Early Evolution Early Evolution of Massive of Massive GalaxiesGalaxies

Romeel DavéRomeel DavéKristian FinlatorKristian Finlator

Ben D. OppenheimerBen D. OppenheimerUniversity of ArizonaUniversity of Arizona

Early (z>2) Massive Early (z>2) Massive Galaxies: Questions for Galaxies: Questions for TheoryTheory

How do galaxies obtain fuel?How do galaxies obtain fuel? Can large stellar masses be produced at early Can large stellar masses be produced at early

times?times? How do they become so metal enriched?How do they become so metal enriched? What are their SFHs as a function of mass What are their SFHs as a function of mass

and environment?and environment?

Understanding high-z galaxy formation Understanding high-z galaxy formation provides the initial conditions for studying provides the initial conditions for studying new issues arising at low-z.new issues arising at low-z.

Physics of Early Massive Physics of Early Massive Galaxies in SimulationsGalaxies in Simulations Cold, smooth accretion.Cold, smooth accretion.

SFR SFR M M** (with little scatter). (with little scatter).

LBGs are LBGs are massivemassive; age ~ ½ t; age ~ ½ tHubbleHubble..

[Merging plays minor role during SF’ing phase.][Merging plays minor role during SF’ing phase.]

Ubiquitous and highly-loaded outflows.Ubiquitous and highly-loaded outflows. Suppress overcooling.Suppress overcooling. Enrich IGM.Enrich IGM. Establish mass-metallicity relation.Establish mass-metallicity relation. Keep galaxies gas-rich.Keep galaxies gas-rich.

Is this plausible, and is this sufficient?Is this plausible, and is this sufficient?

Cold Accretion: Cold Accretion: The Gadget The Gadget storystory Cold mode: No virial Cold mode: No virial

shock; shock; ttdyndyn-limited.-limited. Hot mode: Hot mode: ttcoolcool-limited.-limited. Keres et al: Transition Keres et al: Transition

MMhalohalo~10~1011.411.4MM, hot mode , hot mode dominates at z<2.dominates at z<2.

Gadget’s entropy-Gadget’s entropy-conservative SPH greatly conservative SPH greatly reduces hot modereduces hot mode..

Suppressing hot mode Suppressing hot mode alone unlikely to solve alone unlikely to solve red & dead problem.red & dead problem.

PTreeSPHGadget-2

Rapid accretion Rapid accretion Rapid Rapid early growthearly growth WMAP-3 Gadget WMAP-3 Gadget

simulations vs. simulations vs. z>5.5 z>5.5 galaxiesgalaxies w/IRAC data. w/IRAC data.

In most cases, simul-In most cases, simul-ated galaxies provide ated galaxies provide good fit (good fit (

22<1).<1). Best-fit galaxies…Best-fit galaxies…

are fairly are fairly massivemassive, , have have older starsolder stars, , show show 4000Å break4000Å break..

Typical SFH is Typical SFH is constantly rising.constantly rising.

Finlator, RD, Oppenheimer 06

log M*=8.7

log M*=10.1

log M*=9.2

log M*=10.1

log M*=9.4

log M*=9.3

Quiescent SFH? Quiescent SFH?

SFR SFR M M**: Cold mode.: Cold mode. Mergers/bursts Mergers/bursts notnot a a

significant driver of SF.significant driver of SF. True? At z~0-1, yes. At True? At z~0-1, yes. At

z~2, more confusing.z~2, more confusing. Amplitude too low?Amplitude too low?

Noeske et al. 2006Erb etal 06

The Impact of OutflowsThe Impact of Outflows Needed to regulate SF, Needed to regulate SF,

alleviate overcooling.alleviate overcooling. Common/expected in Common/expected in

z~2+ galaxies.z~2+ galaxies. Defined by outflow Defined by outflow

speed (speed (vvww) and mass ) and mass loading factor (loading factor ().).

Local starbursts: Local starbursts: vvwwvvcirccirc Momentum-Momentum-driven windsdriven winds??

How to constrain? How to constrain? Track the metals!Track the metals!

Martin 2005

Erb et al 2006

Outflows and IGM MetalsOutflows and IGM Metals Gadget + winds from Gadget + winds from

star forming galaxies.star forming galaxies. Momentum-driven Momentum-driven

winds (winds (vvww,,1/1/) vs. ) vs. vvww, , = constant = constant

Must inject metals Must inject metals early: early: High High ..

Must not overheat IGM: Must not overheat IGM: Low vLow vww..

CIV POD CIV POD spatial spatial distribution of metals; distribution of metals; constrains wind model.constrains wind model.

Oppenheimer & RD 06

Mass-Metallicity RelationMass-Metallicity Relation Galaxies enrich Galaxies enrich

early, and early, and MMZZ0.30.3 established early on.established early on.

Why so Why so muchmuch evol- evol-ution observed from ution observed from z=2z=20?0?

Outflows affect Outflows affect (set?) M-Z relation.(set?) M-Z relation.

Momentum-driven Momentum-driven wind scalings work!wind scalings work!

In addition to metals, In addition to metals, winds also keeps winds also keeps galaxies gas-rich.galaxies gas-rich.

Luminosity functionsLuminosity functions

z~6 UVLF seems to z~6 UVLF seems to indicate some SF indicate some SF suppression required: suppression required: OutflowsOutflows??

z~2-4 rest-optical LF’s z~2-4 rest-optical LF’s agree with data.agree with data.

Outflows affect faint Outflows affect faint end of LF: prefers end of LF: prefers higher higher in small in small galaxiesgalaxies..

Data: Bouwens etel 06Models: =0.3, 8=0.9

RD, Finlator, Oppenheimer 06

Baryon Baryon fractionsfractions Winds throw many Winds throw many

baryons outside of halos: baryons outside of halos: Is this real?Is this real?

By z~0, MW-sized halos By z~0, MW-sized halos have only ~½ baryons: have only ~½ baryons: They really are “missing”!They really are “missing”!

Evidence for baryons, not Evidence for baryons, not just metals, being ejected just metals, being ejected at z~3: DLA kinematics at z~3: DLA kinematics (Sungryon Kim et al, in (Sungryon Kim et al, in prep).prep).

z=5 z=2

z=0no winds, 8e-5

constant winds, 0.037

MD winds, 0.51

SummarySummary Galaxy formation at high-z seems to be Galaxy formation at high-z seems to be

driven by (at least):driven by (at least): Cold accretionCold accretion Strong and ubiquitous outflowsStrong and ubiquitous outflows

Direct observational evidence of these are Direct observational evidence of these are sparse, but many indirect constraints.sparse, but many indirect constraints.

Not clear that it’s solved; simulations may Not clear that it’s solved; simulations may be limited by numerics, or physical be limited by numerics, or physical processes may be absent.processes may be absent.

High-z galaxy formation presents own High-z galaxy formation presents own challenges apparently separate from low-z.challenges apparently separate from low-z.