1001.0005.txt

arXiv:1001.0005v1  [astro-ph.CO]  30 Dec 2009Astronomy& Astrophysics manuscriptno.akari˙RXJ1716˙v5 c∝circlecopyrtESO 2018
October30,2018
Environmentaldependenceof 8 µmluminosityfunctionsof
galaxiesatz ∼0.8
Comparison between RXJ1716.4 +6708 andthe AKARI NEP deep field.⋆,⋆⋆
Tomotsugu Goto1,2,⋆⋆⋆, Yusei Koyama3,T.Wada4,C.Pearson5,6,7,H.Matsuhara4,T.Takagi4, H.Shim8, M.Im8,
M.G.Lee8, H.Inami4,9,10,M.Malkan11, S.Okamura3,T.T.Takeuchi12, S.Serjeant7, T.Kodama2, T.Nakagawa4,
S.Oyabu4,Y.Ohyama13, H.M.Lee8, N.Hwang2, H.Hanami14, K.Imai15,and T.Ishigaki16
1Institute for Astronomy, University of Hawaii,2680 Woodla wnDrive, Honolulu, HI,96822, USA
e-mail:[email protected]
2National Astronomical Observatory, 2-21-1 Osawa,Mitaka, Tokyo, 181-8588,Japan
3Department of Astronomy, School of Science,The University of Tokyo, Tokyo113-0033, Japan
4Institute of Space and Astronautical Science, JapanAerosp ace Exploration Agency, Sagamihara,Kanagawa 229-8510
5Rutherford Appleton Laboratory, Chilton, Didcot,Oxfords hire OX110QX, UK
6Department of Physics,Universityof Lethbridge, 4401 Univ ersity Drive,Lethbridge, AlbertaT1J 1B1, Canada
7Astrophysics Group, Department of Physics, The OpenUniver sity, MiltonKeynes, MK76AA, UK
8Department of Physics& Astronomy, FPRD,Seoul National Uni versity, Shillim-Dong,Kwanak-Gu, Seoul 151-742, Korea
9Spitzer Science Center,California Institute ofTechnolog y, Pasadena, CA91125
10Department of Astronomical Science,The Graduate Universi tyfor Advanced Studies
11Department of Physicsand Astronomy, UCLA,Los Angeles, CA, 90095-1547 USA
12Institute for Advanced Research, Nagoya University, Furo- cho, Chikusa-ku, Nagoya 464-8601
13Academia Sinica,Institute of Astronomyand Astrophysics, Taiwan
14Physics Section,Facultyof Humanities and SocialSciences , Iwate University, Morioka, 020-8550
15TOMER&D Inc. Kawasaki, Kanagawa 2130012, Japan
16Asahikawa National College of Technology, 2-1-6 2-joShunk ohdai, Asahikawa-shi, Hokkaido 071-8142
Received September 15, 2009; accepted December 16, 2009
ABSTRACT
Aims.Weaim to reveal environmental dependence of infraredlumin osity functions (IR LFs)of galaxies at z ∼0.8 using the AKARI
satellite. AKARI’s wide field of view and unique mid-IR filter s help us to construct restframe 8 µm LFs directly without relying on
SEDmodels.
Methods. We construct restframe 8 µm IR LFs in the cluster region RXJ1716.4 +6708 at z=0.81, and compare them with a blank
field using the AKARI North Ecliptic Pole deep field data at the same redshift. AKARI’s wide field of view (10’ ×10’) is suitable to
investigate wide range of galaxy environments. AKARI’s 15 µm filter is advantageous here since it directly probes restfr ame 8µm at
z∼0.8, without relyingona large extrapolation based ona SEDfi t,which was the largestuncertainty inprevious work.
Results. We have found that cluster IR LFsat restframe 8 µm have a factor of 2.4smaller L∗and a steeper faint-end slope than that
of the field. Confirming this trend, we also found that faint-e nd slopes of the cluster LFs becomes flatter and flatter with de creasing
local galaxy density. These changes in LFs cannot be explain ed by a simple infall of field galaxy population into a cluster . Physics
that canpreferentiallysuppress IR luminous galaxies inhi gh density regions is requiredtoexplain the observed resul ts.
Keywords. galaxies: evolution, galaxies:interactions, galaxies:s tarburst, galaxies:peculiar, galaxies:formation
1. Introduction
It hasbeenobservedthat galaxypropertieschangeas a funct ion
of galaxyenvironment;the morphology-densityrelation re ports
that fractionof elliptical galaxiesis largerat highergal axyden-
sity(Gotoetal.,2003);thestarformationrate(SFR)ishig herin
lower galaxy density (G´ omezet al., 2003; Tanakaet al., 200 4)
. However, despite accumulating observational evidence, w e
⋆This research is based on the observations with AKARI, a JAXA
project withthe participationof ESA.
⋆⋆Based on data collected at Subaru Telescope, which is operat ed by
the National Astronomical Observatory ofJapan.
⋆⋆⋆JSPSSPDfellowstill do not fully understand the underlying physics govern ing
environmental-dependentevolutionofgalaxies.
Infrared (IR) emission of galaxies is an important
probe of galaxy activity since at higher redshift, a sig-
nificant fraction of star formation is obscured by dust
(Takeuchi,Buat,&Burgarella, 2005; Gotoetal., 2010).
Although there exist low-z cluster studies (Baiet al., 2006 ;
Shimet al., 2010; Tranetal., 2010), not much attention has
been paid to the infrared properties of high redshift cluste r
galaxies, mainly due to the lack of sensitivity in previous I R
satellites such as ISO and IRAS. Superb sensitivity of recen tly
launched Spitzer and AKARI satellites can revolutionize th e
infraredviewofenvironmentaldependenceofgalaxyevolut ion.2 Gotoet al.:Environemental dependence of 8 µm luminosity functions ofgalaxies atz ∼0.8
Fig.1.Restframe 8 µm LFs of cluster RXJ1716.4 +6708 at
z=0.81 in the squares, and those of the AKARI NEP deep
field in the triangles. For RXJ1716.4 +6708, only photometric
and spectroscopic cluster member galaxies are used. For the
NEP deep field, galaxies with photo-z/specz in the range of
0.65< z <0.9are used. The dot-dashed lines are 8 µm LFs
of RXJ1716.4 +6708, but scaled down for easier comparison.
Thethindottedlinesarethebest-fitdoublepowerlaws.Vert ical
arrows show the 5 σflux limits of deep/shallow regions of the
cluster (red) and the NEP deep field (blue) in terms of L8µmat
z=0.81.
In this work, we comparerestframe8 µm LFs between clus-
ter and field regions at z=0.8 using data from the AKARI.
Monochromaticrestframe 8 µm luminosity ( L8µm) is important
since it is known to correlate well with the total IR luminosi ty
(Babbedgeet al., 2006; Huanget al., 2007), andhence,with t he
SFR of galaxies (Kennicutt, 1998). This is especially true f or
star-forminggalaxiesbecausethe rest-frame8 µm fluxare dom-
inatedbyprominentPAHfeaturessuchasat6.2,7.7and8.6 µm
(Desert,Boulanger,&Puget, 1990).
ImportantadvantagesbroughtbytheAKARIareasfollows:
(i) At z=0.8, AKARI’s 15 µm filter (L15) covers the redshifted
restframe 8 µm, thus we can estimate 8 µm LFs without using
a large extrapolation based on SED models, which were the
largest uncertainty in previous work. (ii) Large field of vie w of
the AKARI’smid-IRcamera(IRC, 10’ ×10’)allowsustostudy
wider area including cluster outskirts, where important ev olu-
tionary mechanisms are suggested to be at work (Gotoet al.,
2004; Kodamaet al., 2004). For example, passive spiral gala x-
ies have been observed in such an environment (Gotoet al.,
2003). Unless otherwise stated, we adopt a cosmology with
(h,Ωm,ΩΛ) = (0.7,0.3,0.7)(Komatsuet al., 2009).
2. Data & Analysis
2.1. LFs ofclusterRXJ1716.4 +6708
The AKARI is a Japanese infrared satellite (Murakamiet al.,
2007), which has continuous filter coverage in the mid
IR wavelengths ( N2,N3,N4,S7,S9W,S11,L15,L18Wand
L24). The AKARI has observed a massive galaxy cluster,Fig.2.Restframe 8 µm LFs of cluster RXJ1716.4 +6708 at
z=0.81, divided according to the local galaxy density ( Σ5th).
Thestars,circlesandsquaresareforgalaxieswith logΣ5th≥2,
1.6≤logΣ5th<2,andlogΣ5th<1.6,respectively.
RXJ1716.4 +6708, in N3,S7andL15(Koyamaetal., 2008).
RXJ1716.4 +6708 is at z=0.81 and has σ= 1522+215
−150km s−1,
LXbol= 13.86±1.04×1044ergs−1,kT= 6.8+1.0
−0.6keV.Mass
estimate from weak lensing and X-ray are 3.7 ±1.3×1014M⊙
and 4.35 ±0.83×1014M⊙, respectively (see Koyamaet al.,
2007, forreferences).
An important advantage of the AKARI observation is L15
filter, which corresponds to the restframe 8 µm at z=0.81. With
15 (3) pointings, L15reaches 66.5 (96.5) µJy in deep (shal-
low) regions at 5 σ. Here flux is measured in 11” aperture,
and coverted to total flux using AKARI’s IRC correction table
(2009.5.1)1.ClusterstudieswiththeSpitzerareoftenperformed
in 24µm and thus needed a large extrapolation to estimate ei-
therL8µmor total infrared luminosity ( LTIR,8−1000µm).
Note that we do not claim the L8µmis a better indicator of
thetotalIRluminositythanotherindicators(Brandlet al. ,2006;
Calzetti et al., 2007; Riekeet al., 2009), but it is importan t that
theAKARIcanmeausureredshifted 8µmfluxdirectlyinoneof
thefilters.
Thanks to the AKARI’s wide field of view (10’ ×10’), the
total area coverage around the cluster is 200 arcmin2, which
cover larger area than previous cluster studies with the Spi tzer,
allowingustostudyIRsourcesintheoutskirts,whereimpor tant
galaxyevolutiontakesplace(e.g.,Gotoet al.,2003).Prev iously,
Koyamaet al. (2008) reporteda high fractionof L15sourcesin
the intermediatedensity regionin the cluster,suggesting a pres-
enceofenvironmentaleffectintheintermediatedensityen viron-
ment.
Thissameregionwasimagedwith Suprime-Camin VRi′z′
and has a good photometric redshift estimate (Koyamaet al.,
2007).Usedinthisworkare54 L15-detectedgalaxieswhichare
well identifiedwithopticalsourceswith 0.76≤zphoto≤0.83.
With the L15filter covering the restframe 8 µm, we simply
convert the observed flux to 8 µm monochromatic luminosity
1http://www.ir.isas.jaxa.jp/ASTRO-
F/Observation/DataReduction/IRC/ApertureCorrection 090501.htmlGotoet al.:Environemental dependence of 8 µm luminosity functions ofgalaxies atz ∼0.8 3
Table 1.Best doublepower-lawfit parametersforLFs
Sample L∗
8µm(L⊙)φ∗(Mpc−3dex−1)α β
NEPDeepfield 6.1 ±0.5×10100.0010±0.0003 1.1 ±0.3 5.7 ±1.2
RXJ1716.4 +67082.5±0.1×10100.74±0.04 2.6 ±0.1 5.5 ±0.4
(L8µm) using a standard cosmology. Completeness was mea-
sured by distributing artificial point sources with varying flux
withinthe field andby examiningwhat fractionofthem wasre-
coveredasafunctionofinputflux.Sincewehavedeepercover -
age at the center of the cluster, the completeness was measur ed
separately in the central deep region and the outer regions o f
the field. More detail of the method is described in Wada et al.
(2008).
Oncethefluxisconvertedtoluminosityandcompletenessis
takenintoaccount,it is straightforwardto construct L8µmLFs,
which we show in the squares in Fig.1. Errors of the LFs are
assumedtofollowPoissondistribution.Here,wetakeanang ular
distance of the most distant source from the cluster center a s
a cluster radius ( Rmax= 6.2Mpc). We assumed4
3πR3
maxas
the volume of the cluster to obtain galaxy density ( φ). This is
only one of many ways to define a cluster volume, and thus, a
cautionmustbetakentocompare absolute valuesofourLFsto
other work such as Shimet al. (2010). This cluster is elongat ed
inangulardirection(Koyamaet al.,2007),andthus,thevol ume
mightnotbespherical.Yet,comparisonofthe shapeoftheLFs
isvalid.
2.2. LFs inthe AKARI NEP Deepfield
Our field LFs are based on the AKARI NEP Deep field
data. The AKARI performed deep imaging in the North
Ecliptic Pole Region (NEP) from 2-24 µm, with 4 pointings
in each field over 0.4 deg2(Matsuharaet al., 2006, 2007;
Wada etal., 2008). The 5 σsensitivity in the AKARI IR filters
(N2,N3,N4,S7,S9W,S11,L15,L18WandL24) are 14.2,
11.0, 8.0, 48, 58, 71, 117, 121 and 275 µJy (Wada etal., 2008).
Flux is measured in 3 pix radius aperture (=7”), then correct ed
tototal flux.
AsubregionoftheNEP-Deepfield(0.25deg2)hasancillary
datafromSubaru BVRi′z′(Imaiet al.,2007;Wada etal.,2008),
CFHTu′(Serjeant et al. in prep.), KPNO2m/FLAMINGOs J
andKs(Imaietal., 2007), GALEX FUVandNUV(Malkan
et al. in prep.). For the optical identification of MIR source s,
we adopt the likelihood ratio method (Sutherland&Saunders ,
1992).Usingthesedata,weestimatephotometricredshifto fL15
detectedsourcesintheregionwiththe LePhare (Ilbertet al.,
2006; Arnoutset al., 2007).Themeasurederrorsonthephoto -z
against 293 spec-z galaxies from Keck/DEIMOS (Takagi et al.
in prep.) are∆z
1+z=0.036 at z≤0.8. We have excluded those
sourcesbetterfit with QSO templatesfromtheLFs.
To construct field LFs, we have selected L15sources at
0.65< zphotoz<0.9. There remained 289 IR galaxies with
a median redshift of 0.76. L15flux is converted to L8µmus-
ing the photometric redshift of each galaxy. LFs are com-
puted using the 1/ Vmaxmethod. We used the SED templates
(Lagache,Dole,&Puget, 2003) for k-corrections to obtain the
maximumobservableredshiftfromthefluxlimit.Completene ss
of theL15detection is corrected using Pearsonet al. (2009b).
Thiscorrectionis25%atmaximum,sincewe onlyusethesam-
plewherethecompletenessisgreaterthan80%.
The resulting field LFs are shown in the dotted line and tri-
angles in Fig.1. Errors of the LFs are computed using a 1000Monte Caro simulation with varying zand flux within their er-
rors. These estimated errors are added to the Poisson errors in
eachLFbinin quadrature.
We performed a detaild comparison of restframe 8 µm
LFs to those in the literature in Gotoetal. (2010). Briefly,
there is an oder of difference between Caputiet al. (2007) an d
Babbedgeetal. (2006), reflecting difficulty in estimating L8µm
dominatedbyPAHemissionsusingSpitzer24 µmflux.Ourfield
8µm LF lies between Caputi etal. (2007) and Babbedgeet al.
(2006). Compared with these work, we have directly observed
restframe 8 µm using the AKARI L15filter, eliminating the un-
certaintlyinfluxconversionbasedonSEDmodels.Moredetai ls
andevolutionoffieldIRLFsaredescribedinGotoet al.(2010 ).
3. Results& Discussion
3.1. 8µmIRLFs
In Fig.1, we show restframe 8 µm LFs of cluster
RXJ1716.4 +6708 in the squares, and LFs of the field re-
gion in the triangles. First of all, cluster LFs have by a fact or
of∼700 higher density than the field LFs, reflecting the fact
the galaxy clusters is indeed high density regions in terms o f
infraredsources.
Next, to compare the shape of the LFs, we normalized the
cluster LF to match the field LFs at the faintest end, and show
in the dash-dottedline. In contrast to the field LFs, which sh ow
flattening of the slope at log L8µm<10.8L⊙, the cluster LF
maintainsthesteepslopeintherangeof 10.0L⊙<logL8µm<
10.6L⊙.Thedifferenceissignificant,consideringthesizeofer-
rorsoneachLF.
Wefitadouble-powerlawtobothclusterandfieldLFsusing
thefollowingformulae.
Φ(L)dL/L∗= Φ∗/parenleftbiggL
L∗/parenrightbigg1−α
dL/L∗,(L < L∗) (1)
Φ(L)dL/L∗= Φ∗/parenleftbiggL
L∗/parenrightbigg1−β
dL/L∗,(L > L∗) (2)
Free parameters are: L∗(characteristic luminosity, L⊙),φ∗
(normalization, Mpc−3),αandβ(faint and bright end slopes),
respectively.ThebestfitvaluesforfieldandclusterLFsare sum-
marisedinTable1andshowninthedottedlinesinFig.1.
The bright-end slopes are not very different, but L∗of the
cluster LF is smaller than the field by a factor of 2.4, and the
faint-endtailofclusterLF issteeperthanthatoffieldLF.
To further examine the difference at the faint end of the
LFs, we divide the cluster LF using the local galaxy density
(Σ5th) measuredbyKoyamaet al. (2008). Thisdensityis based
on the distance to the 5th nearest neighbor in the transverse di-
rection using all the optical photo-z members, and thus, is a
surface galaxy density. We separate LFs using similar crite ria,
logΣ5th≥2(dense),1.6≤logΣ5th<2(intermediate), and
logΣ5th<1.6(sparse), then plot LFs of each region in the4 Gotoet al.:Environemental dependence of 8 µm luminosity functions ofgalaxies atz ∼0.8
stars, circles, and squares in Fig.2. A fraction of the total vol-
umeofthe clusteris assignedto eachdensitygroupin invers ely
proportionaltothe sumof Σ3/2
5thofeachgroup.
Interestingly, the faint-end slope becomes flatter and flatt er
with decreasing local galaxy density. This result is consis tent
with our comparison with the field in Fig.1. In fact, the lowes t
densityLF(squares)hasaflatfaint-endtailsimilartothat ofthe
fieldLF.SincetheseLFsarebasedonthesamedata,changesin
the faint-end slope are not likely due to the errors in comple te-
ness correction nor calibration problems. The completenes s of
the deep and shallow regions of the cluster are measured sep-
arately. The changes in the slope is much larger than the maxi -
mumcompletenesscorrectionof25%.Wealsocheckedtheclus -
ter LFs as a function of cluster centric radius, to find no sign ifi-
cantdifference,perhapsduetotheelongatedmorphologyof this
cluster. At the same time, assuming the same cluster volume,
Fig.2 shows that a possible contamination from the field gala x-
ies to cluster LFs is only ∼0.1% in the dense region and ∼1%
eveninthe sparseregion.
It is interesting that not just the change in the scale of the
LFs, but there is a change in the L∗and the faint-end slope ( α)
of the LFs, resulting in the deficit in the 10.2L⊙<logL8µm<
10.8L⊙for cluster LFs. One might imaginea change just in L∗
might explain the difference in Fig.1. However, in Fig.2, th ere
clearlyisachangein theslopeasafunctionof Σ5th.
However,interpretationis rathercomplicated;a shapeofL F
would not change if field galaxies infall into cluster unifor mly
withoutchangingtheirstar-formationactivity.Although inclus-
ter environment,a fractionof MIR luminousgalaxiesis smal ler
than field (Koyamaet al., 2008), uniformand instant quenchi ng
of star-formation activity of field galaxies can only shift a LF,
butcannotaccountforachangein L∗andαoftheLFs.
Two important findings in this work are; (i) L∗is smaller
in the cluster. (ii) the faint-end slopes become steeper tow ard
higher-density regions. To explain these changes in LFs, IR -
luminousgalaxiesneedtobepreferentiallyreduced,witha rela-
tive increase of IR-faint galaxies. However, an environmen tal-
driven physical process such as the ram-pressure stripping or
galaxy-merging would quench star-formation not only in mas -
sivegalaxiesbutinlessmassivegalaxiesaswell,andthusi snot
abletoexplaintheobservedchangesinLFs.
Ontheotherhand,ithasbeenfrequentlyobservedthatmore
massive galaxies formed earlier in the Universe. This downs iz-
ing scenario also depends on the environment,in the sense th at
galaxieswith same mass are moreevolvedin higherdensityen -
vironmentsthangalaxisin less denseenvironments(Gotoet al.,
2005; Tanakaet al., 2005, 2008). Statistically, a good corr ela-
tionhasbeenfoundbetween LTIRandstellarmass(Elbazet al.,
2007). Our finding of the relative lack of IR-luminous galaxi es
in the cluster environmentmay be consistent with the downsi z-
ing scenario, where higher density regions have more evolve d
galaxies and lacks massive star-forming galaxies. In contr ast,
in lower density regions more massive galaxies are still sta r-
forming. However, since the data we have shown is in IR lumi-
nosity, to conclude on this, we need good stellar mass estima te
basedondeepernear-IRdata.
Although a specific mechanism is unclear, the steep faint-
end could also result from the enhanced star-formation in le ss
massive galaxies. In the above scenario, massive galaxies h ave
already ceased their star-formation in the cluster, but les s mas-
sive galaxiesare still formingstars. These less massive ga laxies
may stop star-formation soon to join the faint-end of the red -
sequence(Koyamaet al., 2007).Fig.3.TotalinfraredLFsofclusterRXJ1716.4 +6708atz=0.81
in the solid line, and those of the AKARI NEP deep field in the
dashed line. Overplottedare the LFs of MS1054 from Bai et al.
(2007).
3.2. Total IRLFs
To compare the L8µmLF in Fig.1 to those in the literature, we
needtoconvert L8µmtoLTIR.Weusethethefollowingrelation
byCaputiet al.(2007);
LTIR= 1.91×(νLν8µm)1.06(±55%) (3)
Thisis better tunedfor a similar luminosityrange used here
than the originalrelationby Bavouzetetal. (2008). The con ver-
sion, however, has been the largest source of errors in estim at-
ingLTIRfromL8µm.Caputi etal.(2007)report55%ofdisper-
sion around the relation. It should be kept in mind that the re st-
frame8µm is sensitive to the star-formation activity, but at the
same time, it is where the SED models have strongest discrep-
anciesduetothecomplicatedPAHemissionlines(seeFig.12 of
Caputiet al.,2007; Gotoetal., 2010).
Theestimated LTIRcanbe,then,convertedtoSFRusingthe
followingrelationfor a Salpeter IMF, φ(m)∝m−2.35between
0.1−100M⊙(Kennicutt, 1998).
SFR(M⊙yr−1) = 1.72×10−10LTIR(L⊙) (4)
In Fig.3, we show the LTIRLFs. Symbols are the same as
Fig.1. Inthe topaxis,we showcorrespondingSFR. Overplott ed
asterisks are cluster LF of MS1054 at z=0.83 with ×2 larger
mass by Bai et al. (2007), which show good agreement with
ourLFsofRXJ1716.4 +6708in thesquares.Notethat Bai et al.
(2007) covered only the central region of MS1054 due to the
smaller field of view of the Spitzer. The shape of their LF look s
more similar to our LFs in the highest density bin in Fig.2. A
shift in scale is perhaps due to difference in esimating clus ter
volumes.
Amajordifferenceofourworktothat ofBai etal. (2007)is
that they were not able to compare in detail on the shape of the
LFs between field and cluster regions, due mainly to a smaller
fieldcoverageandlargererrorsonLFs.Theyhadtofixthefain t-
end slope with a local value. The largest source of errors is i nGotoet al.:Environemental dependence of 8 µm luminosity functions ofgalaxies atz ∼0.8 5
converting Spitzer 24 µm flux into 8 µm. Both cluster and field
LFs of this work use L15filter, which measures restframe 8 µm
fluxdirectly,eliminatingthelargestsourceoferrors.Ina ddition,
bothclusterandfiledLFsaremeasuredwithanessentiallysa me
methodology,allowingusa faircomparisonofLFs.
4. Summary
We constructed restframe 8 µm LFs of a massive galaxy cluster
(RXJ1716.4 +6708) and a rarefied field region (the NEP deep
field)at z ∼0.8usingessentially thesame methodanddata from
the AKARI telescope. AKARI’s 15 µm filter nicely covers rest-
frame8µm at z∼0.8,and thuswe do not needa large interpola-
tion based on SED models. AKARI’s wide field of view allows
ustoinvestigatevarietyofclusterenvironmentswith2ord ersof
differenceinlocal galaxydensity.
We found that L∗of the cluster 8 µm LF is smaller than the
field by a factor of 2.4, and the faint-end tail of cluster IR LF s
becomesteeperandsteeperwithincreasinglocalgalaxyden sity.
This difference cannot be explained by a simple infall of fiel d
galaxies into a cluster. Physics that preferentially supre sses IR
luminous galaxes in higer density regions is needed to expla in
theobservedresults.
Acknowledgments
Wethanktheanonymousrefereeformanyinsightfulcomments ,
which significantly improved the paper. We are greateful to
MasayukiTanakaforusefuldiscussion.WethankL.Baiforpr o-
vidingdataforcomparison.
T.G.,Y.K. and H.I. acknowledgesfinancial support from the
JapanSocietyforthePromotionofScience(JSPS)throughJS PS
ResearchFellowshipsforYoungScientists.MIwassupporte dby
the Korea Science and Engineering Foundation(KOSEF) grant
No. 2009-0063616, funded by the Korea government(MEST)”
HML acknowledgesthe supportfrom KASI throughits cooper-
ativefundin2008.
This research is based on the observations with AKARI, a
JAXA projectwiththe participationofESA.
Theauthorswishtorecognizeandacknowledgetheverysig-
nificant cultural role and reverence that the summit of Mauna
Kea has always had within the indigenous Hawaiian commu-
nity. We are most fortunate to have the opportunity to conduc t
observationsfromthissacredmountain.
References
Arnouts S.,et al., 2007, A&A,476, 137
Babbedge T.S.R.,et al., 2006, MNRAS,370, 1159
Bai L.,Rieke G.H.,Rieke M. J.,Christlein D.,Zabludoff A.I .,2009, ApJ,693,
1840
Bai L.,Marcillac D.,Rieke G.H.,etal., 2007, ApJ,664, 181
Bai L.,Rieke G.H.,Rieke M.J.,etal., 2006, ApJ,639, 827
Bavouzet N., Dole H., Le Floc’h E., Caputi K. I., Lagache G., K ochanek C. S.,
2008, A&A,479, 83
Brandl B. R.,Bernard-Salas J.,Spoon H.W.W.,et al., 2006, A pJ,653, 1129
Calzetti D.,Kennicutt R.C.,Engelbracht C.W.,etal., 2007 , ApJ,666, 870
Caputi K.I.,etal., 2007, ApJ,660, 97
Dale D.A.,Helou G.,2002, ApJ,576, 159
Desert F.-X.,Boulanger F.,Puget J.L.,1990, A&A,237, 215
Elbaz D.,DaddiE.,LeBorgneD.,et al., 2007, A&A,468, 33
G´ omezP.L.,Nichol R.C., Miller C. J.,et al., 2003, ApJ,584 , 210
Goto, T.,Takagi, T.,Matsuhara, H.,etal., 2010, submitted to A&A
Goto T.,2005, MNRAS,360, 322
Goto T.,Postman M.,Cross N.J.G.,et al.,2005, ApJ, 621,188
Goto T.,Yagi M.,Tanaka M.,Okamura S.,2004, MNRAS,348, 515
Goto T.,Yamauchi C.,Fujita Y.,etal., 2003, MNRAS,346, 601Goto T.,Okamura S.,Sekiguchi M.,etal., 2003, PASJ,55,757
Huang J.-S.,et al.,2007, ApJ, 664,840
Ilbert O.,et al., 2006, A&A,457, 841
ImaiK.,Matsuhara H.,Oyabu S.et al., 2007, AJ,133,2418
Kennicutt R.C.,Jr., 1998, ARA&A,36,189
Kodama T., Balogh M. L., Smail I., Bower R. G., Nakata F., 2004 , MNRAS,
354, 1103
Komatsu E.,Dunkley J.,Nolta M.R.,et al.,2009, ApJS, 180,3 30
Koyama Y.,KodamaT.,Shimasaku K.,et al., 2008, MNRAS,391, 1758
KoyamaY.,KodamaT.,TanakaM.,Shimasaku K.,OkamuraS.,20 07,MNRAS,
382, 1719
Lagache G.,Dole H.,Puget J.-L.,2003, MNRAS, 338, 555
Matsuhara H.,etal., 2007, PASJ,59,543
Matsuhara H.,etal., 2006, PASJ,58,673
Murakami H.,et al.,2007, PASJ,59, 369
Pearson C.,et al. 2009, A&A,submitted
Rieke G.H.,Alonso-Herrero A.,Weiner B. J.,etal., 2009, Ap J,692, 556
Shim H.,etal. 2010, ApJ, submitted
Sutherland W.,Saunders W.,1992, MNRAS, 259, 413
Takeuchi T.T.,Buat V.,Burgarella D.,2005, A&A,440, L17
Tanaka M.,Finoguenov A.,Kodama T.,et al.,2008, A&A,489, 5 71
Tanaka M.,KodamaT.,Arimoto N.,etal., 2005, MNRAS, 362,26 8
Tanaka M., Goto T., Okamura S., Shimasaku K., Brinkmann J., 2 004, AJ, 128,
2677
Tran,K.,et al. 2010, ApJin press,arXiv:0909.4079
WadaT.,etal., 2008, PASJ,60,517