The Role Of Globular Clusters During ReionizationGlobular Cluster Primer Dense stellar systems Consist of 104 - 106 stars Very, very, very old ~ 13 Gyrs Located in bulge and halo Milky

The Role Of Globular Clusters During Reionization

Brendan GriffenMKI Level 5, Room 582K

Friday, 26 April 13

Collaboration

Michael Drinkwater

Ilian Iliev

PeterThomas

GarreltMellema

Brendan Griffen

University of Queensland University of Sussex

University of Stockholm

Friday, 26 April 13

Outline• What do we know about globular clusters and their formation?

• What is their relationship to reionization?

• Previous numerical work

Background

My Work

Summary

• Identify formation sites & modelling them as sources

• Take home message: Globular clusters are important contributors!

• Extensions: dynamical disruption

Brendan GriffenFriday, 26 April 13

Background

Friday, 26 April 13

Globular Cluster PrimerDense stellar systems

Consist of 104 - 106 stars

Very, very, very old ~ 13 Gyrs

Located in bulge and halo

Milky Way N ~160

BulgeDisk This room

10 kpc

Brendan GriffenFriday, 26 April 13

Globular Cluster PrimerTwo populations: metal-rich & metal-poor

Brendan Griffen

-1.59 -0.59

Friday, 26 April 13

Formation ScenariosMajor Mergers

Brendan Griffen

Antennae Galaxies

Friday, 26 April 13

Formation ScenariosHierarchical

Brendan Griffen

t = 13 Gyrs

t = 0

Friday, 26 April 13

Relation To ReionizationMajority formed between ~10-14 Gyrs ago

Brendan Griffen

Forbes & Bridges (2010)

Friday, 26 April 13

Relation To ReionizationAge consistent with reionization epoch

Barkana & Loeb (2007)

Brendan Griffen

Metal Poor GC

Formation

Friday, 26 April 13

Relation To ReionizationEvidence that metal-poor GCs trace reionization epoch

Brendan GriffenSpitler et al. (2012)

“Num

ber

Den

sity

”

Friday, 26 April 13

Relation To ReionizationPotential to produce high numbers of ionising photons

Brendan Griffen

Schaerer & Charbonnel (2011)

Friday, 26 April 13

Previous WorkModelled formation but not contribution to reionization

Brendan Griffen

Bekki et al. (2003, 2008) Kravtsov & Gnedin (2005)

~20 particles per GCsensitive to halo finder

~100 particles per GConly evolved to z = 3

Friday, 26 April 13

My Work

Friday, 26 April 13

The Aquarius Suite~2000 particles per cluster

Brendan GriffenFriday, 26 April 13

Identifying Formation SitesCooling is most efficient at T~104 K

Brendan Griffen

Coo

ling

Rat

e [e

rg c

m^3

/sec

]

Temperature [K]Loeb (2010)

atomic cooling

molecular cooling

Friday, 26 April 13

Identifying Formation Sites

Brendan Griffen

Assuming dark matter is in equilibrium with the gas, we can use the velocity dispersion to infer a temperature.

We tag all dark matter objects that go above this velocity dispersion threshold.

Friday, 26 April 13

The Radiative Transfer Code

Brendan GriffenFriday, 26 April 13

The Radiative Transfer CodeWell tested against other codes

Brendan GriffenFriday, 26 April 13

The Radiative Transfer CodeRay tracing naturally treats inhomogeneities correctly

Brendan GriffenFriday, 26 April 13

Range Of EfficienciesPhysically motivated from literature

Brendan Griffen

--Tumlinson et al. (2004)--Baumgardt & Makino (2007)--Yajima et al. (2011) Ferrara & Loeb (2011) Wise & Cen (2009)

escape fraction

# photons per baryon

star formation efficiency

Friday, 26 April 13

Photon-Richest

Photon-Poorest

Friday, 26 April 13

Results In-homgeneous Ionization

Brendan Griffen

0246810121416182022240

5

10

15

20

25

30

35

40

Nu

mb

er

redshift

M1_512_ph150 M1_512_ph250 M1_512_ph500 M1_512_ph700 M1_512_ph1000 M1_512_ph5000 potential sources

60% suppression

85% suppression

Griffen+ (2013)

Friday, 26 April 13

Resultsz = 0 radial distributions

Brendan Griffen

0 20 40 60 80 100 120 1400

20

40

60

80

100

120

Nu

mb

er

R/kpc

M1_512_ph150 M1_512_ph250 M1_512_ph500 M1_512_ph700 M1_512_ph1000 M1_512_ph5000 truncated at z = 13 Milky Way GCs [Fe/H] < −1

Shallow when compared to Milky Way metal-poor GC distributionGriffen+ (2013)

Friday, 26 April 13

ResultsContribution To Local Reionization

Brendan Griffen

150250

500700

10005000 1

23

45

6

0

0.25

0.5

0.75

1

box width

[h−1 Mpc]

z = 13

fγ

xv

150250

500700

10005000 1

23

45

6

0

0.25

0.5

0.75

1

box width

[h−1 Mpc]fγ

xm

150250

500700

10005000 1

23

45

6

0

0.25

0.5

0.75

1

box width

[h−1 Mpc]

z = 10

fγ

xv

150250

500700

10005000 1

23

45

6

0

0.25

0.5

0.75

1

box width

[h−1 Mpc]fγ

xm

150250

500700

10005000 1

23

45

6

0

0.25

0.5

0.75

1

box width

[h−1 Mpc]

z = 7

fγ

xv

150250

500700

10005000 1

23

45

6

0

0.25

0.5

0.75

1

box width

[h−1 Mpc]fγ

xm

Griffen+ (2013)

Friday, 26 April 13

Results Contribution To Local Reionization (within 23 h-3 Mpc3)

Brendan Griffen

05101520250

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

redshift

xv

M1_512_ph150 M1_512_ph250 M1_512_ph500 M1_512_ph700 M1_512_ph1000 M1_512_ph5000

05101520250

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

redshift

xm

M1_512_ph150 M1_512_ph250 M1_512_ph500 M1_512_ph700 M1_512_ph1000 M1_512_ph5000

Mas

s Fr

actio

n

Volu

me

Frac

tion

Griffen+ (2013)

Friday, 26 April 13

Model Extensions Dynamical Disruption

Brendan GriffenFriday, 26 April 13

Model Extensions Dynamical Disruption

galaxy rotation velocity

galactocentric radius

scaling parameter orbital eccentricity

mass of cluster

Brendan Griffen

Baumgardt & Makino (2003)

Friday, 26 April 13

Results The effect of disruption

Brendan Griffen

60% of all primordial GCs are destroyed through dynamical disruption alone

0 20 40 60 80 100 120 1400

20

40

60

80

100

120

Nu

mb

er

R/kpc0 20 40 60 80 100 120 140

0

0.2

0.4

0.6

0.8

1

f (<

R)

R/kpc

M1_512_ph150 surv. M1_512_ph150 total M1_512_ph250 surv. M1_512_ph250 total M1_512_ph500 surv. M1_512_ph500 total M1_512_ph700 surv. M1_512_ph700 total M1_512_ph1000 surv. M1_512_ph1000 total M1_512_ph5000 surv. M1_512_ph5000 total Milky Way GCs [Fe/H] < −1

Griffen+ (2013)

Friday, 26 April 13

Future Work

Brendan Griffen

Spitler et al. (2012)

Millennium-II simulation

“Num

ber

Den

sity

”

Friday, 26 April 13

CaterPillar ProjectBeing carried out on level 5/6

Brendan Griffen

Team: Greg Dooley

Alex Ji Phillip ZukinAnna Frebel

Ed Bertschinger

Status:Parent simulation complete

Selecting 60+ Milky Way sized halos

First halo at high resolutionto be completed soon

Avalanche of data to follow...

100

Mpc

/hPa

rent

Sim

ulat

ion

Friday, 26 April 13

Brendan Griffen

• Metal-poor globular clusters formed within small dark matter halos

• Age, stellar properties make them ideal candidates for reionization.

• Combined high-resolution dark matter simulation with radiative transfer.

Model

Take Home Message

• Metal-poor globular clusters are candidate contributors to reionization.

• MPGCs contributed ~50% of ionised mass and volume by z = 10.

• Improve: test against a variety environments + include baryons.

• Characterising their contributions is within reach of JWST.

Summary

Friday, 26 April 13