High-Energy Cosmogenic Neutrinos

Transcription

1 High-Energy Cosmogenic Neutrinos Markus Ahlers UW-Madison & WIPAC HEM 2014 Chicago, June 9 11, 2014 Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 1

2 Cosmogenic neutrinos cos-mo-gen-ic (adj.): produced by cosmic rays but this is true for all high-energy neutrinos... more specifically: not in the source or atmosphere, but during CR propagation most plausibly via pion production in pγ interactions, e.g. p + γ bgr n + π + π + µ + ν µ & µ + e + ν µν e & n pe ν e pγ / pp (e.g. Centaurus A) propagation Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 2

3 Ultra-High Energy (UHE) Cosmic Rays (CRs) sr -1 ] 1.6 m -2 s -1 F(E) [GeV 2.6 E Grigorov JACEE MGU Tien-Shan Tibet07 Akeno CASA-MIA HEGRA Fly s Eye Kascade 15 galactic Kascade Grande IceTop-73 HiRes 1 HiRes 2 Telescope Array Auger Knee 16? 17 E [ev] 2nd Knee 18 Ankle extragalactic [Particle Data Group 13, pdg.lbl.gov] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 3

4 Cosmogenic neutrinos Observation of UHE CRs and extragalactic radiation backgrounds guarantee a flux of high-energy neutrinos, in particular via resonant production in CMB. Guaranteed, but with many model uncertainties and constraints: (low cross-over) proton models + CMB (+ EBL) [Berezinsky & Zatsepin 69] [Berezinsky & Zatsepin 69; Yoshida & Teshima 93; Protheroe & Johnson 96; Engel, Seckel & Stanev 01; Fodor, Katz, Ringwald &Tu 03; Barger, Huber & Marfatia 06; Yuksel & Kistler 07; Takami, Murase, Nagataki & Sato 09, MA, Anchordoqui & Sarkar 09 ] + mixed compositions [Hooper, Taylor & Sarkar 05; Ave, Busca, Olinto, Watson & Yamamoto 05; Allard, Ave, Busca, Malkan, Olinto, Parizot, Stecker & Yamamoto 06; Anchordoqui, Goldberg, Hooper, Sarkar & Taylor 07; Kotera, Allard & Olinto ; Decerprit & Allard 11; MA & Halzen 12] + extragalactic γ-ray background limits [Berezinsky & Smirnov 75; Mannheim, Protheroe & Rachen 01; Keshet, Waxman, & Loeb 03; Berezinsky, Gazizov, Kachelriess & Ostapchenko ; MA, Anchordoqui, Gonzalez Garcia, Halzen & Sarkar ; MA & Salvado 11; Gelmini, Kalashev & Semikoz 12] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 4

5 GZK neutrinos from CMB Greisen-Zatsepin-Kuzmin (GZK) interactions of ultra-high energy CRs with cosmic microwave background (CMB) [Greisen 66;Zatsepin/Kuzmin 66] GZK -neutrinos at EeV energies from pion decay [Berezinsky/Zatsepin 69] three neutrinos (ν µ/ ν µ/ν e) from π + : E νπ 1 4 x Ep 1 20 Ep one neutrino from neutron decay: E νe mn mp E p 3 E p m n log E dn/de, per cm 2.s.ster log E dn/de, per cm 2.s.ster W&B E ν, ev FIG. 4. Fluxes of electron[engel, neutrinos Stanev (dashed& lines) Seckel 01] and antineutrinos (dotted lines) generated in propagation of protons are shown in the upper panel. The lower panel shows the Markus Ahlers (UW-Madison) High-Energy Cosmogenic fluxes Neutrinos of muon neutrinos andjune antineutrinos. 9 11, 2014 Solid lines slideshow 5 ν e ν µ tr th a a b p a a h n su sh A ti b ti o fa th o ti a o b th tr h th sp se c

6 Flavor Composition in general, initial flavor ratio (ν e:ν µ:ν τ ) depend on process and environment mixing between flavor and mass eigenstates ν α = j U αj ν j, flavor oscillations average out over cosmic distances remaining parameter space 50% 25% ν µ 75% 50% 25% pion decay (1:2:0) neutron decay (1:0:0) P να ν β i U αi U βi 2 75% 25% 50% 75% remaining phase space thin black line crossing (1:1:1) ν τ ν e Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 6

7 Extra-galactic background light (EBL) λ [µm] E u(e) [erg cm -3 ] Finke et al. (20) Kneiske et al. (2004) Franceschini et al. (2008) Gilmore et al. (2008) Razzaque et al. (2009) Stecker et al. (2006) E [ev] 1 0 E I(E) [nw m -2 sr -1 ] [Finke, Razzaque & Dermer ] BL models, measurements, and constraints. See Finke et al. for details and re PeV cosmogenic neutrinos via optical-uv background: E ν 8PeV (ev/e γ) Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 7

8 Cosmogenic neutrinos & gamma-rays GZK interactions produce neutral and charged pions p + γ CMB n + π + /p + π b BH p /E b π p/e H 0 Bethe-Heitler (BH) pair production: p + γ CMB p + e + + e BH is dominant energy loss process for UHE CR protons at GeV. energy loss rate [Mpc 1 ] 2 3 EM components cascade in CMB/EBL and contribute to GeV-TeV γ-ray background E [GeV] [Berezinsky&Smirnov 75] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 8

9 Gamma-ray cascades CMB interactions (solid lines) dominate in casade: inverse Compton scattering (ICS) e ± + γ CMB e ± + γ pair production (PP) γ + γ CMB e + + e PP in IR/optical background (red dashed line) determines the edge of the spectrum. this calculation: Franceschini et al. 08 interaction length [Mpc] E [GeV] Γ 1 Γ 1 PP (CMB) PP (IR/opt.) Γ 1 ICS (CMB) E/b syn (pg) Rapid cascade interactions produce universal GeV-TeV emission (almost) independent of injection spectrum and source distribution. cascade bound for neutrinos [Berezinsky&Smirnov 75] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 9

10 Gamma-ray cascades CMB interactions (solid lines) dominate in casade: diffuse γ-ray background inverse Compton scattering (ICS) e ± + γ CMB e ± + γ 5 pair production (PP) γ + γ CMB e + + e PP in IR/optical background (red dashed line) determines the edge of the spectrum. this calculation: Franceschini et al. 08 E 2 J [GeV cm 2 s 1 sr 1 ] Fermi-LAT (Abdo et al. ) Fermi-LAT fit: E 2.4 EGRET (Sreekumar et al. 98) EGRET fit: E E [GeV] Rapid cascade interactions produce universal GeV-TeV emission (almost) independent of injection spectrum and source distribution. cascade bound for neutrinos [Berezinsky&Smirnov 75] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide

11 Cosmogenic neutrinos from CR protons E 2 J [GeV cm 2 s 1 sr 1 ] E min = 17.5 ev HiRes I&II Fermi-LAT p (best-fit) ν, ν (99% C.L.) γ (99% C.L.) maximal cascade E [GeV] [MA, Anchordoqui, Gonzalez-Garcia, Halzen & Sarkar 11] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 11

12 Cosmogenic neutrinos from CR protons E 2 J [GeV cm 2 s 1 sr 1 ] E min = 18 ev HiRes I&II Fermi-LAT p (best-fit) ν, ν (99% C.L.) γ (99% C.L.) maximal cascade E [GeV] [MA, Anchordoqui, Gonzalez-Garcia, Halzen & Sarkar 11] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 12

13 Cosmogenic neutrinos from CR protons E 2 J [GeV cm 2 s 1 sr 1 ] E min = 18.5 ev HiRes I&II Fermi-LAT p (best-fit) ν, ν (99% C.L.) γ (99% C.L.) maximal cascade E [GeV] [MA, Anchordoqui, Gonzalez-Garcia, Halzen & Sarkar 11] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 13

14 Cosmogenic neutrinos from CR protons E 2 J [GeV cm 2 s 1 sr 1 ] E min = 19 ev HiRes I&II Fermi-LAT p (best-fit) ν, ν (99% C.L.) γ (99% C.L.) maximal cascade E [GeV] [MA, Anchordoqui, Gonzalez-Garcia, Halzen & Sarkar 11] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 14

15 Cosmogenic neutrinos from CR protons IC excess (x3) IC excess (x3) [Decerpit & Allard 11] neutrino flux depend on source evolution model (strongest for FR-II ) and EBL model (highest for Stecker model) Stecker model disfavored by Fermi observations of GRBs strong evolution disfavored by Fermi diffuse background Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 15

16 Cosmic rays UHE CR composition 3 )] 2 50 Auger event sr -1 ] s -1 m [GeV F(E) 2.6 E 2 HiRes 1 HiRes 2 Telescope Array Auger energy deposit [PeV/(g/cm photon proton iron E [ev] slant depth [g/cm ] [PDG 13] 27.9: Expanded view of the highest energy portion of the cosmic-ray [Kampert&Unger 12] from data of HiRes 1&2 [1], the Telescope Array [3], andfigure the Auger 9: Example of a longitudinal air shower development as measured with tory [4]. composition The HiRes stereo measurement spectrum [122] is consistent a statistical withfluorescence the HiRes basis telescopes. Data points are taken from [145] (E = (30 ± 2) EeV) nocular results. The differential cosmic-ray flux is multiplied byand E 2.6 compared. The to ten simulated [133] air showers for three di erent primary w indicates first thetwo change moments: in the plotted X data max for& a RMS(X systematic max) shift particle in the types using the hadronic interaction model Epos1.99 [36]. cale of 20%. average mass inferred, e.g. from X max : groups (see e.g. [150]) similar to what is done for surface detectors. energy Xmax data ing the ankle, one possibility is that it is the result of Xmax p a higher ln A = In the following, ln 56however, we will concentrate on the of particles overtaking a lower energy population, for example Xan extragalactic ing to dominate over the galactic flux (e.g. Ref. 0). Anotherpossibilityis max first p two moments X max Fe of the X max -distribution, hx max i and (X max ). ip structure in the region of the ankle is due to γp e + + e energy For the losses determination of the average shower maximum, experiments [8]. Thisbin the recorded events in energy and calculate the actic protons on the 2.7 K cosmic microwave radiation (CMB) re has been cited as a robust signature of both the protonic and extragalactic Markus Ahlers (UW-Madison) High-Energy Cosmogenic mean of the Neutrinos measured shower June maxima. 9 11, 2014 For this averaging slide 16 not 0

17 700 UHE CR composition ] 2 [g/cm ] meas X max QGSJet01 QGSJetII SIBYLL EPJ Web of Conferences proton TA. Data points are shifted to a common energy scale (text 20 for details). iron E [ev] ] 2 [g/cm σ X 70 QGSJetII Fig. 4. The hxmax meas QGSJet01 i (left) and RMS(X max ) (right) as measured by the HiRes experiment. The lines are the corresponding hxmax meas Auger 850 QGSJetII i and s X expectations for proton 70 and iron compositions. The different line SIBYLL2.1 types correspond EPOSv1.99to different Yakutskmodels. 60 ] 2 [g/cm X meas max QGSJet01 QGSJetII SIBYLL proton iron 1 Fig. 3. Measured18 hx max [Mass i (left) Composition and RMS(X Working max ) Group (right) Report for the 13; Auger arxiv: ] and Yakutsk experiments. The lines indicate the hx max i expectations for proton E [ev] and iron compositions using different hadronic lg(e [ev]) interaction Markus models. Ahlers (UW-Madison) Notice that the highest High-Energy energycosmogenic bin for Neutrinos Yakutsk contains only June 9 11, 3 events 2014 (Fig. slide 6) Number of Events E [ev] Fig. 2. hx max i measured by Auger and Yakutsk, together with the hxmax meas i as measured by HiRes and ] 2 [g/cm X max proton iron E [ev] 2 ) [g/cm max RMS(X proton iron E [ev] proton lg(e[ev]) > 18.2 HiRes (798 events) TA (279 events) Auger (5138 events) Yakutsk (412 events) iron E [ev]

18 UHEUHECR2012: composition International Symposium on Future Directions in UHECR Physics (a) using QGSJet-II model. (b) using SIBYLL model. Fig. 11. Comparing the [Mass average Composition composition Working (hlnai) Groupestimated Report 13; using arxiv: ] Auger, HiRes, TA and Yakutsk data. The shaded regions correspond to the systematic uncertainty ranges. To infer the average composition from hx max i, QGSJet-II and SIBYLL models have been used. inferred mass depend on hadronic interactions models large systematic uncertainties! Auger 2 results are consistent within systematic2 uncertainties with TA and Yakutsk, 0 0 but not fully Auger consistent with HiRes. Auger [arxiv: ] lna lna HiRes HiRes Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 18 lna lna

19 Composition dependence of UHE CR sources E 3 J [GeV 2 cm 2 s 1 sr 1 ] 3 2 HiRes I&II (E/1.2) Auger p source Fe source E 2 J [GeV cm 2 s 1 sr 1 ] Fermi-LAT IceCube (IC40) p source Fe source E [GeV] E [GeV] UHE CR emission toy-model: 0% proton: n = 5 & z max = 2 & γ = 2.3 & E max = 20.5 ev 0% iron: n = 0 & z max = 2 & γ = 2.3 & E max = ev Diffuse spectra of cosmogenic γ-rays (dashed lines) and neutrinos (dotted lines) vastly different. [MA&Salvado 11] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 19

20 Approximate scaling law of energy densities ω ν i A 2 γ E i thq 2 i(e th) i 2 γ i } {{ } composition zmax 0 dz (1 + z)n+γ i 4 H(z) } {{ } evolution * disclaimer: source composition Q i with mass number A i and index γ i applies only to models with large rigidity cutoff E max,i A i E GZK previous examples (z max = 2 & γ = 2.3): 0% proton: n = 5 & E max = 20.5 ev ω γ % iron: n = 0 & E max = ev ω γ relative difference: 82. Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 20

21 bution of neutrino events that were detected, showing that the in neutrino angular acceptance spans a range from 5 below ne the horizon to 45 above the horizon, more than 6 steradi- ac ta A Auger Observatory, while probing a similar energy sta TABLE ARA, II: does Expected notnumbers have of asevents highnν afrom neutrino several UHE sensitivity neutrino models, comparing published values from the 2008 ANITA-II asa flight marily with predicted a UHECR events for instrument. a three-year exposure ARAforwill ARA-37. complem m other instruments by making high sensitivity obser tio themodel 0.1- & references EeV energy Nν: range, ANITA-II, matchingara, the peakth (2008 flight) 3 years pected cosmogenic neutrino fluxes. Cosmogenic neutrinos from heavy ans of solid nuclei angle. Yuksel & Kistler 07 ESS 01 strong Kotera et al. max ESS 01 baseline Ahlers et al. 11 re Baseline cosmogenic models: w ar Protheroe & Johnson 1996 [27] al Engel, Seckel, Stanev 2001 [28] ca Kotera,Allard, & Olinto 20 [29] V. CONCLUSIONS Strong source evolution models: ac Engel, Seckel, Stanev 2001 [28] Th We Kalashev havet al. described 2002 [30] the design5.8 and initial 146 perfo tin Barger, Huber, & Marfatia 2006 [32] a new ultra-high energy neutrino detector at the su S Yuksel & Kistler 2007 [33] fe themixed-iron-composition: 16-antenna, self-triggering ARA-testbed, which tim Kotera et al. mid fidelity Ave et prototype al [34] for future ARA0.01 detector6.6 stations. th Kotera et al. low Stanev 2008 [35] operation extending well into the the extreme therma Ave et al. 07 Fe mix Kotera, Allard, & 20 [29] upper al ANITA II (20) mentkotera, of the Allard, austral & Olintowinter 20 [29] indicates lower that radio-freque 4.1 te IceCube 40 (2011) ference Models constrained is infrequent by Fermi cascade and has bound: only a slight impact on am Auger (2009) for our Ahlerstestbed et al. 20 [36] detector, which is0.09 closest20.7 of any fu m ARA 37 3 yrs projected Waxman-Bahcall (WB) fluxes: tu stations to the primary sources of interference at WB 1999, evolved sources [37] PoleWB station. 1999, standard Other [37] than brief periods 0.5 of sporadic 27 int tin the baseline radio noise levels are dominated by the [ARA 11] mal In Table noiseii floor we give ofexpected the ambient neutrinoice, event and totals thefrom thermal a usn wide range of currently allowed cosmogenic neutrino models tim not appear to be correlated to wind velocity. We hav for ARA in three years of operation, compared to recent published strate evolution expectations the ability for models the to best maintain current and limits impulse source to date, trigger from thesensi FIG. 27: Compilation of sensitivity estimates from existing instru- ANITA-II level close flight to [3]. the It isthermal evident that noise. ARA-37We willhave extenddemons in Range of GZK neutrino predictions of various ments, published limits, compositions and a range of GZKrange neutrino models, over two alongorders sensitivity ofabove magnitude! ANITA-2 s sensitivity by factors of two orders of magnitude or more. For strong-source-evolution and impulse propagation of more than 3 km slant rang with the expected 3 year ARA sensitivity. baseline the South models, Pole ARA-37 ice without detects between significant of order loss 50 toofover signaltim c 200 Weevents haveindemonstrated three years of operation, inter-antenna enough to pulse establishtiming the pr m Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos basic order characteristics 0 ps, implying of the June energy 9 11, angular spectrum 2014 and resolutions source slide arrival 21 which Th

22 Nucleon cascade Observed composition is result of source composition and nucleon cascades. A+5 Backtracking conserves energy per nucleon. Bethe-Heitler (BH) loss breaks this approximation b A,BH(E) Z 2 b p,bh(e/a) A+4 Minimal cosmogenic neutrino production from fit to Auger data assuming: maximal backtracking minimal BH loss p He A minimal nucleon emissivity Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 22

23 Nucleon cascade Observed composition is result of source composition and nucleon cascades. A+5 Backtracking conserves energy per nucleon. Bethe-Heitler (BH) loss breaks this approximation b A,BH(E) Z 2 b p,bh(e/a) A+4 Minimal cosmogenic neutrino production from fit to Auger data assuming: maximal backtracking minimal BH loss p He A minimal nucleon emissivity Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 23

24 Guaranteed cosmogenic neutrinos nucleon spectrum for observed mass number A obs: J min N (E N) = A 2 obsj CR(A obse N) dependence on cosmic evolution of sources: no evolution (dotted) star-formation rate (solid) ultimate test of UHE CR proton models with ARA-37 generalization to arbitrary composition via JN min (E N) = i f i(a ie N)A 2 i J CR(A ie N) E 2 J [GeV cm 2 s 1 sr 1 ] IC-86 (yr) p He N Si Fe HiRes TA Auger ARA-37 (3yr) E [GeV] [MA&Halzen 12] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 24

25 Guaranteed cosmogenic neutrinos nucleon spectrum for observed mass number A obs: J min N (E N) = A 2 obsj CR(A obse N) dependence on cosmic evolution of sources: no evolution (dotted) star-formation rate (solid) ultimate test of UHE CR proton models with ARA-37 generalization to arbitrary composition via JN min (E N) = i f i(a ie N)A 2 i J CR(A ie N) E 2 J [GeV cm 2 s 1 sr 1 ] IC-86 (yr) bestfit HSFR H0 HiRes TA Auger ARA-37 (3yr) E [GeV] [MA&Halzen 12] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 25

26 Guaranteed cosmogenic neutrinos nucleon spectrum for observed mass number A obs: J min N (E N) = A 2 obsj CR(A obse N) dependence on cosmic evolution of sources: no evolution (dotted) star-formation rate (solid) ultimate test of UHE CR proton models with ARA-37 generalization to arbitrary composition via JN min (E N) = i f i(a ie N)A 2 i J CR(A ie N) E 2 J [GeV cm 2 s 1 sr 1 ] IC-86 (yr) p@0eev: 0% % 1% HiRes TA Auger SFR evolution ARA-37 (3yr) E [GeV] [MA&Halzen 12] Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 26

27 Summary Cosmogenic neutrinos guarantee a diffuse flux of UHE neutrinos. Present neutrino limits start to constrain optimistic (proton-dominated) model. A cosmogenic origin of the IceCube excess at TeV-PeV energies is very unlikely. Model uncertainties of predictions are large (UHE CR source composition and evolution). Future EeV neutrino observatories (ARA or ARIANNA) will be able to probe proton-dominated CR models. Markus Ahlers (UW-Madison) High-Energy Cosmogenic Neutrinos June 9 11, 2014 slide 27

28 Backup

29 Diffuse CR fluxes spatially homogeneous and isotropic distribution of sources Boltzmann equation of comoving number density (Y = n/(1 + z) 3 ): Ẏ i = E(HEY i) + E(b iy i) Γ i Y i + de j γ jiy j + L i, j H : Hubble rate b i : continuous energy loss γ ji (Γ i) : differential (total) interaction rate power-law proton emission rate: L p(0, E) (E/E 0) γ exp( E/E max) exp( E min/e) redshift evolution of source emission or distribution: L p(z, E) = L p(0, E)(1 + z) n Θ(z max z)θ(z z min)

30 Proton-dominance in UHE CRs? GoF based on Hires-I/II data ( E/E 25%) fixed: E max = 21 ev z min = 0 / z max = 2 priors: 2.1 γ n 6 ω cas ω Fermi range of spectra: 99% C.L. increasing crossover energy from 2nd knee to ankle [MA, Anchordoqui, Gonzalez-Garcia, Halzen & Sarkar 11]

31 Propagation of CR nuclei 52 Ti 53 V 54 Cr 55 Mn 56 Fe 49 Ca 50 Sc 51 Ti 52 V 53 Cr 54 Mn 55 Fe fast photo-disintegration of nuclei (mass number A = N + Z) beyond the giant dipole resonance (GDR): λ GDR 4 A Mpc stable nuclei N longlived nuclei PSB-chain Z 48 Ca 49 Sc 50 Ti 51 V 52 Cr 53 Mn 54 Fe 46 K 47 Ca 48 Sc 49 Ti 50 V 51 Cr 52 Mn 53 Fe 44 Ar 45 K 46 Ca 47 Sc 48 Ti 49 V 50 Cr 51 Mn 52 Fe 43 Ar 44 K 45 Ca 46 Sc 47 Ti 48 V 49 Cr 42 Ar 43 K 44 Ca 45 Sc 46 Ti 47 V 48 Cr 40 Cl 41 Ar 42 K 43 Ca 44 Sc 45 Ti 38 S 39 Cl 40 Ar 41 K 42 Ca 43 Sc 44 Ti 37 S 38 Cl 39 Ar 40 K 41 Ca 36 S 37 Cl 38 Ar 39 K 40 Ca 35 S 36 Cl 37 Ar 38 K 32 Si 33 P 34 S 35 Cl 36 Ar strong influence of mass composition at very high energy BUT: conserves total number of nucleons with nucleon energy E/A! 31 Si 32 P 33 S 28 Mg 29 Al 30 Si 31 P 32 S 27 Mg 28 Al 29 Si 30 P 24 Ne 25 Na 26 Mg 27 Al 28 Si 24 Na 25 Mg 26 Al 22 Ne 23 Na 24 Mg 21 Ne 22 Na 18 O 19 F 20 Ne 17 O 18 F 14 C 15 N 16 O 17 F 13 C 14 N 15 O Be 11 B 12 C 13 N 14 O Neutrino production (mostly) via γ-nucleon interaction! 9 Be B 11 C 7 Li 6 Li 7 Be T 4 He D 3 He p [Puget/Stecker/Bredekamp 76;MA/Taylor ]