[PPT] - TRANSITION FROM GALACTIC TO EXTRAGALACTIC COSMIC RAYS V. Berezinsky PowerPoint Presentation

SLIDE 1

TRANSITION FROM GALACTIC TO EXTRAGALACTIC COSMIC RAYS

V. Berezinsky

INFN, Laboratori Nazionali del Gran Sasso, Italy

SLIDE 2

OBSERVED CR SPECTRUM

Knee

(1 particle per m2-year)

Ankle

(1 particle per km2-year)

Energy (eV) Flux (m2 sr s GeV)-1

109 1011 1013 1015 1017 1019 1021 104

102
10-1

10-4 10-7 10-10 10-13 10-16 10-19 10-22 10-25 10-28

2 nd knee

SLIDE 3

I. TOWARDS THE END OF GALACTIC CR

SLIDE 4

SUPERNOVA-REMNANT PARADIGM:

“Standard Model” for galactic cosmic rays

sources:

supernova remnant

acceleration:

SNR shock acceleration

chemical composition:

rigidity-dependent injection

propagation:

diffusive propagation in magnetic fields

SLIDE 5

DIFFUSIVE SHOCK-ACCELERATION:

spectrum:

At fixed SNR age the spectrum of escaped particles is close to δ-function. but time-averaged spectrum is ∝ E−2 or flatter at highest energies (Ptuskin, Zirakashvili 2006).

Emax :

Acceleration to the highest energies occurs at the beginning of Sedov phase. Non-linear amplification of turbulent magnetic field in the shock precursor due to streaming instability of CR produces magnetic field with strength δB ∼ B ∼ 10−4 G (Bell and Lucek). Emax = 4 × 1015Z B 10−4G W51 ng/cm3 2/5 eV Emax

p

= 4 × 1015B−4 eV, Emax

Fe

= 1 × 1017B−4 eV

SLIDE 6

SM : GALACTIC SPECTRA AND KNEES Berezhko and V¨

lk 2007

SLIDE 7

MASS COMPOSITION VS ENERGY Compilation of H¨

randel 2005

0. 5 1 1. 5 2 2. 5 3 3. 5 4 10

4

10

5

10

6

10

7

10

8

Energy E [GeV] Mean logarithmic mass <ln A>

H H e B e N M g F e ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ∅ ∅ ∅ ∅ ∅ ∅ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ⊗⊗ ⊗⊗ ⊗ ⊗ ⊗⊗⊗ ⊗⊗⊗ ⊗ ⊗⊗ ⊗ ⊗ ⊗ ⊕⊕⊕⊕⊕ ⊕⊕ ⊕ ⊕ ⊕ ⊕⊕ ⊕⊕ ⊕ ⊕ ⊕ ⊕ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✕ ✕ ✕ ✕ ✕ ✕ ∇ ∇ ∇ ∇ ∇ ∇ ✧ CASA-MI A Chacaltaya ✣ EAS-TOP + MACRO ✕ EAS-TOP (e/m ) ∅ HEGRA (CRT) ∇ SPASE/AMANDA ❄ KASCADE (nn) KASCADE (h/m ) ⊗ KASCADE (e/m) QGSJET ⊕ KASCADE (e/m) SIBYLL JACEE direct : RUNJOB

SLIDE 8

CONCLUSION NEEDED FOR ANALYSIS OF TRANSITION In “standard model” the end of Galactic cosmic rays starts at iron knee Eknee

Fe

= ZEknee

p

∼ 1 × 1017 eV Spectrum of Fe-nuclei at E > Eknee

Fe

is steep and it inevitably intersects some- where the more flat extragalactic spectrum .

SLIDE 9

II. FROM UHECR TOWARDS THE KNEE

SLIDE 10

MEASURED FLUXES OF UHECR

SLIDE 11

PROPAGATION OF UHECR THROUGH CMB

SLIDE 12

INTERACTIONS Protons p + γCMB → p + e+ + e− p + γCMB → N + pions Nuclei Z + γCMB → Z + e+ + e− A + γCMB → (A − 1) + N A + γCMB → A′ + N + pions Photons γ + γbcgr → e+ + e−

10

17

10

18

10

19

10

20

10

21

10

22

10

12

10

11

10

10

10

9

10

8

10

7

2 1 e

+e

e

+e

pion-prod.

red-shift

a)

1/E dE/dt, yr

1

E, eV

SLIDE 13

PROPAGATION SIGNATURES Propagation of protons in intergalactic space leaves the imprints on the spectrum most notably in the form: GZK cutoff and pair-production dip These signatures might depend on the distribution of sources and way of propagation.

SLIDE 14

GZK CUTOFF GZK cutoff is modified by discreteness in source distribution and by source local

verdensity/deficit and by different values of Emax.

10

17

10

18

10

19

10

20

10

21

10

23

10

24

10

25

HiRes I - HiRes II

6 5 3 2 1

J(E)E

3, m

2s
1sr
1eV

2

E, eV

SLIDE 15

GZK CUTOFF IN HiRes DATA In the integral spectrum GZK cutoff is numerically characterized by energy E1/2 where the calculated spectrum J(> E) becomes half of power-law extrapolation spectrum KE−γ at low energies. As calculations (V.B.&Grigorieva 1988) show E1/2 = 1019.72 eV valid for a wide range of generation indices from 2.1 to 2.8. HiRes obtained: E1/2 = 1019.73±0.07 eV

log10(E) (eV) J(>E)/KE−γ

0.2 0.4 0.6 0.8 1 1.2 17 17.5 18 18.5 19 19.5 20 20.5 21

SLIDE 16

PAIR-PRODUCTION DIP IN THE DIFFUSE SPECTRUM VB, Grigorieva 1988; Aloisio, VB, Blasi, Gazizov, Grigorieva (2004 - 2007). DEFINITION OF MODIFICATION FACTOR η(E) = Jp(E) Junm

p

(E) where Junm

p

(E) includes only adiabatic energy losses (redshift) and Jp(E) includes total energy losses, ηtot(E) or adiabatic, e+e− energy losses, ηee(E). Since both Junm

p

(E) and Jp(E) include factor E−γg, η(E) depends weakly on γg.

SLIDE 17

DIP IN DIFFUSE SPECTRA

10

17

10

18

10

19

10

20

10

21

10

2

10

1

10

ηtotal

2 1

ηee

2 1 1: γg=2.7 2: γg=2.0

modification factor E, eV

The dotted curve shows ηee, when only adiabatic and pair-production energy losses are included. The solid and dashed curves include also the pion-production losses.

SLIDE 18

DIP IN COMPARISON WITH AKENO-AGASA DATA 10

17

10

18

10

19

10

20

10

21

10

2

10

1

10

Akeno-AGASA ηtotal ηee

γg=2.7

modification factor E, eV

SLIDE 19

DIP IN COMPARISON WITH HIRES DATA 10

17

10

18

10

19

10

20

10

21

10

2

10

1

10

ηtotal HiRes I - HiRes II ηee

γg=2.7

modification factor E, eV

SLIDE 20

DIP IN COMPARISON WITH YAKUTSK DATA 10

17

10

18

10

19

10

20

10

21

10

2

10

1

10

Yakutsk ηtotal ηee

γg=2.7

modification factor E, eV

SLIDE 21

DIP IN COMPARISON WITH AUGER DATA

SLIDE 22

ENERGY CALIBRATION BY DIP : AGASA-HIRES DISCREPANCY

10

17

10

18

10

19

10

20

10

21

10

23

10

24

10

25

Akeno - AGASA HiRes I - HiRes II

J(E)E

3, m

2s
1sr
1eV

2

E, eV

10

17

10

18

10

19

10

20

10

21

10

23

10

24

10

25

Akeno-AGASA HiRes I - HiResII

J(E)E

3, eV 2m

2s
1sr
1

E, eV

AGASA and HiRes spectra calibrated by the dip. The energy shift needed for χ2

min is λAGASA = 0.9 and λHiRes = 1.2.

Both are allowed by systematic errors.

SLIDE 23

DIP AND AGASA-YAKUTSK DISCREPANCY

10

17

10

18

10

19

10

20

10

21

10

23

10

24

10

25

Akeno-AGASA Yakutsk

J(E)E

3, eV 2m

2s
1sr
1

E, eV

10

17

10

18

10

19

10

20

10

21

10

23

10

24

10

25

Akeno-AGASA Yakutsk

J(E)E

3, eV 2m

2s
1sr
1

E, eV

AGASA and Yakutsk spectra calibrated by the dip. The energy shift needed for χ2

min is λAGASA = 0.9 and λYakutsk = 0.75.

Both are allowed by systematic errors.

SLIDE 24

AGASA-HIRES-YAKUTSK DISCREPANCY

10

17

10

18

10

19

10

20

10

21

10

23

10

24

10

25

Akeno - AGASA Yakutsk HiRes I - HiRes II

J(E)E

3, m

2s
1sr
1eV

2

E, eV

10

17

10

18

10

19

10

20

10

21

10

23

10

24

10

25

Akeno - AGASA Yakutsk HiRes I - HiRes II

J(E)E

3, m

2s
1sr
1eV

2

E, eV

AGASA, Hires and Yakutsk spectra calibrated by the dip.

SLIDE 25

COMPARISON OF AUGER WITH CALIBRATED DATA

SLIDE 26

COMPARISON OF AUGER WITH CALIBRATED DATA

SLIDE 27

CONCLUSIONS NEEDED FOR ANALYSIS OF TRANSITION

Very good agreement of the predicted dip energy-shape with the data of

all detectors demonstrates that large fraction of particles observed at 1 × 1018 − 4 × 1019 eV are extragalactic protons propagating through CMB.

The numerical agreement of HiRes data with GZK cutoff implies that at

energy E ≥ 5 × 1019 eV protons dominate, too.

SLIDE 28

III. TRANSITION

SLIDE 29

THREE MODELS OF TRANSITION:

DIP, ANKLE, and MIXED-COMPOSITION MODELS

In the dip model, dip automatically includes ankle.
In ankle model, Ea ∼ 1 × 1019 corresponds to equal fluxes Jgal = Jextr.
In the mixed model, Ea ∼ 3 × 1018 eV is the end of transition.

Necessary assumption for ankle and mixed models: AGREEMENT OF DATA WITH PAIR-PRODUCTION DIP IS ACCIDENTAL

SLIDE 30

THE DIP and ANKLE TRANSITIONS In the dip model transition occurs at Etr < Eb = 1 × 1018 eV, i.e. at second knee. This transition agrees perfectly with the standard galactic model. In the ankle model transition occurs at Ea = 1 × 1019 eV and the galactic flux at this energy is half of the total in contradiction with standard galactic model.

10

7

10

8

10

9

10

1

10

2

10

3

Eb Et r EFe

extr. p
gal. Fe

KASCADE HiRes I HiRes II

J(E)E

2.5, m

2s
1sr
1GeV

1.5

E, GeV

10

9

10

11

10 10

1

10

2

Et r

gal. CR

HiRes I HiRes II

extr. p

J(E)E

2.5, m

2s
1sr
1GeV

1.5

E, GeV

SLIDE 31

THE DIP and ANKLE TRANSITIONS: MASS COMPOSITION In the dip model transition to proton-dominated component is completed at 1 × 1018 eV, while in the ankle model at 1 × 1019 eV. In the range 1 - 10 EeV ankle model predicts iron or mixed composition, while dip model - proton-dominated composition. The elongation rate is most sensitive tool of chemical composition.

500 600 700 800 900 10 8 10 9 10 10 10 11 E0, GeV Xmax, g/cm2

Xmax(E) in the dip model.

500 600 700 800 900 10 8 10 9 10 10 10 11 E0, GeV Xmax, g/cm2

Xmax(E) in the ankle model.

SLIDE 32

MIXED COMPOSITION MODEL Allard, Parizot and Olinto (2005 - 2007)

generation spectrum with γg = 2.1 − 2.3 .
mixed composition at generation.
end of transition at E ∼ 3 × 1018 eV .

Energy spectrum in the mixed model. Xmax(E) in the mixed model.

SLIDE 33

CONCLUSIONS

The galactic CR are well described by the ”standard model” with SNR as

the sources and with diffusive propagation of CR in the Galaxy. The end

f galactic CR corresponds to iron knee Emax

Fe

∼ 1 × 1017 eV, with a sharp steepening above this energy.

The pair-production dip for extragalactic CR at 1×1018 ≤ E ≤ 4×1019 eV

is well confirmed by all existing UHECR detectors and it demonstrates that most of observed particles are extragalactic protons propagating through

CMB. Energy calibration of detectors confirms this conclusion.
The dip model of transition is based on proximity of the end of galactic

CR Emax

Fe

∼ 1 × 1017 eV and the beginning of the dip Eb ≈ 1 × 1018 eV, where transition is completed. The predicted transition from galactic iron to extragalactic protons is very sharp. Observationally transition occurs at the second dip.

The two other models of transition assume agreement of the pair-production

dip (VB & Grigorieva 1988) with the observed dip as incidental and use the two-component dip model by Hill & Schramm 1985.

SLIDE 34

The traditional ankle model assumes transition at Ea ∼ 1 × 1019 eV with

extragalactic generation spectrum ∝ E−2. It needs another component of galactic CR beyond Emax

Fe

≈ 1 × 1017 eV.

The mixed composition model assumes production of extragalactic CR with

flat generation spectrum γg = 2.1 − 2.3. The transition is completed at E ≈ 3 × 1018 eV and the model marginally agrees with “standard model”. The spectral agreement at the dip 1 × 1018 − 4 × 1019 eV is reached using the subtraction procedure and the choice of nuclear composition.

The transition is accompanied by a change in chemical composition, de-

scribed by elongation rate Xmax(E). The dip model predicts fast growth of Xmax(E), while the mixed model - the smooth behaviour. The dip model marginally agrees with the data, while the mixed model gives a good fit.

The energies 1017 − 1018 eV look like the key region for cosmic ray origin.

More precise measurements of Xmax(E) at these energies will be obtained in the nearest future by TALE detector (Utah) and FDs with high elevation angles at Auger detector. They will shed more light not only on transition problem, but also on origin of galactic and extragalactic CR.