

mass (0

−

⊙

) from the Subaru photom-

etry and shallow

Spitzer

/SpUDS data. Here, we at-

tempt to improve upon these estimates and, for the first

time, secure information on dust content and inter-stellar

medium (ISM) metallicity.

3.1.

Far Infrared SED

TABLE 3

Total Magnitudes and Fluxes

of Himiko

Band

Mag/Flux(Total)

′

921

098

125

160

(24

)

([CII])

250

In units of AB magnitudes, if not

specified. The upper limits are 2

and 3

magnitudes in

BV Ri

′

and

−

m bands, respectively.

Total magnitudes from

UKIDSS/UDS DR8 data that

are estimated with a 2

′′

diameter

aperture photometry and the aper-

ture correction in the same manner

as Ono et al. (2010b).

Measurements obtained in

Ouchi et al. (2009a). The contin-

uum magnitudes from

through

′

are defined with a 2

′′

diameter

aperture photometry that matches

to the photometry of the total

magnitudes of NIR bands.

Three sigma upper limits derived

with our ALMA data.

We investigate obscured star-formation and dust prop-

erties of Himiko from its far-infrared (FIR) SED using

the newly available ALMA 1.2mm continuum data. The

SED from the optical to millimeter wavelengths is shown

in Figure 5, together with that of various local starburst

templates. The figure demonstrates that Himiko’s mil-

limeter flux is significantly weaker than that of dusty

Ouchi et al.

Fig. 5.—

The optical to far-infrared SED of Himiko in the ob-

served frame. The filled square shows the upper limit from our

deep

ALMA Band 6 observations and filled circles represent photom

etry

from

HST

/WFC3

125

and

160

photometry and

Spitzer

SEDS 3

and 4

m. Filled pentagons indicate the UKIDSS-UDS DR8

, and

photometry. Cross and plus symbols denote

HST

/WFC3

098

and Suprime-Cam

921 photometry that includes Ly

emission and Gunn-Peterson trough in their bandpasses. Ope

n cir-

cles and arrows are data points and the upper limits taken fro

Ouchi et al. (2009). The open diamond with an arrow shows the

upper limit from the IRAM observations (Walter et al. 2012).

Red,

magenta, green, and blue lines represent the SEDs of local ga

laxies,

Arp220, M82, M51, and NGC6946 (Silva et al. 1998), respectiv

ely,

redshifted to

= 6

595. SEDs of local dwarf irregular galaxies sim-

ilarly redshifted are presented with cyan lines (Dale et al.

2007).

All local galaxy SEDs are normalized in the rest-frame UV, wh

ere

Himiko’s SED is determined reliably.

starbursts in the local universe such as Arp220 and M82,

as well as the spiral galaxy NGC6946; it is more com-

parable to those of dwarf galaxies of much lower mass.

Similarly, Himiko’s rest-frame optical flux derived from

the

Spitzer

/IRAC 3

6 and 4

m photometry is signifi-

cantly weaker than those of dusty starbursts and spiral

galaxies. Given its intense rest-frame UV luminosity and

moderately high stellar mass, Himiko’s dust emission and

evolved stellar flux are remarkably weak. Both proper-

ties imply a low extinction and relatively young stellar

age (

3.3). In this sense Himiko may be similar to many

luminous

∼

3 LBGs whose cold-dust continuum emis-

sion are also comparable to unreddened local starburst

galaxies (Ouchi et al. 1999).

We can estimate a far-infrared luminosity of Himiko

from our 1.2mm continuum limit. Assuming an opti-

cally thin graybody of modified blackbody radiation with

a dust emissivity power-law spectral index of

= 1

and a dust temperature of

= 40K (Eales et al. 1989;

Klaas et al. 1997), we obtain a 3

upper limit of

FIR

⊙

integrated over 8

−

1000

m. We also esti-

mate 3

upper limits of

and

⊙

at 40

−

500

m and 42

−

122

m, respectively. Note

that these upper limits depend upon the assumed dust

temperature and

. For

= 25K and

= 60K, the

upper limit luminosities in 8

−

1000

m are

and

⊙

, respectively. Similarly, for

= 0

and

= 2, the 3

upper limit luminosities in 8

−

1000

are

and

⊙

, respectively.

Fig. 6.—

The

[Cii]

luminosity as a function of SFR. The

filled square indicates the 3

upper limit luminosity of Himiko and

the open diamond presents estimates of HFLS3 (Riechers et al

2013). The solid line is the local scaling relation determin

with the data shown with bars (de Looze et al. 2011). Note that

the bars are obtained by re-calculation the SFR values using

the

data of de Looze et al. (2011), following the formula shown in

de Looze et al. (2011). The shaded region indicates the obser

ved

scatter.

The foregoing upper luminosity limits do depend some-

what on dust temperature and spectral index. Based on

the Herschel measurements, Lee et al. (2012) find that

the average dust temperature is

∼

30 K under

= 1

for a relatively high redshift (

∼

4) LBGs with a lumi-

nosity of

∗

comparable to Himiko. In the local

universe, the median dust temperatures are 33 K, 30 K,

and 36 K, for E/S0, Sb-Sbc, and infrared bright galax-

ies, respectively. (Sauvage & Thuan 1994; Young et al.

1989). Recent numerical simulations have claimed that

LAEs may have a relatively high dust temperature, due

to the proximity of dust to star-forming regions. How-

ever, even in this case the maximum temperature reaches

only

≃

40 K (Yajima et al. 2012a). On the other hand

Himiko’s dust must be heated to some lower limit by

the cosmic microwave background (CMB) whose black-

body temperature scales as

CMB

(1 +

), where

CMB

is the temperature of present-day CMB,

CMB

= 2

K. Assuming local thermal equilibrium between ISM of

Himiko and CMB at

= 6

595 (da Cunha et al. 2013),

this yields a lower limit of

= 21 K. Thus, it is appropri-

ate to consider a range of

≃

−

40 K with

≃

Because the larger assumed dust temperature

= 40 K

with

= 1

5 provides a weaker upper limit, we adopt

a conservative 3

upper limit of

FIR

⊙

−

1000

m). Tables 1 and 3 present the 3

upper limit

of luminosity.

3.2.

ISM metallicity from

[Cii]

Emission

We now turn to estimating the metallicity of the ISM of

Himiko using with

[Cii]

emission as a valuable tracer in

star-forming regions. Despite our significant integration,

no line is seen. Figure 6 (and Table 3) presents the upper

limit to the

[Cii]

luminosity in the context of the correla-

Star-Forming Galaxy Unveiled with ALMA and HST

TABLE 4

Stellar Population of Himiko

Model

∗

(

−

)

∗

Age

SF R

sSF R

dof

(

⊙

)

(mag)

Myr

⊙

−

(1)

(2)

(3)

(4)

(5)

(6)

(7)

stellar+nebular 1

−

182

+22

−

100

2 6

−

pure stellar

−

363

+44

−

2 3

−

Note

. — (1): Models with or without nebular emission. (2): Stella

r mass. (3): Color excess of dust

extinction for stellar continua. (4): Stellar age. (5): Sta

r-formation rate. (6): Specific star-formation

rate. (7): Reduced

. The degree of freedom (dof) is six.

The uncertainty of color excess is smaller than our model-pa

rameter grid of ∆

(

−

) = 0

01.

tion with the star formation rate (SFR) (de Looze et al.

2011). In the case of Himiko, the SFR was obtained by

SED fitting of the rest-frame UV to optical data includ-

ing a correction for dust extinction (Section 3.3). Himiko

clearly departs significantly from the scaling relation; the

deficit amounts to a factor

≃ ×

30. Given the SFRs of

de Looze et al. (2011) for local galaxies are derived in a

similar manner to that for Himiko, including contribu-

tions from dust-free and dusty starbursts with GALEX’s

UV and Spitzer’s infrared fluxes, respectively, it seems

difficult to believe this deficiency arises from some form

of bias arising from comparing different populations.

Graci ́a-Carpio et al. (2011) and Diaz-Santos et al.

(2013) present

[CII]

FIR

ratios for local starbursts

that depend on

FIR

and the FIR and mid-IR surface

brightnesses. As a result, Diaz-Santos et al. (2013) ar-

gue that

[CII]

may not represent a particularly reli-

able indicator of SFR. However, FIR and mid-IR lu-

minosities only trace dusty starbursts and typically ex-

clude dust-free measures such as the UV luminosity.

Because galaxies with fainter FIR/mid-IR luminosities

have a larger ratio of

[CII]

FIR

in the datasets probed

by Graci ́a-Carpio et al. (2011) and Diaz-Santos et al.

(2013), more dust-free star-formation is expected in such

systems. In this sense, the analysis of de Looze et al.

(2011) is perhaps more relevant as a prediction of what

to expect for Himiko. Nonetheless, given the importance

of using

[CII]

as a possible tracer and the discussion that

follows below, independent studies of

[CII]

as a function

of UV luminosity and

FIR

would be desirable. Figure 6

also shows that HFLS3 at

= 6

3 (Riechers et al. 2013)

follows the local scaling relation. However, it should be

noted that the SFR of HFLS3 is derived from the far-

infrared luminosity and thus any contribution from dust-

free star-formation would be missing. In this sense, the

SFR is possibly a lower limit, in which case HFLS3 may

also depart somewhat from the local relation.

The absence of

[Cii]

emission in Himiko is perhaps the

most surprising result from our ALMA campaign. The

emission line is often assumed to be the most robust far-

IR tracer of star formation in high redshift galaxies, such

that it may replace optical lines such as Ly

in securing

spectroscopic redshifts in the reionization era. Our fail-

ure to detect this line in one of the most spectacular

≃

galaxies has significant implications which we discuss in

Section 4.

3.3.

Improved Physical Properties from the

Near-Infrared SED

Although some constraints on the integrated properties

of Himiko were derived in our earlier work (Ouchi et al.

2009a), no

(

−

) estimate and only the lower limit of

SFR with

(

−

)

≥

0 were obtained, due to the large

uncertainties of photometric measurements. We now re-

fine these estimates based on our significantly deeper

HST

and

Spitzer

data. Our near-IR SED is taken us-

ing total magnitudes from the

HST

images (

2.2), the

Spitzer

/IRAC SEDS images (

2.3) and

JHK

DR8 data

from the UKIDSS/UDS survey. We tabulate these total

magnitudes in Table 3 including ground-based optical

data previously given in Ouchi et al. (2009a).

We present the SED of Himiko in Figure 7 and un-

dertake

fitting of a range of stellar synthesis mod-

els in the same manner as Ono et al. (2010b) using

the stellar synthesis models of Bruzual & Charlot (2003)

with dust attenuation formulate given by Calzetti et al.

(2000). We adopt Salpeter initial mass function (IMF;

Salpeter 1955) with lower and upper mass cutoffs of 0.1

and 100

⊙

, respectively. Applying models of con-

stant and exponentially-decaying star-formation histo-

ries with metallicities ranging from

= 0

−

⊙

we search for the best-fit model in a parameter space

(

−

) = 0

−

1 and age= 1

−

810 Myr (where

the latter upper limit corresponds to the cosmic age at

= 6

595). Nebular continuum and line emission, esti-

mated from the ionizing photons from young stars, are

optionally included following the metallicity-dependent

prescriptions presented in Schaerer & de Barros (2009);

Ono et al. (2010b).

For a constant star-formation rate history with no neb-

ular emission and a fixed metallicity of

= 0

⊙

we find our best-fit model has a stellar mass of

∗

−

⊙

, a stellar age of 3

−

yr, a SFR

of 98

−

⊙

−

, and extinction of

(

−

) = 0

15 with

a reduced

of 3

1. This is a significant improvement

over our much weaker earlier constraints which did not

have the benefit of the

HST

/WFC3 or

Spitzer

/SEDS

data(Ouchi et al. 2009a). The new infrared data provide

a critical role in determining the Balmer break thereby

resolving the degeneracy between extinction and age. On

the other hand, the fit itself is not very satisfactory. The

reduced

is large and there is a significant discrepancy

at 3

m. Since the 3

m and 4

m bands sample the

strong nebular lines of H

[Oiii]

and H

, respectively,

= 6

595, this encourages us to include nebular emis-

sion in our fitting procedure. In fact, in Figure 4, we

note the IRAC 4

m emission shows a positional offset

with respect to that at 3

m suggesting the possibility

of contamination by nebular emission.

Adding nebular emission to the stellar SED models

given above, the best fit has a more satisfactory reduced

, 1

6, and we derive a reduced stellar mass of

∗

Ouchi et al.

Fig. 7.—

The optical to near-infrared SED of Himiko newly

obtained by our deep

HST

and

Spitzer

observations, together with

photometry from ground-based observations. The red lines r

epre-

sent the best-fit SEDs of stellar synthesis models with (left

) and

without (right) nebular lines (see Ono et al. 2010 for detail

ed model

descriptions). The filled squares denote HST and Spitzer/SE

fluxes of Himiko defined by the total magnitudes. The open squa

res

show

′

-band fluxes that are not used for the SED fitting, due to

the Ly

line contamination. The large error bars at

≤

m and

m are those obtained by Subaru and Spitzer/SpUDS observa-

tions given by Ouchi et al. (2009a). The red crosses represen

t the

broadband fluxes expected from the best-fit SED models. For va

ri-

ous assumptions the fits indicate that Himiko has a SFR of 100

⊙

−

, stellar mass of 2

−

⊙

, and a selective extinction of

(

−

) = 0

15 (see text for details).

−

⊙

, a younger stellar age of 1

−

yr,

but similar values for the SFR of 100

−

⊙

−

and

extinction of

(

−

) = 0

15. Table 4 summarizes

the results of our SED fitting with the pure stellar and

stellar+nebular models. In the stellar+nebular models,

we assume that all ionizing photons lead to nebular emis-

sion lines corresponding to an escape fraction

esc

=0. If

we allow

esc

to be a free parameter, following Ono et al.

(2012) we find no change from the model above (i.e.

esc

=0) and formally establish that

esc

Labb ́e et al. (2010) and Finkelstein et al. (2010) have

suggested from their pure stellar models that

HST

−

8 dropout galaxies have modest stellar masses (10

−

⊙

) and are quite young (30

−

300Myr), in contrast

with Himiko’s stellar mass (

∗

≃

⊙

) and

age (360Myr) estimated with our pure stellar models.

Of course, Himiko is more massive and energetic than

typical LBGs seen in the small area of Hubble Ultra

Deep Field. Its most notable feature is its high SFR of

≃

100

⊙

−

which is more than an order of magnitude

larger than those of the

HST

LBGs at similar redshifts (

≃

−

⊙

−

; Labb ́e et al. 2010). Himiko’s selective ex-

tinction,

(

−

) = 0

15, is also larger than that of

HST

dropouts, more than half of which are consistent with no

extinction (Finkelstein et al. 2010). On the other hand,

the stellar mass of Himiko is only about 1/10th that of

many submm galaxies (SMGs) at

∼

3 (Chapman et al.

2005). We estimate a specific star-formation rate,

sSF R

defined by a ratio of star-formation rate to stellar mass to

sSF R

= 3

−

and

sSF R

= 6

−

, for the pure stellar and stellar+nebular cases, re-

spectively. Even though the stellar masses are very differ-

ent, Himiko, SMGs and LBGs at

∼

3 share comparable

sSF R

∼

−

(see Figure 12 of Ono et al.

2010a).

3.4.

UV Spectral Slopes on the Spatially Resolved

Images

Fig. 8.—

The UV to FIR luminosity ratio, log(

FIR

1600

), as a

function of the UV-continuum slope,

. The filled square presents

the upper limit of log(

FIR

1600

) and the measurement of

for

the total luminosities of Himiko. Solid line denotes the rel

ation for

local starburts given by Meurer et al. (1999).

The new

HST

data gives us the first reliable measure-

ment of the UV continuum slope for each of the mor-

phological components identified in Figure 4. The UV

spectral slope provides a valuable indicator of the com-

bination of dust extinction, metallicity, the upper IMF

and stellar age. We estimated the UV slope,

, from the

125

and

160

photometry that samples the continua at

the rest-frame wavelengths of

∼

1600 and

∼

2100

A nei-

ther of which is contaminated by either Ly

emission nor

the Ly

-continuum break.

We calculate

via

−

125

−

160

5 log (

/λ

)

−

(1)

where

and

are the central wavelengths of the

125

and

160

filters, respectively. The estimates for

each component are summarized in Table 2. We obtain

−

57 for the entire system of Himiko, com-

parable to the average UV slope of

≃

∗

LBGs,

−

22 (Bouwens et al. 2012, see also Dunlop et al.

2013). Figure 8 shows the UV to FIR luminosity ratio,

log(

FIR

1600

), and the UV-continuum slope,

, for the

entire system of Himiko, and compares these estimates

with the relation of local starbursts (Meurer et al. 1999).

Figure 8 indicates that Himiko has log(

FIR

1600

values comparable with or smaller than those of local

dust-poor starbursts. Since the Small Magellanic Cloud

(SMC) extinction has a smaller log(

FIR

1600

) value

at a given

(see Figure 10 of Reddy et al. 2010) due

to SMC’s steeper extinction curve in

-1

/λ

than

that for local starbursts, it may be more appropriate for

Himiko. Our result also suggests that Himiko is not as-

sociated with additional FIR sources which are invisi-

ble in the rest-frame UV. These implications are con-

sistent with the conclusions of UV-FIR luminosity ratio

discussed in Figure 5.

Star-Forming Galaxy Unveiled with ALMA and HST

Fig. 9.—

The UV-continuum slope,

, as a function of rest-

frame Ly

equivalent width. The filled circle, square, pentagon,

and diamond denote measures for the entire system, Clumps A,

and C, respectively. Blue, cyan, green, and red solid lines r

epresent

predictions based on instantaneous starburst models of Pop

III,

−

⊙

, 0

⊙

, and 0

⊙

with nebular and stellar continua

(Raiter et al. 2010). Thin lines are the same, but for constan

star-formation models. The associated dotted lines show th

e effect

of ignoring nebular emission. The arrow indicates the effect

applying an extinction with

(

−

)

= 0

1 (see the text for

details).

More interestingly, the UV slopes of the individual sub-

structures provide valuable information on the nature of

Himiko. Clumps B and C have

−

47 and

−

28, respectively, comparable to the the

average UV slope of

≃

∗

LBGs. However, Clump A

presents a very blue UV slope,

−

32. Because

this component is detected at the

∼

level in both

125

and

160

, the UV slope is quite reliable. Bouwens et al.

(2012) claim that selection and photometric biases lead

to an error of only ∆

≃

1 for the brightest of their

sources with

∼

photometry (see also Dunlop et al.

2013). Even including such a possible bias, Clump A re-

mains significantly bluer than the average

≃

∗

LBGs at

the

≃

level.

As presented in Section 2.2, Clump A also shows Ly

emission. Together with the blue UV slope, this suggests

a very young and/or metal poor component. However,

the Ly

equivalent width is only

= 68

+14

−

A. To

understand the significance of this, in Figure 9, we com-

pare

and

for the entire Himiko system and the

various clumps with the stellar and nebular models of

Raiter et al. (2010), where Salpeter IMF is assumed. In

Figure 9, the arrow size in

for the stellar extinction of

(

−

)

= 0

1 is calculated with the combination of the

empirical relation,

1600

= 4

43 + 1

(Meurer et al.

1999), and Calzetti extinction,

1600

(

−

)

where

1600

is 10 (Ouchi et al. 2004a). Similarly, the

arrow size in

for

(

−

)

= 0

1 is estimated

from the relation given in Ono et al. (2010a) under the

assumption of

flat continuum and the standard star-

formation rate relations of UV and Ly

luminosities in

the case B recombination. Figure 9 shows that the data

points of Himiko fall on the tracks of star-formation pho-

toionization models (Raiter et al. 2010) within the mea-

surement errors and the dust-extinction correction uncer-

tainties, and indicates that Ly

emission of Himiko can

be explained by the photoionization by massive stars.

DISCUSSION

We now bring together our key results, both from

the earlier Subaru program(Ouchi et al. 2009a) and the

present

HST

and ALMA campaigns, in order to under-

stand the significance of our upper limits on the [C II]

and dust emission and thereby the nature of Himiko.

4.1.

The Low Dust and Metal Content of Himiko

We have shown (Figure 5) that Himiko’s submm emis-

sion is comparable with or weaker than that of local

dwarf irregulars with far lower star-formation rates, in-

dicating intensive star-formation in a dust-poor gaseous

environment. In fact, assuming the local starburst

SF R

−

(FIR) relation of Kennicutt (1998) with the

Himiko’s FIR upper limit luminosity of

⊙

we obtain

SF R

(FIR)

⊙

−

that is far smaller

than not only our best optical-NIR estimate SFR of

≃

100

⊙

−

, but also the UV-luminosity SFR of

SF R

(UV) = 30

⊙

−

with no dust extinction cor-

rection. This is also true under the assumption of the

SF R

−

(FIR) relation (Buat & Xu 1996) valid for local

dust poorer disk systems of Sb and later galaxies, which

provide

SF R

(FIR)

⊙

−

. In this way, Himiko

does not follow the

SF R

−

(FIR) relation of typical lo-

cal galaxies, indicating a dust-poor gaseous environment.

This seems similar to observations which find extended

emission in dust poor low-

galaxies (Hayes et al.

2013) and a high-

QSO (Willott et al. 2013). Based

on numerical simulations, Dayal et al. (2010) find that

∼

−

7 LAEs are dust poor with a dust-to-gas

mass ratio smaller than Milky Way by a factor of 20.

Dayal et al. (2010) predict a 1.4mm continuum flux of

≃

Jy for sources with

(Ly

) = 2

−

erg

−

= 6

6, a result comparable with our ALMA

observations. Deeper ALMA observations could further

test the model of Dayal et al. (2010) and place important

constraints on the dust-to-gas mass ratio.

Similarly, our strong upper limit on the

[Cii]

158

line (Figure 6) places it significantly below the scaling re-

lation of

[CII]

and SFR obeyed by lower redshift galax-

ies. This discovery indicates the following four possi-

bilities: Himiko has a) a hard ionizing spectrum from

an AGN, b) a very high density of photo-dissociation

regions (PDRs), c) a low metallicity, and d) a large col-

umn density of dust. In the case of a), a hard ionizing

spectrum of AGN can produce little

[Cii]

luminosity rel-

ative to FIR luminosity, due to the intense ionization

field (Stacey et al. 2010). As we discuss in (ii) of Sec-

tion 4.2, there are no signatures of AGN; no detections

of X-ray and high-ionization lines as well as extended

sources plus non-AGN like Ly

profile+surface bright-

ness. We can rule out the possibility of a). In the case

of b), a very high density of PDRs gives more rapid col-

lisional de-excitations for the forbidden line of

[Cii]

, and

quench a

[Cii]

emission line. In the case of c), the PDRs

in Himiko are composed of metal poor gas that may be

quite typical of normal galaxies observed at early epochs.