Active Galactic Nuclei: Sources for Ultra High Energy Cosmic Rays!

Home , NGC 1128, NGC 4261, NGC 4651, NGC 4696, NGC 4945

ACTIVE GALACTIC NUCLEI: SOURCES FOR ULTRA HIGH ENERGY COSMIC RAYS!

Peter L. Biermann1,2,3,4, Laurent¸iu Caramete1,5, Alex Curut¸iu1, Ioana Dut¸an1, Ioana C. Mari¸s6, Oana Ta¸sc˘au7, Julia Becker8,9, Ralph Engel6, Heino Falcke10, Karl-Heinz Kampert7, & Todor Stanev11

1 MPI for Radioastronomy, Bonn, Germany 2 Dept. of Phys. & Astron., Univ. of Bonn, Germany 3 Dept. of Phys. & Astr., Univ. of Alabama, Tuscaloosa, AL, USA 4 Dept. of Phys., Univ. of Alabama at Huntsville, AL, USA

1 5 Institute for Space Studies, Bucharest, Romania 6 FZ Karlsruhe, and Phys. Dept., Univ. Karlsruhe, Germany 7 Phys. Dept., Univ. Wuppertal, Germany 8 Institution f¨orFysik, G¨oteborgs Univ., Sweden 9 Dept. of Phys., Univ. Dortmund, Dortmund, Germany 10 Dept. of Astrophys., IMAP, Radboud Univ., Nijmegen, Netherlands 11 Bartol Research Inst., Univ. of Delaware, Newark, DE, USA www.mpifr-bonn.mpg.de/div/theory

2 Abstract: Ultra high energy cosmic rays were discovered over forty years ago (Linsley 1963). The first prediction has been that due to interaction with the microwave background their spectrum should show a turn-down near 5 1019 eV (Greisen 1966, Zatsepin & Kuzmin 1966): this has now been confirmed by both HiRes and Auger (ICRC Mexico 2007). While many sites of origin have been proposed, only one argument demonstrated that protons of order 1021 eV are required in a source region to explain observations, specifically the ubiquitous turnoff in the optical synchrotron spectrum near 3 1014 Hz (Biermann & Strittmatter 1987); this argument led to the second prediction that radio galaxies are sources of ultra high energy cosmic rays, one rare and specific class of active galactic nuclei (AGN). Some other classes of AGN have been discounted already

3 (Becker et al. 2007). However, since radio galaxies are rare, the near isotropy of arrival directions on the sky shows that the particles suﬀer occasionally substantial scattering along their path; in order to avoid a substantial in- crease in the path-length to us, this scattering has to happen near to our Galaxy: presumably all disk galaxies have a magnetic halo-wind, as supported by dynamo arguments and radio observations (Hanasz et al. 2004 +; Chy˙zy et al., 2000 +; Kulsrud & Zweibel 2007); our Galaxy has a wind (Westmeier et al 2005); and quasar absorption lines show absorption by clouds in a halo-wind to very large distances from galaxies (Williger et al. 2006, and Weymann et al.). In a magneto-hydrodynamic wind (Parker 1958) the dominant magnetic ﬁeld is Bφ ∼ 1/r, al- lowing large bending; this wind is probably highly irregular. Therefore a third prediction

4 is that ultra high energy particles show a distribution function of their bending angle from source to us, which is a highly pointed function, with a core of a few degrees (from the hot disk of the Galaxy), and steep tails for large bending from the wind. The jet-disk symbiosis picture (Falcke et al. 1995 +) or the spin-down model (Blandford & Znajek 1977) allow to pre- dict maximum particle energy and maximum particle ﬂux from such radio galaxies, and this yields a fourth prediction. The origin of ultra high energy cosmic rays has now been traced to active galactic nuclei (Auger-Coll. 2007). This is consistent with the second prediction. The third and fourth predictions remain to be tested.

5 Discovery and Prediction 1: The GZK-cutoﬀ

] 1 -1 -1 sr 10 -1

s -2

-2 10 -3 10 -4

GeV cm 10 [

p -5 10

/dE -6 p 10 -7 ·dN

2 10 p

E -8 10 -9 10 2 3 4 5 6 7 8 9 10 11 12 log10(Ep/GeV)

Figure 1 All-particle cosmic ray spectrum from many earlier experiments. Filled cir- cles at the highest energies are recent results from Auger (ICRC 2007), clearly showing the Greisen-Zatsepin-Kuzmin cutoﬀ, which is due to the interaction with the cosmic microwave background. Clearly the particle energies require revision and rescaling. This is the spectrum to explain, and the strongest radio galaxies can provide an explanation.

6 • Linsley (1963): Detection of first event > 20 ∼ 10 eV - uncontainable in magnetic field of Galaxy • Prediction (1966) of GZK-turnoff near 5 × 1019 eV due to interaction with the cosmic microwave background (MWBG: Greisen, Zatsepin & Kuzmin; sometimes called GZK-cutoff). • Turnoff seen by HiRes (2005) and AUGER (ICRC 2007, Mexico), and so prediction confirmed • Now many events near and beyond 5×1019 eV – nearly isotropic (Stanev et al. Phys. Rev. Letters 1995) and yet many correlations with BL Lac type active galactic nuclei claimed = low power radio galaxies, aiming their relativistic jet at us

7 Basics of active galactic nuclei

Figure 2 A sketch of the cylindrically symmetric AGN. The cut shows the r - z-plane, both axes logarithmically scaled to 1 pc, with the black hole at the center providing the symmetry. The basic constituents are the central BH with a surrounding accretion disk, the jet perpendicular to the disk and the torus encircling this conﬁguration. The dark patches in the torus indicate the clouds, made up by stellar winds in the concept proposed (Zier & Biermann 2001, 2002).

Figure 3 This ﬁgure illustrates the change of the direction of the spin of the BH, induced by the merger of 2 massive BHs, and consequently the change of the direction of the jet. Basically the orbital spin wins over the two intrinsic spins. The left panel shows the situation before the merger, when the jet is aligned with the individual spin of the primary black hole of the binary system (Zier & Biermann 2001, 2002).

9 Figure 4 The spin-ﬂip phenomenon in supermassive black hole binary mergers. These three steps show the envisaged temporal evolution of the ﬁnal stages of the merger. L.A. Gergely, P.L. Biermann: [arXiv: 0704.1968]

10 Physics of active galactic nuclei • Almost all galaxies have a central supermassive black hole • Super-massive black hole mass distribution 6 9 from about 10 M to about 3 10 M , with 8 break of slope near 10 M • Activity deﬁned as non-stellar: All supermassive black holes appear to have activity, very few have strong activity: observable in radio emission, and/or emission lines • This activity appears to be always visible in a weak compact radio jet, and in broad emission lines (sometimes only in polarized light, so scattered photons) • Two color diagram of the total ﬂuxes of galaxy plus AGN at 2 keV, 60 micron, and 5 GHz:

11 • Seyfert galaxies, quasars, radio galaxies, and normal and starburst galaxies are clearly separated (Chini et al. 1989): • with Seyfert galaxies and quasars showing a lot more X-ray emission, and radio galaxies showing additionally a lot more radio emission, than normal and starburst galaxies • Activity episodic, driven by minor (merging with another galaxy without a central black hole) and major mergers (merging with another galaxy, also with a central black hole) • Merging activity began early, far beyond redshift 6.4 (800 million years in the con- cordance cosmology): already then black 9 holes with > 10 M ; after redshift 1.5 - 2 merging activity rapidly decreases • Luminosities up to about 1047 erg/s

12 • Highest photon energy observed about 10 TeV • Shortest variability time scale minutes • Highest deduced Lorentz factor for the relativistic jet: 40 • Luminosity function of activity similar shape to mass function, a powerlaw steepening to a steeper slope at some characteristic power • Such behavior typical for multiple merger or agglomeration process, shallow powerlaw until a maximum, and a much steeper law beyond that (Press & Schechter, Silk & Takahashi, ...); the break is then a function of time • Question on ﬁnal spin of merged supermassive black hole

13 90o

360o 0o

-90o

> 7 Figure 5 The sky in black holes, ∼ 10 M : Aitoﬀ projection in galactic coordinates of 5,978 candidate sources in the case of a complete sub sample (the Galactic plane remains obscured). The choice was made from a complete sample of 10,284 candidate brighter than 0.03 Jy at 2 micron, and selected at z< 0.025; this uses the 2 micron all sky survey, limited in a 20 degree band in the Galactic plane. Normal and starburst galaxies were counter-selected using color and FIR/radio ratio (Biermann & Fricke 1977, Kronberg et al. 1985, Chini et al. 1989, and other work). These candidate sources are probably all black holes, with masses near to or above 107 solar masses; the black hole mass was determined with the black hole versus mass spheroidal stellar population correlation, and tested. The color code is Black, Blue, Green, Orange, Red corresponding to redshifts betwen 0, 0.005, 0.01, 0.015, 0.02, 0.025: Caramete et al. 2008

14 • Stellar analogon: microquasars or Gamma Ray Bursts, black holes with relativistic jets; black hole masses from 3 - 10 solar masses • Most noticeable activity from accretion disk, broad line region, and relativistic jet; often shading by torus covering more than half of 4π, if it exists • X-ray emission probably from Inverse Comp- ton emission near foot of jet • Observations strongly inﬂuenced by relativistic boosting, and so aspect angle • Gamma ray emission visible best from boosted sources, when the compact jet dominates the overall emission

15 • Eddington limit (radiation limited infall, or instability in atmosphere): radiation in ultra- violet 46 8 LEdd ' 10 (MBH/10 M ) erg/s • Disk can be thought of as similar to atmo- spheres of upper main sequence stars, so with radiation driven instabilities: Broad line clouds? The disk emission obeys the Eddington limit • In mergers three spins are involved, two intrinsic spins, and orbital spin, orbital spin will win, lead to a spin-ﬂip • Phase with two super-massive black holes, and anisotropic torques: Torus geometry, and absorption by red giant winds? • Z-shaped, super-disk and X-shaped radio galaxies: time sequence in merger?

16 • Strongest interaction with ambient mate- rial after spin-flip, when the jet plows in different direction through fresh gas clouds • Knots (shocks) and final hot spot (shock) accelerate particles to ultra high energy • Jet embedded in shear flow around itself, observable in radio polarization; radio lobe boundary sheath; shear flow in surrounding large scale filament baryonic accretion flow (into sheet, and along sheet to super- cluster): strong scattering of particles? 3 4 • First shock around about 3 10 to 10 Rg

13 8 Rg = 1.5 10 (MBH/10 M ) cm

• Maximum particle energy at ﬁrst shock limited by photon interaction and synchrotron losses: strong neutrino emission

17 • In case of spin-flip of central black hole, through binary black hole merger, limita- tion additionally by p-p collisions: strong neutrino emission • Last shock where flow goes subsonic, hot spot in strong radio galaxies (Cyg A), and or last strong knot in less powerful radio galaxies (e.g. M87, 3C31, 3C449): last shock has low or no adiabatic losses! • Probably source of ultra high energy cosmic rays observed • Ultra high energy cosmic rays possibly scattered in the three surrounding shear flows near the radio galaxy • For observer distributed source possible

18 Discovery and Prediction 2: Particles in radio galaxies

Figure 6 Decomposition of the observed spectrum of 3C 33 south. (K. Meisenheimer et al., Astron. & Astroph. 1989 )

19 • Cutoff in nonthermal spectrum ν? observed in many radio galaxies since 1976 (Rieke et al. 1976+ Nature ; Bregman et al. 1981 Nature ; Stocke et al. 1981 Nature ; Meisen- heimer & Röser 1986+ Nature ): feature of acceleration of protons to 1021 eV, shown by Biermann & Strittmatter (1987 Astro- phys. J. ): • Protons get accelerated in shock, reach loss limit, establish wave field of irregularities in magnetic plasma, non-linear cascading • Cascade equation: ! d I(k) 1 ∂ k4 ∂ I(k) 2 − 2 2 = S(k) dt 4πk k ∂k 3τk ∂k 4πk (1) wave-numer k = 2π/rg , Larmor-radius rg = E/(eB), time scale of diffusion of turbulent energy τk , source term S(k).

20 • Electrons get accelerated in shock, scatter in given wave ﬁeld, go to loss limit, produce cut-oﬀ in non-thermal emission spectrum, maximal frequency: ν? ' 3 × 1014 Hz (2)

• This translates to loss limit ? 1/2 20 νe −1/2 Ep,max ' 1.4×10 eV B 3.1014 Hz (3) • B typically 10−2 to 10−4 Gauß; spatial limit 21 1/2 near 10 L46 eV (Falcke et al. Astron. & Astroph. 1995). Therefore 21 Ep,max ' 1.4 × 10 eV (4)

• Independent of intensity of the turbulence, shock speed, etc.; dependence via magnetic ﬁeld on parameters with the 1/7-power

21 • Prediction: Radiogalaxies sources at energies > 1020 eV! lower energy jet-sources: gamma ray bursts, microquasars, jet-supernovae

• Nearby candidates (Ginzburg & Syrovatskii 1963 Astron. Zh.; before discovery of UHE- CRs): radio galaxies

Cen A (= NGC 5128), Vir A (= M87 = NGC 4486) For A (= NGC 1316)

22 Table 1 Properties of the selection in passband 6cm (5 GHz), redshift z ≤ 0.018 and z ≤ 0.0125 ﬂux density brighter than 0.5 Jy, steep spectrum and no starburst, sample of 21 and 14 candidate sources. The FIR/radio ratio can readily distinguish radio galaxies from normal galaxies and starbursts

Name Morphological Redsh. Dist. MBH Core ﬂux density B-V FIR/Radio 8 type Mpc 10 M mJy mag ratio NGC 5128 S0 pec Sy2 0.001825 3.4 2 133361 0.88 3.39 NGC 4651 SA(rs)c LINER 0.002685 18.3 0.4 700 0.51 8 MESSIER 084 E1;LERG;LINER Sy2 0.003536 16 10 2094.18 0.94 0.17 MESSIER 087 E+0-1 pec;NLRG Sy 0.00436 16 31 9480.75 0.93 0.01 NGC 1399 cD;E1 pec 0.004753 15.9 3 342 0.95 0.04 NGC 1316 (R’)SAB(s)00 LINER 0.005871 22.6 9.2 5651.61 0.06 NGC 2663 E 0.007012 32.5 6.1 628.56 0.08 NGC 4261 E2-3;LINER Sy3 0.007465 16.5 5.2 2662.69 0.97 0.02 NGC 4696 BCG;E+1 pec LINER 0.009867 44.4 3 518.28 0.08 NGC 3801 S0/a 0.011064 50 2.2 300.25 0.9 0.3 IC 5063 SA(s)0+: Sy2 0.011348 44.9 2 321.14 0.93 11.08 NGC 5090 E2 0.011411 50.4 7.4 488.13 .. 0.1 NGC 5793 Sb: sp Sy2 0.011645 50.8 1.4 51.5 0.79 12.76 IC 4296 BCG;E;Radio Galaxy 0.012465 54.9 10 442.22 0.95 0.08 NGC 0193 SAB(s)0-: 0.014657 55.5 2 285.93 0.98 0.76 VV 201 Double galaxy 0.015 66.2 1 450.09 0.05 UGC 11294 E0?;HSB 0.016144 63.6 2.9 254.52 0.33 NGC 1167 SA0-;LINER Sy2 0.016495 65.2 4.6 393.09 0.13 CGCG 114-025 SA0- 0.016885 67.4 1.9 443.39 0.01 NGC 0383 BCG;SA0-: LERG 0.017005 65.8 5.5 414.25 0.21 ARP 308 Double galaxy WLRG 0.018 69.7 1 88.54 0.09

23 Figure 7 a) Map of the Total-power radio emission from Cen A at 6.3 cm; b) Map of the polarized radio emission from Cen A at 6.3 cm (N. Junkes et al., Astron. & Astroph. 1993 )

Figure 8 90 cm image of M87. The central region containing the jet and inner radio lobes is in the red-orange region near the image center. F. N. Owen et al., Astrophys. J. 2000

25 Figure 9 This spectrum shows a best fit, including only three sources, NGC1068, Cen A, and M87. The fit was achieved by setting the ratio of the flux of Cen A relative to M87 to about 20. This is quoted from the M.Sc. thesis of O. Ta¸sc˘au (2004), and widely advertised since then.

26 Particle energy and particle ﬂux predictions Complete samples (Caramete et al., 2008).

Spin-down powered jets (Blandford & Zna- jek 1977, Dut¸an et al., 2005, 2008):

1/2 Emax ∼ M (5) 1/2 2/5 −7/10 FCR ∼ S D M (6)

Accretion powered jets (Falcke et al. 1995, Ta¸sc˘au, M.Sc. thesis 2003, Ta¸sc˘auet al., 2008):

† 1/3 2/3 Emax ∼ S D M (7) † 2/3 2/3 FCR ∼ S D (8) For distances < 50 Mpc usually NGC5128, possibly NGC1316 and a group around M87 dominate in predicted UHECR ﬂux.

27 Shock UKIRT HST Ryle+VLA Jet Nozzle

Optically thick EUVE Post-shock Jet "Disk" BB

RXTE Optically thin Post-shock Jet

Pre-shock Jet + Nozzle Inverse Compton

ξ = 100 ξ = 10

Thermal/Outer Accretion Disk

Figure 10 The Galactic active source J1118, modelled with a jet and disk: Markoff et al., AA 372, L25, 2001. Many tests were performed with jet modelling to reproduce full electromagnetic spectrum (work by H. Falcke, S. Markoff, M. Nowak, J. Wilms, F. Yuan et al.) for many sources; see especially: Markoff, Nowak & Wilms 2005, and Markoff et al. 2008

28 Table 2 Using core flux-density at 5 GHz for the complete sample of 29 steep spectrum sources. Col. 4: (*) Core flux density estimated from the total flux density by using log(Pcore) = 11.01 + 0.47 log(Ptot), cf. Giovannini 1988; Col. 5 & 6: Relative values of the particles maximum energy and UHECR flux by using Eqs. 1 and 2 (spin-down, I. Dut¸an). Col. 7 & 8: (†) Relative values of the particles maximum energy and UHECR flux by using Eqs. 3 and 4 (accretion, O. Ta¸sc˘au).

M87 M87 M87† M87† Source D MBH S5GHz Emax/Emax FCR/FCR Emax/Emax FCR/FCR 9 (Mpc) (×10 M ) (mJy) (1) (2) (3) (4) (5) (6) (7) (8) ARP 308 69.7 0.1 88.53* 0.18 0.31 0.03 0.04 CGCG 114-025 67.4 0.19 2260 0.25 9.37 0.15 0.33 ESO 137-G006 76.2 0.92 631.32* 0.54 0.71 0.51 0.13 IC 4296 54.9 1 214 0.57 0.16 0.31 0.08 IC 5063 44.9 0.2 321.15* 0.25 0.75 0.06 0.12 NGC 0193 55.5 0.2 285.93* 0.25 0.71 0.07 0.09 NGC 0383 65.8 0.55 414.25* 0.42 0.58 0.24 0.11 NGC 1128 92.2 0.2 280.2* 0.25 0.84 0.1 0.07 NGC 1167 65.2 0.46 393.1* 0.39 0.61 0.2 0.1 NGC 1316 22.6 0.92 26 0.54 0.01 0.08 0.03 NGC 1399 15.9 0.3 10 0.31 0.01 0.01 0.02 NGC 2663 32.5 0.61 160 0.44 0.13 0.12 0.09 NGC 3801 50 0.22 635 0.26 1.66 0.09 0.17 NGC 3862 93.7 0.44 1674 0.38 4.15 0.39 0.21 NGC 4261 16.5 0.52 390 0.41 0.32 0.09 0.26 NGC 4374 16 1 168.7 0.57 0.07 0.13 0.15 NGC 4486 16 3.1 2875.1 1 1 1 1 NGC 4651 18.3 0.04 15 0.12 0.04 0 0.03 NGC 4696 44.4 0.3 55 0.31 0.07 0.05 0.04 NGC 5090 50.4 0.74 268 0.49 0.25 0.23 0.1 NGC 5128 5.62 0.2 6984 0.25 13 0.04 3.63 NGC 5532 104.8 1.08 194.58* 0.59 0.18 0.5 0.05 NGC 5793 50.8 0.14 95.38* 0.21 0.23 0.03 0.05 NGC 7075 72.7 0.25 20 0.28 0.03 0.04 0.01 UGC 01841 84.4 0.1 365.46* 0.18 1.81 0.05 0.08 UGC 02783 82.6 0.42 541 0.37 1.06 0.23 0.11 UGC 11294 63.6 0.29 314 0.31 0.64 0.11 0.09 VV 201 66.2 0.1 450.1* 0.18 2.11 0.04 0.11 WEIN 045 84.6 0.27 321.6* 0.29 0.78 0.13 0.08

29 25 1.2, 21.5, spin down 1.2, 21.5, accretion Auger SD

24.5

24 dF/dE, arb. units 3

23.5

log10 E all normalized at Log10(E)=19

23 18 18.5 19 19.5 20 20.5 log10 E, eV

Figure 11 The results after propagation for maximum energy (exponential cutoﬀ) at 21 10 eV and acceleration spectral index gamma=1.2, log10(Emax) = 21.5, spin down and accretion. Both models normalized to the Auger experimental value at 1019 eV. T. Stanev

30 Discovery and Prediction 3: Galactic winds

• Events are not isotropic: • However, very few radio galaxies are viable • All black holes in our cosmic neighborhood extremely anisotropic in their distribution: no known and identiﬁed sub-class of sources (low power radio galaxies, Seyfert galaxies, Liners, ..) , which would explain the arrival distribution of the events on the sky • Therefore, bending of orbits required for consistency • What is the distribution of bending angles θ (A. Curut¸iu)?

31 NGC 4569 4.86 GHz + PI B-vectors + optical DSS Blue

13 16

10 DECLINATION (J2000) 08

04 12 37 15 00 36 45 30 RIGHT ASCENSION (J2000) . Chyzy 2007 in prep. (VLA obs.) . Chyzy et al. 2006, A&A 447, 465 (Effelsberg obs.)

Figure 12 VLA observations at 4.86 GHz in contours, B-vectors proportional to polarized intensity, DSS Blue image. Chy´zyet al. 2007

Figure 13 Sky plot of Rotation Measures, using Han et al., and Sun et al.; courtesy R. Wielebinski

33 • Our galaxy should have a wind: Many pa- pers by Parker from the 1960ies to 1990ies (also cf. Biermann & Davis 1960) • Our galaxy has an observed wind, head- tail structure of high velocity clouds (West- meier et al. 2005 Astron. & Astroph. ). Many other galaxies have a magnetic wind (Cracow group). Quasar absorption lines show absorption by clouds to very large distances from galaxies (Williger et al. 2006 Astrophys. J. , and Weymann et al.) • The Parker’s (1992) cosmic ray driven wind conﬁrmed in simulations (Cracow group) • At high energy only single or very few scat- terings, scattering angle θ distribution (A. Curut¸iu)

1/θ2

34 Simulation of correlations • Test in Auger paper was to identify with active galactic nuclei from the Véron-Cetty catalogue. • Using a Monte-Carlo approach we can re- peat the exercise with a) a source list (L. Caramete), b) a flux prediction (I. Dut¸an), and c) a scattering model (A. Curut¸iu). • We can then also differentiate between the Auger sky and the HiRes sky. • Result favors radio galaxies, selected at 5 GHz and with steep spectrum, using their compact components for estimating the power in ultra high energy cosmic rays, and adopt- ing the point of view, that spin-down (I. Dut¸an) or accretion mode (O. Ta¸sc˘au)are key to understanding the powering of the jet, hot spots and cosmic ray acceleration.

Figure 14 Virtual events found in the V´eron-Cetty & V´eron catalogue (2006) using a core of 3 degrees, A. Curut¸iu

36 90o

NGC 4631 MESSIER 094 MESSIER 051 MESSIER 051a MESSIER 066 NGC 3628 NGC 3690 NGC 3079 NGC 2903 MESSIER 082

NGC 2146 NGC 4945 NGC 6946 NGC 3256 NGC 1569 ESO 173- G 015 360o 0o Circinus Galaxy

NGC 0891 MESSIER 031 MESSIER 033 LMC NGC 1808

NGC 0660 SMC MESSIER 077 NGC 1365 NGC 1097 NGC 7582 NGC 7552 NGC 0055 NGC 0253 -90o

Figure 15 Aitoﬀ projection in galactic coordinates of the selection from NED in 60µm, redshift z ≤ 0.0125, ﬂux density brighter than 50 Jy, starburst selected, sample of 32 candidate sources and 100 virtual events from this sources and weighted contribution. Double Monte-Carlo to simulate the intermittent nature of Gamma Ray Bursts

37 Table 3 The fraction of the 100 virtual events from sources at redshift z ≤ 0.0125 for Auger and HiRES sky coverage and the correlation with the V´eron-Cetty & V´eroncatalogue (2006). We test radio galaxies, normal and starburst galaxies assuming that Gamma Ray Bursts are the sources, and simple isotropy. Here with z < 0.0125

Model for CR contribution Auger Possible HiRES Extreme HiRES Sky Coverage Sky Coverage Sky Coverage 5GHz, steep spectrum weighted contribution 92 with 62 correlation 28 with 15 correlation 41 with 19 correlation from I. Dut¸an spin-down model 5GHz, steep spectrum weighted contribution 94 with 53 correlation 34 with 20 correlation 58 with 24 correlation from O. Ta¸sc˘au accretion model 5GHz, steep spectrum uniform contribution 93 with 60 correlation 41 with 20 correlation 56 with 28 correlation

For estimated core ﬂux 5GHz, steep spectrum weighted contribution 89 with 67 correlation 15 with 8 correlation 28 with 13 correlation from I. Dut¸an spin-down model 5GHz, steep spectrum weighted contribution 92 with 63 correlation 20 with 9 correlation 38 with 16 correlation from O. Ta¸sc˘au accretion model 60micron, starburst selected uniform contribution 67 with 30 correlation 46 with 30 correlation 67 with 37 correlation double Monte-Carlo for GRBs 60micron, starburst selected weighted contribution 58 with 33 correlation 55 with 32 correlation 70 with 39 correlation double Monte-Carlo for GRBs 100 random events 73 with 20 correlation 53 with 19 correlation 76 with 23 correlation across the sky

38 Prediction 4: Test on source spectra

• All sky samples of radio sources, active galactic nuclei • Using jet-disk symbiosis concept (Falcke et al. 1995 +) or spin-down model (Blandford & Znajek 1976 +; Dut¸an et al. 2005) to pre- dict maximal particle energy, and maximal ﬂux of ultra high energy cosmic rays • Take each event and rank plausible sources with predicted UHECR-ﬂux divided by an- gular distance squared, and then take the top as ”source” – uses probability • So check on scattering distribution; allow for extended source, central shift, and temporal variability • Check on source spectrum vs prediction

39 Ultra high energy neutrinos

Attempt to scale the neutrino production from the sources of ultra high energy cosmic rays (J. Becker):

neutrino production proton emission p Disk γ −> ... ν p p pp p FSRQ p p log(B/Gauss)

r −2

−1 r SSRQ

−3 −2 −1 0 +1 +2 +3 log(r/pc)

Figure 16 Neutrino production in AGN

40 ]

-1 1 sr -1 s -2 cm

1.7 -1 10 GeV [ 2.7

-2 dN/dE*E 10

8 8.5 9 9.5 10 10.5 11 11.5 12 log10(E/GeV)

Figure 17 Fitting to the ultra high energy cosmic ray data

41 ]

-2 Atmos.-conv ① StSa② MPR -5 -2 -2.3

cm ③ E ④E

-1 10 s

-1 -6 10 -7 AMANDA (4yr)

GeV sr 10 [

2 ②

ν -8 10 *E ④ ν -9

/dE 10 ① ν

dN -10 10 ③ -11 10 -12 10 3 4 5 6 7 8 9 10 log(Eν/GeV)

Figure 18 Neutrino ﬂux prediction from UHECR correlation

42 90o

1345+12

1226+02 1253-05

1127-14 0923+39 1641+39

0851+20 0831+55

0742+10 0834-20 360o 0o

2200+42 0316+41 1549-79 0537-44 + 0637-75 0440 00 0521-36 2251+15 2134+00 04280438-53-43 2223-05 2203-18 0208-51 2345-16

-90o

Figure 19 Aitoff projection in galactic coordinates of the 25 flat and inverted radio spectrum sources highest in flux density at 2.7 GHz. These sources are prime candidates to be ultra high energy neutrino sources

43 90o

Messier 87

Centaurus A

360o 0o

Fornax A

-90o

Figure 20 Aitoﬀ projection in galactic coordinates of the three strongest radio sources in > 19 the sky and the arrival directions of the 27 cosmic rays with the highest energy ∼ 6 10 eV detected by the Pierre Auger Observatory

44 Key Concepts • Quantitative theory: Radio galaxies able to produce particles to 1021 eV: Test: Scatter- ing distribution and spectra (spin-down vs current accretion) • Disk galaxies: Magnetic winds: Scattering: Point spread function with steep tails • Correlation with Véron-Cetty & Véron catalogue (2006) reproduced • Ultra high energy cosmic rays from last shock in jet • All only consistent with the assumption: Protons - but injection plasma physics? • Neutrino emission beamed and from first shock in jet near black hole • Neutrino emission from recently merged black holes with spin-flip

45 Future

The sites of acceleration and interaction

physics laboratory to understand the

fundamental physics of the structure of matter at the highest energies, so at the deepest level

46 Acknowledgements PLB would like to acknowledge intense dis- cussions about these topics with many, especially with E,.J. Ahn, V. Berezinsky, T. Gaisser, L. Gergely, D. Harari, T. Kneiske, P.P. Kron- berg, A. Meli, F. Munyaneza, E. Roulet, D. Ryu, M. Teshima, P. Tinyakov, G. Thomson, St. Westerhoﬀ, and R. Wielebinski. Support for PLB is coming from the AUGER mem- bership and theory grant 05 CU 5PD 1/2 via DESY/BMBF.