- - PowerPoint PPT Presentation

カリフォルニア⼤夨学サンタバーバラ校 現実的なシミュレーション ライマン - α スペクトル マックス・グロンケ

Ly α observables contain unique information stars & emitting gas …about the high-z Universe. …to understand the behaviour of cold, neutral gas. Lots of Ly α data… neutral … hydrogen …to constrain galactic subgrid models!

Ly α spectra from hydrodynamical simulations The plan: correlate spectral properties with subgrid parameters! The SILCC simulations • 4pc fixed resolution • radiative heating & cooling • chemistry (CO & H2) • ionizing RT • several “subgrid” feedback mechanisms (SN, winds, …) Hydro-sim: Walch, Girichidis, Naab, Gatto, Glover et al. - https://hera.ph1.uni-koeln.de/~silcc/

Ly α spectra from hydrodynamical simulations The plan: correlate spectral properties with subgrid parameters! The SILCC simulations • 4pc fixed resolution • radiative heating & cooling • chemistry (CO & H2) • ionizing RT • several “subgrid” feedback mechanisms (SN, winds, …) Hydro-sim: Walch, Girichidis, Naab, Gatto, Glover et al. - https://hera.ph1.uni-koeln.de/~silcc/

Ly α spectra from hydrodynamical simulations The plan: correlate spectral properties with subgrid parameters! Hydro-sim: Walch, Girichidis, Naab, Gatto, Glover et al. - https://hera.ph1.uni-koeln.de/~silcc/

Ly α spectra from hydrodynamical simulations

Ly α spectra from hydrodynamical simulations “Green Peas” (Henry et al. 2015; Yang et al. 2016, …); z~0 Erb et al. (2014); z~2-3

Ly α spectra from hydrodynamical simulations “Green Peas” (Henry et al. 2015; Yang et al. 2016, …); z~0 Erb et al. (2014); z~2-3 Do not match observations! Ly α photons escape through cleared channels! (similar problems with toy models of multiphase media) MG & Dijkstra (2016)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs ‣ too symmetric ➞ diminishing constraining power of Ly α .

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? 1. IGM Laursen et al. (2011)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) Hayes et al. (2010); Trainor et al. (2015); Sobral et al. (2017) Laursen et al. (2011)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) Hayes et al. (2010); Trainor et al. (2015); Sobral et al. (2017) 2. CGM

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) Hayes et al. (2010); Trainor et al. (2015); Sobral et al. (2017) ➞ does not affect Ly α spectrum by much (?) 2. CGM Wisotzki et al. (2016); Kakiichi & Dijkstra (2017)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs The impact of the CGM on Ly α spectra ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) Hayes et al. (2010) unrealistic! ➞ does not affect Ly α spectrum by much (?) 2. CGM Wisotzki et al. (2016); Kakiichi & Dijkstra (2017)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs The impact of the CGM on Ly α spectra ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? coarsen the 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) resolution Hayes et al. (2010) unrealistic! for |z| > 1 kpc ➞ does not affect Ly α spectrum by much (?) 2. CGM Wisotzki et al. (2016); Kakiichi & Dijkstra (2017)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs The impact of the CGM on Ly α spectra ‣ too symmetric n o i t u l o s e r ➞ diminishing constraining power of Ly α . Solutions? coarsen the 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) resolution Hayes et al. (2010) “realistic”? unrealistic! for |z| > 1 kpc ➞ does not affect Ly α spectrum by much (?) 2. CGM Wisotzki et al. (2016); Kakiichi & Dijkstra (2017)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) Hayes et al. (2010); Trainor et al. (2015); Sobral et al. (2017) ➞ does not affect Ly α spectrum by much (?) 2. CGM Wisotzki et al. (2016); Kakiichi & Dijkstra (2017) 3. a new shape of cold gas

MG, Dijkstra, McCourt, Oh (2016, 2017) Escape of Ly α photons through a multiphase medium 1 Neufeld (1990) Hansen & Oh (2006) Distribution 0.1 0.01 0.001 -40 -30 -20 -10 0 10 20 30 40 ∼ x/x esc ∆ V/V dop esc ∼ Bx esc ( N HI ) esc ∼ f c B t h t mp c c

MG, Dijkstra, McCourt, Oh (2016, 2017) Escape of Ly α photons through a multiphase medium N HI = 4 / 3 × 10 19 cm − 2 1 Neufeld (1990) Hansen & Oh (2006) Distribution 0.1 0.01 0.001 -40 -30 -20 -10 0 10 20 30 40 ∼ x/x esc ∆ V/V dop esc ∼ Bx esc ( N HI ) esc ∼ f c B t h t mp c c

MG, Dijkstra, McCourt, Oh (2016, 2017) Escape of Ly α photons through a multiphase medium N HI = 4 / 3 × 10 19 cm − 2 1 Neufeld (1990) Hansen & Oh (2006) Distribution 0.1 0.01 0.001 -40 -30 -20 -10 0 10 20 30 40 ∼ x/x esc ∆ V/V dop esc ∼ Bx esc ( N HI ) esc ∼ f c B t h t mp c c

MG, Dijkstra, McCourt, Oh (2016, 2017) Escape of Ly α photons through a multiphase medium N HI = 4 / 3 × 10 19 cm − 2 1 Neufeld (1990) Hansen & Oh (2006) Distribution 0.1 0.01 0.001 -40 -30 -20 -10 0 10 20 30 40 ∼ x/x esc ∆ V/V dop esc ∼ Bx esc ( N HI ) esc ∼ f c B t h t mp c c

MG, Dijkstra, McCourt, Oh (2016, 2017) Escape of Ly α photons through a multiphase medium N HI = 4 / 3 × 10 19 cm − 2 1 Neufeld (1990) Hansen & Oh (2006) Distribution 0.1 0.01 0.001 -40 -30 -20 -10 0 10 20 30 40 ∼ x/x esc ∆ V/V dop Transition at esc ∼ Bx esc ( N HI ) esc ∼ f c B f c , crit ∼ x esc t h t mp c c

MG, Dijkstra, McCourt, Oh (2016, 2017) Ubiquitous droplets lead to realistic Ly α spectra N HI = 4 / 3 × 10 19 cm − 2

MG, Dijkstra, McCourt, Oh (2016, 2017) Ubiquitous droplets lead to realistic Ly α spectra N HI = 4 / 3 × 10 19 cm − 2 (asymmetry of spectra) more stratification leads to: ✓ less flux at v~0 ✓ more asymmetry …in agreement with observations.

MG, Dijkstra, McCourt, Oh (2016, 2017) Ubiquitous droplets lead to realistic Ly α spectra N HI = 4 / 3 × 10 19 cm − 2 (asymmetry of spectra) more stratification leads to: ✓ less flux at v~0 ✓ more asymmetry …in agreement with observations. increasing resolution?

Other evidence for tiny “droplets” Arav et al. (1997) Dv ~ 6km/s + many more… Prochaska et al. (2017)

Other evidence for tiny “droplets” • smooth absorption & Arav et al. (1997) emission profiles over v Dv ~ 6km/s ≳ 1000 km/s in BLR & CGM • dense cold gas out to in galactic halos (large areal but low volume filling fraction; ➞ e.g., Ting- Wen Lan’s talk ) + many more… Prochaska et al. (2017)

Other evidence for tiny “droplets” • smooth absorption & Arav et al. (1997) emission profiles over v Dv ~ 6km/s ≳ 1000 km/s in BLR & CGM • dense cold gas out to in galactic halos (large areal but low volume filling fraction; ➞ e.g., Ting- Wen Lan’s talk ) • “shattering” instability (theory predicting droplets of size ~0.1 n cgs-1 pc) + many more… Prochaska et al. (2017) McCourt et al. (2018)

The “Ly α puzzle” Problem: Simulated Ly α spectra “look different” ‣ too much flux at line center vs ‣ too symmetric ➞ diminishing constraining power of Ly α . Solutions? 1. IGM ➞ not at low-z (and problematic at high-z due to f esc ≳ 0.5) Hayes et al. (2010); Trainor et al. (2015); Sobral et al. (2017) ➞ does not affect Ly α spectrum by much (?) 2. CGM Wisotzki et al. (2016); Kakiichi & Dijkstra (2017) 3. a new shape of cold gas 4. another form of feedback

Girichidis et al. (2018) “old” “new”

Girichidis et al. (2018) “old” “new”

Girichidis et al. (2018) no CRs with CRs no CRs with CRs

MG, Girichidis, Naab, Walch (in prep.) Impact of cosmic-ray feedback on Ly α spectra • non-thermal feedback moves gas without heating it • doesn’t clear channels Ly α can escape through