Neutrino Astronomy: Exploring the Universe with High-Energy Neutrinos

1. AMANDA and IceCube: Neutrino Astrophysics at the South Pole Brennan Hughey University of Wisconsin – Madison Hope College November 12th, 2004

2. Outline • Neutrinos and Neutrino Astronomy • Overview of AMANDA and IceCube • Rolling Search for Gamma Ray Burst Signal • A Quick Look at Several AMANDA Analyses • Life at South Pole

3. Neutrinos “Neutrinos they are very small. They have no charge and have no mass And do not interact at all…..” - John Updike Actually, they do have a very small mass, but it’s really really tiny. Of all the forces in nature (Electromagnetic, Gravitational, Strong and Weak Nuclear forces) neutrinos interact only through the weak force, and then only very rarely. On average, roughly one neutrino decays in a human body in a lifetime, even though trillions of them are passing through your body every second.

4. Neutrinos Pauli realized neutrinos had to exist because something had to be taking extra energy away when a neutron decays into a proton and electron (called beta decay) “I have done a terrible thing. I have postulated a particle that cannot be detected.” – WolfgangPauli

5. Neutrinos Since they occasionally interact through the weak force, you actually can detect them Detected sources include: nuclear reactors the sun Supernova 1987 A atmospheric neutrinos The sun in neutrinos as seen by Superkamiokande

6. Neutrino Flavors e: electron neutrino : muon neutrino : tau neutrino Each flavor also has a corresponding antiparticle Produces corresponding particle when it interacts: Electron neutrino makes electron, Muon neutrino produces muon, Tau neutrino produces tau particle

7. So why are we studying high energy neutrinos? Solving mystery of where high energy cosmic rays (charged particles like protons and Iron nuclei) come from: Finding high energy neutrinos from a source indicates there are processes there which could create high energy cosmic rays as well.

8. So why are we studying high energy neutrinos? The same thing that makes neutrinos hard to detect makes them unique cosmic messengers. High energy photons get absorbed by photon fields and ambient matter. Cosmic rays get deflected by magnetic fields Neutrinos point straight back to their source

9. Provides a new way to view universe: we normally see photons, what if we could “see” with neutrinos?

10. Neutrinos and Neutrino Astronomy • Overview of AMANDA and IceCube • Rolling Search for Gamma Ray Burst Signal • A Quick Look at Several AMANDA Analyses • Life at South Pole

11. The AMANDA Detector AMANDA-B10 (inner core of AMANDA-II) 10 strings 302 OMs Data years: 1997-99 AMANDA-II 19 strings 677 OMs Trigger rate: 80 Hz Data years: 2000+ PMT noise: ~1 kHz “Up-going” (from Northern sky) “Down-going” (from Southern sky) South Pole ice: Very transparent, but contains dust layers which affect absorption and scattering Optical Module

12. South Pole Dome road to work AMANDA Summer camp 1500 m Amundsen-Scott South Pole station 2000 m [not to scale]

13. IceTop AMANDA South Pole Runway 1400 m 2400 m IceCube First strings deploy in January 2005 Completed in 2010 70-80 strings Up to 4800 OMs in deep ice 1 km3 instrumented volume IceTop: 320 OM, 1 km2 surface array

14. The AMANDA Collaboration Europe VUB-IIHE, Brussel ULB-IIHE, Bruxelles Université de Mons-Hainaut Imperial College, London DESY, Zeuthen United States SouthAmerica Bartol Research Institute UC Berkeley UC Irvine Pennsylvania State UW Madison UW River Falls LBNL Berkeley U. Simón Bolivar, Caracas Mainz Universität Wuppertal Universität Stockholms Universitet Uppsala Universitet Kalmar Universitet Antarctica South Pole Station IceCube Oxford University, England Chiba University, Japan Christchurch, New Zealand

15. ANTARES La-Seyne-sur-Mer, France AMANDA, IceCube South Pole, Antarctica High EnergyNeutrino Telescope Projects BAIKAL Russia NEMO Catania, Italy DUMAND Hawaii (cancelled 1995) NESTOR Pylos, Greece

16. Cherenkov light cone muon interaction Detector • The muon radiates blue light in its wake • Optical sensors capture (and map) the light neutrino

17. Cherenkov Radiation The products of a neutrino-ice interaction emit a cone of blue Cherenkov radiation as they travel faster than the speed of light in ice*. You can see the same thing as a wake in water (or hear it as a sonic boom for sound waves) www.brantacan.co.uk/ cherenkov.htm *Yes, Einstein would be okay with this.

18. Optical Modules Photomultiplier: 10 inch Hamamatsu Active PMT base Glass sphere: Nautillus Mu metal magnetic shield

19. Photomultiplier Tubes Each OM contains a photomultiplier tube Cherenkov photon hits collecting plate and releases a photoelectron Photoelectron arrives in first dynode, which emits multiple electrons for each electron received Signal is amplified as it passes into each dynode in series Total amplification is ~109 for AMANDA, ~107 for ICECUBE Photon Dynode Photoelectron

20. Backgrounds Downgoing muons – created by cosmic ray interactions with atmosphere Outnumber neutrinos 1 million to 1 Cut out by direction Atmospheric neutrinos – Also created by cosmic rays hitting atmosphere AMANDA sees about 1 per day Distinguished from extraterrestrial neutrinos by energy Atmosphere Earth

21. Reconstruction cascade channel muon channel

22. Reconstruction Timing and intensity used to reconstruct muon path or cascade Multiple computer fits available – Muon, cascade, “first guess”

25. Deployment

26. Neutrinos and Neutrino Astronomy • Overview of AMANDA and IceCube • Rolling Search for Gamma Ray Burst Signal • A Quick Look at Several AMANDA Analyses • Life at South Pole

27. Outshine galaxies for short periods: Event Ergs Getting hit by a truck 1010 “Killer” asteroid striking the earth 1026 Emitted by our sun in 1 second 1033 Emitted by a supernova in 1 second 1041 Emitted by a GRB in 1 second 1052 Gamma Ray Bursts

28. Gamma Ray Bursts Originally discovered by Vela U.S. military network in 1969 while it monitored Soviet compliance with the nuclear test ban treaty Not evidence of an alien civilization, but still pretty interesting Eventually determined to be isotropic – hence extragalactic in origin http://cossc.gsfc.nasa.gov/batse/

29. Gamma Ray Bursts Possible sources: stellar collapse: in coincidence with Supernova (March 2003) merger between two objects: Black Hole -Black Hole, Black Hole -Neutron Star, et cetera http://jilawww.colorado.edu/www/gallery/captions/supernova.html Chandra X-ray image

30. Gamma Ray Bursts “When you've seen one GRB, you've seen one GRB” However, they have been roughly grouped into two classes

31. -1 hour +1 hour 10 min Preliminary AMANDA 1997-2000 Low background analysis due to space and time coincidence! Year Detector NBursts NBG, Pred NObs Event U.L. 1997 B-10 78 (BT) 0.06 0 2.41 1998 B-10 94 (BT) 0.20 0 2.24 1999 B-10 96 (BT) 0.20 0 2.24 2000 A-II (2 analyses) 44 (BT) 0.83/0.40 0/0 1.72/2.05 97-00 B-10/A-II 312 (BT) 1.29 0 1.45 2000 A-II 24 (BNT) 0.24 0 2.19 2000 A-II 46 (New) 0.60 0 1.88 2000 A-II 114 (All) 1.24 0 1.47 97-00 Flux Limit at Earth*: E2≤4·10-8GeV cm-2 s-1 sr-1 *For 312 bursts w/ WB Broken Power-Law Spectrum (Ebreak= 100 TeV, ΓBulk= 300) Search for  coincident with GRBs Blinded Window Data required to be stable within an hour on either side of GRB Background taken ±5 minutes around burst (BT = BATSE Triggered BNT = BATSE Non-Triggered New = IPN & GUSBAD)

32. Rolling Search Satellites missing a large portion of GRBs, especially since loss of BATSE on CGRO in early 2000 Scans through entire year (2001) of data to look for excess number of events In a bin compared to background expectation 1 second and 15 second time windows for short and long bursts Time

33. Blindness AMANDA analyses are done “blind” – meaning all cuts to get rid of data and other procedures are done on Computer simulations or a part of real data that won’t be used in actual Analysis This prevents massaging data to get a signal to show up My rolling search uses 5 days from throughout the year to determine cuts

34. Data Reduction A major part of every AMANDA analysis is getting rid of background events while keeping as much of the signal you’re looking for as possible 3 steps: 1. High Energy cut (number of OMs hit and number with at least 2 hits) Keeps 1% of background and 2/3 of signal 2. Cut on number of direct hits (hits that agree with muon hypothesis) Keeps 7% of remaining background and 99% of signal 3. Six variable support vector machine cut

35. Data Reduction ----------------- Real Data Background Computer Simulation (should match data) Computer Simulation of Signal

36. ----------------- Real Data Background Computer Simulation Computer Simulation of Signal

37. Support Vector Machine Program is fed training data of signal and background events “Learns” best six-dimensional cut to keep signal and throw away background “Cost factor” can be varied so that more or less data is thrown away Keep Keep

38. What Counts as a signal? Data follows poissonian distribution (time between events follows exponential pattern) Cuts selected so that there is less than 1% chance of a false detection in both searches for whole year: 3 events in a bin needed

39. Broken Power Law Spectrum ______ 100 TeV break _ _ _ _ 300 TeV break ........... 700 TeV break

40. Detector Stability

41. Cascade Effective Volume/ Neutrino Effective Area

42. Upcoming…. Finish rolling search – unblind Do satellite coincident search taking advantage of framework already laid Do coincidence studies between AMANDA and Milagro (an air shower array that studies gamma rays)

43. Neutrinos and Neutrino Astronomy • Overview of AMANDA and IceCube • Rolling Search for Gamma Ray Burst Signal • A Quick Look at Several AMANDA Analyses • Life at South Pole

44. Preliminary AMANDA 2000 E2(E) < 2.87·10–7 GeV cm-2 s-1 sr-1 1997 results PRL 90, 251101 (2003) Atmospheric Neutrinos Atmospheric spectrum provides test of detector Matches lower energy Frejus data Downgoing muon background rejected using a neural network First spectrum above 1 TeV Used to set limit on extraterrestrial E-2 diffuse flux in the range 100-300 TeV Includes 33% systematic uncertainty

45. E2all (E) < 8.6·10–7 GeV cm-2 s-1 sr-1 flavor mixing e::=1:1:1 50 TeV < E < 5 PeV Nobs = 1 event Natm μ = 0.90 Natm  = 0.06 ± 25%norm +0.69 -0.43 +0.09 -­0.04 Accepted for Publication Astroparticle Physics Diffuse Flux Search using Cascades 2000 data 197.2 days livetime Cascade analysis has 4 coverage Event selection based on -energy -topology Signal MC: E-2 energy spectrum

46. Nobs = 5 events Nbgr = 4.6 ± 36% events horizontal events E2all (E) < 9.910-7 GeV cm-2 s-1(e::= 1:1:1) Paper in progress Ultra-High Energy Search (PeV - EeV) neutrino effective area vs. energy Earth opaque above 1016 eV Look at downgoing muons and events near horizon Characteristics: few 1 p.e. Peaks long muon tracks and bright events average all angles

47. PRELIMINARY Diffuse Results Summary AMANDA 90% CL upper limits to a diffuse E-2 all neutrino flux obtained from : • High energy tail of atmospheric neutrino spectrum • search for cascade events • search for UHE events Several Models of AGN neutrino emission ruled out

48. Preliminary AMANDA 2000-2003 x 10-6 GeV-1 cm-2 s-1 Search for High Energy Point Sources No Excess observed results consistent with atmos. background 3369 events observed 3438 events expected background Skymap in equatorial coordinates Search in sky for clustering of events : • Grid search : sky subdivided in 300 bins • Shift grid to cover boundaries • Pointing resolution ~ 2.5 ° • Optimized in each declination band • Optimized for E-2 and E-3 spectra sin()

49. Preliminary AMANDA 2000-2003 Data scrambled in right ascension Neutrino Point Sources: Unbinned Analysis Significance of local fluctuations compared to expectation of all being atmospheric neutrinos Max 3.4 sigma: consistent with background fluctuation

50. Other Analyses WIMP search – Search for Weakly Interacting Massive Particles (a candidate for dark matter) annihilating in the sun or Earth Magnetic Monopoles search Supernova Monitoring – MeV neutrinos from Supernovae are too low-energy to be detected individually, but lead to an overall increase in detector noise