Stellar Population Synthesis: Which Uncertainties Matter?

1. Stellar Population Synthesis: Which Uncertainties Matter? Charlie Conroy w/ Jim Gunn (Princeton)

2. SPS Ingredients Stellar evolution e.g., mass-loss, rotation, binaries, horizontal branch, blue stragglers, WR stars, TP-AGB stars Stellar spectral libraries IMF Dust attenuation

3. An Introduction to TP-AGB stars • Thermally-pulsating AGB stars • Double shell burning • Important especially for ages of ~0.5-2 Gyr • Carbon star production • Metallicity-dependent • Dynamic atmospheres • Very uncertain • How can we constrain these uncertainties?? Maraston 2005 with TP-AGB

4. TP-AGB stars in the LMC Conroy & Gunn 2010 Compare SSP predictions to colors of `super-star clusters’ as a function of cluster age Latest Padova tracks produce near-IR colors too red BaSTI tracks are OK Flexible SPS allows simple re-calibration of Padova tracks to fit the LMC cluster data What about calibration at other Z??

5. TP-AGB stars in post-starburst galaxies: I • Post-starburst (“K+A”) are galaxies that have a strong A star component in their spectrum, without H emission • Implies a significant population of ~0.3-1 Gyr old stars • Excellent testbed for TP-AGB stars • No stochastic effects • Model testing at ~solar Z FSPS+Padova Bruzual & Charlot 2003 Maraston 2005 Conroy & Gunn 2010 Post-starbust spectrum

6. TP-AGB stars in post-starburst galaxies: II Predictions for Zsol coeval populations Dn4000=1.3 (t=0.44 Gyr) Near-IR spectra of post-starburst galaxies can jointly constrain: 1. the overall luminosity in the TP-AGB phase 2. the presence of carbon stars Other near-IR features such as VO (1.06m) and H2O (1.33m) will provide additional constraints on the fraction of carbon stars Near-IR data currently being taken for a sample of post-starburst galaxies from SDSS (w/ Mariska Kriek) Dn4000=1.4 (t=0.79 Gyr) Dn4000=1.2 (t=0.25 Gyr)

7. Dust Attenuation Curves • Attenuation curve depends on: • Size distribution of grains • Metallicity • Intensity of UV radiation • Geometry • In general, we do not know what the attenuation law “should” be • UV bump at 2200A is an important feature in MW and LMC extinction curves, and shows considerable variation amongst environments • Not seen in local starbursts (Calzett et al. 1994), nor in the SMC • seen in z~2 star forming galaxies, (Noll et al. 2009)

8. face-on edge-on Dust in disk-dominated, star-forming galaxies • Consider a sample of disk-dominated star-forming galaxies at z<0.05 • GALEX+SDSS+2MASS • Consider average photometric properties in bins of inclination • Trends with inclination will depend only on dust attenuation curve

9. UV dust attenuation at z~0 • SEDs of star-forming, disk-dominated galaxies at z<0.05 • GALEX+SDSS photometry as a function of inclination • Clear signature of the UV bump • First detection at low redshift • Based on stacking several hundred galaxies per inclination bin, so must be generic • Implies that dust corrections in the UV is significantly more complicated than traditionally assumed UV bump Conroy, Schiminovich, & Blanton 2010

10. Summary Despite impressive progress in SPS model-building, significant uncertainties remain However, it is possible to constrain the uncertain SPS inputs directly from observations of galaxies Two Examples: TP-AGB stars in post-starburst galaxies Dust attenuation curves in disk-dominated star-forming galaxies Additional Directions: IMF in passive Ellipticals Wing-Ford band at 9910A is sensitive to the dwarf/giant ratio High mass-loss rates at super-solar metallicity? If (some) stars bypass the HB/AGB phase it should be detectable in the integrated light