Physics of KM3NET- Version 0
Work Package 2, 20/10/07
1.0 Introduction
After several decades large-scale neutrino astronomy is reaching a state of experimental maturity,
as demonstrated by the ANTARES and AMANDA/IceCube experiments. Nevertheless, despite
all efforts, up to now no high-energy neutrino from outside the Earth’s atmosphere has been
identified. It is generally recognised that even the discovery of a few astrophysical neutrino
events would be a major discovery. In this respect, KM3NET will allow to explore with
unprecedented sensitivity and angular resolution a large region of our Universe, including the
Galactic Centre that is not accessible to similar detectors at the South Pole. Its full-sky coverage
and 24hr/24hr operation are well suited to detection of transient or unexpected phenomena.
The undisputed galactic origin of cosmic rays at energies below the so-called knee implies an
existence of a non-thermal population of galactic objects which effectively accelerate protons and
nuclei to TeV-PeV energies. The distinct signatures of these cosmic accelerators are high energy
neutrinos and gamma rays produced through hadronic interactions with ambient gas or
photoproduction on intense photon fields near the source. In either case, the neutral pions will
decay to photons and the charged pions will include neutrinos in their decay products. While
gamma rays can be produced also by directly accelerated electrons, high-energy neutrinos
provide unambiguous and unique information on the hadronic nature of the accelerated particles.
Unlike -rays, neutrinos are not fragile; they interact only weakly with the ambient medium - gas,
radiation and magnetic fields, and thus carry information about high energy processes occurring
in ”hidden” regions where the particle accelerators could be located. This concerns, first of all,
the regions associated with compact objects - black holes, pulsars, the initial epochs of
supernovae explosions, etc. The penetrating potential of neutrinos is important not only for
extremely dense environments in which -rays are dramatically absorbed, but also moderately
opaque sources from which we do see -rays, but after significant distortion due to internal and
external absorption. Ironically, this nice (from an astrophysical point of view) feature of neutrinos
makes, at the same time, their detection extremely difficult. This explains why, over several
decades high energy neutrino astronomy has remained essentially a theoretical discipline with
many exciting ideas and predictions but without the detection of a single VHE neutrinos source.
However, it is expected that, with arrival of the km3-volume class scale detectors like IceCube
and KM3NeT (see e.g. [1, 2]), the status of the field will be changed dramatically.
Presently extragalactic objects like Active Galactic Nuclei (AGN) and sources of Gamma Ray
Bursts (GRBs) are believed to be the most likely objects to be detected as neutrino sources. The
current models of AGN and GRBs indeed contain many attractive components (concerning the
conditions of particle acceleration and their interactions) which make these objects potentially
detectable sources of VHE neutrinos. On the other hand, the poor understanding of many aspects
of the physics of AGN and especially GRBs, as well as the lack of constraints on neutrino
productions rates from -ray observations (because of intrinsic and intergalactic absorption of
VHE -rays), give rise to large uncertainties.
1
On the other hand, the models of potential galactic neutrino sources, in particular the shell type
Supernova Remnants (SNRs), Pulsar Wind Nebulae (PWNe), Star Formations Regions and the
dense molecular clouds related to them, are robustly constrained by -ray observations of the
galactic disk in very-high energy (>1 TeV) [5, 6] and ultra-high energy (>100 TeV) [7] domains.
Typically, the expected fluxes from these objects are below the detection threshold of the planned
neutrino detectors. However, the recent HESS discoveries of several TeV -ray sources at the flux
level of ”1 Crab”, which can be interpreted within the hadronic models of gamma-ray emission,
sustain a hope that that the first TeV galactic sources will be detected in foreseeable future by
km3-volume class instruments like IceCube and Km3NeT.
KM3NET is a multi-purpose detector. Besides high-energy neutrino astronomy, a number of
other important physics questions can be addressed, for example dark matter annihilations,
magnetic monopoles etc.
Figure 1: One year flux limits as a function of the source declination for the reference KM3NET
detector, compared to other experiments.
2.0 Galactic sources [5, 6]
Recent performance studies [1,2,3,4] of the km3-volume scale detectors show that the detection
of a persistent point-like neutrino source for a realistic exposure time (typically, a few years of
continuous observations) is limited to a flux F(>1TeV) ~10−11 /cm2/s. The corresponding energy
flux is ~10−10erg/cm2/s or somewhat less, depending on the spectrum in the most relevant energy
band between 1-100 TeV. The accompanied gamma-ray flux of ~1 “crab” can be considered as
the detection threshold for galactic neutrino astronomy with km3-volume class detectors.
2
The recent progress in Galactic gamma ray astronomy using the HESS instrument provides a
comprehensive list of bright Galactic TeV gamma ray sources for which the neutrino flux can be
reliably estimated within the framework of a hadronic model. These sources are located in the
southern hemisphere and are therefore not visible to south-pole detectors looking for upgoing
neutrinos.
3
2.1 Supernova Remnants (SNRs) and Pulsar Wind Nebulae
Explosions of massive stars (supernovae) produce an expanding shell of material which is known
from radio observations to accelerate high-energy particles. In some cases, the residue of the
supernova is a neutron star which is detectable as a pulsar. Protons inside supernova shells can be
accelerated by a first-order Fermi mechanism if (as seems likely) the shell is turbulent. If a pulsar
is present there are additional acceleration mechanisms: in the magnetosphere of the pulsar, or at
the front of the shock wave produced by the magneto-hydrodynamic wind in the shell. The
interaction of these protons with the matter of the shell gives rise to neutrinos and photons (from
charged and neutral pion decays respectively).
Young SNRs are considered the best candidate for CR acceleration. HESS has observed a
number of SNRs of which two, RXJ1713.7−3946 [7] and RXJ0852.0−4622 [8], are young and
very bright. Their hard, intense TeV gamma-ray spectrum, are best explained as being hadronic
in nature. HESS has also measured a TeV spectrum from Vela X, the pulsar wind nebula (PWN)
associated with the larger Vela SNR [9]. If, as proposed in Ref. [10], this spectrum is hadronic
rather than leptonic, the accompanying neutrino flux would be detectable [4].
2.2 Compact binary systems
This class of binary stars, which are among the brightest cosmic X-ray sources, consists of
compact objects, such as neutron stars or black holes, which accrete matter from their normal
companion stars. The accretion process leads to plasma waves in the strong magnetic field of the
compact object, which bring protons to high energies by stochastic acceleration. Interactions of
the accelerated particles with the accreting matter or with the companion star would then produce
a neutrino flux comparable to that in high-energy particles with a spectral index close to 2.
The recent detections of TeV gamma rays from two binary systems, tentatively called
microquasars, LS5039 [11] and LSI 61 303 [12] clearly demonstrate that galactic binaries, are
sites of effective acceleration to multi-TeV energies for which hadronic models are well
motivated [13]. In this case, the severe internal absorption expected in the source would allow up
to a factor 10-100 increase in neutrino flux over that observed in gamma-rays. If the spectrum of
accelerated protons continues to 100 TeV and beyond, LS5039 can be probed by km3-scale
detectors [4].
2.3 Galactic Center Diffuse Emission
HESS has also recently discovered a region of diffuse TeV emission from the Galactic Center
ridge [14]. The large extent of the emission, hardness of the spectrum, high gas density (which is
well-correlated with the TeV emission) and strong magnetic fields in the region indicate a
hadronic origin. In addition, the total flux from this region is actually about twice as intense as
that of the previously discovered source coincident with Sgr A*. Measurement of an
accompanying neutrino flux would provide an independent confirmation of the production
mechanism.
2.4 No Counterpart
HESS has reported the discovery of several TeV sources in the Galactic Plane which have no
apparent counterparts at other wavelengths [15]. The lack of x-ray synchrotron emission is
indicative of a hadronic interpretation. In addition to these, EGRET previously discovered
4
numerous unidentified sources of GeV gamma rays in the MilkyWay which are also potential
sources for neutrinos [discuss Semikoz arguments here].
2.5 Hidden sources
The possibility of proton accelerators, completely shielded from us by very thick gas material is
certainly not excluded [16]. For example the “converter mechanism” (multiple conversions of
protons to neutrons through photomeson reactions) can significantly enhance proton acceleration
and in relativistic flows with large aspect angles could naturally lead to the appearance of
“orphan” neutrino sources [17].
3.0 Extragalactic Sources [18]
Models for the origin of the highest energy cosmic-rays typically predict associated neutrino
fluxes. A requirement on the sources is that they must provide sufficient power to supply the
derived energy in the extragalactic component of the cosmic radiation. One natural scenario, that
gives comparable energy in neutrinos and cosmic rays, occurs when protons remain trapped in
the acceleration region until they suffer inelastic collisions, while secondary neutrons escape and
decay to become the cosmic-ray protons.
The observations of the diffuse fluxes of gamma-rays and cosmic rays have been used to set
theoretical upper bounds on the diffuse neutrino flux, the cumulative flux of very high-energy
neutrinos from unresolved cosmic sources. Most predictions were made using a source model to
determine the spectral shape, which was normalized to the bolometric total flux of extragalactic
gamma rays background to obtain upper bound of the neutrino flux of the order of [6.5, 6.13].
E2   10 6
GeVcm 2 s 1 sr 1
(6.1)
The limit on the diffuse neutrino flux from the corresponding cosmic ray sources was first
discussed by Waxman and Bahcall (WB) [6.14]. They claimed that the measured flux of
ultrahigh-energy cosmic rays provides the most restrictive limit on extragalactic diffuse neutrino
fluxes for a broad class of sources. This limit is two orders of magnitude lower than the bound
from the extragalactic gamma ray background. The WB limit is mainly based on the three
assumptions that (i) high energy neutrons (En1019 eV) produced in photohadronic interactions
can escape freely from the astrophysical source; (ii) magnetic fields in the Universe do not affect
the observed flux of extragalactic cosmic rays, and (iii) the overall injection spectrum of
extragalactic cosmic rays is dN/dEE-2. The WB upper bound from cosmic ray observation is:
1
E 2    1.5  10 8  z  GeVcm  2 s 1 sr 1 ,        e (6.2)
2
The quantity z in eq. (5.2) is 1 if the redshift energy loss of neutrinos is neglected and if the
cosmic-ray generation rate per unit volume is assumed independent of cosmic time. If also the
evolution and the redshift losses were considered, the authors estimate z =3.
The WB limit was extensively discussed by Mannheim, Protheroe and Rachen (MPR) [6.15,
6.16]. MPR obtained a bound for a possible flux of extragalactic neutrinos in the energy range
1014 eV E 1020 eV based on the observed flux of cosmic rays and gamma rays. They
considered also two extreme kinds of sources: opaque and transparent to neutrons. The bound
obtained for transparent sources (shown as n<1 in Fig. 6.1) is in agreement with the WB upper
bound in the range 1016 eV E 1018 eV, but is higher at lower and higher energies. At energies
lower than 1016 eV, and higher than 1018 eV, the bound is at the same level of the bound obtained
5
from extragalactic gamma rays (eq. 6.1). The limit for sources opaque to neutrons, not considered
GeVcm2 s 1 sr 1 .
by the WB bound (n»1 in Fig. 6.1) is E2   2  10 6
Figure 6.1: The MPR muon neutrino upper bounds for optically thin pion photoproduction
sources (curve labelled n<1) and optically thick sources (curve labelled n»1). The hatched
range between the two curves can be considered the allowed region for upper bounds for
sources with n>1. The bold line is the WB bound for an evolving source distribution. For
comparison, two predictions for optically thin sources are also shown: proton-blazar model
[6.5] (dotted curve), and GRB sources [6.4] (dashed curve). Also shown is the atmospheric
background [6.17] (adapted from [6.15]).
From the experimental point of view, since there is no directional information, the only way to
detect diffuse neutrinos against the background of the atmospheric neutrinos is looking from an
excess of high-energy events in the measured energy spectrum. This approach has been adopted
by existing detectors (MACRO [6.6], Frejus [6.7], SuperKamikande [6.8], AMANDA [6.9],
Baikal [6.10]) and experiments in construction (ANTARES [6.11], ICECUBE [6.12]).
6
Figure 2: Limits on diffuse neutrino fluxes for an example KM3NET detector compared with
limits from other neutrino telescopes.
Presently, extragalactic objects such as Gamma Ray Bursts (GRBs) and Active Galactic Nuclei
(AGN) are believed to be the most likely origin of the extra-galactic cosmic rays. The
observation in TeV gamma rays of the most distant of these objects is severely hampered by
absorption, via pair production, on the extragalactic background; possible associated neutrinos
would not be affected.
3.1 Gamma Ray Bursts
Gamma-ray bursts (GRBs), short flashes of gamma rays lasting for a few seconds, are observed
about once a day, randomly distributed across the sky. Most bursts occur at cosmological
distances. Possible progenitor models include massive fast rotating star whose iron core collapses
into a black hole which accretes fast spinning surrounding material, or close compact binary
systems where one companion accretes material fast from the other. In both scenarios, fast
accretion releases huge amount of energy, a fireball rushes out near the speed of light, and due to
the surrounding stellar pressure, it collimates into a relativistic jet. A large fraction (~10%) of the
fireball energy is converted into a burst of neutrinos of energy ~100 TeV, created by the
interaction in the relativistic jet between gamma rays and accelerated protons with energies up to
1020 eV [5.1]. Collision of the expanding fireball with the surrounding medium results in the
GRB afterglow, during which neutrinos of energies up to 1017-1019 eV can be created, following
the burst by ~10s. The GRB neutrino flux is not expected to extend above ~1019 eV [5.2]. The
expected flux spectra are illustrated in Figure 5.1. As these events are correlated in time and
direction with gamma-ray signals detected by satellite alert systems (SWIFT, INTEGRAL etc.),
such searches are essentially background free.
Quote some expected rates here
7
3.2 Active Galactic Nuclei
AGNs are believed to be supermassive black holes accreting matter from the nucleus of the host
galaxy. They exhibit highly collimated jets of relativistic plasma with opening angles of a few
degrees or less. Blazars have their jet axis aligned close to the line of sight of the observer, giving
a significant flux enhancement through Doppler broadening. A total of 66 blazars have been
detected as GeV gamma ray sources by EGRET. In more than 10 of these sources the gamma ray
spectrum is observed to extend above a TeV [22]. In the framework of a hadronic model, the
sources, 1ES1101-232, H2356-309 and PKS2155-304 are very promising [23].
In addition, the blazar emissions can be highly variable on short time scales, for example HESS
recently observed a two order of magnitude flux increase of PKS 2155-304, reaching ~10*crab,
during a one hour period [24]. Such flaring episodes are well motivated targets of opportunity for
neutrino telescopes.
Table 1: Upper limits for the integral event rate from candidate Blazars above 1 TeV and 5 TeV
including the expected atmospheric neutrino rate and the statistical significance of source over
the background for 5 years of observing with the standard KM3NeT geometry [23].
3.3 GZK
At energies higher than 5.1019 eV, cosmic rays experience important energy losses due to
photopion production reactions with the CMB photons. The decay of charged mesons generated
in these interactions leads to the production of neutrinos, generally called “cosmogenic
neutrinos”. This mechanism, which was first proposed in 1969 by Berezinsky and Zatsepin [7.5],
provides a guaranteed ultra high-energy neutrino flux.
Cosmogenic neutrino fluxes are generally calculated with detailed CRs propagation simulation
codes, taking into account the different energy losses that CRs experience along their journey.
Secondary charged pion decay produces neutrinos whereas neutral pion decay produces highenergy photons that will cascade by e+e- pair production on CMB photons to the GeV energy
range
The resulting neutrino flux mainly relies on the properties of the injected CR spectrum. The
implementation of the evolution of the CRs sources luminosity with redshift gives neutrino fluxes
up to a factor of 10 greater compared to the case of no evolution with redshift. The EGRET data
on diffuse  ray in the GeV range constraint the cosmogenic neutrino flux, since the total amount
of energy emitted in neutrinos is related to the total energy emitted in . The authors of [7.7]
derived a maximal neutrino flux from UHECRs protons allowed by the EGRET observations.
The results are presented in Fig. 7.1, together with the current most stringent limits.
8
Figure 7.1: Maximal cosmogenic neutrino flux per flavor in agreement with EGRET data,
obtained in [6.7] by tuning the parameters of the proton injection spectrum. Red line: -ray flux,
green line: neutrino flux, blue-dotted line: proton spectrum, pink points: observed UHECRs
spectrum.
As explained above, the UHECRs can produce neutrino fluxes at very high energy, through pion
photoproduction on CMB photons. This process can also occur at lower energies, when the CRs
interact with the extragalactic infrared and optical background (IRB/Opt in the following)
photons. In 2004, Stanev [7.8] gave an estimate of the cosmogenic neutrino flux generated in
interactions of CRs on the IRB, but did not account properly for the cosmological evolution of
the IRB. In Ref [7.9], a more realistic model of the IRB evolution with redshift is used to
compute the neutrino fluxes produced from the interaction of UHECR protons with the IRB
photons.
9
Figure 7.2: Predicted cosmogenic neutrino flux for pure proton and for mixed composition in the
case of CRs interactions with CMB and IR/Opt/UV background photons. The sensitivities of
Auger (for ) and ANITA at high energy are also represented together with the limit of Ice Cube
(for the  channel) after 3 years of observation with 90% CL.
3.4 Top-Down Models
Topological defects, Z bursts, from antares CDR
An exciting possibility is that UHECRs result from the decay of massive particles, rather than
being accelerated up from low energies. The most popular models in this context are based on the
annihilation or collapse of topological defects (TDs) such as cosmic strings or monopoles formed
in the early universe [17]. When TDs are destroyed, their energy is released as massive gauge and
Higgs bosons with masses of O(1025) eV if such defects have formed at the GUT-symmetry
breaking phase transition. The decays of these bosons can generate cascades of high energy
nucleons, g-rays and neutrinos. These models are constrained both by considerations of the
cosmological evolution of TDs [18] and observational bounds on the extra-galactic g-ray
background [19]. These require the mass of the decaying bosons to be less than 1021 eV, i.e. well
below the GUT scale. Since only GUT scale TDs have independent motivation, e.g. to provide
‘seeds’ for the formation of large-scale structure, the above constraint thus disfavours TDs as the
source of UHECRs.
A more recent suggestion is that UHECRs arise from the decays of metastable relics with masses
exceeding 1021 eV which constitute a fraction of the dark matter [20, 21]. Such particles can be
naturally produced with a cosmologically interesting abundance during re-heating following
inflation [22]. A lifetime exceeding the age of the universe is natural if they have only
gravitational interactions, e.g. if they are ‘cryptons’— bound states from the hidden sector of
string theory [23, 21]. This interpretation also naturally accounts for the required mass. A detailed
study of the fragmentation of such heavy particles has been performed [21] in order to calculate
the expected spectra of nucleons, -rays and neutrinos. Such particles would, like all ‘cold dark
matter’ particles, be strongly clustered in the Galactic halo, i.e. within 100 kpc. Therefore, the
extra-galactic contribution to the cosmic ray flux would be negligible in comparison and the
observed flux fixes the ratio of the halo density to the lifetime. As an example, if such particles
comprise all of the halo dark matter then the required lifetime is 1020 yr, with a proportionally
shorter lifetime for a smaller contribution. Most of the energy in the cascade ends up as neutrinos
and the predicted flux is then F(> E_) _ 108(E_=1 TeV )−1 km−2y−1sr−1, through normalisation to
the observed UHECR flux. Furthermore, the neutrinos should be well correlated in both time and
arrival direction with the UHECRs, given the relatively short propagation length in the halo. A
small departure from isotropy of O(20%) should also be observed, since our location is _ 8 kpc
from the galactic centre [24]. This anisotropy will be less than that for UHECRs, however, since
there is no GZK cutoff to reduce the extra-galactic (isotropic) flux of neutrinos [25].
The essential point is that whatever process creates the UHECRs, it is exceedingly likely that
there is a concomitant production of very high energy neutrinos. Measurement of the neutrino
flux will, at the very least, provide important clues as to the origin and may even provide
dramatic evidence for new physics.
10
3.5 Neutrino Flavour Ratios
4.0 Neutrinos from dark matter annihilations
While the usual picture of cold dark matter is in good agreement with astrophysical and
cosmological observations, its true quantum nature is still unknown; the consensual particle
description is that of a stable weakly interacting and massive particle (WIMP). Among possible
signatures of dark matter candidates, the assumption of its thermal origin in the primordial
universe without matter/antimatter asymmetry leads to expected annihilation processes today,
whose rate scales like the square of its local density. In this context, the WMAP experiment [3.1]
has measured a dark matter density of dm~0.25, indicating a mean annihilation cross-section of
<v>~3.10-26cm3/s. Astrophysical regions with very high dark matter densities would be natural
sites of WIMP annihilation and therefore potential sites for neutrino production.
High-energy neutrinos are very interesting messengers that could reveal annihilation of dark
matter, being produced either directly or resulting from the fragmentation (respectively decay) of
quarks (massive gauge bosons,  and  leptons), depending on the particle physics model. The
former case is favoured in universal extra-dimensional (UED) theories, whereas the later case is a
standard feature of supersymmetric (SUSY) theories (see [3.2] for a review). The most interesting
source of such exotic neutrinos is the Sun, in the sense that it is expected to be an efficient and
very close gravitational trap for dark matter. Wimps are expected to loose kinetic energy by
elastic collisions inside high matter density regions like the sun, resulting in their capture, and the
annihilation rate is then governed by the capture rate as soon as the equilibrium is reached. As the
capture rate is determined by the elastic cross-section of wimps off nuclei inside the Sun
(dominantly spin independent), this indirect detection is complementary to direct searches that
focus on energy deposits on deep-underground targets (dominantly spin dependent).
The typical energy of the neutrino-induced muons would be of the order of ~25% of the
neutralino mass. Since the muon threshold for KM3NET is ~100 GeV, we expect KM3NET to be
sensitive mainly to neutralinos heavier than about 400 GeV. In contrast to south-pole detectors,
the location of KM3NET is very well suited for detection of neutrinos produced in the sun or
around the black hole at the galactic centre. For example, a 400 GeV WIMP, coupling largely by
spin-dependent scattering and having a cross-section near the current direct detection limit, could
produce thousands of detected neutrinos [25].
The Sun is a point-like source of neutrinos from neutralino annihilation, but the Earth is not,
especially for lower mass neutralinos. Despite the smearing produced by measuring the muon
flux rather than the neutrinos, a detector with good angular resolution would be able to use the
observed angular distribution to constrain the neutralino mass, as shown in figure 2.3 from [40].
11
Figure x. neutrino detection rates in the reference km3net detector for a scan of the mSUGRA
parameter space. Points in blue are excludable, dark blue points are those favoured by WMAP.
5.0 Magnetic monopoles
Any early universe phase transition occurring after inflation has the potential to populate the
Universe with a flux of magnetic monopoles. Observations of galactic magnetic fields, as well as
observations matched with models for extragalactical fields lead to the conclusion that
monopoles of masses below 1015 GeV can be accelerated in these fields to relativistic velocities
[ref]. Monopoles may be at the origin of baryon number violating processes [43] and have been
searched for in proton decay experiments. A magnetic monopole with unit magnetic Dirac charge
g = 137=2 _ e and a velocity _ close to 1 would emit Cerenkov radiation along its path, exceeding
that of a bare relativistic muon by a factor of 8300 for =1. Due to the production of -electrons,
a monopole produces light even below its own Cerenkov threshold.
Figure X summarizes the limits obtained until now. Note that these limits are below the so-called
Parker bound (1015 cm-2 s-1 sr-1). This bound is derived from the very existence of galactic
magnetic fields which would be destroyed by a too high flux of magnetic monopoles. A cube
kilometer detector could improve the sensitivity of this search by about two orders of magnitude
compared to the present AMANDA limit. As mentioned above, the search could be extended to
down to velocities =0:5 by detecting the  electrons generated along the monopole path.
12
Figure 53: Limits on the flux of relativistic monopoles achievable with IceCube compared to
existing limits.
6.0 Other physics
In addition, the intrinsic properties of neutrinos can also be studied (neutrino oscillations, Lorentz
and CPT invariance etc.) [27].
7.0 New phenomena
Whenever a new window is opened on the Universe, unexpected phenomena are observed. This
has been verified in numerous cases throughout history, from the observation of the moons of
Jupiter by the first Galilean telescopes to the discovery of pulsars with radio-based astronomy.
Neutrino astronomy provides an exciting new probe of the Universe to the highest possible redshifts in an energy window so far not observable by other techniques. There is plenty of scope for
surprises.
References
[1] Vissani F., 2006 Astropart. Physics 26 310.
[2] Kistler M. and Beacom, J. F., Phys. Rev. D 74 (astro-ph/067082 v1).
[3] Distefano C., 2007 Astrophys Space Sci, in press, astro-ph/0608514.
[4] Kappes A., Hinton J., Stegmann Ch. and Aharonian F., 2007 Astrophys. J. 656 870.
[5] W. Bednarek, G. F. Burgio, T. Montaruli, Astro-ph/0404534.
[6] F. Aharonian, astro-ph/0702680.
[7] HESS collaboration 2004 Nature 432 75; 2006 Astron. Astrophys. 449 223.
[8] HESS collaboration 2005 Astron. Astrophys. 437 L7.
[9] HESS collaboration 2006 Astron. Astrophys. 448 L43.
[10] Horns D, Aharonian F, Santangelo A, Hoffmann A and Masterson C, 2006 Astron.
Astrophys 451 L51.
[11] HESS collaboration, 2005 Science 309 746; Astron. Astrophys 460 743.
[12] Albert J et al (MAGIC collaboration), 2006 Science 312 1771.
[13] Aharonian F., Anchordoqui L., Khangulyan D. and Montaruli T., 2006 J of Physics: Conf.
Ser. 39 408.
[14] HESS collaboration, nature 439, 695 (2006).
[15] HESS collaboration, Astro-ph/0510397.
[16] Berezinsky, Dokuchaev, Astro-ph/0002274.
13
[17] Derishev E. V., Aharonian F. A., Kocharovsky V. V. and Kocharovsky Vl. V., 2003 Phys.
Rev. D 68 id.043003.
[18] E. Waxman, astro-ph/0502159.
[19] E. Waxman & J. N. Bahcall, Phys. Rev. D59, 023002 (1999); J. N. Bahcall & E. Waxman,
Phys. Rev. D64, 023002 (2001).
[20] K. Mannheim, R.J. Protheroe, and J.P. Rachen, Phys. Rev. D63, 023003 (2001).
[21] Waxman E, 2007 Science 315 63.
[22] R. Mukherjee et al., Astrophys. J. 490 (1997) 116.
[23] R. White., Proc. ICRC2007, Merida.
[24] W. Benbow, HESS collaboration, arXiv:0709.4608 [astro-ph].
[25] F. Halzen & D. Hooper, hep-ph/0510048 v1.
[26] V. Popa, NIM A567, (2006) 480.
[27] Anchordoqui et al., hep-ph/0510389.
14