Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.mrao.cam.ac.uk/galevol/Cambridge2013_Yabe_v2.pdf Äàòà èçìåíåíèÿ: Mon Oct 7 11:48:03 2013 Äàòà èíäåêñèðîâàíèÿ: Fri Feb 28 00:10:41 2014 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: star

"Galaxy Evolution over Five Decades", 3-6 Sep. 2013, Cavendish Laboratory, University of Cambridge

NIR Spectroscopy of star-forming galaxies at z~1.4 with Subaru/FMOS
(Yabe et al. 2012, PASJ, 64, 60, Yabe et al. 2013, MNRAS submitted)

Kiyoto Yabe
(National Astronomical Observatory of Japan)

Collaborators: Kouji Ohta, Fumihide Iwamuro (Kyoto Univ.), Suraphong Yuma (Tokyo Univ.), Masayuki Akiyama (Tohoku Univ.), Naoyuki Tamura (Tokyo Univ.), Gavin Dalton (Oxford Univ.), Japan-UK FMOS GTO Team, and John Silverman (Tokyo Univ.)


Introduction: Mass-Metallicity Relation up to z~3
· Gas-phase metallicity (hereafter, metallicity) Metallicity is a key to understand the galaxy evolution as an independent ·
tracer of the past star-formation activity. Correlation between stellar mass (luminosity) and metallicity Firstly reported by Lequeux+1979 for nearby Irr, blue compact galaxies Massive (bright) galaxies tend to show larger metallicities Now mass(luminosity)-metallicity (MZ) relation is well established at z~0 What is the origin of MZ relation?
Tremonti+2004

· MZ relation up to z~3 (e.g., Erb+06, Maiolino+08) Evolution of the MZ relation from z~3 to z~0? Still controversy as to the MZ relation at z~2 We need larger sample at z>1

metallicity

stellar mass

Maiolino+08


ines) 2.3AV (SED) according to Calzetti et al. (2000)]. We (1) - 0.017m3 + 0.046m4 , checked that the inclusion of this effect would have little effect e final relations and on the conclusions of this paper. where m = log(M ) - 10 in solar units. b et al. (2006) have observed a large sample of 91 galaxies at We have computed the median metallicity of SDSS galaxies for .2. Metallicities have been measured only on average spectra different values of SFR. Median has The , wh relation at z~0.1 has a ed according to MMZich has the results of mixing galaxies ofscatter (e.g., Tremonti+04)been computed in bins of mass and SFR of ent SFRs.What this problem, nparameters can explain this0.15 dex width in both quantities. On average, each bin Despite physical o systematic differences in contains 760 scatter?only bins containing more than 50 galaxgalaxies, and llicity are detected with respect to the other galaxies measured ies are (Ellison+2008), idually, and theSFR al. (2006) galaxies are included in the Erb et (Mannucci+2010), specific SFR considered. The left-hand panel of Fig. 1 also shows these median metallicities as a function of M . It is evident t at a sy tem redshift sample,half light radius (Ellison+2008), galaxy interaction (Rupke+2008),hand ssoalthough without binning them with the rest atic segregation in SFR is present in the data. While galaxies with galaxies. high and M > 10.9] show no correlation between metallicity The intrinsic scatter of the MZ relation M [log(its) dependence of physical and SFR, at low M more active galaxies also show lower metallic= 3­4parameters is still unknown at high-z ity. The same systematic dependence of metallicity on SFR can be seen nificant saWe of 16 galaxiesgeedshiftbetweeat high aredshift in the right-hand panel of Fig. 1, where metallicity is plotted as mple need lar at r sample n 3and 4w s a function of SFR for different values of mass. Galaxies with high ved by Maiolino et al. (2008) and Mannucci et al. (2009) for the

Introduction: Scatter of the Mass-Metallicity Relation
· · · ·

Fundamental Metallicity Relat
Mannucci+2010 Mannucci+2010

on

· Mannucci+10 suggested that the scatter of th They proposed the Fundamental Metallici They claimed that this FMR are unchanged

Mannucci+2010

lower SFR higherSFR

metallicity metallicity

stella mas stellarrmasss

SFR SFR


Introduction: FMOS on the Subaru Telescope
· What's FMOS (Fibre Multi-Object Spectrograph)? Fibre-fed NIR multi-object spectrograph on the Subaru Telescope Collaboration among Japan, UK, and Australia Multi-object spectrograph in NIR (0.9-1.8µm) w/ 400 fibers and FoV of 30' Low Resolution (LR; R~650) and High Resolution (HR; R~3000) mode Details are in Kimura et al. 2010, PASJ, 62, 1135 We conduct large NIR spectroscopic surveys with FMOS Initial results were published by Yabe et al. 2012, PASJ, 64, 60 and the
subsequent results will appear in Yabe et al. 2013, MNRAS, submitted
Two spectrographs Prime Focus Unit

Fibre cable

Fiber positioner on prime focus

FMOS on the Subaru Telescope

Optical design of FMOS including OH-mask mirror


Sample Selection and Observations:

· Target Sample Field : SXDS/UDS (effective area~0.7 deg2) K-selected catalogue with zphot, M*, and other properties 1.2109.5 Msun, F(H)exp>5.0x10-17 cgs · Observations Mainly FMOS/GTOs in 2010-2011 Typical exposure time is 3-4 hrs per FoV · Data Reduction FMOS standard reduction pipeline FIBRE-pac (see Iwamuro+2012) Emission line fitting including the FMOS OH-mask effects (see Yabe+2012) H of ~340 objects was detected
1e-14

109.5 Msun
Expected Ha Flux (cgs) Expected H flux (erg/s/cm2)

1e-15

Targeted Objects

z=1.442 F(H)=1.1x10-16 F([NII])=3.8x10-17 FWHM=390 km/s M*=4.6x1010 Msun 12+log(O/H)=8.644

5 hrs integration H [NII]

~340 H detection!

1e-16

5x10-17 cgs
1e-17

[NII]

1.2 1e-18 1e+09 1e+10 Stellar Mass (M
sun

1e+11 )

1e+12

Stellar Mass (Msun)


AGN contamination:
· · · · ·
AGN diagnostics by using the BPT diagram ([NII]/H vs. [OIII]/H) Most objects are placed in the SF region in the BPT diagram 21 objects are AGN candidates (BPT, extremely large [NII]/H ratio and line width) Stacking analysis shows that our sample is in the SF region on average Possible AGN candidates are excluded from the sample

BPT

: Stacking analysis

· Some objects fall in the "Composite

region" of the BPT diagram · The data points from the stacking analysis show higher [OIII]/H or [NII]/H ratio than the local SF sequence Effect of hidden AGN? Effect of shock? Different ISM conditions? · See theoretical works by Kewley, Maier, Yabe, et al. 2013 ApJ, 774, L10 · Further follow-ups would be desirable
Kewley+01


Mass-Metallicity Relation at z~1.4:
· · · ·
Possible AGN candidates are excluded by using BPT diagram 12+log(O/H) from [NII]/H line ratio (N2 method; Pettini & Pagel 2004) No significant [NII] emission (S/N<3.0) from ~70% Stacking analysis Massive galaxies tend to show larger metallicity (MZ relation)

· A large scatter (~0.1 dex) in the
MZ relation at z~1.4 · Comparison to previous works up to z~3 IMF and metallicity calibration conversion for the fair comparison Our results at z~1.4 are between those at z~0.8 and z~2.2 · The MZ relation appears to evolve smoothly from z~3 to z~0

Metallicity calibration and IMF of other works are all the same as ours


Comparison to Theoretical Models:
· Comparison with recent theoretical predictions (DavÈ et al. 2011) N-body + SPH cosmological simulations 4 wind models (no wind; constant wind; slow wind; mass dependent
wind) implemented
For theoretical models at z~1.4, we average the results at z=1 and z=2 presented by DavÈ et al. 2012.

· DavÈ et al. implemented the

Metallicity calibration and IMF of other works are all the same as ours

following wind models: No wind (nw) model Slow wind (sw) :dMwind/ dt=2xSFR, vwind=340 km/s Constant wind (cw) : dMwind/ dt=2xSFR, vwind=680 km/s Momentum conserved wind (vzw) : the wind velocity depends on the velocity dispersion (=mass) · Our result agrees with cw or vzw model · Ubiquitous, strong galactic winds have been reported at high redshift (e.g., Weiner+2009)


Second Parameter Dependence:
· There exists a large scatter (~0.1 dex) in the MZ relation at z~1.4 · Dependence of the MZ relation on other parameters SFR : derived from H luminosity corrected for the dust extinction The dependency of SFR on the MZ relation is not clear But averaged points are close to the local FMR by Mannucci+10 Our sample has narrow range of SFR parameter?
extinction corrected SFR(H)


Second Parameter Dependence:
· Dependence of the MZ relation on the galaxy size and color We take deconvolved-half light radius (R50) as the size (from K-band image) Galaxies with smaller R50 tend to show higher metallicities Galaxies with redder R-H (and B-R) color tend to show higher metallicities Small galaxies may have high surface gas density, and star-formation and
metal enrichment might occur effectively, leading to the higher metallicity
R-H Color Half Light Radius (r50)


Morphology Dependence:

· Morphology provides information about structural properties of galaxies · 33 objects in the CANDELS/UDS field are observed with FMOS · For these objects, the morphology can be examined as well as metallicity
Color composites with HST/ACS+WFC3 images

· Diffuse and disk
dominated galaxies tend to show lower metallicity than compact and bulge dominated galaxies?

Compact? Bulge Dominated?

Diffuse? Disk Dominated?


Morphology Dependence:

· Morphology provides information about structural properties of galaxies · 33 objects in the CANDELS/UDS field are observed with FMOS · For these objects, the morphology can be examined as well as metallicity
Color composites with HST/ACS+WFC3 images

· Diffuse and disk
dominated galaxies tend to show lower metallicity than compact and bulge dominated galaxies?

Compact? Bulge Dominated? Compact?

· CAS parameterization
Diffuse? Diffuse? Disk Dominated?

(Conselice+03) for compactness (= 5log (r80/ r20)) shows that galaxies with higher compactness show higher metallicity at fixed mass, which is well consistent with eye inspection


Summary:
· · · · · · ·
We observed star-forming galaxies at z~1.4 are measured with Subaru/FMOS We detected H line from ~340 objects with significance of S/N>3 Gas-phase metallicity is derived from [NII]/H line ratio The mass-metallicity (MZ) relation at z~1.4 with the largest sample ever By comparing previous results: The MZ relation evolves smoothly from z~3 to z~0 They agree with theoretical models with strong galactic winds The MZ relation at z~1.4 may have a large scatter of ~0.1 dex We examined the dependence of the MZ relation on other physical properties. Not so clear trend for SFR: Disagrees with that at z~0.1 by Mannucci+10 Clear trend for size: Galaxies with larger R50 tend to show lower metallicity Clear trend for color: Bluer galaxies tend to show lower metallicity Our results may show the morphology dependence Bulge-dominated galaxies located in the upper region on the MZ relation? Disk-dominated galaxies located in the lower region on the MZ relation? Scenarios of differing star-formation efficiency would be plausible? Larger sample with wider parameter range may be needed for clearer trends.

· · ·