Reionization constraints post Planck-15

Transcription

1 Reionization constraints post Planck-15 Tirthankar Roy Choudhury National Centre for Radio Astrophysics Tata Institute of Fundamental Research Pune CMB Spectral Distortions from Cosmic Baryon Evolution RRI, Bangalore, India 12 July

2 Studying the epoch of reionization 3 Figure courtesy Raghunath Ghara universe getting ionized by the first stars aim is to study the neutral hydrogen fraction x HI (x,z) as it decreases from 1 to 0 get insights on the nature of the first stars 1

3 Quasar absorption spectra at z 6 Fan, Carilli & Keating (2006) ( F obs = F cont e τ xhi ) GP, τ GP

4 Thomson scattering τ el from CMBR Planck Collaboration (2016) τ el = σ T c z[t] 0 dt n e (1+z) 3 3

5 Thomson scattering τ el from CMBR Planck Collaboration (2016) τ el = σ T c z[t] 0 dt n e (1+z) 3 3

6 Quasar absorption + CMBR Quasar absorption spectra measure e τgp, not straightforward to convert into x HI CMBR measures the integrated effect Need theoretical models to interpret the two data set simultaneously 4

7 Reionization models Formation of (dark matter) haloes: Analytical: Press-Schechter/Sheth-Tormen formalism: dn(m, z) 2 dm = ρ m δ c(z) dσ(m) π M σ 2 (M) dm Simulations: DM only N-body codes e δ2 c (z)/2σ2 (M) 5

8 Reionization models Formation of (dark matter) haloes: Analytical: Press-Schechter/Sheth-Tormen formalism: dn(m, z) 2 dm = ρ m δ c(z) dσ(m) π M σ 2 (M) dm Simulations: DM only N-body codes Photon production ṅ γ e δ2 c (z)/2σ2 (M) 5

9 Reionization models Formation of (dark matter) haloes: Analytical: Press-Schechter/Sheth-Tormen formalism: dn(m, z) 2 dm = ρ m δ c(z) dσ(m) π M σ 2 (M) dm Simulations: DM only N-body codes e δ2 c (z)/2σ2 (M) Photon production ṅ γ Galaxy/star formation: cooling, fragmentation, feedback (radiative, mechanical, chemical) 5

10 Reionization models Formation of (dark matter) haloes: Analytical: Press-Schechter/Sheth-Tormen formalism: dn(m, z) 2 dm = ρ m δ c(z) dσ(m) π M σ 2 (M) dm Simulations: DM only N-body codes e δ2 c (z)/2σ2 (M) Photon production ṅ γ Galaxy/star formation: cooling, fragmentation, feedback (radiative, mechanical, chemical) Radiation from stars: population synthesis. 5

11 Reionization models Formation of (dark matter) haloes: Analytical: Press-Schechter/Sheth-Tormen formalism: dn(m, z) 2 dm = ρ m δ c(z) dσ(m) π M σ 2 (M) dm Simulations: DM only N-body codes e δ2 c (z)/2σ2 (M) Photon production ṅ γ Galaxy/star formation: cooling, fragmentation, feedback (radiative, mechanical, chemical) Radiation from stars: population synthesis. Escape of photons f esc: neutral hydrogen within the host galaxy 5

12 Reionization models Formation of (dark matter) haloes: Analytical: Press-Schechter/Sheth-Tormen formalism: dn(m, z) 2 dm = ρ m δ c(z) dσ(m) π M σ 2 (M) dm Simulations: DM only N-body codes e δ2 c (z)/2σ2 (M) Photon production ṅ γ Galaxy/star formation: cooling, fragmentation, feedback (radiative, mechanical, chemical) Radiation from stars: population synthesis. Escape of photons f esc: neutral hydrogen within the host galaxy Radiative transfer in the IGM: evolution of ionization fronts Simulations, semi-numerical, analytical 5

13 Analytical models: basic formalism Choudhury & Ferrara (2005, 2006) Average the radiative transfer equation over large volumes = evolution of volume filling factor of ionized regions dq HII dt = ṅγ n H Q HII C HII n e a 3α R(T) can be extended to account for density-dependent reionization Miralda-Escúde, Haehnelt & Rees (2000) 6

14 Analytical models: basic formalism Choudhury & Ferrara (2005, 2006) Average the radiative transfer equation over large volumes = evolution of volume filling factor of ionized regions dq HII dt = ṅγ n H Q HII C HII n e a 3α R(T) can be extended to account for density-dependent reionization Miralda-Escúde, Haehnelt & Rees (2000) Supplemented by temperature and species evolution equations 6

15 Analytical models: sources Choudhury & Ferrara (2005, 2006) dq HII dt = ṅγ n H Q HII C HII n e a 3α R(T) Assumption: reionization driven primarily by galaxies. Photon production rate: ṅ γ = N ion ( Ωb Ω m ) df coll dt Number of ionizing photons in the IGM per baryons Collapse rate of dark matter haloes N ion = f esc ǫ number of photons per baryons in stars 7

16 Analytical models: sources Choudhury & Ferrara (2005, 2006) dq HII dt = ṅγ n H Q HII C HII n e a 3α R(T) Assumption: reionization driven primarily by galaxies. Photon production rate: ṅ γ = N ion ( Ωb Ω m ) df coll dt Number of ionizing photons in the IGM per baryons Collapse rate of dark matter haloes N ion = f esc ǫ number of photons per baryons in stars Predict observables, e.g., τ el (or C l ), photoionization rate (or mean transmitted flux),... 7

17 Analytical models: sources Choudhury & Ferrara (2005, 2006) dq HII dt = ṅγ n H Q HII C HII n e a 3α R(T) Assumption: reionization driven primarily by galaxies. Photon production rate: ṅ γ = N ion ( Ωb Ω m ) df coll dt Number of ionizing photons in the IGM per baryons Collapse rate of dark matter haloes N ion = f esc ǫ number of photons per baryons in stars Predict observables, e.g., τ el (or C l ), photoionization rate (or mean transmitted flux),... full MCMC analysis accounting for N ion (z) and other free parameters 7

18 Data constrained models Mitra, Choudhury & Ferrara (2015) Constraints based on Planck15 data on τ el quasar absorption line measurements at z 6 (either Γ HI or τ eff ) prior on x HI at z based on dark pixel fraction McGreer, Mesinger & D Odorico (2015) 8

19 Data constrained models Mitra, Choudhury & Ferrara (2015) reionization starts at z % ionized at z 6 10 large uncertainties at 7 z 10 8

20 How to constrain reionization at z 7? Galaxy luminosity function: uncertain escape fraction Quasar absorption spectra (damping wings/near zones) IGM temperature Lyman-α emitters (number density, also clustering) Future: 21 cm experiments 9

21 How to constrain reionization at z 7? Galaxy luminosity function: uncertain escape fraction Quasar absorption spectra (damping wings/near zones) IGM temperature Lyman-α emitters (number density, also clustering) Future: 21 cm experiments 9

22 Lyα emitters and reionization Dijkstra, Mesinger & Wyithe (2011) 10

23 Fraction of galaxies having Lyα emission Lyα emitter fraction x Lyα, M UV > This work M UV < Stark et al. (2011) A Redshift Sharp change in behaviour at z > 6. Schenker et al (2014) 11

24 Fraction of galaxies having Lyα emission 11

25 Uncertainties and challenges decrease in the space density of Lyα emitters at z > 6. intrinsic, or damping wing of the surrounding neutral medium? modelling challenges: reionization topology, optically thick (super-) Lyman-limit systems use high (effective) dynamic range numerical simulations Choudhury, Puchwein, Haehnelt & Bolton (2015), Mesinger et al (2015), Kakiichi et al (2015) 12

26 Calibrating the reionization simulations Early (τ = 0.084) Late (τ = 0.068) Very Late (τ = 0.055) 13

27 Matching the data default model late model late reionization seems to explain the decrease in Lyα visibility consistent with other studies Choudhury, Puchwein, Haehnelt & Bolton (2015) 14

28 21 cm maps ν (MHz) ν [MHz] cmpc/h Early (0.084) cmpc/h log10 (Tb /mk) cmpc/h Late (0.068) Very Late (0.055) z redshift Kulkarni, Choudhury, Puchwein & Haehnelt (in prep) 15

29 21 cm power spectra 10 5 Early HM12 (0.084) Late/Default (0.068) Very Very Late Late (0.055) GMRT (k) [mk 2 ] PAPER k [h/cmpc] k [h/cmpc] z = 7 z = 8 z = 10 z = 7 z = 8 z = 10 (dashed: no self-shielding) k [h/cmpc] Kulkarni, Choudhury, Puchwein & Haehnelt (in prep) 16

30 Summary Good progress in modelling the reionization, possible to construct models consistent with available data Uncertainties at z 7, the Lyα emitters could put some constraints Future lies in the 21 cm experiments Currently operating telescopes (e.g., LOFAR) may be able to detect the statistical signal, else have to wait till the SKA1-low 17

31 Summary Good progress in modelling the reionization, possible to construct models consistent with available data Uncertainties at z 7, the Lyα emitters could put some constraints Future lies in the 21 cm experiments Currently operating telescopes (e.g., LOFAR) may be able to detect the statistical signal, else have to wait till the SKA1-low Thank you 17

32 UV luminosity function at z > 6 18

33 Self-consistent reionization from simulations Assume Q(z) to be given. Choose a z: mean free path λ mfp emissivity ṅ ion trial photoionization rate Γ HI dq/dt clumping factor C ionization field, self-shielding invert Γ HI ṅ ion λ mfp solve dq/dt = ṅ ion /n H Cn H α rec 19

34 Galaxy luminosity function N ion = f esc ǫ number of photons per baryons in stars Φ(M AB,z) [mag -1 Mpc -3 ] ϵ = * z = ϵ = * z = ϵ = * z = M AB ϵ = * z = max ϵ = * z = Mitra, Choudhury & Ferrara (2015) 20

35 Constraints on f esc N ion = f esc ǫ number of photons per baryons in stars best-fit 0.3 f esc z Mitra, Choudhury & Ferrara (2015) 21

36 f esc at lower redshifts Escape fraction (f esc) σ upper limits Grazian et al Vanzella et al Siana et al Cowie et al HM z Γ HI in s Becker+2013 Wyithe+2011 Bolton+2007 Calverley+2011 HM12 with f esc = z Planck Collaboration Q H II Schenker McGreer z τ el f esc = 0.14 f esc = 0.18 f esc = 0.22 f esc = z f esc z Khaire, Srianand, Choudhury & Gaikwad (2016) 22

37 Reionization driven by quasars? ϵ 912 (erg s -1 Hz -1 Mpc -3 ) Hopkins+07 Madau+15 Giallongo z 22 faint quasar candidates detected through multi-wavelength observations Giallongo et al (2015) leads to higher number of ionizing photons contributed by quasars 23

38 Constraints on the galaxy contribution τ el Hopkins+07 Madau+15 Giallongo z dn LL /dz z Parameters best-fit with 2-σ errors H07 MH15 G15 ǫ II < f esc < x HI τ el z what about helium reionization? 24