New results from K2K experiment Rik Gran University of Washington for the K2K collaboration Seminar for SLAC, 6 January 2005 1. Neutrino oscillations – atmospheric and beam 2. The K2K experiment and oscillation results 3. Future directions Simple two flavor neutrino oscillation Flavor (e,m,t) eigenstates may not be the same as mass eigenstates If the eigenstates of the Hamiltonian evolve, and non-zero Dm, nm will change to nt 2 2 2 Probability(nm to nt) = sin 2q sin (1.27 Dm L/E) units: L in km, E in GeV, m in GeV/c2 Neutrino oscillation allowed regions LSND unconfirmed MiniBoone Fall 2005? (does not fit standard oscillation model) Super-K, K2K Soudan2,MACRO (Kamiokande, IMB) SNO, KamLand (Homestake, Kamiokande, IMB) Plot prepared by M. Sorel Oscillation of atmospheric neutrinos Look at the ~1 GeV muon and electron neutrinos from Picture of SuperK cosmic ray interactions in earth's atmosphere Super-Kamiokande Soudan2 MACRO Kamiokande IMB underground detectors see muon neutrino deficit consistent with nm to nt oscillation, and no ne deficit. Reference: PRL 81 (1998), PRL 93 (2004) Oscillation of atmospheric neutrinos at Super-K 1489.2 days FC+PC Null oscillation Best fit expectation w/ MC systematic errors Best-ft expectation First dip is seen as expected from neutrino oscillation Range of L is 10 km to 10,000 km Range of E is 400 MeV to 10 GeV best fit Dm2 = 2.4 x 10-3, sin22q = 1.0 Reference: PRL 93 (2004) 101801 Features of K2K beam experiment Reduce the uncertainty in L and E values inherent in cosmic ray source by building a muon-neutrino beam. Use Super-K detector and KEK accelerator. L is fixed to 250 km Based on Super-K result, we expected oscillation near 700 MeV. Make a wide-band beam that includes these energies. Build a set of near detectors near the production target to measure the neutrino energy spectrum The K2K Collaboration JAPAN: High Energy Accelerator Research Organization (KEK) Institute for Cosmic Ray Research (ICRR), University of Tokyo / Kobe University / Kyoto University Niigata University / Okayama University / Tokyo University of Science / Tohoku University KOREA: Chonnam National University / Dongshin University / Korea University / Seoul National U. U.S.A.: Boston University / University of California, Irvine / University of Hawaii, Manoa Massachusetts Institute of Technology / State University of New York at Stony Brook University of Washington at Seattle POLAND: Warsaw University / Solton Institute Since 2002 JAPAN: Hiroshima University / Osaka University U.S.A.: Duke University, Lousiana State University CANADA: TRIUMF / University of British Columbia ITALY: Rome FRANCE: Saclay SPAIN: Barcelona / Valencia SWITZERLAND: Geneva RUSSIA: INR-Moscow K2K experiment Overview At Super-K ~106 νµ / 2.2 sec (40 m x 40 m) ~100 interactions 12GeV protons νµ + SK ν π µ+ τ Target+Horn 200m 100m ~250km π decay pipe monitor at near neutrino detectors µ monitor ~1011 νµ / 2.2 sec monitor the beam center (10 m x 10 m) ~100,000 interactions Accumulated protons on target 89 x 1018 POT for analysis Accumulated protons on target x 1018 K2K-1 K2K-2 Protons/Pulse x 1012 Jan 99 Jan 00 Jan 01 Jan 02 Jan 03 Jan 04 The far detector: Super Kamiokande K2K-I: 11,146 PMT's K2K-II: 5,282 PMT's 22.5 kTon fiducial volume (50 kTons total volume) Detect Cerenkov ring from charged particles. Use all observed beam neutrino events AND ALSO use the subset of events for which we can estimate the muon momentum (1-ring muon-like events) and look for a distortion in the spectrum. Super-K events GPS from time of flight SK Tspill TOF=0.83msec TSK Decay electron cut. ±500µsec 20MeV Deposited Energy No Activity in Outer Detector Event Vertex in Fiducial Volume More than 30MeV Deposited Energy ±5µsec Analysis Time Window 107 events background: 1.6 events within 500µs 2.4×10 -3 events in 1.5µs) TDIFF. (ms) Analysis overview KEK Observation Measurement #ν, pµ and θµ Φ(Eν), ν int. ν interaction MC Far/Near Ratio (beam MC with π mon.) SK Observation Expectation rec. #ν and Eν #ν and Eν rec (sin22θ, ∆m2) Near detector measurements Super-K expected flux from KT measurement SK (En )(En )dE n M N exp N obs SK SK SK KT (E ) (E )dE M KT n n n KT KT Far/Near Ratio, ~1×10-6 (from MC) M: Fiducial mass MSK=22,500ton, MKT=25ton ε : efficiency SK-I(II)=77.0(78.2)%, KT=74.5% exp +12 obs NSK =151 -10 NSK =107 NEUT: K2K neutrino interaction Monte Carlo Charged current quasi-elastic Smith and Moniz with MA = 1.1 /E (10-38cm2/GeV) CC (resonance) single pion Rein and Sehgal with MA = 1.1 Total (NC+CC) CC Total Deep inelastic scattering GRV94 + JETSET with CC quasi-elastic Bodek and Yang correction DIS CC coherent pion CC single π Rein and Sehgal with cross-section π0 rescaled by J. Marteau NC single Neutral current events En (GeV) Nuclear effects for oxygen and carbon Two track QE and non-QE enhanced samples Both SciFi and SciBar data (Scibar shown here) CCQE Candidate CCQE 2 track events non-QE n p DATA CC QE m CC 1 CC coherent- ∆θ p CC multi- p m 25° ∆θ p (degree) Hint of forward muon deficit from K2K data Seen in data from all near detectors SciBar non-QE event sample (small angle is also small q2) preliminary Most clear in non-QE samples DATA CC 1 CC coherent- We do not identify the cause but make two phenomenological models for this suppression: reconstructed q2 (GeV/c)2 CC-1 suppression of q2/0.1 for events with q2 < 0.1 (GeV/c)2 -- OR -- Zero coherent The oscillation analysis is insensitive to this choice. reconstructed q2 (GeV/c)2 Near detector spectrum measurements 1 KT Cerenkov detector (water target) Fully contained one-ring muon-like events SciFi fiber tracker (water target) one-track, two-track QE, two-track non-QE (two-track non-QE cut is Dqp > 30 degrees, QE is Dqp < 25 degrees) SciBar active scintillator (scintillator target) one-track, two-track QE, two-track non-QE (two-track non-QE and QE cuts are at Dqp = 25 degrees) For each event we measure pm and qm. After the low q2 suppression, all data samples agree with the MC. Example: Pm and qm from 1KT detector before and after applying 1 suppression after before 0 800 1600 after pm (MeV/c) 1KT 1-ring muon-like events Black is quasi-elastic, Red is single-pion. qm (degrees) How we turn pm and qm into energy spectrum Free parameters and other aspects of the fit Flux for eight energy regions (E ). n (relative to our standard MC) Ratio nQE/QE. Detector uncertainties such as pm scale, track counting efficiency. Nuclear effect uncertainties such as proton and pion rescattering. Strategy 1. Fit (En) without low angle data to avoid discrepancy. 2. Apply either 1- suppression or coherent- suppression at low q2. 3. With (En) fixed in step one, fit for nQE/QE ratio. 4. Use these results to calculate the spectrum at Super-K. Example: performing the pm and qm fit MC templates 1KT data Pm vs qm distribution QE integrated ees) En < 0.5 GeV nQE degr ( µ θ 0.5 to 0.75 0.75 to 1.0 Pm (MeV/c) Free parameters: 1.0 to 1.5 relative flux (En) (8 energy regions) ratio nQE/QE Flux measurement results (En) at KEK 1 ( En < 0.50) GeV = 0.78 ± 0.46 preliminary 2 (0.50 ≤ En< 0.75 ) GeV = 1.01 ± 0.09 3 (0.75 ≤ En < 1.00) GeV = 1.12 ± 0.06 4 (1.00 ≤ En < 1.50) GeV = 1.00 fixed 5 (1.50 ≤ En < 2.00) GeV = 0.90 ± 0.05 6 (2.00 ≤ En < 2.50) GeV = 1.07 ± 0.06 7 (2.50 ≤ En < 3.00) GeV = 1.33 ± 0.13 8 (3.00 ≤ En < ) GeV = 1.04 ± 0.17 nQE/QE = 1.02 ± 0.10 En c2 is 638.1 for 609 degrees of freedom. The nQE/QE error is determined from the variation with different fit criteria. nQE/QE = 0.95 ± 0.04 with theta cut, nQE/QE = 1.02 ± 0.03 with single pion suppression, nQE/QE = 1.06 ± 0.03 with no coherent pion. Example: Scifi data with measured spectrum θ Pµ 1trk µ 1trk flux measurement Pµ 2trk θµ 2trk QE QE Pµ 2trk θµ 2trk non-QE non-QE 0 2 (GeV/c) 0 10 40 (degree) Example: SciBar data with measured spectrum Pµ 1trk θµ 1trk fflluxux measurementmeasurement Pµ 2trk QE θµ 2trk QE θ Pµ 2trk nQE µ 2trk nQE 10 Super-K oscillation analysis Total number of beam neutrino events Reconstructed En shape for 1 ring, muon like events Systematic error terms, fx L(Dm2 ,sin 2q , f x ) 2 x 2 x x Lnorm (Dm ,sin 2q , f ) Lshape (Dm ,sin 2q , f ) Lsyst ( f ) Systematic error terms include normalization, flux, nQE/QE ratio, pion monitor and beam MC constraints, and Super-K systematic uncertainties. K2K beam events in Super-K detector preliminary K2K-all DATA MC (K2K-I, K2K-II) (K2K-I, K2K-II) (K2K-I, K2K-II) Fully contained 107 150.9 22.5kt fiducial V. (55, 52) (79.1, 71.8) 1-ring 67 94.0 (33, 34) (48.9, 45.1) m-like 57 85.4 rec for EEn (30, 27) (44.9, 40.5) e-like 10 8.6 (3, 7) (4.0, 4.5) Multi Ring 40 56.9 (22, 18) (30.2, 26.7) Ref; K2K-I(47.9×1018POT), K2K-II(41.2×1018POT) Compare each with null-oscillation hypothesis x x Lnorm ( f ) Lshape ( f ) #SK Events KS probability=0.11% Toy MC Expected shape (No Oscillation) 107 151 0 1 2 3 Enrec[GeV] CC-QE assumption 2 2 rec (mN V )Em mm 2 mNV V 2 En (mN V ) Em pm cos qm Best fit oscillation parameters Best fit values: sin2(2q) = 1.51, Dm2 = 2.19 x 10-3 eV2 Best fit values in the physical region sin2(2q) = 1.00, Dm2 = 2.79 x 10-3 eV2 The difference in these two fits is 2 m A toy MC D log L = 0.6 ∆ 14.4% A value sin2(2q) > 1.53 can occur due to a statistical fluctuation with a probability of 14% 1.00 1.51 2 sin 2q Data are consistent with best fit result 2 c) ● obs / NSK =107 eV exp ( ● NSK (best fit)=103.8 2 m D Best Fit -2 10 KS prob.=36% 10-3 10-4 sin22q Based on DlnL Enrec[GeV] 90% CL Dm2 is between 1.87 and 3.58 x 10-3 (eV/c)2 nm disappearance vs En shape distortion N (#νµ) E shape SK 2 ν 2 c) c) / / eV eV ( ( 2 2 m m D D 10-2 10-2 10-3 10-3 10-4 10-4 sin22q sin22q Both disappearance of nm and the distortion of En are consistent.