1. Neutrinos and the origin of the cosmic rays TexPoint fonts used in EMF: AAA Walter Winter DESY, Zeuthen, Germany ICRC 2015 The Hague, Netherlands July 30-Aug 6, 2015 … or: what neutrinos can tell us about the origin of the cosmic rays

2. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 2 Contents > Introduction > How many of the observed cosmic neutrinos come from cosmic ray interactions in the Milky Way? > Can the observed neutrinos come from the sources of the ultra-high energy cosmic rays? > Are gamma-ray bursts the sources of the ultra-high energy cosmic rays – even if they cannot describe the observed neutrino flux? > Summary

3. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 3 2015: 54 high energy cosmic neutrinos IceCube: Science 342 (2013) 1242856; Phys. Rev. Lett. 113, 101101 (2014); Halzen at WIN 2015 No evidence for Galactic origin, no significant clustering: diffuse extragalactic flux? + Cascades × Muon tracks The Earth is intransparent for E >> 10 TeV

4. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 4 Guaranteed contribution: Neutrinos from CR interactions in the Milky Way > Cosmic rays interact with hydrogen in our Galaxy > Production region can be inferred from diffuse gamma-ray observations (very narrow around Galactic plane) > Complication: the CR composition changes non-trivially in relevant range: Fermi-LAT, ApJ 750 (2012) 3 (see also arXiv:1410.3696) Gaisser, Stanev, Tilav, 2013 UHECRs prima- ries How many of the observed neutrinos come from these interactions? (see discussions in Evoli, Grasso, Maccione, 2007; Ahlers, Murase, 2014; Joshi, Winter, Gupta, 2014; Kachelrieß, Ostapchenko, 2014; Neronov, Semikoz, Tchernin, 2014; Ahlers, Bai, Barger, Lu, 2015; …)

5. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 5 > Assumptions: single component composition directly inferred from data, cosmic ray density everywhere directly from observation at Earth > Does not rely on modeling sub-leading components (e.g. hydrogen if large-A dominates), which could give additional contributions > Includes interactions of extra-galactic cosmic rays in Milky Way > Conclusion from  Fit to data  Required hydrogen densities  Cosmic ray composition  Gamma-ray constraints (Fermi, etc) > About 0.6 events expected > Robust (from observations only), but model-dependent estimates typically somewhat higher (few events) How many of the neutrinos come from CR interactions? Joshi, Winter, Gupta, MNRAS, 2014 (all-sky averaged prediction)

6. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 6 Neutrinos and the origin of the cosmic rays (?) CR   Require hadron acceleration in sources Connection depends on hadronic loading Long outstanding issues (highest E): Wherefrom? Composition? The new player in town: New ways to identify the sources of the cosmic rays?! Waxman, Bahcall, 1999 Energy input required to power UHECRs + efficient neutrino production  Waxman-Bahcall bound. Matches obs.: Coincidence???

7. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 7 A simple model describe the astrophysical neutrinos > Example: Neutrinos from Ap interactions; sources cosmologically distributed > Find possible fits to data: Winter, PRD 90 (2014) 103003, arXiv:1407.7536 Protons  =2 B ~ 10 4 G (magnetic field effects on sec. pions, muons, kaons) Nuclei  =2, E max =10 10.1 GeV Composition at source with  =0.4 Protons  =2 E max =10 7.5 GeV Protons,  =2.5 (agrees with arXiv:1507.03991) Challenge: Murase-Ahlers- Lacki bound, 2103

8. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 8 Can these neutrinos come from the sources of the UHECRs? … conceptual insights WW, arXiv:1407.7536 Protons  =2 B ~ 10 4 G Nuclei  =2, E max =10 10.1 GeV Composition at source with  =0.4 Protons  =2 E max =10 7.5 GeV Protons,  =2.5 Yes, but: Energy input per decade very different in neutrino-relevant and UHECR energy ranges (Energetics seem to favor  ~2 – Waxman/Bahcall!) see e. g. Katz et al, 1311.0287 for generic discussion; for GRBs specifically: extremely large baryonic loadings implied: Baerwald et al, Astropart. Phys. 62 (2015) 66 Yes, but: Synchrotron losses limit maximal proton energies as well. Need large Doppler factors (  GRBs?) Yes. Need energy-dependent escape timescale leading to break/cutoff within source (diff. from ejection!) see e.g. Liu et al, PRD, 2014; arXiv:1310.1263 or starburst galaxies Yes, but: A(E) change somewhat too shallow to match observation; difference source-observation from propagation?

9. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 9 > Idea: Use timing and directional information to suppress atm. BGs > Strong constraints from GRB stacking IceCube, Nature 484 (2012) 351; see arXiv:1412.6510 for update > Not the dominant source of observed diffuse flux! > Current limit close to prediction from gamma-rays; however: many assumptions (e.g. baryonic loading f e -1, , z) How about neutrinos from long gamma-ray bursts? (Source: NASA) GRB gamma-ray observations (e.g. Fermi, Swift, etc) (Source: IceCube) Neutrino observations (e.g. IceCube, …) Coincidence! (Hümmer, Baerwald, Winter, PRL 108 (2012) 231101) Observed

10. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 10 Can GRBs be the sources of the UHECRs? One zone model. (  p =2, fit range 10 10... 10 12 GeV, protons assumed) > Combined source-propagation model: f e -1 is obtained from UHECR fit > Neutrino connection depends on mechanism for cosmic ray escape; only if escaping neutrons (from p  ) dominate, the connection is strong! (Baerwald, Bustamante, Winter, Astropart. Phys. 62 (2015) 66; here figures with TA data)  =300  =800

11. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 11 > The different messengers originate from different collision radii of the same GRB > Direct escape (leakage) dominates UHECR escape for larger R C. Neutrinos are produced mostly close to the photosphere The new paradigm: Multiple collision models (Bustamante, Baerwald, Murase, Winter, Nat. Commun. 6 (2015) 6783, arXiv:1409.2874; see also Globus et al, 2014)

12. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 12 Consequences for neutrino production > The observed gamma-rays come (by definition) from beyond the photosphere > Use these only to predict a solid “minimal” neutrino flux from  -ray observations > Result: E 2  ~ 10 -11 GeV cm -2 s -1 sr -1 (much lower than one zone predictions) > Most importantly: The prediction is robust, i.e., hardly depends on , baryonic loading, … [as it is dominated by few collisions close to photosphere and  p  ~  Th ~1/R C 2 in the same way] E iso =10 53 erg per GRB Bustamante, Baerwald, Murase, Winter, Nature Commun. 6 (2015) 6783

13. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 13 Summary and conclusions > How many of the observed cosmic neutrinos come from cosmic ray interactions in the Milky Way? Only a few, at most. Maybe a few addl. ones from Galactic sources > Can the observed neutrinos come from the sources of the ultra-high energy cosmic rays, conceptually? This is possible, even in different spectral fit scenarios. Perhaps energy- dependent escape timescale most “natural” model > Are gamma-ray bursts the sources of the ultra-high energy cosmic rays?  They are not the main source of the observed cosmic neutrinos. Yet, they could be the source of the UHECRs  A key issue is the UHECR mechanism for the sources; another one that estimators from gamma-rays may not be applicable to neutrinos in (more realistic) multi-zone collision models  Neutrinos will play an important role in establishing the UHECR paradigm for GRBs, as the GRB sensitivity in IceCube is the best to any object class

14. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 14 BACKUP

15. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 15 UHECR fit of single-collision model Example: Efficient escape by Bohm-like diffusion,  =0.1 (Baerwald, Bustamante, Winter, Astropart. Phys. 62 (2015) 66) Direct/diffu sive escape significant Neutron model dominates IceCube excluded (current); Neutron model ruled out! IceCube excluded (15 years) Baryonic loading log 10 f e -1 UHECR fit

16. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 16 Challenges for neutrino flux prediction from CR interactions > Multi-population fit models reproduce the observed cosmic ray composition as fct. of E > Neutrino challenge: sensitive to lighter elements, mostly > However: these lighter elements may be subdominant components of the model > Neutrino flux strongly depends on those model assumptions without direct (observational) evidence Gaisser, Stanev, Tilav, 2013

17. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 17 Why is the new GRB prediction robust? > Neutrino flux comes from a few collisions at photosphere > Photospheric radius (  Thomson optical depth) and photohadronic interactions both depend on particle densities (scale in same way) > Consequence: Pion production efficiency at photosphere does not depend on  : (  : overall dissipation efficiency: dissipated/initial kinetic energy) > Changing the energy in electrons/photons also hardly affects results (if baryons dominate:  p ~ 1): Compare to Guetta et al, Astropart. Phys. 20 (2004) 429 (Bustamante, Baerwald, Murase, Winter, arXiv:1409.2874, Nature Commun. 6:6783 doi: 10.1038/ncomms7783 (April 2015))

18. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 18 (Bustamante, Baerwald, Murase, Winter, Nature Commun. 2015, arXiv:1409.2874; see also Globus et al, arXiv:1409.1271) Consequences for neutrino-UHECR connection > The different messengers originate from different regimes of the GRB > Are parameters derived from gamma-ray observations ( , t v, etc) really appropriate indicators for the neutrino flux? (protons)

19. Walter Winter | ICRC 2015 | July 30 – Aug 6, 2015 | Page 19 Parameter space dependence: Numerical cross-check! (Bustamante, Baerwald, Murase, Winter, Nature Commun.)