Santiago de Compostela: Winter School 2007

1. Santiago de Compostela: Winter School 2007 Three Lectures on Cosmic Rays Alan Watson University of Leeds a.a.watson@leeds.ac.uk

2. Cosmic Rays form the roots for many areas:- Particle Physics (e+, μ, π+/-, K+/-, Λ) Astrophysics (e.g. radio astronomy) Climate (Jasper Kirkby)

3. Outline Plan: • 1. Some History • How extensive air showers develop • Overview of data up to 1016 eV: • Demonstrate the hadronic model dependence • Why study UHECR? • Results from AGASA and HiRes • 3. The Pierre Auger Observatory • Building the detector • Our New Results: model independence

4. Lecture 1 By Johnny Hart

5. To 5 km without oxygen

6. Hess Data Kolhoster data

7. Clay’s Results taken by Berlage Amsterdam Genoa Jakarta Jakarta

8. Vallarta 1932 • East-West effect: • Rossi • Compton • Alvarez From Hillas ‘Cosmic Rays’

9. Why is the origin of cosmic rays still unknown ~ 95 years after discovery? Magnetic Fields: While gamma-rays and neutrinos are ‘blind’ to magnetic fields, most cosmic rays are charged particles, the nuclei of atoms. BUT the highest energy particles are expected to be almost undeflected by the fields → cosmic ray astronomy Detection of these relies on a phenomenon found unexpectedly in 1938

10. Known energy scale extended by ~106 Observed Rate was found to be much higher than the Calculated Chance Rate(2N1N2τ)– even when the counters were as far as 300 m apart

11. Wilson and Lovell, Nature 1938

12. p + Ar  p + fragments + many pions proton interacting with argon in cloud chamber

13. 1.3 cm Pb Shower initiated by particle in lead plates of cloud chamber What is the particle? Fretter: Echo Lake, 1949

14. Incoming particle is highly likely to be a proton • Level of ionisation excludes heavier nucleus (dE/dx) ~Z2 • Traversal of the particle through 6 Pb-plates • (about 88.5 g cm-2 or 13.9 rad. lengths) strongly • excludes an electron. • Reasonable to have point of interaction in 7th plate. • pCR + p p + p + N(+ + - + 0) • Also K, , , , …. are undoubtedly created • What is the energy of the particle?

15. 1.3 cm Pb Shower initiated by particle in lead plates of cloud chamber Fretter: Echo Lake, 1949

16. The following discussion and slides are due to Jim Matthews (LSU). The treatment is an approximation intended to exhibit some of the physics driving the main features of air showers plainly and simply. It does not replace full simulations It is a useful pedagogic tool from which we can learn a lot. J Matthews: Astropart. Phys. 22 (2005) 387.

17. The Heitler Model E0 W. Heitler, The Quantum Theory of Radiation ,3rd Ed., (1954), p.386.

18. “Heitler’s Model” E0 E0/2 W. Heitler, The Quantum Theory of Radiation ,3rd Ed., (1954), p.386.

19. Heitler Model E0/4 W. Heitler, The Quantum Theory of Radiation ,3rd Ed., (1954), p.386.

20. Heitler Model } d = mean distance } d λr= 37 g cm-2 in air (radiation length)

21. After “n” splits, there are “N” particles (e+, e-, and photons):

22. After “n” splits, there are “N” particles (e+, e-, and photons): • Energy is evenly split between two secondaries. • Cascade stops after “nc” splits when individual energies are too low: critical energyξc. • (ξc is when collision losses > radiative losses, 85 MeV in air)

23. Things the Heitler Model does well: Nmax~ Eo - but not constant of proportionality Xmax~ log Eo = 2.3 λr = (85 g cm-2 )/decade Things it does not do: - relative numbers of photons/electrons - attenuation, especially after maximum

25. Nch=Nπ±=10 , Nπo = 5 Extension to hadronic cascade is Jim Matthews’s original contribution Nμ = Nπ±

26. For pions, the distance between events uses the interaction length λI = 120 g cm-2 The hadronic critical energy is reached when the distance to the next interaction exceeds the (dilated) lifetime d = λI ln2 ξc = 20 GeV

27. After n generations, there are Nπ pions: Nπ = (Nch)n Total energy carried by all these pions: (2/3)n Eo e.g. ~ 10% after 5 generations So each pion has:

28. When pions drop below their critical energy : Eπ = ξc = 20 GeV all π± decay to muons

29. (Full simulations give β = 0.85 – 0.92) (n.b.: logarithmic dependence on Nch )

30. The growth of Nμwith Eo is less-than-linear (β < 1). Lower energy showers are more “efficient” in muon production This is why Fe primaries make more muons than protons do (superposition model: 56 showers each with E = Ep/56) β depends on the (logarithmic) ratio of charged to neutral pions

31. The primary energy of the shower is divided into EM and hadronic channels: Use observedNe = Nmax / g, g≈10:

32. CASA-MIA

33. Depth of shower-maximum must be treated a little more carefully because it strongly depends on the first interaction. • Do an EM shower with (1/3 Eo)/Nch • Use increasing multiplicity Nch~ Eo1/5 • Use energy dependent p-air λI

34. Depth of shower-maximum must be treated a little more carefully because it strongly depends on the first interaction. • Do an EM shower with (1/3 Eo)/Nch • Use increasing multiplicity Nch~ Eo1/5 • Use energy dependent p-air λI Not deep enough by ~100 g cm-2

35. Express in terms of EM-shower Xmax : Elongation rate is in very good agreement with detailed simulations: (Note Linsley’s “elongation rate theorems” from ’77 Plovdiv ICRC: Λγis an “upper limit”)

36. p, Fe lines shifted (up) by 100 g/cm2 gamma line not shifted (FULL SIMS) slopes are in good agreement We will come back to this diagram in lecture 3 “Full Sims” from: Heck et al., Hamburg ICRC (p.233); Fowler et al., Ast.P.Ph. 15 (2001),49.

37. Inelasticity Inelasticity – a leading particle carrying away a significant fraction of energy – can be treated in this model κ= fraction of energy available for pion production Increases β by about 10% :

39. Summary • Eo ~ (Ne + 25 Nμ).Linear relation, independent of primary type, from partition of energy into hadronic and EM channels. Weighting from the ratio of characteristic energies where EM and hadronic multiplications cease. • Nμ~ Eoβ , β = 0.85-0.95. The value of β is from the relative numbers of charged and neutral secondaries. Inelasticity and overall multiplicity cause 10% effects. • Elongation rate is less than pure EM. This difference arises from the growth of the p-air cross section and the increase in multiplicity

40. Kinetic energy per nucleus Energetics of cosmic rays • Energy density: rE ~ 10-12 erg/cm3 ~ B2 / 8p • Power needed: rE / tesc galactic tesc ~ 3 x 106 yrs Power ~ 10-26 erg/cm3s • Supernova power: 1051 erg per SN ~3 SN per century in disk ~ 10-25 erg/cm3s • SN model of galactic CR Power spectrum from shock acceleration, propagation BUT - no DIRECT evidence for SNR

41. Knee Importance of knowing the Hadronic Physics Ankle air-showers >1019 eV 1 km-2 sr-1 year-1 after Gaisser

42. LHC Forward Physics & Cosmic Rays Models describe Tevatron data well - but LHC model predictions reveal large discrepancies in extrapolation. ET (LHC) E(LHC) η = - log (tan ) James L. Pinfold IVECHRI 2006 13

43. LHCf: an LHC Experiment for Astroparticle Physics: wisdom of Kakenhi agency in Japan LHCf: measurement of photons and neutral pions and neutrons in the very forward region of LHC Adding an EM calorimeter at 140 m from the Interaction Point (IP1 ATLAS) For low luminosity running

46. Prospects from LHCf

47. The p-p total cross-section LHC measurement of sTOT expected to be at the 1% level – useful in the extrapolation up to UHECR energies 10% difference in measurements of Tevatron Expts: (log s) James L. Pinfold IVECHRI 2006 14

48. 1.3 cm Pb Shower initiated by particle in lead plates of cloud chamber Fretter: Echo Lake, 1949

49. Shower Detection Methods ~1° Nitrogen fluorescence 300 – 400 nm Fluorescence in UV → OR Array of water-Cherenkov or scintillation detectors 11