G ü nter Sigl Astroparticule et Cosmologie, Université Paris 7

IDAPP 2006, Munich, May 2006 Günter Sigl, Astroparticules et Cosmologie, Paris

Günter SiglAstroparticule et Cosmologie, Université Paris 7GReCO, Institut d’Astrophysique de Paris, CNRShttp://www2.iap.fr/users/sigl/homepage.html

Gamma-Rays, Cosmic RaysAnd Neutrinos

Introduction to the observational situation Introduction to the propagation of cosmic rays and electromagnetic cascades Applications to diffuse and point source fluxes of gamma-rays and neutrinos Neutrino flux sensitivities, detection techniques, and neutrino cross sections Testing neutrino properties with astrophysical neutrinos


The structure of the spectrum and scenarios of its origin

supernova remnants wind supernovae AGN, top-down ??

toe ?


Supernova Remnants and Galactic Cosmic and -Rays

Supernova remnants have been seen by HESS in γ-rays: The remnant RXJ1713-3946has a spectrum ~E-2.2: => Charged particles have been accelerated to > 100 TeV.Also seen in 1-3 keV X-rays (contour lines from ASCA)

Aharonian et al., Nature 432 (2004) 75


Hadronic versus leptonic model of SN remnant HESS J1813-178:both are still possible

Albert et al., ApJ 637 (2006) L41


These are either electromagnetic or hadronic accelerators

If hadronic, then the γ-rays are produced by pp -> ppπ0 -> γγ

σpp ~ [35.49 + 0.307 ln2(s/28.94 GeV2)] mb

This cross section is almost constant -> secondary spectra roughly thesame shape as primary fluxes.


HESS sources: X-ray binary LS 5039

Expected neutrino fluxes above TeV 10-9-10-7 GeV cm-2s-1

γ-ray injection fluxdepends on locationof γ-ray production

F.Aharonian et al., astro-ph/0508658


Hadronic Interactions and Galactic Cosmic and -Rays

HESS has observed γ-rays from objectsaround the galactic centre which correlatewell with the gas density in molecularclouds for a cosmic ray diffusion time ofT ~ R2/D ~ 3x103 (θ/1o)2/η years whereD = η 1030 cm2/s is the diffusion coefficientfor protons of a few TeV.

Aharonian et al., Nature 439 (2006) 695


Given the observed spectrumE-2.3, this can be interpretedas photons from π0 decayproduced in pp interactionswhere the TeV protons havethe same spectrum and couldhave been produced in a SNevent.

Note that this is consistent with the source spectrum both expected fromshock acceleration theory and from the cosmic ray spectrum observed in thesolar neighborhood, E-2.7, corrected for diffusion in the galactic magneticfield, j(E) ~ Q(E)/D(E).


conventional hard electron injection

optimized

Diffuse γ-ray spectra predicted and observed by EGRET

Strong, Moskalenko, and Reimer, ApJ 613 (2004) 962

Above 100 MeVdominated bypp induced γ-rays


The galactic neutrino flux is comparable to the galactic diffuse γ-ray flux

Candia and Beacom, JCAP 0411 (2004) 009


Optimized astrophysical modelleaves no room for new component.

In contrast, conventional astrophysicalmodel allows for a dark matterannihilation component.

de Boer et al., astro-ph/0408272


Lowering the AGASA energyscale by about 20% brings itin accordance with HiRes upto the GZK cut-off, but notbeyond.

May need an experiment combining ground array with fluorescence such asthe Auger project to resolve this issue.

HiRes collaboration, Phys.Lett.B 619 (2005) 271

The Highest Energy Cosmic Rays


The Greisen-Zatsepin-Kuzmin (GZK) effect

Nucleons can produce pions on the cosmic microwave background

nucleon

-resonance

multi-pion production

pair production energy loss

pion production energy loss

pion productionrate

sources must be in cosmological backyardOnly Lorentz symmetry breaking at Г>1011

could avoid this conclusion.

eV1044

2 192

th

mmm

E N


M.Boratav

1st Order Fermi Shock Acceleration

The most widely accepted scenarioof cosmic ray acceleration

u1

u2

Fractional energy gain per shockcrossing on a time scalerL/u2 .Together with downstream lossesthis leads to a spectrum E-q withq > 2 typically.When the gyroradius rL becomescomparable to the shock size L,the spectrum cuts off.

21 uu

upstreamdownstream


A possible acceleration site associated with shocks in hot spots of active galaxies


Rough estimate of neutrino flux from hadronic AGN jets

Following Halzen and Zas, Astrophys.J. 488 (1997) 669


Diffuse Neutrino fluxes from AGN jets

Halzen and Zas, Astrophys.J. 488 (1997) 669


Loeb and Waxman, astro-ph/0601695

A “guaranteed” flux from starburst galaxies:

Note: this is TWICE the fluxper flavor

E-2

E-2.25

Idea: protons loose most of their energy in form of pions => secondary electrons produce radio synchrotron => can be related to secondary neutrinos


70 km

Pampa Amarilla; Province of Mendoza3000 km2, 875 g/cm2, 1400 mLat.: 35.5° south Surface Array (SD):

1600 Water Tanks1.5 km spacing3000 km2

Surface Array (SD):1600 Water Tanks1.5 km spacing3000 km2

Fluorescence Detectors (FD):4 Sites (“Eyes”)6 Telescopes per site (180° x 30°)

Fluorescence Detectors (FD):4 Sites (“Eyes”)6 Telescopes per site (180° x 30°)

Southern Auger Site


First Auger Spectrum !!107% AGASA exposureStatistics as yet insufficient to draw conclusion on GZK cutoff

Deviation from best fit power law


Ultra-High Energy Cosmic Rays and the Connection to-ray and Neutrino Astrophysics

rays

neutrinoso

XN

p

accelerated protons interact:

during propagation (“cosmogenic”)or in sources (AGN, GRB, ...)

=> energy fluences in -rays and neutrinos are comparable due to isospin symmetry.

Neutrino spectrum is unmodified,-rays pile up below pair productionthreshold on CMB at a few 1014 eV.

Universe acts as a calorimeter fortotal injected electromagneticenergy above the pair threshold.=> neutrino flux constraints.


The universalphoton spectrum


Low energy photon target: Diffuse fluxes


The EGRET background


quasar evolution

Theoretical Limits, Sensitivities, and “Realistic” Fluxes: A Summary

Neutrino flux upper limit for opaque sourcesdetermined by EGRETbound

Neutrino flux upper limitfor transparent sourceslimited by primary cosmicrays: Waxman-Bahcall bound

No source evolution

Armengaud and Sigl


Neutrino-Nucleon Cross Section and Required Detector Size


Ultra-High Energy Neutrino Detection: Traditional and New Ideas

Mostly uses the charged-current reactions: ,, , eiNlN ii

1.) detect Cherenkov radiation from muons in deep sea or ice AMANDA, ANTARES, BAIKAL, NESTOR aims at 1 km3 for E> 100 GeV to 1 TeV

2.) horizontal air showers for electron and τ-neutrinos PIERRE AUGER, MOUNT for E> 1018 eV, increased efficiency for τ-neutrinos if surrounded by mountains on 100 km scale which is decay length of produced taus.

3.) detection of inclined showers from space for E>1020 eV EUSO, OWL

4.) detection of radio emission from negative charge excess of showers produced in air, water, ice, or in skimming rock. RICE (in South-pole ice), GLUE (radio-telescope observing the moons rim)

6.) Earth-skimming events in ground arrays or fluorescence detectors.

5.) acoustic detection in water: hydrophonic arrays


• Neutrinos coming from above are secondary from cosmic rays

• Neutrino coming from below are mixture of atmospheric neutrinos and HE neutrinos from space

• Earth is not transparent for neutrinos E>1015eV

• Former/current experiments: MACRO, Baikal, AMANDA

• Future experiments: ANTARES, ICECUBE, NEMO,

NESTOR…

Experimental Detection of E<1017eV Neutrinos


4-string stage (1996)

First underwater telescopeFirst neutrinos underwater


Neutrino detection in polar ice

O(km) long muon tracks

~15 m

7.0)TeV/(7.0 E

South Pole ice: (most?) transparent

natural condensed material

Longer absorption length → larger effective volume

cascades

event reconstruction byCherenkov light timing


Mediterranean Projects

4100m

2400m

3400mANTARES

NEMO NESTOR


IceCube

1400 m

2400 m

AMANDA

South Pole

IceTop

- 80 Strings- 4800 PMT - Instrumented volume: 1 km3

- Installation: 2004-2010

~ 80.000 atm. per year


Air Shower Detection of UHE (E>1017eV) Neutrinos

• Neutrinos are not primary UHECR

• Horizontal or Earth-skimmingair showers – easy way to detectneutrinos

• Former/current experiments:Fly’s Eye, AGASA

• Future experiments:Pierre Auger, OWL/EUSO…


Neutrino penetration depth

Curved shower => neutrinoFlat shower => hadron


Telescope Array


MOUNT/ASHRA


OWL/EUSO


e + n p + e-

e- ... cascade

relativist. pancake ~ 1cm thick, ~10cm

each particle emits Cherenkov radiation

C signal is resultant of overlapping Cherenkov cones

for >> 10 cm (radio) coherence

C-signal ~ E2

nsec

negative charge is sweeped into developing shower, which acquires a negative net chargeQnet ~ 0.25 Ecascade (GeV).

Threshold > 1016 eV

Radio Detection of Neutrinos


GLUE Goldstone Lunar Ultra-high Energy Neutrino Experiment

E2·dN/dE < 105 eV·cm-2·s-1·sr-1

Lunar Radio Emissions from Inter-actions of and CR with > 1019 eV

1 nsec

moon

Earth

Gorham et al. (1999), 30 hr NASA Goldstone70 m antenna + DSS 34 m antenna

at 1020 eV

Effective target volume~ antenna beam (0.3°) 10 m layer

105 km3


RICE Radio Ice Cherenkov Experiment

firn layer (to 120 m depth)

UHE NEUTRINO DIRECTION

300 METER DEPTH

E 2 · dN/dE < 105 eV · cm-2 · s-1 · sr-1

20 receivers +transmitters

at 1017 eV


SalSA Salt Dome Shower Array


ANITA Antarctic

Impulsive

Transient

Array

Flight in 2006


d

R

Particle cascade ionization heat pressure wave

P

t

s

Attenuation length of sea water at 15-30 kHz: a few km(light: a few tens of meters)

→ given a large initial signal, huge detection volumes can be achieved.

Threshold > 1016 eV

Maximum of emission at ~ 20 kHz

Acoustic detection of Neutrinos


More on Cosmogenic Neutrino Fluxes

Uncertainties due to source evolution Dependence on Cosmic Ray Composition Contributions of Infrared backgrounds


Parameter dependencies of cosmogenic neutrino fluxes: Optimistic cases

Emax=1023 eV, m=3, α=1.5 zmax=2, m=3, α=1.5

Emax=3x1022 eV, zmax=2, α=1 Emax=3x102é eV, zmax=2, m=3


Parameter dependencies of cosmogenic neutrino fluxes:Optimistic cases continued

Emax=1023 eV, zmax=2, m=3, α=1.5

Maximized over all parameters exceptinjection spectrum

Kalashev, Kuzmin, Semikoz, Sigl, PRD 66 (2002) 063004


Influence of infrared background on cosmogenic neutrinos for proton injection

E-2.5 up to 1021.5 eV

Up to ~1 event/year/km3

T.Stanev and D.De Marco, Phys.Rev.D73 (2006) 043003

Low cross-over scenario

E-2.7 up to 1021.5 eV


Influence of Composition on Cosmogenic Neutrinos

p

4He

16O

56Fe

D>Hooper, A.Taylor, S.Sarkar, Astropart.Phys.23 (2005) 11

E-2 up to 1021.5 eVThe highest ratesare 1 event/yearin ICECUBE forprotons,but suggestsrates 10-30 timeslower for iron.

See, however,results by Aveet al.


Influence of Composition on Cosmogenic Neutrinos

M.Ave at al., Astropart.Phys.23 (2005) 19

Standard AGN evolution, E-2 injectionup to 4 Zx1020 eV

Suggests that rates don’t depend much oncomposition, contrary to Hooper et al.

Dependence on maximalenergy


Chemical Composition, Magnetic Fields, Nature of the Ankle

knee2nd knee

Scenario of Berezinsky et al.:Galactic cosmic rays level out at the 2nd knee at ~4x1017 eV where dominatedby heavy nuclei..The ankle at ~5x1018 eV is due to pair production of extragalactic protonson the CMB. Requires >85% protons at the ankle and E-2.6 injection.

“Conventional Scenario”:The ankle at ~5x1018 eV is a cross-over from a heavy Galactic to a lightextragalactic component. Best fit injection spectrum E-2.3


The low cross-over scenario where flux is dominated by extragalactic protonsabove 4x1017 eV may be close to be ruled out by AMANDA.This, however, assumes transparent sources which cosmic rays have to leaveas neutrons which each come with one π+ decaying into neutrinos.

Ahlers et al., Phys.Rev. D72 (2005) 023001


Cosmic Rays, Gamma-Rays,Neutrinos, and Magnetized Sources

Various connections:Magnetic fields influence propagation path lengths. This influences:

photo-spallation and thus observable composition, interpretation of ankle

production of secondary gamma-rays and neutrinos, thus detectability oftheir fluxes and identification of source mechanisms and locations.


Example: Source in a magnetized galaxy cluster injecting protons up to 1021 eV

Armengaud, Sigl, Miniati, astro-ph/0511277


Pair production by protons can dominate the GeV-TeV photon flux ifinjection spectrum is steep. Example for a cluster at 100 Mpc.


Source at 20 Mpc, E-2.7 proton injection spectrum with 4x1042 erg/s above 1019 eV

possible enhancementdue to magnetic fields

Note that the 3d structure of the field matters and leads to furtherenhancement of GeV γ-ray fluxes.


The source magnetic fields can give rise to a GeV-TeV γ-ray halothat would be easily resolvable by instruments such as HESS

In case of previous example, γ-rays above 1 TeV:


This is quite relevant for γ-ray astronomy in the GeV-TeV band

Rowell et al., astro-ph/0512523


The GZK neutrino flux can also be enhanced by magnetic fields


Maximal diffuse neutrino flux from magnetized galaxy clusters

Shortcomings: - overproduces UHECR flux by factors 10-40 if not blocked by magnetic horizon effects - neglects neutrinos produced by photo-reactions outside the clusters

De Marco et al., Phys.Rev.D73 (2006) 043004

Upper bound of ν-fluxfrom pp reactions


Short Advertizement: CRPropa a public code for UHE cosmic rays,Neutrinos and γ-Rays

Eric Armengaud, Tristan Beau, Günter Sigl, Francesco Miniati,astro-ph/0603675


Speculative Mechanisms and Sources

Dark Matter Annihilation Top-Down Mechanism Z-burst Mechanism


Indirect Dark Matter Detection with Intermediate Mass Black Holes

IMBHs of mass 105 Msun

in the Milky Way.

Bertone, astro-ph/0603148


Application: Flux calculations in Top-Down scenarios

a) Assume mode of X-particle decay in GUTs

jets hadronic

:Example qqlX

c) fold in injection history and solve the transport equations for propagation.

b) Determine hadronic quark fragmentation spectrum extrapolated from accelerator data within QCD: modified leading log approximation (Dokshitzer et al.) with and without supersymmetry versus older approximations (Hill). More detailed calculations by Kachelriess, Berezinsky, Toldra, Sarkar, Barbot, Drees: results not drastically different.

Fold in meson decay spectra into neutrinos and -rays.

QCD

SUSY-QCD


3

-1213

with sources shomogeneou

,G 10, GeV102 ,

t

BmqqX

X

X

A typicalexample:

In general, the ratio of γ-rays to nucleons istoo high to explain highest energy cosmic rayswithout overproducing GeV γ-ray background.

Semikoz, Sigl, JCAP 0404 (2004) 003


Future neutrino flux sensitivities and top-down models



Farrar, Biermann radio-loud quasars ~1%

Virmani et al. radio-loud quasars ~0.1%

Tinyakov, Tkachev BL-Lac objects ~10-4

G.S. et al. radio-loud quasars ~10%

Correlations with extragalactic Sources

Surprise: Deflection seems dominated byour Galaxy. Sources in direction of voids?

BL-Lac distances poorly known: Arethey consistent with UHECRenergies ?

Tinyakov, Tkachev, Astropart.Phys. 18 (2002) 165Tinyakov, Tkachev, hep-ph/0212223


1.) Neutrino primaries but Standard Model interaction probability in atmosphere is ~10-5.

resonant (Z0) secondary production on massive relic neutrinos: needs extreme parameters and huge neutrino fluxes.

2.) New heavy neutral (SUSY) hadron X0: m(X0) > mN increases GZK threshold. but basically ruled out by constraints from accelerator experiments.

3.) New weakly interacting light (keV-MeV) neutral particle electromagnetic coupling small enough to avoid GZK effect; hadronic coupling large enough to allow normal air showers: very tough to do.

In all cases: more potential sources, BUT charged primary to be accelerated toeven higher energies.

strong interactions above ~1TeV: only moderate neutrino fluxes required.

Avoiding the GZK Cutoff

If correlated sources turn out to be farther away than allowed by pionproduction, one can only think of 4 possibilities:

4.) Lorentz symmetry violations.

The Z- burst eff ect

A Z-boson is produced at theneutrino resonance energy

mE

eVeV 104 21res

“Visible” decay products haveenergies 10-40 times smaller.

Fargion, Weiler, Yoshida

Main problems of this scenario:* sources have to accelerate upto ~1023eV.

* -rays emitted f rom thesources and produced byneutrinos during propagationtend to over-produce diff usebackground in GeV regime.


The Z-burst mechanism: Relevant neutrino interactions


The Z-burst mechanism: Sources emitting neutrinos and -rays

Sources with constant comoving luminosity density up to z=3, with E-2 -rayinjection up to 100 TeV of energy fluence equal to neutrinos, mν=0.5eV, B=10-9 G.

Kalashev, Kuzmin, Semikoz, Sigl, PRD 65 (2002) 103003


The Z-burst mechanism: Exclusive neutrino emitters

Sources with constant comoving luminosity up to z=3, mν=0.33eV, B=10-9 G.



For homogeneous relic neutrinos GLUE+FORTE2003 upper limits onneutrino flux above 1020 eV imply (see figure).

eV 3.0 im

Cosmological data including WMAP imply

eV 6.0 im

Solar and atmospheric neutrino oscillations indicate near degeneracyat this scale

eV 2.0i

m

For such masses local relic neutrino overdensities are < 10 on Mpc scales.This is considerably smaller than UHECR loss lengths => required UHEneutrino flux not significantly reduced by clustering.

Even for pure neutrino emitters it is now excluded thatthe Z-burst contributes significantly to UHECRs


Testing Neutrino Properties with Astrophysical Neutrinos

Oscillation parameters, source physics, neutrino decay and decoherence Neutrino-nucleon cross sections Quantum Gravity effects


Observed Flavor Ratios can be sensitive to source physics

Kashti and Waxman, Phys.Rev.Lett. 95 (2005) 181101

pγ

pp


Serpico, Phys.Rev.D 73 (2006) 047301

Observed Flavor Ratios can be sensitive to oscillation parameters

For a source optically thick to muons but not to pions: Pions decay rightaway, but muons loose energy by synchro before decaying


Oscillation phase is( L ∆m2 / 4 En )Numbers indicate∆m2/eV2.

Sensitivity of astrophysical neutrinos to oscillations: The Learned Plot


Probes of Neutrino Interactions beyond the Standard Model

Note: For primary energies around 1020 eV:Center of mass energies for collisions with relic backgrounds ~100 MeV – 100 GeV ―> physics well understood

Center of mass energies for collisions with nucleons in the atmosphere ~100 TeV – 1 PeV ―> probes physics beyond reach of accelerators

Example: microscopic black hole production in scenarios with a TeV string scale:

For neutrino-nucleon scattering withn=1,…,7 extra dimensions,from top to bottom

Standard Model cross section

Feng, Shapere, PRL 88 (2002) 021303

This increase is not sufficientto explain the highest energycosmic rays, but can be probedwith deeply penetrating showers.


However, the neutrino flux from pion-production of extra-galactic trans-GZK cosmic rays allows to put limits on the neutrino-nucleon cross section:

Future experiments will either close the window down to the Standard Modelcross section, discover higher cross sections, or find sources beyond thecosmogenic flux. How to disentangle new sources and new cross sections?

Comparison of this N- (“cosmogenic”) flux with the non-observation ofhorizontal air showers results in the present upper limit about 103 above theStandard Model cross section.

Ringwald, Tu, PLB 525 (2002) 135


Solution: Compare rates of different types of neutrino-induced showers

Deeply penetrating (horizontal)

Earth-skimmingupgoing

Figure from Cusumano


Earth-skimming τ-neutrinos

Air-shower probability per τ-neutrino at 1020 eV for 1018 eV (1)and 1019 eV (2) threshold energy for space-based detection.

Kusenko, Weiler, PRL 88 (2002) 121104

Comparison of earth-skimming and horizontal shower rates allows tomeasure the neutrino-nucleon cross section in the 100 TeV range.


Sensitivities of LHC and the Pierre Auger project tomicroscopic black hole production in neutrino-nucleon scattering

Ringwald, Tu, PLB 525 (2002) 135

LHC much more sensitive than Auger, but Auger could “scoop” LHC

MD = fundamental gravity scale; Mbhmin = minimal black hole mass


Sensitivities of future neutrino telescopes tomicroscopic black hole production in neutrino-nucleon scattering

Ringwald, Kowalski, Tu, PLB 529 (2002) 1

Contained events: Rate ~ Volume Through-going events: Rate ~ Area


Probes of Quantum Gravity Effects with Neutrinos

Dispersion relation between energy E, momentum p, and mass m may bemodified by non-renormalizable effects at the Planck scale MPl,

n

nn m

EEmp

Pl1

222 1

where most models, e.g. critical string theory, predict ξ=0 for lowest order.

1 Pl

1)(

2

2

1

2 n

ni

ni

i m

E

E

mEp

For the i-th neutrino mass eigenstate this gives

The « standard » oscillation term becomes comparable to the new termsat energies

GeV 10x7.1 , 10x8.1 , 10x2 , 2.0 9742

1

2Pl

2

Pl

n

nm

mmE

for n=1, 2, 3, 4, respectively, and ∆m2=10-3 eV2, for which ordinaryOscillation length is ~2.5(E/MeV) km.

See, e.g., Christian, Phys.Rev.D71 (2005) 024012


Other possible effects: Decoherence of oscillation amplitude withexp(-αL):Assume galactic neutron sources, L~10 kpc, giving exclusivelyelectron-anti-neutrinos before oscillation. After oscillation theflavor ratio becomes 1:0:0 -> 0.56:0.24:0.20 without decoherence,but 0.33:0.33:0.33 with decoherence.

At E~1 TeV one has a sensitivity of α~10-37 GeV (somewhat dependenton energy dependence of α)

Hooper, Morgan, Winstanley, Phys.Lett.B609 (2005): 206


Conclusions1

2.) There are many potential high energy neutrino sources including speculative ones. But the only guaranteed ones are due to pion production of primary cosmic rays known to exist: Galactic neutrinos from hadronic interactions up to ~1016 eV and “cosmogenic” neutrinos around 1019 eV from photopion production. Flux uncertainties stem from uncertainties in cosmic ray source distribution and evolution.

4.) The highest neutrino fluxes above 1019 eV are predicted by top-down models, the Z-burst, and cosmic ray sources with power increasing with redshift. However, extragalactic top-down models and the Z-burst are unlikely to considerably contribute to ultra-high energy cosmic rays.

1.) Pion-production establishes a very important link between the physics of high energy cosmic rays on the one hand, and -ray and neutrino astrophysics on the other hand. All three of these fields should be considered together.

3.) Sources are likely immersed in (poorly known) magnetic fields of fractions of a microGauss. Such fields can strongly modify spectra and composition as well as secondary γ-ray and neutrino fluxes.


Conclusions2

6.) The coming 3-5 years promise an about 100-fold increase of ultra-high energy cosmic ray data due to experiments that are either under construction or in the proposal stage. This will constrain primary cosmic ray flux models.

7.) Many new interesting ideas on a modest cost scale for ultra-high energy neutrino detection are currently under discussion, see experimental talks.

5.) At energies above ~1018 eV, the center-of mass energies are above a TeV and thus beyond the reach of accelerator experiments. Especially in the neutrino sector, where Standard Model cross sections are small, this probes potentially new physics beyond the electroweak scale, including possible quantum gravity effects.

G ü nter Sigl Astroparticule et Cosmologie, Université Paris 7

Documents

Transcript of G ü nter Sigl Astroparticule et Cosmologie, Université Paris 7