Galactic Archaeology with SuMIRe PFS

Wako Aoki (NAOJ)Masashi Chiba (Tohoku University)

Nobuo Arimoto (NAOJ)

Galaxy formation in CDM paradim

Bright parts arecentrally concentrated

Galactic dark haloAssembly of CDM

Main issues in GA1. Merging history of the Milky Way?

Formation process of each Galactic component Is Mily Way different from M31? If so, why?

2. Nature of galactic dark matterMissing satellites problem?

Properties of luminous satellites? How many there?

Dark matter profiles? Cuspy or cored?

All are recorded in old stellar populations

(Belokurov+ 2006)

Substructures in the MW halo(tidal debris of building blocks)

SDSS

Metal-poor stars inangular momentum space（Hipparcos sample）• measurement error of a few 100 (kpc km/s) smears out substructures

clump

Astrometry with Gaiaprecise distances and proper motions• resolves each of substructures（Building blocks of the stellar halo)

Lz

L

Substructures in angular momentum space

Galactic astronomy through resolved stars

Photometry : mag., color (→ color-mag. diagram)

Spectroscopy : metallicity (→ age), Vrad (kinematics)abundance pattern (→ SF & chemical evol.)

Astrometry : proper motion, distance (→ 6d phase space)

Structure, dynamics, star formationand chemical evolution

⇒ galaxy formation and evolution

PFS

HSC (+SDSS)

Gaia

Stream

Northern Spur

G1 ClumpM32

Substructures in the M31 halo

Ferguson+02

180 pointings

HSC photometric survey of M31’s halousing optimized NB515 filter (g<22.5)

Metallicities and RVs of substructuresand satellites with PFS

Tanaka+10

ITRGB =20.5

Dark Matter distribution in dwarf galaxies around the Milky Way

● Radial velocity and proper motion measurements for a large sample of dwarf galaxy stars

　~ 0.03mas/yr for stars of 10km/s at D=70kpc↓

● Mass distribution of the central regions of dwarf galaxies

　Mass within 300pc / Cuspy or cored ？↓

● Nature of dark matter (CDM or WDM ?)● Relation with stellar chemical composition and star

forming history

Velocity dispersion of dwarf galaxy starsGilmore et al. 2007, ApJ, 663, 948

Globular cluster

velo

c ity

di s

pers

ion

(old data)

M/L ratio for Mass enclosed within stellar Extent (~4x107Msun) in dSph galaxies

Luminous→←faint

mas

s/lu

min

osit

y

Ultra-faintDwarf galaxies

Ultra Faint dwarf Galaxies (UFdG) studied with Suprime-Cam

Ursa Major I Bootes I Canes Venatici I96kpc 60kpc 220kpc

(Okamoto+2008)

Mass within 300pc in dwarf galaxies

Strigari et al. 2008

Ultra-faint dwarfs

How is the mass distribution in the center of dark matter halo? cusp or core?

Comparisons with GA proposal for WFMOS

•GA proposal for WFMOS:“Chemical tagging” for halo/thick disk stars observed by a very wide survey (with high resolution)→comprehensive, but time consuming (~500

nights?)

•Our current proposal for PFS:spectroscopic studies for Galactic

substructures (streams) and dwarf galaxies→more feasible→future extension by high resolution mode?

Ability to measure radial velocities and [Fe/H] for many stars at the same time

•Accurate determination of velocity dispersion of dSphs and streams, i.e. ΔVrad <2 km/s•Best synergy with Gaia, i.e., 15<V<20•Enable to observe M31 stars with ITRGB =20.5,or horizontal branch stars in dwarf galaxies

(V~22 mag.) i.e., Vlim ~ 22

Requested performance of PFS

Estimate of velocity accuracy for R=3000/5000

Simulation using synthetic spectra for 5000-6000A

S/N R=3000 R=500010 4.2km/s 1.7km/s20 2.2km/s 1.0km/s50 1.0km/s 0.4km/s

A preliminary result

“template” (S/N=100)“object” (S/N=20)

Ca II HK as [Fe/H] indicator

KP([Fe/H],B-V)

Valid for both dwarfs and giantswith -4.0<[Fe/H]<0.5

Beerrs+99

[Fe/H]=-4.0

[Fe/H]=-0.5

Requested performance of PFS1. Ability to measure RVs and [Fe/H] for many stars at the same time

-R=3000-5000-coverage of full optical range (3900-9000A)-number of fibers?

2. Ability to follow up high-res. spectroscopy for reasonable number of stars at the same time

R=2-40000, λ<9000A, a few 100 fibers, V<17