The influence of environment on galaxy populations Michael Balogh University of Waterloo, Canada.

The influence of environment The influence of environment on galaxy populationson galaxy populations

Michael BaloghUniversity of Waterloo, Canada

Outline

• Low redshift– Simple trends encompass most of

what we know of as environmental influences

• Models: what works and what doesn’t

• Redshift evolution• The future: what’s next?

The influence of The influence of environment environment on galaxy on galaxy populationspopulations

Populations• Current star formation

rate• Recent star formation • Stellar mass (average

SFR)• Morphology (of stars,

neutral gas, ionized gas)

• AGN • Gas content

Environment• Mass of dark

matter halo• Position within

halo• Local density• Large-scale

density

The The influenceinfluence of environment of environment on galaxy populationson galaxy populations

• Nature vs. nurture?• Entangled in current models

– Gas accretion, merger, and feedback history scale with halo mass.

– No longer the right question?• A better question: what physics

operates in haloes of a given mass, at a given epoch?– Today’s population is the result of

different environments at different epochs: cannot try to isolate one mechanism as responsible for the observed trends.

The local Universe

Colour-magnitude distribution

• Nearby galaxies seem to fall into two surprisingly well-defined, smoothly varying distributions.

• Colour, luminosity, concentration, star formation rate

Blanton et al. 2004

Colour-magnitude distribution

• Colour distribution in 0.5 mag bins can be fit with two Gaussians

• Mean and dispersion of each distribution depends strongly on luminosity

• Dispersion includes variation in dust, metallicity, SF history, and photometric errors

• At bright magnitudes, significant fraction of “blue” population “contaminates” red: c.f. talk by Wolf.

(u-r)

Bright

Faint

Baldry et al. 2003

• Fraction of red galaxies depends strongly on density. This is the primary influence of environment on the colour distribution.

• Mean colours depend weakly on environment: transitions between two populations must be rapid (or rare at the present day)

Balogh et al. 2004

• Fraction of red galaxies depends strongly on density. This is the primary influence of environment on the colour distribution.

• Mean colours depend weakly on environment: transitions between two populations must be rapid (or rare at the present day)

• Trend is not completely absent for fainter galaxies; but never dominant

Balogh et al. 2004

The star-forming population

• Rines et al. 2005: H distribution in virial, infall and field regions nearly identical.

• Carter et al. (2001) – 3150 nearby galaxies

• H for SF galaxies does not depend on environment

– Triggering of SF occurs on small spatial scales

•Hard to explain with simple, slow-decay models (e.g. Balogh et al. 2000)

Halo mass dependence

• Environment: halo mass– Use luminosity

as tracer of mass. Compare with theoretical mass function

• At fixed mass the late-fraction depends weakly on luminosity

• Late-type fraction depends most strongly on halo mass

Weinmann et al. 2005

[-21,-22][-22,-23]

[-20,-21][-19,-

20][-18,-19]

R luminosity

Halo mass dependence

colo

ur

SF

Rco

nce

ntr

ati

on

• Average properties of galaxies in either peak is independent of halo mass– But depends

on luminosity

[-21,-22][-22,-23]

[-20,-21][-19,-

20][-18,-19]

R luminosity

Weinmann et al. 2005

Local effects?

• Still a (weak) trend with radius in haloes of fixed mass

• Dependence on luminosity (surprisingly?) weak

Weinmann et al. 2005

1014<M<1015

1013<M<1014

Conformity• Properties of “satellite” galaxies appear to be

connected with properties of “central” (actually brightest) galaxy

Weinmann et al. 2005

Similar to effect seen in 2PIGG groups? See Vince Eke’s talk.

Definition of central?

Implications

• Simple dependence of “late-type” fraction on environment characterizes much of observed trends (e.g. SFR-density, morphology-density, colour-density etc.).

• Interpretation?1. Two modes of formation. Within each

peak is variance due to dust, metallicity (second-order effects).

2. Transitions: Where do S0, E+A fit in? 3. Burst vs. continuous SFR (Kauffmann et

al. 2005)

Signs of Nurture: Virgo spirals

Kenney et al. 2003Vollmer et al. 2004

• Ram-pressure stripping in Virgo

H for Virgo galaxy

H for normal galaxy

• Truncated H disks in clusters

Koopmann & Kenney 2004also: Vogt et al. 2004

Signs of Nurture: morphology and SFR

• Passive Spirals• E+A galaxies?• S0, dSph, UCDs• Wolf’s dusty

spirals? Peak in infall region?

• e.g. Christlein & Zabludoff (2005)– Residual [OII] after subtracting

expectation for given B/T, D4000 and Mstar.

• SFR gradient is not entirely:– Consequence of MDR– Consequence of change in mass

function– Effect of initial conditions

AGN• AGN fraction independent of density

– Surprising?

Carter et al. (2001)

Miller et al. (2003)

Models

Semi-analytic approach

• Trace merger histories with N-body simulations (cannot use Press-Schechter because you need to know where the galaxies are)

• More massive haloes form earlier: longer merger history.– There is also a larger-scale bias: haloes of a

given mass form earlier in denser environments (Sheth & Tormen 2004; Abbas & Sheth 2005; Harker et al. 2005)

• Make simple assumptions about gas accretion (e.g. no accretion onto satellites) and feedback (supernova, AGN)

General trends: successes

Springel et al. 2001: morphology-density relation

Okamoto & Nagashima (2003)SFR-radius

Diaferio et al. (2001)colour-radius

0.0 0.5 1.0 1.5 2.0R/R200

Bimodality?•Springel et al. 2001; Diaferio et al. 2001

–Bimodality in field not clear

–All cluster galaxies are red

Okamoto & Nagashima 2003

–SFR is suppressed in all galaxies: blue peak is distorted

-24 -22 -20 -18 -16

MV-logh-24 -22 -20 -18 -16

MV-logh

Cole et al. 2000Supernova feedback prescription does not produce bimodal colour distribution at faint magnitudes. Spirals

Ellipticals

All

cluster

DataModel

SPH simulationsKeres et al. (2005): SPH simulations

reproduce trend of decreasing SFR with increasing density (see also Berlind et al. 2004).

Confirm this is due to reduced accretion of hot gas

But colour-distribution of galaxies doesn’t look quite right…

SFR

Hot accretion

Cold accretion

SPH

Observed

Improving the colour distribution

• Springel, Di Matteo & Hernquist (2005)– Including black hole

feedback terminates star formation more quickly. Leads to rapid reddening of merger remnants

• Sijacki & Springel 2005• AGN feedback removes

young population in cD galaxiesNo feedback

Magorrian-AGN

BH accretion rate

Improving the colour distribution

• Croton et al. (2005)• Radio-feedback most

efficient in large groups.• Proportional to Mgas×MBH

Cooli

ng

rate

(M

sun

/yr)

Models: summary

• When feedback parameters are tuned to reproduce the field luminosity function and colour distribution, what will we find as a function of environment?– General trends will be reproduced.

But will it be for the right reasons?– Any differences in detail: will they

signify “nurture” processes? Or just that feedback parameters need further tuning?

Back to observations: Evolution

Evolution: clusters(briefly)

• Morphology-density relation (see talks by Postman, Dressler)– Fewer S0 in z=1 clusters, but non-zero – Little evolution in MDR z=1 to z=0.5– Suggests high-z MDR is primordial, with

z<0.5 environment-driven evolution

• SFR and colour gradients– Radial gradients steeper in the past

(Ellingson et al. 2001; Kodama & Bower 2001)

– Can be related to truncation of star formation in an infalling field population

Clusters• Tanaka et al. 2005 (see poster)

– tight CMR in place in clusters to z=0.8– Faint end of CMR in groups formed z~0.5– No CMR in field at z=0.8– Also De Lucia (2004): faint end of red sequence disappears at

z>0.5

Clusters

Field

2dF

Nakata et al. 2005

Postman, Lubin & Oke 2001van Dokkum et al. 2000

Fisher et al. 1998

Czoske et al. 2001

Cluster galaxy evolution

• Supported by observed evolution in [OII]-emission fraction (Nakata et al. 2005)– Field evolves much more

strongly than clusters (for bright galaxies)

Evolution: photo-z surveys

• Similar rate of increase in red fraction in the field and clusters – average field red sequence

galaxy came into the sample later

All galaxies

CFHTLS: Nuijten et al. (2005)

Re

d g

ala

xy fr

act

ion

0

0.2

0

.4

0.6

0

.8

1.0

0.2

0

.4

0.6

0

.8

MV < -20

High density

Low density

RedshiftR

ed

ga

laxy

fra

ctio

n

COMBO-17: E. Bell et al.

Luminosity, density and redshift dependence of red

fraction

RCS z>0: Yee et al. (2005)SDSS z=0: Balogh et al. (2004)

Luminosity, density and redshift dependence of

colour

RCS z>0: Yee et al. (2005)SDSS z=0: Balogh et al. (2004)

Luminosity, density and redshift dependence of

colour• Little evolution in red peak colour

RCS z>0: Yee et al. (2005)SDSS z=0: Balogh et al. (2004)

Luminosity, density and redshift dependence of

colour• Little evolution in red peak colour• Colours of bright blue galaxies evolve strongly

RCS z>0: Yee et al. (2005)SDSS z=0: Balogh et al. (2004)

Galaxy groups at z=0.4

• Selected from CNOC2 survey• >30 nights Magellan spectroscopy

(better completeness, depth)• ACS image of ~30 groups• GALEX data rolling in slowly• Spitzer (IRAC and shallow MIPS)

data from GTO programs• Collaborators: Dave Wilman (MPE),

Richard Bower (Durham), Gus Oemler, John Mulchaey (Carnegie), Ray Carlberg (Toronto)

Groups at z=0.4: Morphologies

Spiral-dominated group=270 km/s

E/S0-dominated group=226 km/s

Morphologies: early results

• There are fewer spiral galaxies in groups than in the field, at the same redshift.

• No evidence for more disturbance/irregularities in group galaxies

Field

Sp

iral

fract

ion

E/S

0 f

ract

ion

Groups

Groups

Groups

FieldS

pir

al

fract

ion

Vel. Dispersion (km/s)

The connection between star formation rate, morphology and environment

Like clusters, groups contain passive spirals: disk morphology but low star formation rates

FieldGroups

Elliptical

Early spiral Late spiral

S0

Distributions are corrected for differences in luminosity function between group and field

Stellar mass-SFR

• Stellar masses from archival Spitzer (IRAC) data

• Significant star formation seen in more massive galaxies than locally: downsizing?

• No significant difference between group and field for this subsample.

Rosati? z=1

SDSS (Kauffmann et al.)

Evolution in groups

Wilman et al. (2004)

Fra

ctio

n o

f n

on

-SF

g

ala

xies

• Use [OII] equivalent width to find fraction of galaxies without significant star formation

• most galaxies in groups at z~0.4 have significant star formation – in contrast with local groups

• cf. Gonzalez talk: supergroup

Wilman et al. 2004

• Fraction of non-SF galaxies increases with redshift

• for both groups and field

• Insensitive to aperture effects

• Evolution cannot be account for by passive-evolution models. Require truncation of star formation (both groups and field)

Fra

ctio

n o

f n

on

-SF

g

ala

xies

Groups

Group SFR evolutionF

ract

ion

of

non

-SF

g

ala

xies

Field

Clusters

Field

2dF

Nakata et al. 2005

Postman, Lubin & Oke 2001van Dokkum et al. 2000

Fisher et al. 1998

Czoske et al. 2001

Group Evolution

Groups: Wilman et al. (2005)

High redshift• Spectroscopic survey: ~100 redshifts 1.48<z<2.89• Overdense region has more massive, older galaxies• Consistent with expectations for earlier formation time

(1600 Myr vs 800 Myr)

Steidel et al. (2005)

High redshift

• UV-selected LBG survey

• No environmental dependence of SFR

• Can be consistent: cluster galaxies get head start, but instantaneous SFR the same

• Even at z=0 it seems star-forming galaxies have a distribution independent of environment

Bouché & Lowenthal (2005)

The future

• Theory: still has a lot of catching up to do– Thus we are in discovery mode rather

than testing mode• Observations:

– Dust-obscured SF (Spitzer, Herschel)– AGN/SF connection at z>0– Lower luminosities– Spatial dependence of SFR (i.e. IFU

spectroscopy)– Transitional galaxies