arXiv:1001.0007v1  [astro-ph.CO]  30 Dec 2009Cosmicstarformation history
revealedby the AKARI
& Spatially-resolvedspectroscopyofan E+A(Post-starbur st)system
Tomotsugu GOTO∗, the AKARINEPDteam†,M.Yagi∗∗andC.Yamauchi†
∗InstituteforAstronomy,Universityof Hawaii,2680Woodla wnDrive, Honolulu,HI,96822,USA
†JapanAerospaceExplorationAgency,Sagamihara,Kanagawa 229-8510,Japan
∗∗NationalAstronomicalObservatory,2-21-1Osawa,Mitaka, Tokyo,181-8588,Japan
Abstract. We reveal cosmic star-formation history obscured by dust us ing deep infrared observa-
tionwiththeAKARI.Acontinuousfiltercoverageinthemid-I Rwavelength(2.4,3.2,4.1,7,9,11,
15, 18, and 24 µm) by the AKARI satellite allows us to estimate restframe 8 µm and 12 µm lumi-
nositieswithoutusingalargeextrapolationbasedonaSEDfi t,whichwasthelargestuncertaintyin
previouswork. We found that restframe 8 µm (0.38<z<2.2), 12µm (0.15<z<1.16), and total
infrared (TIR) luminosity functions (LFs) (0 .2<z<1.6) constructed from the AKARI NEP deep
data, show a continuous and strong evolution toward higher r edshift. In terms of cosmic infrared
luminosity density ( ΩIR), which was obtained by integrating analytic fits to the LFs, we found a
goodagreementwithpreviousworkat z<1.2,withΩIR∝(1+z)4.4±1.0.Whenweseparatecontri-
butionsto ΩIRby LIRGs and ULIRGs, we foundmore IR luminoussourcesare inc reasinglymore
importantathigherredshift.WefoundthattheULIRG(LIRG) contributionincreasesbyafactorof
10(1.8)from z=0.35toz=1.4.
Keywords: galaxies:evolution,galaxies:starburst
PACS:98.70.Lt
Introduction .Revealingthecosmicstarformationhistoryisoneofthemaj orgoals
of the observational astronomy. However, UV/optical estim ation only provides us with
alowerlimitofthestarformationrate(SFR) duetotheobscu rationbydust.Astraight-
forward way to overcome this problem is to observe in infrare d, which can capture the
starformation activityinvisiblein the UV. The superb sens itivitiesofrecently launched
SpitzerandAKARI satellitescan revolutionizethefield.
However,most of theSpitzer work relied on a large extrapola tionfrom 24 µm flux to
estimate the 8, 12 µm or total infrared (TIR) luminosity, due to the limited numb er of
mid-IR filters. AKARI has continuous filter coverage across t he mid-IR wavelengths,
thus, allows us to estimate mid-IR luminosity without using a largek-correction based
on the SED models, eliminating the largest uncertainty in pr evious work. By taking
advantage of this, we present the restframe 8, 12 µm TIR LFs, and thereby the cosmic
starformationhistoryderivedfrom theseusingtheAKARINE P-Deep data.
Data&Analysis .TheAKARIhasobservedtheNEPdeepfield(0.4deg2)in9filters
(N2,N3,N4,S7,S9W,S11,L15,L18WandL24) to the depths of 14.2, 11.0, 8.0, 48, 58,
71, 117, 121 and 275 µJy (5σ)[14]. This region is also observed in BVRi′z′(Subaru),
u′(CFHT), FUV,NUV(GALEX), and J,Ks(KPNO2m), with which we computed
photo-zwithΔz
1+z=0.043. Objects which are better fit with a QSO template are re movedFIGURE 1. (left) Restframe 8 µm LFs. The blue diamonds, purple triangles, red squares, and orange
crosses show the 8 µm LFs at 0 .38<z<0.58,0.65<z<0.90,1.1<z<1.4, and 1.8<z<2.2,
respectively. The dotted lines show analytical fits with a do uble-power law. Vertical arrows show the
8µm luminosity corresponding to the flux limit at the central re dshift in each redshift bin. Overplotted
are Babbedge et al. [1] in the pink dash-dotted lines, Caputi et al. [2] in the cyan dash-dotted lines,
and Huang et al. [6] in the green dash-dotted lines. AGNs are e xcluded from the sample. (middle)
Restframe 12 µm LFs. The blue diamonds, purple triangles, and red squares s how the 12 µm LFs at
0.15<z<0.35,0.38<z<0.62, and 0 .84<z<1.16, respectively. Overplotted are Pérez-González
et al. [11] at z=0.3,0.5and 0.9 in the cyan dash-dottedlines, and Rush, Mal kan, & Spinoglio [12] at z=0
inthegreendash-dottedlines. (right)TIRLFs.
from the analysis. We compute LFs using the 1/ Vmaxmethod. Data are used to 5 σwith
completeness correction. Errors of the LFs are from 1000 rea lization of Monte Carlo
simulation.
8µm LF.Monochromatic 8 µm luminosity ( L8µm) is known to correlate well with
the TIR luminosity [1, 6], especially for star-forming gala xies because the rest-frame
8µmfluxaredominatedbyprominentPAHfeaturessuchasat6.2,7 .7and8.6 µm.The
leftpanelofFig.1showsastrongevoltuionof8 µmLFs.Overplottedpreviousworkhad
torelyonSEDmodelstoestimate L8µmfromtheSpitzer S24µmintheMIRwavelengths
whereSEDmodelingisdifficultduetothecomplicatedPAHemi ssions.Here,AKARI’s
mid-IR bands are advantageous in directly observing redshi fted restframe 8 µm flux in
one of the AKARI’s filters, leading to more reliable measurem ent of 8µm LFs without
uncertaintyfromtheSED modeling.
12µm LF.12µm luminosity ( L12µm) represents mid-IR continuum, and known to
correlate closely with TIR luminosity [11]. The middle pane l of Fig.1 shows a strong
evoltuion of 12 µm LFs. Here the agreement with previous work is better becaus e (i)
12µm continuum is easier to be modeled, and (ii) the Spitzer also captures restframe
12µm inS24µmat z=1.
TIRLF.Lastly,weshowtheTIRLFsintherightpanelofFig.1.Weused Lagache,
Dole, & Puget [8]’s SED templates to fit the photometry using t he AKARI bands
at>6µm (S7,S9W,S11,L15,L18WandL24). The TIR LFs show a strong evolution
comparedto localLFs. At 0 .25<z<1.3,L∗
TIRevolvesas ∝(1+z)4.1±0.4.FIGURE2. Evolutionof TIRluminositydensitybasedon TIRLFs (redcir cles),8µmLFs (stars), and
12µm LFs (filled triangles). The blue open squares and orange fill ed squares are for LIRG and ULIRGs
only,alsobasedonour LTIRLFs.Overplotteddot-dashedlinesareestimatesfromtheli terature:LeFloc’h
et al. [9], Magnelli et al. [10] , Pérez-González et al. [11], Caputi et al. [2], and Babbedge et al. [1] are
in cyan, yellow, green, navy, and pink, respectively. The pu rple dash-dotted line shows UV estimate by
Schiminovichet al.[13].Thepinkdashedlineshowsthe tota lestimateofIR(TIRLF)andUV [13].
Cosmic star formation history .We fit LFs in Fig.1 with a double-power law, then
integrate to estimate total infrared luminosity density at various z. The restframe 8
and 12µm LFs are converted to LTIRusing [11, 2] before integration. The resulting
evolution of the TIR density is shown in Fig.2. The right axis shows the star formation
densityassumingKennicutt[7].We obtain ΩIR(z)∝(1+z)4.4±1.0. Comparisonto ΩUV
[13] suggests that ΩTIRexplains 70% of Ωtotalatz=0.25, and that by z=1.3, 90% of
the cosmic SFD is explained by the infrared. This implies tha tΩTIRprovides good
approximationofthe Ωtotalatz>1.
In Fig.2, we also show the contributions to ΩTIRfrom LIRGs and ULIRGs. From
z=0.35 to z=1.4,ΩIRby LIRGs increases by a factor of ∼1.6, andΩIRby ULIRGs
increases byafactorof ∼10. Moredetailsarein Gotoet al. [3].
Spatially-Resolved Spectroscopy of an E+A (post-starburs t) System .We per-
formed a spatially-resolved medium resolution long-slit s pectroscopy of a nearby E+A
(post-starburst) galaxy system with FOCAS/Subaru [4]. Thi s E+A galaxy has an obvi-
ous companion galaxy 14kpc in front (Fig.3, left) with the ve locity difference of 61.8
km/s.
WefoundthatH δequivalentwidth(EW)oftheE+Agalaxyisgreaterthan7Å gal axy
wide (8.5 kpc) with no significant spatial variation. We dete cted a rotational velocity in
the companion galaxy of >175km/s. The progenitor of the companion may have beenFIGURE 3. (left) The SDSS g,r,i-composite image of the J1613+5103. The long-slit position s are
overlayed.The E+A galaxy is to the right (west), with bluer c olour. The companion galaxy is to the left
(east). (right) H δEW is plotted against D4000. The diamonds and triangles are f or the E+A core/north
spectra, respectively. The squares and crosses are for the c ompanion galaxy’s core/north spectra. Gray
lines are population synthesis models with 5-100% delta bur st population added to the 10G-year-old
exponentially-decaying( τ=1Gyr)underlyingstellarpopulation.SalpeterIMFandmet allicityof Z=0.008
areassumed.Onthe models,burstagesof0.1,0.25,0.5and2 G yraremarkedwiththefilled circles.
a rotationally-supported, but yet passive S0 galaxy. The ag e of the E+A galaxy after
quenching the star formation is estimated to be 100-500Myr, with its centre having
slightly younger stellar population. The companion galaxy is estimated to have older
stellarpopulationof >2 Gyrs ofagewithnosignificantspatialvariation(Fig.3, ri ght).
Thesefindingsareinconsistentwithasimplepicturewheret hedynamicalinteraction
createsinfallofthegasreservoirthatcausesthecentrals tarburst/post-starburst.Instead,
ourresultspresentanimportantexamplewherethegalaxy-g alaxyinteractioncantrigger
agalaxy-widepost-starburstphenomena.
REFERENCES
1. BabbedgeT.S.R., et al.,2006,MNRAS, 370,1159
2. CaputiK.I.,et al.,2007,ApJ,660,97
3. GotoT.,et al. 2010,A&AAKARI specialissue
4. GotoT.,YagiM.,YamauchiC., 2008,MNRAS, 391,700
5. HopkinsA.M.,ConnollyA. J.,HaarsmaD. B.,CramL. E.,200 1,AJ, 122,288
6. HuangJ.-S.,et al.,2007,ApJ, 664,840
7. KennicuttR. C.,Jr., 1998,ARA&A,36,189
8. LagacheG., DoleH.,PugetJ.-L.,2003,MNRAS, 338,555
9. LeFloc’hE.,etal., 2005,ApJ,632,169
10. MagnelliB., et al.2009,A&A,496,57
11. Pérez-GonzálezP. G.,etal., 2005,ApJ,630,82
12. RushB., MalkanM. A.,SpinoglioL.,1993,ApJS,89,1
13. SchiminovichD.,et al.,2005,ApJ, 619,L47
14. Wada T.,et al.,2008,PASJ, 60,517