Accelerator Neutrino Oscillation Physics Lecture III

1. Accelerator Neutrino Oscillation PhysicsLecture III Deborah Harris Fermilab SUSSP St. Andrews, Scotland August 17, 2006

2. In the words of Ken Peach “When I was on an experiment to determine e’/e, once we were close to getting the result out, I realized something: All the theorists asked ‘what value did you measure?’ and All the experimentalists asked ‘what uncertainty on the measurement did you end up getting?’ ” This talk will try to speak to both theorists and experimentalists… but remember who wrote the talk… Accelerator Neutrino Oscillation Physics Lecture III

3. Goals of Long Baseline Oscillation Measurements • Measurements of “atmospheric neutrino oscillation parameters”: nm disappearance as a function of neutrino energy • Searches for CP violation and understanding the neutrino mass hierarchy: ne appearance • Verify Oscillation Framework: nt appearance • Search for Sterile Neutrinos: Neutral Current disappearance, looking for three distinct Dm2 Accelerator Neutrino Oscillation Physics Lecture III

4. Oscillation Experiments: Beams past, present, and near future… MiniBooNE OPERA T2K MINOS NOvA K2K Accelerator Neutrino Oscillation Physics Lecture III

5. Oscillation Experiments: Detectors past, present, and near future… OPERA e-,15 GeV, pT=1.16 GeV Vertex: 10,2p,3n,2 ,1e- ICARUS Super-K ne+A→p p+ p- e- nm CC ne CC NOvA MINOS Accelerator Neutrino Oscillation Physics Lecture III

6. Systematic Uncertainties Problem: uncertaintiesall affect the near and far detector both, you can’t always separate one from the other • Neutrino Flux • Hadron Production • p/K ratio • x and pt spectrum of produced Pions/Kaons • Beamline Geometry • Focusing uncertainties • Alignment Uncertainties • Neutrino Interactions: Background and Signal! • Quasi-elastic Uncertainties • Resonance (low W) Uncertainties • DIS (high W) • Nuclear Effects • Event Selection • Event Energy Resolution • Important especially for measurements versus neutrino energy • Narrow Band beams: energy resolution is key to background rejection Accelerator Neutrino Oscillation Physics Lecture III

7. Two detector experiment (in theory) N(NC)n=FnsNC N(CC)n=FnsCC Far detector Near detector F(L)=F0/L2 • Make two detectors as identical as possible - same scintillator, water, steel etc. • Measure n spectrum in the near detector • Predict the n spectrum in the far detector • Cross section uncertainties should cancel… • Detector efficiency uncertainties should cancel… • Simple, right? N(NC)f=FfsNC N(CC)f=FfsCC Accelerator Neutrino Oscillation Physics Lecture III

8. Two Detector Experiment (in practice) • Near Detector sees a line source of neutrinos, far detector sees a point source • Think: where do nm’s in beam decay compared to ne’s? • Near Detector will have different event rate • Beam-induced rates differ by 104 to 105 • Cosmic ray rates differ due to different shielding and detector size • Near Detector Design is different • Different electronics, PMT’s, active area coverage… • nm→nt may be large: nt CC suppression large • nm CC energy distribution is very different Accelerator Neutrino Oscillation Physics Lecture III

9. Near Detector Design • Far detector must be massive: the more instrumented it is, the more $/kton… • Tradeoff between segmentation and far detector mass • Near Detector Design options: • “Identical” to far detector • Argue that detector efficiencies and cross sections are the “same”, you just need independent flux measurements • Much more segmented and fine-grained • Try to measure fluxes and cross sections as best you can, make far detector prediction • Ideally, you would do both… Accelerator Neutrino Oscillation Physics Lecture III

10. K2K near neutrino detectors (K2K-II) SciBar detector Scintillating fiber tracker Full active scintillator tracker CH target (9.38t fid. vol.) (6t fid. vol.) water target CCQE identification Muon range detector Iron target (330t fid. vol.) nbeam • beam monitor (momentum & direction.) 1kt water Cherenkov detector water target (25t fiducial volume) Accelerator Neutrino Oscillation Physics Lecture III

11. T2K Near Detector Suite • What’s interesting here is how it differs from the K2K Near Detector suite: no Cerenkov detector Nakadaira, n2006 Accelerator Neutrino Oscillation Physics Lecture III

12. T2HK Near Detector Addition: 2km Detectors located 2km from target sees point sourceof neutrinos, like Far Detector Question: what about nm→ nt at the 2km detector? 2km detectors: Liquid Argon Water Cerenkov Muon Range Detector n direction Accelerator Neutrino Oscillation Physics Lecture III

13. MINOS Near Detector (cf Far) • 1040m from target (735km) • 103m underground(705m) • 980 ton mass(5400 ton mass) • 3.8m x 4.8m x 16m(8m octagon) • 282 steel + 153 scintillator planes(484 planes) • Two distinct sections: Front: Calorimeter • Every plane instrumented Back: Spectrometer • One in five planes instrumented • Fast QIE electronics • Continuous (19ns) sampling in spill Accelerator Neutrino Oscillation Physics Lecture III

14. MINOS Near Rates Low Medium High Low Signal Time (nsec) Medium Events in 10ms at Near Detector, for 1013 Protons on Target Accelerator Neutrino Oscillation Physics Lecture III High

15. NOvA Near Detector • Same segmentation and structure as far detector, but • sees line source • Needs very tight fiducial cuts • Designed to operate at several different angles Can operate between 4-21 mrad Off axis (Far Detector is 14mrad) Far Courtesy Peter Shanahan, n2006 Accelerator Neutrino Oscillation Physics Lecture III

16. Near Detector Summary Accelerator Neutrino Oscillation Physics Lecture III

17. Remainder of Talk • nm disappearance • K2K Near Detector Analysis and Result • MINOS Near Detector Analysis and Result • ne Appearance • K2K Result • MINOS, NOvA, T2K, OPERA Sensitivity and Background Comparison • What will we need to take advantage of more statistics? • Hadron Production Experiments • Dedicated Cross Section Measurements • Reward for working hard: combining NOvA and T2K • Mass Hierarchy Accelerator Neutrino Oscillation Physics Lecture III

18. m- m- pm(GeV/c) 2 track 2 track non-qe 1 track Constraining Cross Section Model in K2K Near Detector 1-ring events at 1kT QE non- QE You can derive this… hep-ex/0606032 Accelerator Neutrino Oscillation Physics Lecture III

19. Measurement ofnm  nmsurvival in K2K Use both Number of events + Spectrum shape Null oscillation probability is 0.003% (4.19s) Reconstructed En Allowed regions Best fit parameters (in physical region) sin22q = 1.0 Dm2 = (2.76  0.36)x10-3eV2 No oscillation Best fit point preliminary preliminary En(visible) (GeV) Accelerator Neutrino Oscillation Physics Lecture III

20. LE-10 LE-10 Measurement ofnm  nmsurvival in MINOS See 10%-40% data-MC differences in near detector: how to extrapolate to Far? Error envelopes include uncertainties in cross-sections, beam and detector Accelerator Neutrino Oscillation Physics Lecture III

21. Near Detector Tuning at MINOS LE-10/170kA LE-10/185kA • By taking data at several horn currents and target positions, MINOS isolated the problem to Hadron Production Model = pions focused by horns LE-10/200kA pME/200kA Weights applied vs pz & pT Horn off pHE/200kA Accelerator Neutrino Oscillation Physics Lecture III

22. MINOS nm Systematic Errors • Systematic shifts in the fitted parameters are computed using MC “fake data” samples for Dm2=2.7x10-3 eV2 and sin22q=1.0 • The uncertainties considered and shifts obtained: • Magnitude of systematic error is ~40% of statistical error for Dm2 • Several systematic uncertainties are data driven → improve with more data and study Chris Smith, FNAL Seminar Accelerator Neutrino Oscillation Physics Lecture III

23. MINOS nm Survival Results Accelerator Neutrino Oscillation Physics Lecture III

24. N X N X Challenges to ne Appearance Problem: looking for a ne in a beam of nm’s • nm charged current events • nt charged current events • Intrinsic beam ne • K decays • m decays • Neutral Current events m nm lost K→pene p→m→enenm Accelerator Neutrino Oscillation Physics Lecture III

25. Probabilities f=flux, s= cross section e=efficiency M=mass Bfar= Backgrounds at far detector, from any flux Accelerator Neutrino Oscillation Physics Lecture III

26. Probabilities, continued 2 Regimes: Problem: Don’t always know a priori which regime you are in ---depends on Dm2, ---depends on sin22q13 Accelerator Neutrino Oscillation Physics Lecture III

27. Near Detector Strategy Backgrounds come from several sources Build near detector with same e Simulations better at predicting ratios absolute levels Accelerator Neutrino Oscillation Physics Lecture III

28. Near Detector Strategy (cont’d) • But ratios don’t cancel everything • Underlying problem: fluxes are different • Near detector: line source, far detector: point source • But even if that is solved, still nmCC oscillations • All of these terms are functions of energy • Uncertainties in energy dependence of cross sections translate into far detector uncertainties… Accelerator Neutrino Oscillation Physics Lecture III

29. Search fornmneoscillation in K2K As a result, # of expected BG 1.63 events (1.25 fromnm& 0.38 from beamne) # of observed events 1 event Signal candidate event RUN: 21858 EVENT: 2240771 Eg1: 266.7MeV Eg2: 170.8MeV qgg: 22.5 deg. Mgg: 83.1MeV/c2 Though, this event looks like multi-ring… Slide courtesy Y. Hayato Accelerator Neutrino Oscillation Physics Lecture III

30. Chooz limit Search fornmneoscillation in K2K Expected # of electron candidates (NSK) Expected BG: 1.63 Observed: 1 preliminary upper limit onsin22qme (90% CL) 0.18@Dm2=2.8x10-3eV2 ( 0.25@Dm2=2.0x10-3eV2 0.16@Dm2=3.0x10-3eV2) Slide courtesy Y. Hayato Accelerator Neutrino Oscillation Physics Lecture III

31. Search fornmneoscillation in MINOS • How to discriminate between electrons and p±m±? • Longitudinal, transverse event shape… • How to discriminate between electrons and p0? • Less obvious in MINOS… Neural Net MC example • Oscillation parameters: sin2(2q13) = 0.1 |m32|2 = 2.710-3eV2 sin2(2q23) = 1 • POT = 16x1020 (12 what has already collected) Neural Net Particle ID Accelerator Neutrino Oscillation Physics Lecture III

32. Today’s signal is tomorrow’s background… • OPERA: main goal is to see nt CC events through t→e decay channel, so should be sensitive to 5 years, 4.5x1019POT/year Accelerator Neutrino Oscillation Physics Lecture III

33. Sensitivity versus Time • As Kai Zuber mentioned, people writing proposals much prefer to be in the situation where you have <1 event background. • Then sensitivity  time • Note: ne searches are already at 1 or more background events: • K2K: 1.6 background events per 1020 • MINOS: 3.6 background events per 1020 POT • OPERA: 13 background events per 1020 POT • Need to improve • Intrinsic c in the beam: use Off-axis trick • nm peaked in energy. Electron neutrinos over a broad spectrum. • ne CC/ NC event separation: use lower energy, or better detector, or off axis beam (since NC events reconstruct with energy lower than the peak) • Statistics: more detector mass or proton power or both • Next Generation ne searches: • T2K: 23 background events in 5-year run • NOvA: 19 background events in 6-year run Accelerator Neutrino Oscillation Physics Lecture III

34. How well do new designs do? Assume sin2q13=0.1, d=0, *normal hierarchy, but not all same Dm2 References: K2K PRL(96)2006, MINOS: Smith, FNAL SeminarOPERA: JPhysG(29) 2003, T2K: Nakadaira n2006, NOvA:Shanahan,n2006 Accelerator Neutrino Oscillation Physics Lecture III

35. How does MiniBooNE compare? B. Fleming, n2006 Assume Dm2=1eV2, sin22qme=0.004 MiniBooNE Signal assumes mixing angle is factor of 12 lower than the that of the other experiments Note: MiniBooNE numbers extracted from plot at left and assume 90% mis-ID is NC (n2006 talk). Accelerator Neutrino Oscillation Physics Lecture III

36. Now that we have reduced backgrounds and increased mass… • Remember, just because your simulation says it is true, that doesn’t mean the simulation is right • Need to measure neutrino interactions better • What really comes flying out of the nucleus when it is hit by a neutrino? • Need to measure hadron production better • What really comes flying out of the target when it is hit by protons? Accelerator Neutrino Oscillation Physics Lecture III

37. ne Appearance Analysis Summary: Event Samples are different Near to far, so Uncertainties In cross sections Won’t cancel If signal is small, Worry about background Prediction (ne flux and nc xsection), if signal is Big, worry about signal cross sections Accelerator Neutrino Oscillation Physics Lecture III

38. How much do cross section errors cancel near to far? • Toy analysis: start with old NOvA detector simulation, which had same ne/NC ratio, mostly QE & RES signal events accepted, more nmCC/NC accpeted • Near detector backgrounds have ~3 times higher nmcc! • Assume if identical ND, can only measure 1 background number: hard to distinguish between different sources Assume post-MINERnA, s’s known at: DQE = 5%, DRES = 5, 10% (CC, NC) DDIS = 5%, DCOHFe = 20% For large sin22q13, statistical=8% For small sin22q13 , statistical=16% Accelerator Neutrino Oscillation Physics Lecture III

39. Nuclear effects at MINOS m • Visible Energy in Calorimeteris NOT n energy! • p absorption, rescattering • final state rest mass p Nuclear Effects Studied in Charged Lepton Scattering, from Deuterium to Lead, at High energies, but nuclear corrections may be different between e/m and n scattering Toy MC analysis: Accelerator Neutrino Oscillation Physics Lecture III

40. Dedicated Neutrino Interaction Measurements • MINERvA: exclusive final state measurements, 3 nuclear targets, to run in NuMI beamline in time for MINOS and NOvA (and T2K’s) data • T2K 280m Off axis detector: inclusive p0 measurements and some exclusive states, water target • SciBooNE: use SciBAR in MiniBooNE beam to look at anti-n’s NOW! Accelerator Neutrino Oscillation Physics Lecture III

41. Need Dedicated Hadron Production Experiments! Example: NuMI: Absolute rates known only to 20% in high energy tail, Far/near ratio known better, but still only at 5% ratio without MIPP Accelerator Neutrino Oscillation Physics Lecture III

42. Hadron Production Experiment Case Study: MIPP FNAL expt E907, ran with NuMI Target for MINOS Will run with thin targets as well HARP is CERN H.P. experiment that looked at K2K and MiniBooNE targets Figures courtesy M.Messier Accelerator Neutrino Oscillation Physics Lecture III

43. P(nm→ne) on one slide (3 generations) P(nm→ne)=P1+P2+P3+P4 P(nm→ne)% Minakata & Nunokawa JHEP 2001 The ± is n or n Accelerator Neutrino Oscillation Physics Lecture III

44. 4 Equations (NOvA, T2K, n and n ), 3 unknowns: sinQ13, d, hierarchy Simplifications… Note: this is for Dm122<<Dm232, and for L/E such that sin2 (Dm232L/4E)=1 Accelerator Neutrino Oscillation Physics Lecture III

45. How are T2K and NOvA Complimentary? • Remember the example biprobability plot? • For an experiment more sensitive to matter effects (NOvA) ellipses are split • For an experiment less sensitive to matter effects (T2K), they are degenerate Accelerator Neutrino Oscillation Physics Lecture III

46. NOvA Alone, T2K Alone NOvA curves are for The assumption that They run in n and n: That is why they aren’t as Sensitive to d. T2K curves assume only n running. Sensitivity to q13>0 at 3s Line: CHOOZ Limit at 90%CL Accelerator Neutrino Oscillation Physics Lecture III

47. Combining NOvA and T2K: Mass Hierarchy With NOvA alone, and NOvA and T2K With 6xNOvA and 4xT2K Accelerator Neutrino Oscillation Physics Lecture III

48. What should you take away from this course? Accelerator Neutrino Oscillation Physics Lecture III

49. Last Words… Accelerator Neutrino Oscillation Physics Lecture III