Neutrino Astrophysics with IceCube

1. Neutrino Astrophysics with IceCube Kael Hanson UniversitéLibre de Bruxelles 12th Marcel Grossmann Meeting 13 – 18 July 2009 Paris

2. Two-minute IceCube quiz Easy Question: Is IceCube a TeV-scale neutrino observatory? Answer: Yes, of course. IceCube was designed to optimize the response to ν-induced µ in energy range between 1-100 TeV (Aeff,ν ≈ 10 m2 at 10 TeV). Muon energy loss related to Eµ above 1 TeV. Angular resolution of 1°. Harder Question: Is IceCube a GeV-scale neutrino observatory? Answer: Also yes. With the addition of the DeepCore detector, the threshold energy is lowered from 100 GeV to 10 GeV. In addition, 4π solid angle acceptance possible due to shielding power of surrounding detector. Trick Question: Is IceCube an MeV-scale neutrino observatory? Answer: Yes and no. By turning the array into a simple photon counting system, it is possible to detect bursts of low-energy neutrinos emitted by supernova explosions. Discrete events are lost – you are left with a rate-vs-time. However, the effective volume is extremely large and with the resulting high statistics quite a bit of information can be extracted.

3. Motivation: why the supernova connection? • Only a handful of confirmed astrophysical sources of neutrinos • Neutrinos from the Sun • Atmospheric neutrinos • Neutrinos from 1987A supernova explosion • Measurement of temporal profile of neutrino burst from SNe in our galaxy would be invaluable data for explosion models. • Realtime monitoring into worldwide burst alert network can give hours of advance warning to optical observers. • Finally, de gustibus non estdisputandum

4. The IceCube Collaboration R. Abbasi24, Y. Abdou18, T. Abu-Zayyad29, J. Adams13, J. A. Aguilar24, M. Ahlers28, K. Andeen24, J. Auffenberg35, X. Bai27, M. Baker24, S. W. Barwick20, R. Bay7, J. L. Bazo Alba36, K. Beattie8, J. J. Beatty15,16, S. Bechet10, J. K. Becker17, K.-H. Becker35, M. L. Benabderrahmane36, J. Berdermann36, P. Berghaus24, D. Berley14, E. Bernardini36, D. Bertrand10, D. Z. Besson22, M. Bissok1, E. Blaufuss14, D. J. Boersma24, C. Bohm30, J. Bolmont36, O. Botner33, L. Bradley32, J. Braun24, D. Breder35, T. Castermans26, D. Chirkin24, B. Christy14, J. Clem27, S. Cohen21, D. F. Cowen32,31, M. V. D'Agostino7, M. Danninger30, C. T. Day8, C. De Clercq11, L. Demirörs21, O. Depaepe11, F. Descamps18, P. Desiati24, G. de Vries-Uiterweerd18, T. DeYoung32, J. C. Diaz-Velez24, J. Dreyer17, J. P. Dumm24, M. R. Duvoort34, W. R. Edwards8, R. Ehrlich14, J. Eisch24, R. W. Ellsworth14, O. Engdegård33, S. Euler1, P. A. Evenson27, O. Fadiran4, A. R. Fazely6, T. Feusels18, K. Filimonov7, C. Finley24, M. M. Foerster32, B. D. Fox32, A. Franckowiak9, R. Franke36, T. K. Gaisser27, J. Gallagher23, R. Ganugapati24, L. Gerhardt8,7, L. Gladstone24, A. Goldschmidt8, J. A. Goodman14, R. Gozzini25, D. Grant32, T. Griesel25, A. Groß13,19, S. Grullon24, R. M. Gunasingha6, M. Gurtner35, C. Ha32, A. Hallgren33, F. Halzen24, K. Han13, K. Hanson24, Y. Hasegawa12, J. Heise34, K. Helbing35, P. Herquet26, S. Hickford13, G. C. Hill24, K. D. Hoffman14, K. Hoshina24, D. Hubert11, W. Huelsnitz14, J.-P. Hülß1, P. O. Hulth30, K. Hultqvist30, S. Hussain27, R. L. Imlay6, M. Inaba12, A. Ishihara12, J. Jacobsen24, G. S. Japaridze4, H. Johansson30, J. M. Joseph8, K.-H. Kampert35, A. Kappes24,a, T. Karg35, A. Karle24, J. L. Kelley24, P. Kenny22, J. Kiryluk8,7, F. Kislat36, S. R. Klein8,7, S. Knops1, G. Kohnen26, H. Kolanoski9, L. Köpke25, M. Kowalski9, T. Kowarik25, M. Krasberg24, K. Kuehn15, T. Kuwabara27, M. Labare10, S. Lafebre32, K. Laihem1, H. Landsman24, R. Lauer36, D. Lennarz1, A. Lucke9, J. Lundberg33, J. Lünemann25, J. Madsen29, P. Majumdar36, R. Maruyama24, K. Mase12, H. S. Matis8, C. P. McParland8, K. Meagher14, M. Merck24, P. Mészáros31,32, E. Middell36, N. Milke17, H. Miyamoto12, A. Mohr9, T. Montaruli24,b, R. Morse24, S. M. Movit31, R. Nahnhauer36, J. W. Nam20, P. Nießen27, D. R. Nygren8,30, S. Odrowski19, A. Olivas14, M. Olivo33, M. Ono12, S. Panknin9, S. Patton8, C. Pérez de los Heros33, J. Petrovic10, A. Piegsa25, D. Pieloth17, A. C. Pohl33,c, R. Porrata7, N. Potthoff35, P. B. Price7, M. Prikockis32, G. T. Przybylski8, K. Rawlins3, P. Redl14, E. Resconi19, W. Rhode17, M. Ribordy21, A. Rizzo11, J. P. Rodrigues24, P. Roth14, F. Rothmaier25, C. Rott15, C. Roucelle19, D. Rutledge32, D. Ryckbosch18, H.-G. Sander25, S. Sarkar28, S. Schlenstedt36, T. Schmidt14, D. Schneider24, A. Schukraft1, O. Schulz19, M. Schunck1, D. Seckel27, B. Semburg35, S. H. Seo30, Y. Sestayo19, S. Seunarine13, A. Silvestri20, A. Slipak32, G. M. Spiczak29, C. Spiering36, M. Stamatikos15, T. Stanev27, G. Stephens32, T. Stezelberger8, R. G. Stokstad8, M. C. Stoufer8, S. Stoyanov27, E. A. Strahler24, T. Straszheim14, K.-H. Sulanke36, G. W. Sullivan14, Q. Swillens10, I. Taboada5, A. Tamburro29, O. Tarasova36, A. Tepe35, S. Ter-Antonyan6, C. Terranova21, S. Tilav27, P. A. Toale32, J. Tooker5, D. Tosi36, D. Turčan14, N. van Eijndhoven34, J. Vandenbroucke7, A. Van Overloop18, B. Voigt36, C. Walck30, T. Waldenmaier9, M. Walter36, C. Wendt24, S. Westerhoff24, N. Whitehorn24, C. H. Wiebusch1, A. Wiedemann17, G. Wikström30, D. R. Williams2, R. Wischnewski36, H. Wissing1,14, K. Woschnagg7, X. W. Xu6, G. Yodh20, S. Yoshida12 RWTH Aachen University University of Alabama University of Alaska Anchorage Clark-Atlanta University Georgia Institute of Technology Southern University University of California, Berkeley Lawrence Berkeley National Lab Humboldt-Universitätzu Berlin UniversitéLibre de Bruxelles VrijeUniversiteitBrussel Chiba University University of Canterbury University of Maryland Ohio State University Ohio State University TU Dortmund University University of Ghent Max-Planck-InstitutfürKernphysik University of California, Irvine ÉcolePolytechniqueFédérale University of Kansas University of Wisconsin, Madison University of Wisconsin, Madison University of Mainz University of Mons-Hainaut Bartol Research Institute University of Oxford University of Wisconsin, River Falls Stockholm University Penn State University Penn State University Uppsala University Utrecht University University of Wuppertal DESY Zeuthen

5. The IceCube Detector • When complete 2012 • 80 in-ice strings • 6 deep core strings • 160 surface airshower tanks • 2009 “IC59” status • 58 normal in-ice strings • 1 DeepCore string • 118 surface tanks • 3730 channels in the DAQ • 1.8 kHz trigger rate (CR µ) • 15 MB/sec raw data to tape • 50 GB per day filtered data over satellite link from Pole • 300 atmospheric neutrinos per day at trigger level • 2008 IC40 run complete now analyzing data from this period (5/08 – 5/09)

6. IceCube DeepCore • Increase in detector eff. at 10 – 100 GeV • IceCube strings form veto shield around core for 4π acceptance • Ice extremely clear at depth (λatt > 150 m) • Low-energy topics • Atmospheric neutrinos • Neutrino mass hierarchy • Dark matter • Low energy astrophysical fluxes

7. Drilling and deployment

8. Digital Optical Module technology DOM Optical Large Area Photocathode10” (500 cm2) Hamamatsu R7081-02 bialkali PMT (peak QE 24% @ 420 nm); High QE variant (peak QE 35% @ 420 nm) used in DeepCore DOMs Low noise< 300 Hz background counting rate in-ice (with artificial deadtime - see later) Glass / Gel ImprovementsBetter transmission in 330 - 400 nm relative to AMANDA OM Optical calibrationEach DOM is calibrated ε(λ) in the lab to about 7%; in-situ flasher board additionally permits in-ice measurements DOM Electronics “Smart” sensor digital technologyVersatile FPGA design with option to expand / change programming at any point in lifecycle. Core of supernova DAQ resides inside DOM itself. Array TimingHandled in DOM logic - DOM-to-DOM timing good to 2-3 ns using RAPCal method. Low power - 3.75 W / DOM 15 of 3776 DOMs are useless, 35 more have serious problems. As of June 2009 all DOMs have been produced.

9. Part II TeV Neutrino Astro-PARTICLE PHYSICS WITH IceCube

10. Cosmic ray acceleration Model of CR acceleration in shocks of SN remnants fits observation but not confirmed. TeVγ emission now established for many sources but could be from EM processes. Neutrino emission would be “smoking gun” for hadronic acceleration in these sources.

11. The neutrino sky

12. Detecting TeVν in IceCube Neutrino undergoes CC or NC interaction with nuclear material, produces charged particles which emit Cherenkov radiation CC NC linear tracks CASCADES “Double-bang” νμ(or UHE ντ) produces μ or τ via CC scatter. which can travel for many kilometers along linear track radiating Cherenkov photons in conical wavefront about track. The extended range of taus and muons means vertex can lie far outside detector volume – detector effective area is key performance parameter. Angular resolution of 1° is achievable with sophisticated maximum likelihood track reconstruction. νe CC or νX NC nuclear interactions produce either EM or hadronic cascades. These produce enormous amounts of Cherenkov photons (108 photons per TeV) radiated over 4π. The extent of the cascade is ~10 m longer at UHE due to LPM. Detector effective volume is operational parameter for cascades. Good energy resolution – ice is caloric medium. Poor angular resolution VHE ντ interacting inside the detector produces the primary recoil cascade and a τ which can propagate many 100’s of m at HE. τ decay produces a secondary cascade – leaving a very distinct event. Other topologies as well: “lollipop” and “sugar daddy.” τ channel has no ATM background. UHE ντ fluxes can regenerate and are not absorbed by passage through Earth.

13. 900 PeV cosmic ray event

14. Atmospheric Neutrinos Energy and baseline of atmospheric neutrinos: able to probe regions of parameter space for Lorentz violation and quantum decoherence completely inaccessible D. Chirkin ICRC 2009 New techniques developed for unfolding the energy spectrum of atmospheric neutrinos – here from IC22 data. Heulsnitz & Kelley ICRC 2009

15. Point Sources ½ year of IC40 data – 175.5 d live time 17777 evts – 6796 up, 10981 down IceCube Preliminary Dumm ICRC 2009

16. Diffuse Neutrinos (IC22) • Extraterrestrial neutrino flux harder than atmospheric neutrinos – look for HE excess of events • Tricky analysis – very sensitive to systematics in Monte Carlo simulation • IC22 diffuse results are just now being released (Hoshina 2009 ICRC) • Use 3 ‘simple’ energy estimators • Nch : # of hit channels • Npe: integral of reconstructed Q from DOM waveforms • µ dE/dX : muon energy loss from photon tables • Nch and Npe showing significant excess at high multiplicity while µ dE/dX is consistent with atmopheric neutrino background • (Continuing) investigation of systematics associated with very primitive channel, charge counting • Limit from dE/dX analysis:

17. Neutrinos from Gamma-Ray Bursts • GRB fireball model predicts HE neutrinos from pγ interactions in GRB jet. • Satellite-triggers (Fermi/SWIFT) used to pinpoint the burst search windows and reduce background (looser cuts possible) Right after deployment of the 40 strings last year the theoretically-visible-to-the-naked-eye GRB 080319B went off. Unfortunately IceCube was in maintenance mode at the time and only 9 strings were active. Search for high-energy muon neutrinos from the “naked-eye” GRB080319B with the IceCube neutrino telescopearXiV:0902.0131accepted by ApJ

18. Neutrino-triggered optical follow-up • IceCube ‘trigger’ to optical network (ROTSE) on neutrino multiplet: 2 or more neutrino events inside 100 s and 4° space angle (25 accidentals / year for M=2) • Motivation: GRBs, gamma-poor bursts, SNe with jets and TeV neutrino emission • Program initiated end of 2008

19. Dark Matter

20. Part III MeV Neutrino Astro-PARTICLE PHYSICS WITH IceCube

21. Supernovae • For star of mass > 8 M☉ it is possible to develop Fe core ~ 1.4 M☉. • Burning of Fe not exothermic – core collapses under it’s own gravity. • Almost all (99%) of gravitational energy of collapsing core is radiated away as neutrinos – approx 1053 erg. • Not all supernovae are core collapse supernovae. Also Type Ia (used as standard candles in redshift measurements) which do not produce strong neutrino emission. • Core collapse are Type Ib/c and Type II

22. Observation of neutrinos from SN1987A

23. Detection of MeVν’s from SNe Primary detection channel is inverse βdecay Note this is only sensitive to the electron anti-neutrinos. Electron neutrinos (in particular those from the de-leptonization burst) detected primarily from or The weak Cherenkov signal from any particular neutrino-nucleus interaction for IceCube-scale detector seen by at most one PMT. However, for sufficiently intense burst of finite duration, the counting rates of many such interactions recorded in many PMTs may be combined in manner to discriminate from background count rate.

24. IceCube DOM Effective Volume (MeV neutrinos) • SNe models give <E3> ~ 15 MeV • One may easily derive approx photon yield from positrons: 2500 • This gives per PMT, effective volume of 450 m3 • 2 Mton target mass for IC86 • 500,000 counts above background expected for SN1987A at galactic center • cf. rms noise fluctuation of 3700 •  S/N = 130 : 1 Effective volume scales as photocathode area, attenuation length, and E3, thus dependent on temperature of SNe. We compute in E-independent manner the single photon eff volume, Veff,γ(note PMT response already folded into this expression): <Veff,γ> = 0.185 m3

25. Photon counting in IceCube • PMT counts pulse crossing discriminator threshold of 0.25 pe • DOM logic maintains virtual scalers: 4-bit counters with integration time 1.64 ms. • Edges of bins defined in DOM clock frame – translation to global time via RAPCal method • Introduction of artificial deadtime important to optimize S/N in presence of optical artifacts: PMT after-pulsing and other late light effects. • The scalers from each channel collected and sent to assembly phase where they are chronologically sorted and then written to disk file (3 MB/sec) for • handoff to realtime supernova alert system • permanent storage

26. Supernova detection in realtime • Realtime supernova alert system at pole consumes data files emitted by DAQ system • Merge and globally align scaler bins coming from DOMs • Compute the following statistics to search for excess counts Signal + error Background elimination Internally we generate multiple triggers per day for monitoring purposes. Of these, high significance triggers sent to SNEWS via 24/7 Iridium satellite messaging system. SNEWS alert rate < 2 / week – average latency is approx. 10 min.

27. SNeν’s as probes of the supernova explosion mechanism • Various phases of evolution signaled by change in neutrino flux • infall • neutronization burst • accretion • cooling • Theoretical models still have trouble producing explosion – neutrinos may be critical ingredient restarting the stalled shockwave • For very massive stars > 25 M☉ supernova explosion may be interrupted by formation of a black hole: optical burst not present – neutrino fluxes characterized by increase in temperature and eventual truncation.

28. Particle physics with SN ν

29. Conclusions • Add these

30. Backup please! Supplementary Material

32. RAPCal

33. RAPCal Analog Waveform

34. Neutrino Effective Area

35. UHE Neutrino Cross Sections

36. Tau neutrino events