Population Synthesis at the Crossroads Claus Leitherer (STScI) and Sylvia Ekström (Geneva Obs.)

9/5/2011Claus Leitherer:

Population Synthesis 1

Population Synthesis at the Crossroads

Claus Leitherer (STScI)and

Sylvia Ekström (Geneva Obs.)

9/5/2011Claus Leitherer:

Population Synthesis 2

Cosmology

Galaxy Evolution

Stellar Populations

Stellar Astrophysics Nuclear and Atomic Physics

Popularity • Mobility • Front Page News Errors • I

nformatio

n Flow

Evolutionary Population Synthesis: Basics

o Goal: model SEDs of nearby and distant galaxies

o Star formation law (IMF, SF history, clustering…)o Stellar evolution models (conversion of mass into

luminosity and introduction of the time-scale)o Spectral libraries (empirical and theoretical)o Other (dust, geometry…)

9/5/2011 3Claus Leitherer:

Population Synthesis

Most widely used, public model packages:

o GALEV (Schulz et al. 2002; Kotulla et al. 2009)o GALEXEV (Bruzual & Charlot 1993; 2003)o PEGASE (Fioc & Rocca-Volmerange 1997; Le

Borgne et al. 2004) o STARBURST99 (Leitherer et al. 1999; Leitherer

& Chen 2009) generally good agreement

9/5/2011 4Claus Leitherer:

Population Synthesis

Broadened scope beyond stellar populations:

o Panchromatic SEDs including stellar and gaseous contributions (Dopita et al. 2005; 2006a; 2006b; 2008)

o SEDs with self-consistent inclusion of dust (Dwek & Scalo 1980; Dwek 2009)

o Chemical evolution due to stars and galactic outflows and infall (Matteucci 2009; 2010)

9/5/2011 5Claus Leitherer:

Population Synthesis

New Developments: the New Normal

o Stochasticityo Rapid eruptive evolutionary phaseso Stellar multiplicityo Convective overshootingo Stellar rotation

9/5/2011 6Claus Leitherer:

Population Synthesis

Stochasticityo Bruzual (2002):o Cluster simulations of

sparsely populated evolutionary phases due to stochastic fluctuations of the IMF

o Cerviño et al. (2002):o Relevant for clusters with

M < 105 M⊙, depending on wavelength

9/5/2011 7Claus Leitherer:

Population Synthesis

9/5/2011Claus Leitherer:

Population Synthesis 8

• SLUG synthesizes stellar populations using a Monte Carlo technique with stochastic sampling, including the effects of clustering, the stellar IMF, star formation history, stellar evolution, and cluster disruption

Da Silva et al. (2011): SLUG – Stochastically Light Up Galaxies

9/5/2011Claus Leitherer:

Population Synthesis 9

o Application: comparison with sample of 300 star-forming galaxies within 11 Mpc (Lee et al. 2009)

o SFR(H)/SFR(UV) declines for small SFR(UV)o Photon leakage?o Eldridge (2011): can also be accounted for by stochastic effects

Eldridge (2011): BPASS – Binary population and spectral synthesis

9/5/2011Claus Leitherer:

Population Synthesis 10

o Four stars have masses between 150 and 300 M⊙

o These stars account for ~50% of the ionizing photon output of R136

o Correspondingly, they account for ~10% of the ionizing luminosity of the entire LMC

Crowther et al. (2010): the most massive stars in R136 (LMC)

Rapid Eruptive Evolutionary Phases

o Maraston (2011)o Contribution to LBol of

SSP by TP-AGB starso Evolution/atmospheres

still poorly understood

9/5/2011 11Claus Leitherer:

Population Synthesis

9/5/2011Claus Leitherer:

Population Synthesis 12

o SPS code combines stellar evolution calculations with stellar spectral libraries to produce simple stellar populations

o Manually adjust stellar phases and estimate errors

o Allows estimate of uncertainties of, e.g., the AGB phase

Conroy & Gunn (2010): FSPS – flexible stellar population synthesis

9/5/2011Claus Leitherer:

Population Synthesis 13

o Continuum in M(peak) from novae to SNe

o AGB in the red, vs. LBV in the blue

o AGB important for mass loss and luminosity

o LBV only important for mass loss

o LBVs are critical predecessors of WR stars

Smith et al. (2011): transient, eruptive phases in the upper HRD

Stellar Multiplicity

o Sana & Evans 2011: >50% of all massive stars in binaries; significant fraction in close binaries

9/5/2011 14Claus Leitherer:

Population Synthesis

9/5/2011Claus Leitherer:

Population Synthesis 15

1. On the main-sequence: spin-up and higher rotation; mixing and evolution to higher L and Teff.

2. Post-main-sequence: Roche-lobe overflow and mass transfer to secondary: “rejuvenation”.

de Mink (2010): evolution of a massive close binary

9/5/2011Claus Leitherer:

Population Synthesis 16

o Evolution of Hβ equivalent width with time in SF galaxyo Solid: single stars with different Z; dashed: binaries with different Zo Hot, ionizing population appears after ~10 Myro Relevance: age spread of star formation? LINERs?

Eldridge & Stanway (2009): rejuvenation effect in SSP

Convective Overshooting

Energy production of massive stars in convective core

9/5/2011 17Claus Leitherer:

Population Synthesis

Stellar Rotation

o Hunter et al. (2009): N/H and v sin i in LMC OB starso Significant N enrichment on the main sequence

9/5/2011 18Claus Leitherer:

Population Synthesis

o Stellar winds to remove outer, stellar layers

o Mass loss and mass transfer in close binaries

o Enhanced convective overshooting

o Correlates with rotation rotationally induced mixing

9/5/2011 19Claus Leitherer:

Population Synthesis

How to reproduce the N enhancement

9/5/2011Claus Leitherer:

Population Synthesis 20

1. M < 2 M⊙: rotation negligible because of magnetic braking

2. 2 M⊙ < M < 15 M⊙: Teff decrease caused by centrifugal forces

3. M > 15 M⊙: larger convective core with higher Teff and L

Brott et al. (2011): evolutionary models with rotation

9/5/2011Claus Leitherer:

Population Synthesis 21

o 0.8 M⊙ < M < 120 M⊙

o Z = Z⊙ (other Z in preparation)

o vrot = 0.4 vbreakup on ZAMS

o Calibrated extensively via local stars and star clusters

o Implemented in Starburst99

Ekström et al. (2012): full set of evolution models with rotation

9/5/2011Claus Leitherer:

Population Synthesis 22

o Models with rotation are more luminous by ~0.4 mag because of the higher L/M and Teff of individual stars

MBol vs. time (SSP, 106 M⊙, Z⊙, Kroupa IMF)

9/5/2011Claus Leitherer:

Population Synthesis 23

o The number ionizing photons increases by a factor of ~4 when hot, massive stars are present

Lyman photons vs. time (SSP, 106 M⊙, Z⊙, Kroupa IMF)

9/5/2011Claus Leitherer:

Population Synthesis 24

o W(H) increases by ~0.2 dex. If used as an IMF indicator in late-type galaxies, the new models change the IMF exponent from, e.g, 2.3 to 2.6

H equivalent width vs. time (continuous SF, Z⊙, Kroupa IMF)

9/5/2011Claus Leitherer:

Population Synthesis 25

o Applied to IR-luminous galaxies: models with rotation lead to, e.g., SFR = 100 M⊙ yr-1, whereas models without give SFR = 175 M⊙ yr-1.

Br luminosity vs. time (SFR = 100 M⊙ yr-1, Z⊙, Kroupa IMF)

9/5/2011Claus Leitherer:

Population Synthesis 26

Take-Away Pointso Replacemement of traditional population synthesis

models by a new generation of modelso Ready for quantitative testing, beyond mere

concept studieso Stochastic effects, stellar multiplicity, rotationo Significant decrease of M/L, and therefore revision

of IMF and SF rateso The new models need to be tested by comparison

with galaxy properties

9/5/2011Claus Leitherer:

Population Synthesis 27

Cosmology

Galaxy Evolution

Stellar Populations

Stellar Astrophysics Nuclear and Atomic Physics

Inform

ation Flow