Schematic History of the Universe

1. Mark Dijkstra, PSU, June 2010 Seeing Through the Trough: Detecting Lyman Alpha from Early Generations of Galaxies ‘Mark Dijkstra (ITC, Harvard)based on Dijkstra & Wyithe ‘10, arXiv:1004.2490, MNRAS accepted

2. Mark Dijkstra, PSU, June 2010 Schematic History of the Universe 400 kyr 1 Gyr 14 Gyr CMB Reionization ‘Dark Ages’ Now Big Bang Time

3. First generation of galaxies: lower metallicity=hotter stars. Mark Dijkstra, PSU, June 2010 Ly From ‘First’ Galaxies Decreasing Z Tumlinson & Shull `00

4. Hotter stars produce more ionizing radiation-> Stronger nebular emission from HII regions around hotter stars, I.e more Ly. Mark Dijkstra, PSU, June 2010 Ly From First Galaxies T=1e5 K T=3e4 K Ly line much stronger at higher z, EW higher by factor of ~ 15: Lya

5. Ly line emission strongest spectral feature of first generation of galaxies (e.g. Partridge & Peebles ‘67, Bromm+01, Schaerer ‘02, Johnson+09). EW~1500 A (restframe). Ly luminosity ~ 20% of bolometric luminosity. H flux lower by factor of ~ 8 (deeper in IR). HeII H (= 1640 A) flux can be comparable to HI H, but this depends sensitively on stellar initial mass function (Johnson+09). Can we detect Ly line? Mark Dijkstra, PSU, June 2010 Ly From First Galaxies

6. First galaxies surrounded by neutral intergalactic medium. Optical depth to Lya is ~ Gunn-Peterson (GP) optical depth. The GP-optical depth is smaller for photons that first enter neutral intergalactic medium with some velocity off-set v (redward of line center). (this is why Lya may be detected from galaxies that reside in large HII regions during later stages of the EoR, and why Lya emitting galaxies probe the EoR.) Mark Dijkstra, PSU, June 2010 Can we Detect Ly From First Galaxies?

7. Mark Dijkstra, PSU, June 2010 Redshifted Lya from Winds in Galaxies • Winds appear present in all galaxies (Steidel+10) and affects Lya spectrum. Lya scattered from backside Absorption by metals

8. Mark Dijkstra, PSU, June 2010 Winds in Galaxies & Ly Transfer • Because winds can Doppler boost photons out of resonance, it promotes escape of Lya from galaxies (photons no longer resonantly `trapped’) Lya scattered from backside • Handful of local starburst galaxies suggest that Lya f_esc does not depend on dust content when outflows are present; otherwise f_esc decreases with dust content (Kunth+94, Atek+08, also Giavalisco+96).

9. Mark Dijkstra, PSU, June 2010 Winds in Galaxies & Ly Transfer • Lya line shape in observed galaxies (z=0-5) can be reproduced using spherical shells of outflowing HI gas, with column density NHI and outflow speed vshell . (see Verhamme+06,08,Schaerer & Verhamme+08,Vanzella+10.) . Vanzella+10 NHI vshell There can be flux at |v|> 1000 km/s.

10. Observations at z<6 typically require log NHI = 19-22 and vshell=0-500 km/s (Verhamme+08). Compute Lya spectrum emerging from Lya source surrounded by spherical shell with column density log NHI = 20-21 and outflow speed vshell=0-200 km/s. Assume no dust. Compute what fraction is transmitted through neutral IGM. Mark Dijkstra, PSU, June 2010 Winds & Ly From the First Galaxies RedBlue Example: z=10, vshell=200 km/s, log NHI = 20 Line center

11. Observations at z<6 typically require log NHI = 19-22 and vshell=0-500 km/s (Verhamme+08). Compute Lya spectrum emerging from Lya source surrounded by spherical shell with column density log NHI = 20-21 and outflow speed vshell=0-200 km/s. Assume no dust. Compute what fraction is transmitted through neutral IGM. Mark Dijkstra, PSU, June 2010 Winds & Ly From the First Galaxies RedBlue Example: z=10, vshell=200 km/s, log NHI = 20 Line center 4% directly transmitted to observer.

12. Directly transmitted fraction ftrans as a function of NHI and vshell . . Mark Dijkstra, PSU, June 2010 Winds & Ly From the First Galaxies ftrans=3% transmitted fraction MD & Wyithe

13. Consider a suite of models with different NHI and vshell. Mark Dijkstra, PSU, June 2010 Winds & Ly From the First Galaxies ftrans=3% transmitted fraction MD & Wyithe

14. It may be possible to directly transmit ftrans> 3% of Ly flux directly to observer through fully neutral IGM at z=10-15 (without HII `bubbles’!). Translates to observed restframe EW~45(ftrans/0.03)(EWint/1500) A -> Strong Lyman Alpha emitters. Ly provides opportunity for spectroscopic confirmation. E.g. SFR~ 1Msun/yr Line flux: NIRSPEC, R=1000, integration time 1e5 s, S/N ~3 Continuum NIRCAM wide filter, same integration time, S/N~2. Mark Dijkstra, PSU, June 2010 Ly From the First Galaxies

15. Mark Dijkstra, PSU, June 2010 Ly From First Galaxies HII HI HII HI

16. The first galaxies were strong Ly emitters (LAEs). Restframe EW~ 1500 A. Ly luminosity can be 20% of bolometric luminosity. Observational evidence that outflows affect transport and facilitate escape of Lya from galaxies at z <6. Similar radiative transfer effects can cause >3% of emitted Lyman Alpha to be observed from galaxies at z>>6 even through a fully neutral IGM. This may facilitate the detection/spectroscopic confirmation of z >> 6 galaxies. Understanding outflows and their impact on Lya is crucial when assessing the impact of IGM (as well as of dust) on the Lya emission line, also at z<7 (outflows can strongly affect how one interprets existing data, e.g. through the `Zheng effect’). Mark Dijkstra, PSU, June 2010 Conclusions

17. But no one-to-one relation between dust content and Lya escape fraction. Dusty galaxies can be strong Lya emitters (Ono+10, arxiv 0911.2544). Blue galaxies can emit no Lya at all (Kunth+98). `We …find that the velocity structure of the neutral gas in these galaxies is the driving factor that determines the detectability of Lyalpha in emission.’ Galaxies become bluer towards higher redshift, which implies dust attenuation is less important (Bouwens+10). Mark Dijkstra, PSU, June 2010 Dust.