Recent results from Super-Kamiokande

Recent results from Super-Kamiokande ~ atmospheric neutrino ~ Yoshinari Hayato ( Kamioka, ICRR, U-Tokyo ) for the Super-Kamiokande collaboration 1

41.4m Super-Kamiokande detector 50000 tons Ring imaging Water Cherenkov detector Fiducial volume : 22.5 ktons 1000m under the ground Inner detector Outer detector 11129 20 PMTs 1885 8 PMTs 39m About 40% of the inner detector is covered by the sensitive area of PMT. Operation started in Apr. 1996. 2

Super-Kamiokande detector History of the SK detector SK-I April 1996 ~ June 2001 SK-II October 2002 ~ October 2005 SK-III June 2006 ~ September 2008 SK-IV September 2008 ~ running 11146 ID PMTs (40% coverage) SK-I SK-II SK-III SK-IV Acrylic (front) + FRP (back) 5182 ID PMTs (19% coverage) 11129 ID PMTs (40% coverage) Electronics Upgrade 3

41.4m Physics in Super-Kamiokande Neutrino oscillation Accelerator neutrinos Atmospheric neutrinos Solar neutrinos GUT Proton decay p e + + p 0 p K + + n New physics WIMP search n-n oscillation dinucleon decay etc. Neutrino astrophysics Super nova burst neutrino Super nova relic neutrino 39m 4

Neutrino mixing parameter measurements Normal hierarchy Inverted hierarchy Δm 2 21 = 7. 49 +0.19 0.17 10 3 2 2 ev/c 2 Δm 32 ~2. 5 ± 0. 2 10 3 2 2 ev/c 2 Δm 32 ~2. 5 ± 0. 2 10 3 2 2 ev/c Δm 2 21 = 7. 49 +0.19 0.17 10 3 2 2 ev/c Remaining questions sin 2 q 12 ~ 0.305±0.013 sin 2 2q 13 = 0.098 ± 0.013 1) q 23 is really 45 o or < 45 o or >45 o? Current uncertainty of sin 2 q 23 is still large ~ 10% level sin 2 q 23 = 0. 514 ± 0. 055 ( T2K 2014 ) 2) CP is violated or not ( d = 0 or not )? 3) Mass hierarchy ~ which is heavier? ( Dm 2 32 > 0 or < 0?) 5

Neutrino oscillation probability ~ n m to n e oscillation q 13 Leading term CPV a a For anti neutrinos, a -a, d -d a Now, q 13 is known to be ( quite ) large. There are chances to observe the contributions from mass hierarchy and CPV ( d )! 6

Recent results from Reactor & T2K n e appearance clearly observed in T2K Reconstructed n energy Data Best fit No oscillation Allowed region of δ CP for each sin 2 2θ 13 Extracted sin 2 2q 31 is slightly larger than the ones from reactor experiments sin 2 2q 13 = +0.038 0.140 0.032 +0.044 0.170 0.037 ( normal hierarchy ) ( inverted hierarchy ) Indication of non-zero d CP? 7

Characteristics of atmospheric neutrino Primary cosmic ray ( p, He.. ) atmosphere Generation height 10~30km p ±, K ± μ ± n m e ± Atmospheric n energy spectrum n m /n e ~ 2 ( < ~1 GeV ) n m /n e > 2 ( > ~1 GeV ) n m n e Atmospheric neutrino energy spectrum Peaked at ~ several hundreds of MeV, Extended > TeV Neutrino travel length from ~ 10 km to 13,000 km Zenith angle corresponds to travel length of neutrinos. 8

Atmospheric neutrino ~ event topology in SK Fully Contained (FC) Average energies FC: ~ 1 GeV PC: ~ 10 GeV UpMu: ~ 100 GeV Partially Contained (PC) Upward-going Muons (Up-m)

Atmospheric neutrino ~ Analysis samples in SK Fully Contained (FC) Zenith angle distribution of each sample Partially Contained (PC) Upward-going Muons (Up-m) In total 19 analysis samples (classified by ν flavors, event topologies, energies, ) Fit to the data in bins of cosq zenith and momentum Dominated by ν µ ν τ oscillations Interested in sub-dominant contributions Three-flavor effects, Sterile Neutrinos, LIV,

Neutrino oscillation studies using atmospheric n Evidence for n t Appearance at Super-K SK-I+II+III : 2806 days Phys. Rev. Lett. 110, 181802 (2013) Fitted Excess Atm n BKG MC Search for events ~ hadronic decay of t lepton Multi-ring e-like events ~ mostly DIS interactions Actual event selection performed by neural network Total efficiency ~ 60% Negligible primary n t flux ~ n t must be oscillation-induced Expected only in upward-going b = 0 : no n t Result Background DIS (g) Signal SK-I+II+III 0.94 0.02 1.10 0.05 1.42 0.35 180.1 ±44.3 (stat) +17.8-15.2 (sys) events ~ 3.8 s excess (Expected 2.7 s significance )

Neutrino oscillation studies using atmospheric n High statistics atmospheric neutrino data ~ Possibility in observing small distortion in n e Matter effect ~ from mass hierarchy Possible n e enhancement in several GeV passed through the earth core Solar term ~ from q 23 octant degeneracy Possible n e enhancement in sub-gev Interference CP phase could be studied. n e flux ratio ( f (osc.) / f (no-osc.) ) sin 2 q 23 = 0.5 sin 2 2q 13 = 0.1 Interference Difference in # of electron events: Solar term Matter effect Solar term Interference Matter 12

Neutrino oscillation studies using atmospheric n Expected Sensitivity Hierarchy Sensitivity ( True : NH ) Owing to the broad energy spectrum and baseline, atmospheric neutrino has sensitivity to the mass hierarchy. However, sensitivity to the mass hierarchy has rather strong relation with the other oscillation parameters. As a function of the true value of sin 2 q 23, this plot shows the ability to reject the inverted mass hierarchy hypothesis assuming the normal hierarchy

Neutrino oscillation studies using atmospheric n Changes and Updates to Oscillation Analyses 1775 days of SK-IV data: 4581.4 days total (282.2 kton yrs) Addition of a new analysis sample Fully contained Multi-Ring e-like other = not-n e & not-n e Multi-Ring e-like Sample Purities Purity CC n e CCn m CCn t NC n-like 72.2% 8.3% 3.2% 16.1% n -like 75.0% 6.5% 2.8% 15.6% other 30.9% 33.4% 5.1% 30.5% Prediction n m n t Osc. n/ n Separated Samples New Sample Improved systematic error treatments Updates to cross-section, FSI, detector systematics, 2p-2h (MEC)

Neutrino oscillation studies using atmospheric n q 13 Fixed Analysis (NH+IH) SK Only q 13 fixed to PDG average uncertainty is included as a systematic error Preliminary SK Normal Hierarchy SK Inverted Hierarchy Fit (517 dof) c2 q 13 d cp q 23 Dm 23 (x10-3 ) SK (NH) 559.8 0.025 3.84 0.57 2.6 SK (IH) 560.7 0.025 3.84 0.57 2.5 Offset in these curves shows the difference in the hierarchies

Neutrino oscillation studies using atmospheric n q 13 Fixed Analysis (NH+IH) SK Only Normal hierarchy favored but c 2 IH c 2 NH = - 0.9 Not a significant preference Previous results (2013 Summer) favored inverted hierarchy by Dc 2 ~ 1.5 Preliminary SK Normal SK Inverted Preliminary PDG 2013 Driven by excess of upward-going e-like consistent with the effects of q 13 Primarily in SK-IV data New multi-ring e-like sample also pulls the fit towards the NH Fit for q 13 now weakly favors q 13 > 0 Rejection of d cp ~ 60º driven by excess in Sub-GeV electron events Constraint is consistent with sensitivity Inverted Normal Preliminary

Neutrino oscillation studies using atmospheric n Introduction of External Constraint Constraints from the other experiments, ( especially, Dm 2 and sin 2 q 23 ), make it possible to improve sensitivity to mass hierarchy. Fit the T2K n m and n e data sets with SK atmospheric neutrino data. Use publicly available T2K information and results. Simulate T2K using SK tools ( not a joint result of the T2K and the SK collaborations ) Hierarchy Sensitivity NH True SK + T2Kn m, n e SK Alone

Neutrino oscillation studies using atmospheric n q 13 Fixed SK + T2K n m, n e (External Constraint) Normal Hierarchy Preliminary SK Atmospheric n T2K n m,n e Constraint SK+T2K n m,n e Constraint Fit (543 dof) c2 q 13 d cp q 23 Dm 23 (x10-3 ) SK + T2K (NH) 578.2 0.025 4.19 0.55 2.5 SK + T2K (IH) 579.4 0.025 4.19 0.55 2.5 c 2 IH c 2 NH = -1.2 (-0.9 SK only ) sind cp = 0 is still allowed at (at least) 90% C.L. for both hierarchies.

Neutron tagging ~ For further improvements ~ Effective neutron tagging make it possible to 1) discriminate n and n interactions more efficiently 2) reduce background of proton decay Compare: 204.8 ms Phys Rev. C 15 (1977) 1636 Electronics in SK-IV store all PMT hits for 500 msec after a neutrino-like trigger. Search for the 2.2 MeV gamma from p(n,g)d Actual search is performed with a neural network ( 16 variables ) Data and MC show good agreement on atmospheric neutrino sample t0 = n interaction time Preliminary Geant3 2.2 MeV g Selection Efficiency 20.5% Background / Event 0.018 19

Sterile neutrino oscillations in atmospheric neutrino Sterile Neutrino searches at SK Independent from the sterile Dm 2 and the # of sterile neutrinos. MNS Sterile a) 3+1 and 3+N models have the same signatures b) For Dm s2 ~ 1 ev 2 oscillations appear fast < sin2 Dm 2 L/E > ~ 0.5 U m4 2 Induces a decrease in event rate of m-like data of all energies and zenith angles U t4 2 Shape distortion of angular distribution of higher energy m-like data U m4 2 at 5s sensitivity U t4 2 at 5s sensitivity Limits on sterile neutrino mixing using atmospheric neutrinos in Super-Kamiokande K. Abe et al., PHYSICAL REVIEW D 91, 052019 (2015)

Sterile neutrino oscillations in atmospheric neutrino Assuming sterile neutrinos experience vacuum oscillation, ( = turning off sterile matter effects ), while preserving standard three-flavor oscillations provides a pure measurement of U m4 2 (3) P ee = P ee, (3), P eμ = 1 U μ4 2 Peμ 2 (3) P μe = 1 U μ4 Pμe, m survival probability P μμ = 2 1 U 2 (3) μ4 P μμ + U μ4 4 Suppression due to existence of sterile n Black Without Sterile Red With Sterile U μ4 2 = 0.0058 *) Infinite width for same L/E ~ n e CC matter effects

Sterile neutrino oscillations in atmospheric neutrino Sterile vacuum oscillation Limits on sterile neutrino mixing using atmospheric neutrinos in Super-Kamiokande K. Abe et al., Phys. Rev. D 91, 052019 (2015) U m4 2 < 0.041 at 90% C.L.(*) SK-I~IV data ( 4438 days) MiniBooNE + SciBooNE PRD86, 052009 (2012) SK CCFR PRL. 52, 1384 (1984) (*) Limit is valid for Dm 41 > 0.1 ev 2

Summary n t appearance 180.1 ±44.3 (stat) +17.8-15.2 (sys) events( 3.8s ) Three-Flavor neutrino oscillation analysis Using data from SKI to IV ( 4538 days ), ~1 s preference for the NH, and second octant Neutron tagging Data taken with SK-IV electronics, we developed new neutron tagging analysis tools. Efficiency ~ 20.5 % Will be used for atmospheric neutrino oscillation analyses and proton decay search. No indication of oscillations into sterile neutrinos For 3+N models U m4 2 < 0.041 at 90% C.L. 23

fin. 24

Neutrino oscillation studies using atmospheric n Searching for Three-Flavor Effects: Oscillation probabilities ~100 km P(n m n m ) P(n m n e ) ~10,000 km Sub-GeV Multi-GeV Resonant oscillations between 2-10 GeV ( Multi-GeV region ) for n or n depending upon MH, No oscillations from downward-going neutrinos above ~5 GeV No oscillations above 200 GeV, No n m n e appearance above ~20 GeV, Expect effects in most analysis samples, largest in upward-going n e

Neutrino oscillation studies using atmospheric n Oscillation Effects on Analysis Subsamples P(n m n m ) ~100 km Ratio to two-flavor oscillations Sub-GeV m-like 0 dcy-e sin 2 q 23 = 0.6 sin 2 q 23 = 0.5 sin 2 q 23 = 0.4 Multi-ring m-like 0 dcy-e ~10,000 km sin 2 q 23 = 0.6 sin 2 q 23 = 0.5 sin 2 q 23 = 0.4

Neutrino oscillation studies using atmospheric n Oscillation Effects on Analysis Subsamples P(n m n e ) ~100 km Ratio to two-flavor oscillations Sub-GeV e-like 0-dcy e d cp = 3p/2 d cp = p/2 Multi-ring e-like n e Sub-GeV ~10,000 km Multi-GeV Appearance effects are roughly halved for the inverted hierarchy. sin 2 q 23 = 0.6 sin 2 q 23 = 0.5 sin 2 q 23 = 0.4

Atmospheric neutrino ~ Analysis samples in SK FC Sub-GeV 1-ring e-like 0 decay electron 1 decay electron p 0 like 0 decay electron m-like 1 decay electron 2 decay electron 2-rings p 0 like Multi-GeV 1-ring e-like 0 decay electron m-like with decay electron multi-ring e-like n e -like n e -like others m-like 28

Atmospheric neutrino ~ Analysis samples in SK PC OD stopping OD through going Up-going m Stopping Through going showering Through going non-showering 29

Sterile neutrino oscillations in atmospheric neutrino Assuming q 13 = q 12 = 0, P μμ = 1 U μ4 2 2 1 sin 2 2θ m sin 2 (E m L) + U μ4 4, E 2 m = Δ 32 2 4E 2 + A s 2 + 2A 32 A s cos 2θ 23 2θ s, sinθ m = A 32 sin 2θ 23 +A s sin 2θ s E m, A s = U μ4 2 + Uτ4 2 2 m survival probability U μ4 2 = 0.0018, Uτ4 2 = 0.33 Suppression