Probing Reionization with Lyman Alpha Emitters Pratika Dayal SISSA / International School for Advanced Studies Collaborators : Stefano Borgani, Andrea Ferrara, Simona Gallerani, Hiroyuki Hirashita, Ruben Salvaterra, Alex Saro, Luca Tornatore… MNRAS, 2008, 389, 1683 arXiv:0907.0337 arXiv:0908.4989 arXiv:0908.1571 19th August, 2009 Albanova Center

OUTLINE A brief introduction to Reionization (a) the theoretical model (b) the scenarios Linking LAEs and reionization (a) What are they? (b) Why are they important? (c) How can they be used to constrain reionization? (d) What do we know of their nature? (e) How much do they contribute to reionization? Summary

Reionization at a glance Barkana & Loeb, 2007 QSOs WMAP 5

Building a consistent reionization model Choudhury & Ferrara 2006 Vol averaged fraction of HI Electron sc. Optical depth Evolution of LL systems with z SFR density GP Ly optical depth GP Ly optical No. counts of High-z sources Temperature Hi photo- Ionization rate

The possible Reionization scenarios Early Reionization Model - reionization ends at z ~ 7. Late Reionization Model - reionization ends at z ~ 6. Filling factor of HI Fraction of HI Gallerani et al. 2008

The focus of the talk - LAEs and Reionization Q1: What are LAEs? Q2: Why are they important ? Q3: How can they be used to probe reionization ? Q4: What is the nature of LAEs? Q5: How much do they contribute to reionization ?

What are LAEs ? Galaxies with an observable Lya line. Observed EW > 20 A Lya line very strong and narrow Sharp cut-off blueward of the Lya line Kashikawa et al. 2006

Why are LAEs important ? Highest redshift galaxies with good statistics. Since Ly sensitive to HI, excellent probes of ionization state of the IGM. Relative damping of Ly and UV used to constrain dust amounts/distribution. Deficit due to mass fn. evolution or reionization? Inc. dust/ dec. SFR at z=5.7 Cumulative Ly LF UV LF Kashikawa et al. 2006

The LAE model : SPH+semi-analytics Cosmological SPH simulations 75/h cMpc For each galaxy, obtain (a) SFR, (b) age, (c) metallicity. The LAE model : SPH+semi-analytics Use intrinsic properties of each galaxy to obtain spectra. Clustered galaxies Build a Stromgren Sphere around each galaxy and include photoionization boost due to clustered galaxies. Transmission through IGM Use the value of HI at each point along the LOS to calculate the transmission for the ERM. Spectra from SB99+lum from cooling of collisionally excited HI

The Lya Luminosity Function Clustering effects negligible for a highly ionized IGM, Dayal et al. arXiv:0907.0337 Including clustering Excluding clustering Including/ Excluding clustering Clustering effects important for a highly ionized IGM, Only about 30% of the Ly photons must escape the galaxy, undamped by dust

The UV Luminosity Function 22% of UV photons escape the galaxy undamped by dust E(B-V) = 0.15 37% of UV photons escape the galaxy undamped by dust E(B-V) ~ 0.1 Dust distribution / inhomogeneity evolving with redshift. Dayal et al. arXiv:0907.0337

Clustering Results Boost imparted by clustering <2 Dayal et al. arXiv:0907.0337 Boost imparted by clustering <2 Boost imparted by clustering ~ 58 UVB photoionization rate evolves 3 orders of magnitude between z=7.6 and 6.6. Contribution from clustered galaxies dominates transmission for a highly neutral IGM.

Validating the model with observed SEDs Dayal et al. arXiv:0907.0337, SEDs from Lai et al., 2007 Choose galaxies matching most closely in observed Ly luminosity and attenuate the spectra of the galaxy with E(B-V)=0.15. Predict intrinsic properties of these LAEs :

The Nature of LAEs - I Dayal et al. arXiv:0907.0337 Halo mass Age Stellar metallicity Halo mass range narrows with increasing z due to mass function Evolution. LAEs have ages between 50-350 Myr. Stellar metallicity between 2-55% of the solar value.

The Nature of LAEs- II Dayal et al. arXiv:0907.0337 Star Formation rate SFR efficiency suppressed in low mass halos due to mechanical feedback. z = 5.7 z = 6.6 z = 7.6 * For bulk of galaxies, I is less than 1, i.e. average SFR was higher in the past.

The Nature of LAEs - Dust Dust mass Escape fraction of Ly photons Since SNII are primary dust factories, dust mass increases with SFR (halo mass). z = 5.7 z = 6.6 Escape fraction of Ly photons decreases by an order of magnitude from ~ 1 due to increasing dust. Dayal et al. arXiv:0907.4989

A new window on LAEs - Sub-millimeter with ALMA 353 GHz 220 GHz 100 GHz 0.02 mJy 0.05 mJy 0.12 mJy Dayal et al. arXiv:0908.1571 UV photons (910-4000 A) absorbed are re-emitted in the Far Infra red (FIR) in the rest frame of the galaxy (submm in observer frame). About 3% of LAEs visible with a 5 1 hour integration limit in 353 GHz band. If observations do not match, exotic scenarios need to be considered: top heavy IMF, popIII stars, ages <10 Myr ?

LAEs are PASSIVE REIONIZATION TRACERS ! How much do LAEs contribute to reionization LAEs contribute about 1% of the ionizing photons budget at z ~ 6.6. LAEs are PASSIVE REIONIZATION TRACERS !

SUMMARY We use cosmological SPH simulations coupled with a Ly/continuum production/ transmission model to use LAEs to probe reionization. The ERM reproduces the observations well; the LRM can not still be completely ruled out. Escape fractions of Ly, continuum photons are (0.3,0.3), (0.22,0.37) at z=(5.7,6.6). The dust inhomogeneity / clumping is evolving with redshift. Ly escape fractions decrease with increasing halo mass (SFR), going from ~1-0.1 for halo mass between 10^9 and 10^11.2 solar masses. LAEs can be detected with ALMA for a 5 1 hour integration. LAEs are ‘Passive Reionization Tracers’.

Fraction of ionizing power from halo mass > M LAEs