High Energy Radiation and Cosmic Rays
from Clusters of Galaxies
Susumu Inoue(Nat. Astron. Obs. Japan)
GeV100keV
HESS Auger
GLASTSuzaku
F. Aharonian (MPIK), N. Sugiyama (NAO), P. Coppi (Yale)G. Sigl, E. Armengaud (IAP), F. Miniati (ETH)
collaborators:
ZeV
TeV
current evidence for nonthermal emission: Coma1. introduction
radio Giovannini et al. 93 hard X-rayFusco-Femianoet al. 04
Rossetti &Molendi 04
4.8 detection
no detection
gamma-rayno clear evidence yet!GeV Reimer et al. 03
EUV
Bowyeret al. 04
TeV Perkins et al. 06
large scale structure formation (SF) shocks
formation of galaxies, groups, clusters...= hierarchical, dark matter-driven mergers and accretion→ shock formation → gas heating + nonthermal particle acceleration → nonthermal radiation
cosmological hydro simulations by Ryu et al. 03
shock velocitiesthermal emission
clusters are forming this very moment!
QuickTime˛ Ç∆TIFFÅià≥èkǻǵÅj êLí£ÉvÉçÉOÉâÉÄ
ǙDZÇÃÉsÉNÉ`ÉÉÇ å©ÇÈÇΩÇflÇ…ÇÕïKóvÇ≈Ç∑ÅB
shock Mach numbers & particle spectra
major merger
minor merger, accretion
p=2(M2+1)/(M2-1)Not all SF shocks are equal
also Gabici & Blasi 03 Inoue & Nagashima 05
(test particle)
strong (high M) shock-> hard spectra (p~2)
weak (low M) shock-> soft spectra (p>2)
Ryu et al. 03
Vazza’s poster
primary electron inverse Compton
LIC >~ Lsyn ifB < 3G(1+z)2
keV MeVeV GeV
merger
accretion
TeV
Ec(tIC~tshock)
Emax(tIC~tacc)
E
L
E-2
schematicspectrum
accretion(minor merger)
2. high energy emission processes in clusters
traces shock promptly(annular distribution for accretion shock)
e.g. Hwang 97 Ensslin & Biermann 98 Waxman & Loeb 00 Totani & Kitayama 00shock-accelerated e-+CMB→ e-+
tIC<<tshock
electron relativistic bremsstrahlunggenerally not so efficient
proton-proton 0 and secondary pairs
tloss~(nICMppppc)-1~100 Gyr (n/10-3 cm-3)-1
tconf ~R2/6D(E)~200 Gyr (R/Mpc)2 (D/1029cm2s-1) (E/GeV)1/3
pCR+pICM→ ,
traces gas (centrally peaked) +cosmic rays accumulated over cluster history
e.g. Völk, Aharonian & Breidschwerdt 96 Berezinsky, Blasi & Ptuskin 97
t-integrated Mach no. distribution for individual clusters?steep (p>2) spectra?
0→2+-→e+e-+B(~G)→ syn, e+e-+CMB→ IC
>>tH
GeV
0 merger
TeV E
L schematicspectrum
0 accretion (+ radio galaxy, SN-driven wind)
E-2
e+- IC
MeV
origin of radio halos: primary electron vs p-p secondaryprimary electron
p-p secondary
e.g. Brunetti et al. 01,04turbulent stochastic acceleration?need >GeV “seed” particlesmany uncertain parameters
short trad, but no clear sources
long tpp, large-scale injection Dennison 80
but no spectral steepening requires extremely distributed injection0 gamma-rays close to EGRET upper limits
hybrid (p-p secondary injection + turbulent acceleration) Brunetti & Blasi 05
11
10
9
8
720.520.019.519.018.518.017.5
log E [eV]
UHE proton-induced pair emission from cluster accretion shocks
accel. vs CMB losses, lifetime
photopion
lifetimeescape accel.
Bs=0.1 G
photopair
accel.Bs=1 G
e.g. Coma-like clusterM=2x1015 M(T=8.3 keV)WMAP cosmo. parameters
Inoue, Aharonian & Sugiyama 2005 ApJ 628, L9proton Emax
c.f. Kang, Rachen & Biermann 97
Rs~3.2 MpcVs~2200 km/sBs,eq~ 6 G
Emax~1018-1019 eVphotopair important
tacc=(20/3) rgc/Vs2
shock radius, velocity, etc.
Bohm limit shock accel. time
SNR observations ~1e.g. Völk et al. 05
escape timetesc~R2/D(E=Emax)~R/V~2 Gyr shock lifetimetsl~R/V~2 Gyr < tadiab~6 Gyr
proton injection luminosity in accretion shocks
accretion rate & luminosityM(M,z)=fgasfaccVs
3/GLacc(M,z)=fgasfaccGMM/Rs
~2.7x1046 (fgas/0.16) (facc/0.1) (M/ 2x1015 M)5/3 erg/s
proton luminosity & spectrumLp(M,z)=fpLacc(M,z) fp=0.1Fp(E,M,z) ∝ E-2 exp(-E/Emax)
facc=0.1 normalized from simulationKeshet et al. 03
secondary production and emission processesp+CMB→ p+ e+e-
Ep~1018eV E+-~+-Ep~1015eV, L+-~t+-/tinj f(E>1017.7eV) Lp
→ e+e-+B(~G)→ syn. E~keV-MeV e+e-+CMB→ IC E~TeV-PeV
Aharonian 02 code solve proton & pair kinetic eq.
..
-15
-14
-13
-12
-11
-10
14121086420 log E [eV]
e+e- syn
Lp=2x1045
erg/s
Tinj=2x109 yr, R=3 Mpc
n=10-6
cm-3
D=100 Mpc
B=0.1 (G Ep=8 1x 17 )eV
=.3 B (G Ep=2.5 1x 18 )eV
=1 B (G Ep=8 1x 18 )eV
pp (x1)
e+e- IC
emitted flux
> TeV absorption by IRB, CMBComa-like cluster at D=100 Mpc
- large radiative efficiency from protons- hard (~-1.5) spectrum + rollover- sensitive to B
<-> primary IC, pp 0
(~-2)
emitted flux & detectability Coma-like cluster at D=100 Mpcsensitivities for 1 deg2 extended source diameter ~3 deg
TeV : HESS (~0.1° FOV~5°), MAGIC, CANG. III, VERITAShard X: e.g. NeXT, maybe Suzaku/HXD & XIS
implications UHE p-induced emission in cluster accretion shocks
• UHE proton acceleration maximum E• accretion shock direct obs. evidence still tentative• magnetic fields at cluster outskirts emission sensitive to B <-> pp 0
info on B in LSS filaments, cluster B origin• IR background intrinsic spectra to ~PeV, steady <-> TeV blazars
probe of :
cascade emission: pair halo, background
pre-“absorbed” fluxcascade down to GeV-TeV
cluster pair halos- isotropic (much stronger than beamed sources)- hard spectrum
Inoue, Coppi & Aharonian, in prep.
also for p-p 0 from core
probe of IRB, TeV-PeV power
Aharonian, Coppi & Völk 94Coppi & Aharonian 97
3. UHECRs from cluster accretion shocks?
Norman, Achterberg & Melrose ‘95Kang, Ryu & Jones 96Kang, Rachen & Biermann 97
GRB
AGN jet
clusters
energetic requirements
Lcluster~1046 erg/sncluster~10-6 Mpc-3
Pcluster~~1040 erg s-1Mpc-3
UHECR@1020 eVuCR ~10-20 erg cm-3
CR ~0.3(1) Gyr for p (Fe)PCR ~3x1037 erg s-1Mpc-3 massive clusters (~1015 M)
energetically plausiblebut proton Emax insufficient
oblique shocks do not helpOstrowski & Siemieniec-Ozieblo 00
nuclei from cluster accretion shocks as UHECRs
107
108
109
1010
1011
1018
2 3 4 5 6
1019
2 3 4 5 6
1020
2 3 4 5 6
1021
log E [eV]
56Fe
photodisint
photopair
Bs=0.1 G
Bs=1 G
heavy nuclei Emax
for Bs~1 G, EFe, max>~1020 eV
Inoue, Sigl, Armengaud & Miniatiin prep.
UHE nuclei propagation calculations
56Fe
lifetimeescape
log E [eV]lo
g t ac
c, t lo
ss [y
r]
Bs~1 G Johnston-Hollitt & Ekers 05
- structured IGB models based on cosmological simulations- source density ns~10-6 Mpc-3 ∝ baryon density- source power LCR(M)~ 3x1045 erg/s (fCR/0.1)(M/2x1015 M)5/3
- spectral index p=2, Emax(Z) from tacc vs. tloss, tlife- Galactic CR-like source composition (nFe/np~10-3 at fixed E/A)- CMB+FIRB losses, IGB deflections inc. all secondary nuclei
Feretti & Neumann 06
UHECRs: energy losses during propagation
QuickTime˛ Ç∆ǻǵ êLí£ÉvÉçÉOÉâÉÄ
ǙDZÇÃÉsÉNÉ`ÉÉÇ å©ÇÈÇ…ÇÕïKóvÇ≈Ç∑ÅB
p+CMB→ p+ e+e- Ep>~5x1017eVp+CMB→ p+ Ep>~7x1019eV
Lp, 20eV <~100 Mpc
A+CMB→ A+ e+e-
A+FIRB→ A-iN +iNNagano & Watson 00
Fe
p
E
Lloss
LFe, 20eV <~300 Mpc
protons: photopair+photopion
nuclei: photopair+photodisint.
e.g. Stecker &Salamon 99
E>2x1019 eV no data at all (too low statistics)E<2x1019 eV light dominant but greater uncertainties than commonly believed?
current data on composition HiRes stereo Xmax
Watson astro-ph/0408110
intergalactic B fieldse.g. shock generation modelsbased on cosmological simulations
Sigl, Miniati & Ensslin 03,04
normalized to cluster B fields
should be correlated with large scale struc.
Dolag et al 04, 05Brüggen et al 05
different assumptions, numerical methods-> important quantitative differences
quantitatively very uncertain theoretically and observationally
Galactic B: also importante.g. Yoshiguchi et al, Takami et al
Ryu, Kang & Biermann 98
small box size -> periodic boundaryunconstrained -> average over realizationsno Galactic B
UHE nuclei from clusters: results
with IGB
no IGB
spectra composition
anisotropyspectra, anisotropy, composition• consistent with current HiRes but not AGASA? higher Bs?• predictions: - “GZK” cutoff >1020 eV - heavy dominant >1019 eV - large scale aniso. toward few nearby sources
Auger, TA, EUSO
1020 eV1019 eV 1019 eV 1020 eV
fCR~0.03
fCR~0.005
source composition “Galactic CR-like” (solar metallicity)metallicity at cluster outskirtstentatively observed ~0.1 solar
e.g. Finoguenov et al 03
hard spectra at high E p<~1.5
Drury, Meyer & Ellison 99
nonlinear acceleration effects?
Kang & Jones 05
rigidity selection (heavy enhancement)stronger effects for accretion shocks?
UHE nuclei induced pairs and emission
107
108
109
1010
1011
1018
2 3 4 5 6
1019
2 3 4 5 6
1020
2 3 4 5 6
1021
log E [eV]
56Fe
photodisint
photopair
Bs=0.1 G
Bs=1 G
107
108
109
1010
1011
1018
2 3 4 5 6
1019
2 3 4 5 6
1020
2 3 4 5 6
1021
log E [eV]
16O
nuclei photopair+photodisint. loss importantadditional hard X-ray and -ray emission, broader spectra
56Fe16O
lifetimeescape
direct proof of nuclei accelerationconstrain source compositionpotentially
Ee+e-,A ~ (me /Amp ) Z Ep Ee-,ndec ~ (mn-p /Amp ) Z Ep
Inoue, Sigl & Armengaud, in prep.
hard-X/gamma-ray emission from individual clusters: roundup
MeV, GeV and TeV should look different
outskirts
core
GeV0 merger
TeVE
L
0 accretion(+ radio galaxySN-driven wind)
pri. IC
UHE p- pair IC (+UHE Z)
detailed study of cluster emission through simulations warranted
MeV
e+- IC
halo
summary
high energy emission from clusters• primary inverse Compton outskirts, MeV-GeV• p-p 0 core, GeV-TeV• p-p e+- core, MeV• UHE p (+UHE nuclei) photopair emission outskirts, MeV+TeV• cascade emission (pair halo, background) larger scales, GeV-TeV
different components dominate at MeV, GeV, TeVmerger and accretion shock contribute different spectra
probe of structure formation, non-gravitational effects
potentially very rich informationfertile new field of high energy astrophysics!
Inoue & Nagashima; Inoue, Nagashima & Völk, in prep.