Particle Acceleration and Radiation from Galaxy Clusters

Transcription

1 Particle Acceleration and Radiation from Galaxy Clusters 1,2 in collaboration with Anders Pinzke 3, Torsten Enßlin 4, Volker Springel 4, Nick Battaglia 1, Jon Sievers 1, Dick Bond 1 1 Canadian Institute for Theoretical Astrophysics, Canada 2 Kavli Institute for Theoretical Physics, Santa Barbara 3 Stockholm University, Sweden 4 Max-Planck Institute for Astrophysics, Germany 2 Oct 29 / KITP Conference on Astrophysical Plasmas

2 Outline 1 Cosmological simulations with cosmic rays Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters 2 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus 3 The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence

3 Outline Cosmological simulations with cosmic rays Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters 1 Cosmological simulations with cosmic rays Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters 2 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus 3 The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence

4 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters How universal is diffusive shock acceleration (DSA)? What can galaxy clusters teach us about the process of shock acceleration? Cosmological structure formation shock physics complementary to interplanetary and SNR shocks: probing unique regions of DSA parameter space: Mach numbers M with infinitely extended (Mpc) and lasting (Gyr) shocks (observationally z = ) plasma-β factors of β origin and evolution of large scale magnetic fields and nature of turbulent models in a cleaner environment (1-phase medium) consistent picture of non-thermal processes in galaxy clusters (radio, soft/hard X-ray, γ-ray emission) illuminating the process of structure formation history of individual clusters: cluster archeology

5 Shocks in galaxy clusters Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters 1E 67-6 ( Bullet cluster ) (X-ray: NASA/CXC/CfA/M.Markevitch et al.; Optical: NASA/STScI; Magellan/U.Arizona/D.Clowe et al.; Lensing: NASA/STScI; ESO WFI; Magellan/U.Arizona/D.Clowe et al.) Abell 3667 (radio: Johnston-Hollitt. X-ray: ROSAT/PSPC.)

6 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Radiative simulations with GADGET flowchart CP, Enßlin, Springel (28)

7 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Radiative simulations with cosmic ray (CR) physics CP, Enßlin, Springel (28)

8 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Diffusive shock acceleration Fermi 1 mechanism (1) Spectral index depends on the Mach number of the shock, M = υ shock /c s : log f strong shock weak shock kev 1 GeV log p

9 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Diffusive shock acceleration efficiency (2) CR proton energy injection efficiency, ζ inj = ε CR /ε diss : CR energy injection efficiencyζinj Mach numberm 11/3 3 kt 2 = 1 kev kt 2 =.3 kev kt 2 =.1 kev α inj 3. 4.

10 Cosmological simulations with cosmic rays Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Radiative cool core cluster simulation: gas density y [ h -1 Mpc ] δgas x [ h -1 Mpc ]

11 Cosmological simulations with cosmic rays Mass weighted temperature 1 1 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters 1 8 y [ h -1 Mpc ] Tρgas / ρgas [K] x [ h -1 Mpc ] 1

12 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Mach number distribution weighted by ε diss y [ h -1 Mpc ] - M εdiss/ εdiss x [ h -1 Mpc ] 1

13 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Mach number distribution weighted by ε CR,inj y [ h -1 Mpc ] - M εcr,inj / εcr,inj x [ h -1 Mpc ] 1

14 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters CR pressure P CR y [ h -1 Mpc ] PCRρgas / ρgas [erg cm 3 h 2 7 ] x [ h -1 Mpc ] 1-17

15 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Relative CR pressure P CR /P total y [ h -1 Mpc ] PCR/Ptot ρgas / ρgas x [ h -1 Mpc ] 1-2

16 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Multi messenger approach for non-thermal processes

17 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Multi messenger approach for non-thermal processes

18 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Multi messenger approach for non-thermal processes

19 Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters Multi messenger approach for non-thermal processes

20 Non-thermal emission from clusters Exploring the memory of structure formation Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters primary, shock-accelerated CR electrons resemble current accretion and merging shock waves CR protons/hadronically produced CR electrons trace the time integrated non-equilibrium activities of clusters that is modulated by the recent dynamical activities How can we read out this information about non-thermal populations? new era of multi-frequency experiments, e.g.: GMRT, LOFAR, MWA, LWA, SKA: interferometric array of radio telescopes at low frequencies (ν (1 24) MHz) Simbol-X/NuSTAR: future hard X-ray satellites (E (1 1) kev) Fermi γ-ray space telescope (E (.1 3) GeV) Imaging air Čerenkov telescopes (E (.1 1) TeV)

21 Non-thermal emission from clusters Exploring the memory of structure formation Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters primary, shock-accelerated CR electrons resemble current accretion and merging shock waves CR protons/hadronically produced CR electrons trace the time integrated non-equilibrium activities of clusters that is modulated by the recent dynamical activities How can we read out this information about non-thermal populations? new era of multi-frequency experiments, e.g.: GMRT, LOFAR, MWA, LWA, SKA: interferometric array of radio telescopes at low frequencies (ν (1 24) MHz) Simbol-X/NuSTAR: future hard X-ray satellites (E (1 1) kev) Fermi γ-ray space telescope (E (.1 3) GeV) Imaging air Čerenkov telescopes (E (.1 1) TeV)

22 Outline Cosmological simulations with cosmic rays Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus 1 Cosmological simulations with cosmic rays Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters 2 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus 3 The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence

23 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus CR proton and γ-ray spectrum (Pinzke & CP 29) 1. f(p) p Fγ(> Eγ)Eγ [GeV cm 2 s 1 ] p = P p / (m p c) sic pic Pion decay E γ [ GeV ]

24 Cosmological simulations with cosmic rays Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Hadronic γ-ray emission, E γ > 1 GeV y [ Mpc ] S γ (E γ > 1 GeV) [ ph cm -2 s -1 ] x [ Mpc ] 1-12

25 Cosmological simulations with cosmic rays Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Inverse Compton emission, E IC > 1 GeV y [ Mpc ] S IC,total (E γ > 1 GeV) [ ph cm -2 s -1 ] x [ Mpc ] 1-12

26 Cosmological simulations with cosmic rays Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Total γ-ray emission, E γ > 1 GeV y [ Mpc ] S total (E γ > 1 GeV) [ ph cm -2 s -1 ] x [ Mpc ] 1-12

27 ] ] Cosmological simulations with cosmic rays Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Correlation between thermal X-ray and γ-ray emission 1-4 merger, 1 1 h -1 M O : merger, 1 1 h -1 M O : S γ (> 1 MeV) [ γ cm -2 s -1 h correlation space density [arbitrary units] S γ (> 1 MeV) [ γ cm -2 s -1 h correlation space density [arbitrary units] S X [ erg cm -2 s -1 h 3 7 ] S X [ erg cm -2 s -1 h 3 7 ] Correlation with pion decay/sec. IC emission, X-ray and secondary emission n 2 (CP, Enßlin, Springel 28) Correlation with primary IC emission, correlation space substructure oblique curved shocks; B-generation!

28 Cosmological simulations with cosmic rays Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Photon index Γ - variations on large scales y [ Mpc ] Γ (1 MeV, 1 GeV) y [ Mpc ] Γ (1 GeV, 1 TeV) x [ Mpc ] x [ Mpc ] 1. Γ 1 1 GeV MeV (Fermi): pion bump (center) transition to pic (strong accretion shocks) Γ 1 1 TeV (IACT s): pion-decay (center) GeV pic (accretion shocks, cutoff E max)

29 Universal CR spectrum in clusters Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus 1 1 Fermi: α p ~ 2. < f(p) p 2 >.1 IACT: α p ~ p Normalized CR spectrum shows universal concave shape governed by hierarchical structure formation and the implied distribution of Mach numbers that a fluid element had to pass through in cosmic history (Pinzke & CP 29).

30 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus An analytic model for the cluster γ-ray emission Comparison: simulation vs. analytic model, M vir (1 14, 1 1 ) M pion decay profile: simulation analytic model pion decay spectrum: simulation analytic model s -1 ] λ γ (r, Eγ = 1 MeV) [ph cm g914 g72a Fγ(> Eγ)Eγ [GeV cm 2 s 1 ] g914 g72a λ γ / λ γ R / R vir Spatial γ-ray emission profile Fγ/Fγ E γ [ GeV ] Pion decay spectrum

31 Gamma-ray scaling relations Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus 1 46 total γ-ray emission, excl. galaxies, R<R vir pion-decay γ-ray emission primary IC emission secondary IC emission L γ (E γ > 1 MeV) [ ph s -1 ] M vir [ M O ] Scaling relation + complete sample of the brightest X-ray clusters (HIFLUGCS) predictions for Fermi and IACT s

32 Cosmological simulations with cosmic rays Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Predicted cluster sample for Fermi and IACT s Fγ(Eγ> 1 MeV) CRs with galaxies CRs witout galaxies FORNAX A74 A16 A326 COMA AWM7 PERSEUS 3C129 M49 A3627 OPHIUCHU Fγ(Eγ> 1 GeV) extended HIFLUGCS cluster ID black: optimistic model, including galactic point sources that bias γ-ray flux high; red: realistic model, excluding galactic point sources

33 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Predicted cluster sample for Fermi brightest objects 14 excl. galaxies 12 1 Nclusters C129 A3626, A16 Fornax A3627, Perseus Coma Ophiuchus F γ [ph cm 2 s 1 ]

34 MAGIC observations of Perseus Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus

35 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Upper limit on the TeV γ-ray emission from Perseus radiative physics w/ gal. x 3. radiative physics w/ gal. radiative physics w/o gal. F γ, min (B = 1 µg, ε B < ε th / 3) F γ ( >E ) [ph s -1 cm -2 ] E [GeV] The MAGIC Collaboration: Aleksic et al. & CP et al. 29

36 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Results from the Perseus observation by MAGIC assuming f p α with α = 2.1, P CR P th : E CR <.17E th most stringent constraint on CR pressure! upper limits consistent with cosmological simulations: F upper limits (1GeV ) = 3. F sim (optimistic model) simulation modeling of pressure constraint yields P CR / P th <.7 (.14) for the core (entire cluster) 3 physical effects that resolve the apparent discrepancy: concave curvature hides CR pressure at GeV energies galactic point sources bias γ-ray flux high and pressure limits low (partly physical) relative CR pressure increases towards the outer parts (adiabatic compression and softer equation of state of CRs)

37 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Minimum γ-ray flux in the hadronic model jν,ic(b)/ jν,ic(bcmb) jic(b)/ jic(bcmb) jν(b)/ jν(bcmb) For jν(b) : αν= 1 αν= 1.1 αν= 1.3 αν= B[µG] Synchrotron emissivity of highenergy, steady state electron distribution is independent of the magnetic field for B B CMB! Synchrotron luminosity: ε (αν+1)/2 B L ν = A ν dv n CR n gas ε CMB + ε B A ν dv n CR n gas (ε B ε CMB ) γ-ray luminosity: L γ = A γ dv n CR n gas minimum γ-ray flux: F γ,min = A γ L ν A ν 4πD 2

38 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Minimum γ-ray flux in the hadronic model jν,ic(b)/ jν,ic(bcmb) jic(b)/ jic(bcmb) jν(b)/ jν(bcmb) For jν(b) : αν= 1 αν= 1.1 αν= 1.3 αν= B[µG] Synchrotron emissivity of highenergy, steady state electron distribution is independent of the magnetic field for B B CMB! Synchrotron luminosity: ε (αν+1)/2 B L ν = A ν dv n CR n gas ε CMB + ε B A ν dv n CR n gas (ε B ε CMB ) γ-ray luminosity: L γ = A γ dv n CR n gas minimum γ-ray flux: F γ,min = A γ L ν A ν 4πD 2

39 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Minimum γ-ray flux in the hadronic model: Fermi Minimum γ-ray flux (E γ > 1 MeV) for the Coma cluster: CR spectral index F γ [1 1 ph cm 2 s 1 ] These limits can be made even tighter when considering energy constraints, P B < P gas /3 and B-fields derived from Faraday rotation studies, B = 3 µg: F γ,coma ( ) 1 9 γ cm 2 s 1 F Fermi, 2yr Non-detection by Fermi seriously challenges the hadronic model. Potential of measuring the CR acceleration efficiency for diffusive shock acceleration.

40 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus Minimum γ-ray flux in the hadronic model: IACT s Minimum γ-ray flux (E γ > 1 GeV) for the Coma cluster: CR spectral index F γ [1 14 ph cm 2 s 1 ] These limits can be made even tighter when considering energy constraints, P B < P gas /3, FRM B-fields with B = 3 µg, and α p < 2.3 (caution: this assumes a power-law scaling): F γ,coma ( ) 1 13 γ cm 2 s 1 Potential of measuring the CR spectrum, the effective acceleration efficiency for diffusive shock acceleration, and relate this to the history of structure formation shock waves (Mach number distribution).

41 Outline Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence 1 Cosmological simulations with cosmic rays Motivation and observations Cosmological galaxy cluster simulations Non-thermal processes in clusters 2 Spectra and morphology Predictions for Fermi and IACT s MAGIC observations of Perseus 3 The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence

42 Cosmological simulations with cosmic rays Cosmic web: Mach number 1 The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence 1 1 y [ h -1 Mpc ] - M εdiss/ εdiss x [ h -1 Mpc ] 1

43 ] Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence Radio gischt (relics): primary CRe (1.4 GHz) y [ h -1 Mpc ] S ν,primary [ mjy arcmin -2 h x [ h -1 Mpc ] 1-8

44 ] Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence Radio gischt: primary CRe (1 MHz) y [ h -1 Mpc ] S ν,primary [ mjy arcmin -2 h x [ h -1 Mpc ] 1-8

45 ] Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence Radio gischt: primary CRe (1 MHz) y [ h -1 Mpc ] S ν,primary [ mjy arcmin -2 h x [ h -1 Mpc ] 1-8

46 The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence ] Radio gischt: primary CRe (1 MHz), slower magnetic decline y [ h -1 Mpc ] S ν,primary [ mjy arcmin -2 h x [ h -1 Mpc ] 1-8

47 Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence Radio gischt illuminates cosmic magnetic fields y [ h -1 Mpc ] 2 M εdiss / εdiss y [ h -1 Mpc ] x [ h -1 Mpc ] Structure formation shocks triggered by a recent merger of a large galaxy cluster x [ h -1 Mpc ] red/yellow: shock-dissipated energy, blue/contours: 1 MHz radio gischt emission from shock-accelerated CRe

48 The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence Diffuse cluster radio emission an inverse problem Exploring the magnetized cosmic web Battaglia, CP, Sievers, Bond, Enßlin (28): By suitably combining the observables associated with diffuse polarized radio emission at low frequencies (ν 1 MHz, GMRT/LOFAR/MWA/LWA), we can probe the strength and coherence scale of magnetic fields on scales of galaxy clusters, the process of diffusive shock acceleration of electrons, the existence and properties of the WHIM, the exploration of observables beyond the thermal cluster emission which are sensitive to the dynamical state of the cluster.

49 Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence Population of faint radio relics in merging clusters Probing the large scale magnetic fields Finding radio relics in 3D cluster simulations using a friends-of-friends finder with an emission threshold relic luminosity function y [ h -1 Mpc ] S ν [ mjy arcmin -2 ] y [ h -1 Mpc ] S ν [ mjy arcmin -2 ] x [ h -1 Mpc ] radio map with GMRT emissivity threshold x [ h -1 Mpc ] theoretical threshold (towards SKA) 1-3

50 Relic luminosity function theory The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence Relic luminosity function is very sensitive to large scale behavior of the magnetic field and dynamical state of cluster: F ν [mjy] for a z ~ F ν [mjy] for a z ~ ν = 1MHz, B = µg α B =.3 α B =. α B =.7 α B =.9 1 ν = 1MHz, α B =. B = 1 µg B = µg B = 2. µg Number of relics Number of relics ), Jν= 1 erg/hz/s/ster log(jν/jν varying magnetic decline with radius ), Jν= 1 erg/hz/s/ster log(jν/jν varying overall normalization of the magnetic field

51 Rotation measure (RM) The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence.4.2 RM maps and power spectra have the potential to infer the magnetic pressure support and discriminate the nature of MHD turbulence in clusters: arcmin [rad 2 m -4 ] P [RM](k) k 2 1 P [RM] (k) P [Bz ] (k) 1-6 P [Bz ](k) k 2 [µg 2 Mpc] y [ Mpc ] arcmin -7 RM [ rad m -2 ] P [RM](k) k 2 [rad 2 m -4 ] 1 1 MFR1 MFR2 MFR x [ Mpc ] k [h Mpc -1 ] Left: RM map of the largest relic, right: Magnetic and RM power spectrum comparing Kolmogorow and Burgers turbulence models.

52 Conclusions Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence In contrast to the thermal plasma, the non-equilibrium distributions of CRs preserve the information about their injection and transport processes and provide thus a unique window of current and past structure formation processes and diffusive shock acceleration! 1 Universal distribution of CR protons determined by maximum shock acceleration efficiency ζ max and adiabatic transport: mapping between the hadronic γ-ray emission and ζ max cosmological simulations are indispensable for exploring this (non-linear) map spectral shape illuminates the process of structure formation 2 Primary radio (gischt) emission traces the magnetized cosmic web; sensitive to electron acceleration efficiency Faraday rotation on polarized Mpc-sized relics allows determining the nature of the intra-cluster turbulence

53 Conclusions Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence In contrast to the thermal plasma, the non-equilibrium distributions of CRs preserve the information about their injection and transport processes and provide thus a unique window of current and past structure formation processes and diffusive shock acceleration! 1 Universal distribution of CR protons determined by maximum shock acceleration efficiency ζ max and adiabatic transport: mapping between the hadronic γ-ray emission and ζ max cosmological simulations are indispensable for exploring this (non-linear) map spectral shape illuminates the process of structure formation 2 Primary radio (gischt) emission traces the magnetized cosmic web; sensitive to electron acceleration efficiency Faraday rotation on polarized Mpc-sized relics allows determining the nature of the intra-cluster turbulence

54 Conclusions Cosmological simulations with cosmic rays The cosmic magnetized web Properties of cluster magnetic fields Cluster turbulence In contrast to the thermal plasma, the non-equilibrium distributions of CRs preserve the information about their injection and transport processes and provide thus a unique window of current and past structure formation processes and diffusive shock acceleration! 1 Universal distribution of CR protons determined by maximum shock acceleration efficiency ζ max and adiabatic transport: mapping between the hadronic γ-ray emission and ζ max cosmological simulations are indispensable for exploring this (non-linear) map spectral shape illuminates the process of structure formation 2 Primary radio (gischt) emission traces the magnetized cosmic web; sensitive to electron acceleration efficiency Faraday rotation on polarized Mpc-sized relics allows determining the nature of the intra-cluster turbulence