1. Deflections of Ultra High Energy Cosmic Rays by Intergalactic Magnetic Fields Based on astro-ph/0310902 with Klaus Dolag (Dipartimento di Astronomia - Padova Volker Springel (Max Planck Institute for Astrophysics - Munich) Igor Tkachev (CERN )

2. Ultra High Energy Cosmic Rays UHECR Cosmic Rays spectrum

3. Why UHECRs are so interesting? Likely, they are of extra-galactic origin The acceleration mechanism is unknown and it may involve new physics Galactic magnetic field are not expected to deflect them significantly  UHECR may point their sources if intergalactic magnetic fields are not too strong

4. Plan of the talk Observational situation: UHECR energy spectrum and angular distribution Why, most likely, UHECR come from astrophysical sources How far UHECR can travel; why they are nuclei Galactic and intergalactic magnetic fields and their origin Reconstructing the IGMF by numerical simulations Maps of deflections and general results Conclusions

5. The observational situation

6. Extensive Air Showers Ground Arrays either scintillators or water cherenkov tanks are deployed on a large region. The energy of the particle is obtained by measuring the density of particles at 600 m from the core of the shower. Fluorescence Telescope antennas are used to detect the isotropic 300-400 nm radiation emitted by fluorescence from the nitrogen molecules in the atmosphere. Energy and composition are measured from the position of the maximum of the shower.

7. EAS experiments EAS experiments THE PRESENT: AGASA: ground detector [acceptance 160 km 2 sr, angular resolution ~ 2 0 ] HiRes (monocular): fluorescence detector [ 350 km 2 sr, angular resolution ~ 1 o (binocular)] THE FUTURE:  AUGER (ground + florescence) [accep. 7000 km 2 sr, res. 1.5 o ] EUSO ( fluorescence from the space) [ accep. 35000-70000 km 2 sr] 

8. Energy spectrum: AGASA vs HiRes AGASA ~ 900 events with E > 10 19 eV –~ 100 “ “ 4  10 19 eV –~ 15 “ “ 10 20 eV HiRes 2 events with E > 10 20 eV (expected 20 by assuming AGASA flux)

9. Energy spectrum: AGASA vs HiRes - discrepancy reduces at ~ 2  between the two experiments - AGASA is compatible with the GZK cut-off at 2.3  Hopefully this will be settled by AUGER which combine both techniques ! By correcting for possible systematic in the energy determination - 15 % + 15 % Blasi, De Marco & Olinto ‘03

10. The composition of UHECR From the study of EAS development Haverah Park [Ave et al. 2001] No more than 48% of the inclined showers can be photons above 10 19 eV No more than 54% can be Iron above 10 19 eV No more than 50% of the showers can be photons above 4 10 19 eV Neutrinos are excluded (if the cross section is not enhanced) Fly's Eye [Dawson et al. 98] Transition from heavy (at 10 17.5 eV) to light composition (at ~10 19 eV) Similar limits from AGASA

11. Angular distribution Galactic plane Supergalactic plane Doublets and triplets Fly’s Eye + AGASA: excess of events (4 %) for 0.4 < E < 1.0  10 18 eV in the direction of the galactic center For E > 4  10 19 eV AGASA and HIRes find no evidence of a large scale anisotropy UHECR must be extragalactic or be produced in an extended halo !! No evidence of an excess along the supergalactic plane

12. The Local Super Cluster (LSC) ~ 40 Mpc ~ 10 Mpc

13. How far UHECRs may come from ? The GZK cut-off and the composition of UHECR

14. PROTONS Energy losses due to

15. The GZK cut-off AGASA ONLY !! may be due to a systematics or to a local excess of sources (top-down models) or To a local confinement of UHECRs due to strong magnetic fields in the LSC

16. NEUTRONS They decay after travelling a distance COMPOSITE NUCLEI (Less interesting in this context since they smaller Larmor radius) Same energy losses as for the proton + photodisintegration At 10 20 eV they cannot travel more than ~ 10 Mpc

17. PHOTONS (electromagnetic showers) Protheroe & Johnson ‘96 (NO MAGNETIC FIELDS)

18. The large scale isotropy is not a problem for protons For E ≤ 5  10 19 eV l E ≥ 1 Gpc The flux is dominated by far source  No expected excess along the Supergalatic plane ! The some doesn’t hold for photons since l E (  ) < 10 Mpc A large component of UHECRs is composed by protons or nuclei

19. Why UHECR sources are astrophysical: The spectrum of CR is practically a single power law There are evidence for small scale clustering Hints for a correlation between UHECR and BL-Lacs

20. Small scale clustering AGASA: 92 events with E > 4  10 19 eV - 8 doublets - 2 triplets Chance probability to reproduce them with a homogeneous distribution of sources: P chance < 1 % AGASA + Yakutsk [Tiniakov & Tkachev, 01] P chance < 10 -5 (different treatment of triplets) SIGNIFICANT INDICATION THAT SOURCES MAY BE POINT LIKE

21. The BL-Lacs - UHECR connection Chance probability to reproduce that with uniform sources P < 10 -4 –From a sample of 22 BL-Lacs 5 coincide with UHECR arr.dir. Crossing BL-Lacs catalogue with EGRET sources gives 12 BL-Lacs –4 UHECR source candidates are among these 12 ! –The remaining 10 correlate with UHECR after correcting for the galactic field ( if Q = + 1). Gobunov et al. astro-ph/0204360 Tinyakov & Tkachev astro-ph/0102476

22. Galactic and intergalactic magnetic fields Observations

23. Galactic magnetic field

24. Nuclei Deflections in the Milky Way – the regular field give rise to undetectable deflections at 10 20 eV the galactic disk signature should be visible if sources are galactic and UHECR are not heavy nuclei – Sizeable deflections at 10 19 eV which, however, can be disentangled ! [See e.g. Tinyakov & Tkachev ‘03] – The random field produce ~ 1 o deflections

25. Magnetic fields in galaxy clusters Coma cl. : visibleX-raysradio Synchrotron emission X-ray non-thermal emissions

26. Faraday Rotation Measurments (RMs) Coma RMs

27. MF in the Inter-Galactic Medium (IGM) Only upper limits are available based on the Faraday Rotation Measurements of Quasars’ radio emission at cosmological distance (z ~ 1) LCLC First limit by Kronberg’94 who, however assumed n e = const. and  b = 1 !! Then Blasi, Burles and Olinto, `99 accounted for n e inhomogen. in Ly-alpha clouds UHECR deflections may be large if these limits are saturated !

28. Possible origin of magnetic fields 1.Primordial seed + dynamo amplification in galaxies + pollution of the IGM by galactic winds 2.Ejection from AGN’s + amplification during the cluster accrection 3.Battery + amplification during the cluster accrection 4.Primordial origin ( phase transitions in the early universe, generation at neutrino decoupling, the inflation) + amplification during the cluster accrection [For a review see D.G. & Rubinstein ‘01] MHD SIMULATIONS FAVOUR 2-4

29. MSPH simulations in galaxy clusters N-body simulations of DM + gas hydrodynamics (SPH) + MHD MSPH: Magnetic Smooth Particle Hydrodynamics [Dolag, Bartelmann & Lesch, astro-ph/0109541, 0202272]. gas “particle” + magnetic field DM “particle” The initial DM fluctuations are fixed at z = 20 compatibly with  CDM Smoothing length

30. Magnetic field evolution The MF is evolved starting from a seed (AGNs, battery or primordial) The electric conductivity of the IGM is practically infinite The magnetic field is amplified by: ADIABATIC COMPRESSION: MHD amplification by shear flows: The memory of the initial power spectrum is erased !

31. Simulated RMs RMs profile is reproduced for

32. Reconstructing the magnetic structure of the local universe Basic assumption: magnetic fields in galaxy clusters are originated by a uniform primordial seed generated at high redshift Motivations: - this is easier to be implemented numerically - it should give the largest deflections of UHECRs Approach: - we combine MSPH simulations previously performed for galaxy clusters with constrained simulations of the local universe LSS

33. Simulations of IGMF in the local universe We need a realistic 3D simulation of the LSS such that the size and position of simulated structures (clusters) coincide with those observed on the sky. We need to know the observer position in the magnetized structure. The size of the simulation volume has to be > 50 Mpc such to enclose the Local Super Cluster (LSC) the GZK sphere REQUIREMENTS

34. Constrained simulations Initial conditions (density fluctuations) are chosen randomly from a Gaussian field with a power spectrum compatible with  CMD cosmology but constrained so that the smoothed density field coincide with that observed. Simulations of this kind have been performed, successfully, to simulate the local flow [Mathis, Springel, White et al., astro-ph/0111099]

35. OUR SIMULATION Simulation volume: 80/h Mpc = 107/h 70 Mpc gas mass resolution 5 x 10 8 M sun ; max spatial resolution ~ 10 kpc Initial redshift: z = 60 Initial magnetic field: B 0 = 1 x 10 -11 G l C = ∞ CONSTRAINED SIMULATION OF DM AND GAS + MSPH We use IRAS 1.2 galaxy catalogue to constrain initial conditions

37. Results of the MHD simulation CLUSTERS: B c = 1 - 10  G - OK with RMs - L C << L: no memory of the initial field structure High density FILAMENTS: B c = 10 -2 - 10 -3  G ; (no memory) Low density FILAMENTS: B c = 10 -4  G ; field aligned with the fil. axis

38. RAY TRACING We sum deflections produced by every gas particle If  > 5 o trajectories are ignored Proton energy: E = 4  10 19 eV  Energy losses can be neglected up to ~ 1 Gpc ( deflections at larger energies can only be smaller !! ) Smoothed gas particle

39. Deflections up to 107 Mpc

46. What we learn from these maps ? UHECR cannot be isotropized !! Large deflections are produced only in gal. clusters which cover only a tiny fraction of the sky Filaments produce small deflections (< 2 o ) At large distances the bulk of deflections is due to filaments No observable deflections in the local region and in the voids

47. Sky fraction covered by observable deflections Self-similarity is consistent with a uniform density of deflectors (filaments) For d   a large number of filaments is crossed, each time giving a random     0.5 There is a considerable fraction of the sky where correlation with source is preserved !! OK with small scale clustering and UHECR-BL Lacs correlation

48. The effect of a possible unclustered field Observable only if l C > 20 Mpc Quite hard to get unless the field is primordial and generated during the inflation To be expected only if the field is of primordial origin

49. Conclusions We performed the first simulation of the magnetic structure in the local universe The simulation reproduce MF observed in galaxy clusters and give new hints on the MF in less dense regions like filaments According to our results the LSC is weakly magnetized UHE protons undergoes large deflections only when they pass trough clusters and ~ 1 o deflections in filaments Our results are compatible with small scale clustering and UHECR-BL Lacs correlation CHARGE PARTICLE ASTRONOMY SHOULD BE POSSIBLE