Atmospheric Neutrino Results from Super-Kamiokande

Zepeng Li, Duke University For The Super-Kamiokande Collaboration

ICHEP 2016 1 Super-Kamiokande

• Super-K is a 50 kton water Cherenkov detector with 22.5 kton of fiducial volume at 2,700 m.w.e underground. • The detector is optically separated into ID and OD. • Excellent in detection of atmospheric neutrinos. • 20 years since the start of data taking in 1996, >47,000 atmospheric neutrino events. Four Run Periods: • A Nobel prize winning experiment! SK-I (1996-2001) SK-II (2003-2005) SK-III(2005-2008) SK-IV(2008-Present) ICHEP 2016 2 Atmospheric neutrinos in SK cosmic ray

ν ~10 km Event categories cosmic rays

π±, K± Fully contained

ν Partially contained ± νu μ ν ~13,000 km

e±

νu cosmic ray Upward muon νe

ICHEP 2016 3 Atmospheric neutrinos in SK cosmic ray

ν ~10 km cosmic rays

π±, K±

± νu μ ν ~13,000 km

e±

νu cosmic ray ν e Sub-GeV to more than 10 TeV neutrino energy.

ICHEP 2016 4 Atmospheric neutrinos measurement in SK

-1 ] 10 1.6 -1 ν 1.4 e Ratio

sr 1.2 -1 10-2 1 0.8 sec 0.6 HKKM11 Bartol -2 -3 HKKM07 FLUKA 10 1.6 νµ

Ratio 1.4 -4 1.2 10 1 0.8 [GeV cm [GeV -5 0.6 Φ 10 Super-Kamiokande I-IV ν µ 2 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 Frejus ν µ E IceCube ν µ unfolding Log (E / GeV) 10 ν -6 IceCube ν µ forward folding 10 AMANDA-II ν µ unfolding AMANDA-II ν µ forward folding ANTARES ν µ HKKM11 ν +ν (w/ osc.) 10-7 µ µ SK measures the flux of atmospheric Super-Kamiokande I-IV ν e Frejus ν e neutrino with energy of sub-GeV to IceCube/DeepCore 2013 ν -8 e IceCube 2014 ν 10 e ~10 TeV. The measurement is consistent HKKM11 ν e+ν e (w/ osc.) with model prediction. 10-9 -1 0 1 2 3 4 5 Log (E /GeV) 10 ν With the large sample, SK can test the effects on atmospheric neutrino flux of geo-magnetic field, solar activity, etc. arXiv:1510.08127

ICHEP 2016 5 Atmospheric neutrinos measurement in SK

700 e Cosmic rays are mostly 650 600 positive charged. 550 500 μ-like 450 400S W N E 350 0 60 120 180 240 300 360

850 µ 800

750 East West 700 650

600 550 e-like 500 0 60 120 180 240 300 360

Reconstructed Azimuth φ rec [deg] Azimuthal asymmetries of SK neutrino events

3 α SK observes clear azimuthal dependence 2 1 for both μ-like(6.0σ) and e-like(8.0σ) 0 samples due to the effect of geo-magnetic -1 field, and no significant solar modulation of -2 -3 SK-I SK-II SK-III SK-IV e up e down µ up µ down atmospheric neutrino flux. Sub-sample Fitted α as degree of solar modulation, arXiv:1510.08127 α=0: no correlation, 1:expected

ICHEP 2016 6 Neutrino oscillations at Super-K ~10 km P(�� → ��) P(�� → ��)

~13,000 km

Leading effect is νμ disappearance(� → �). ν� appearance from neutrino oscillations could be detected by charged current ν� interaction in SK. ICHEP 2016 7 Tau neutrino appearance in SK

ντ �± Tau lepton decays in ~10-13s, ντ e,μ or tau lepton track is undetectable hadrons. in SK detector.

• Tau lepton production is rare in SK due to 3.5GeV energy threshold of charged-current ν� interaction. • Multiple light-producing particles from tau decay. • Neural network(input variables in backup) to select hadronic tau decay events.

ICHEP 2016 8 Tau neutrino appearance in SK

Downward sample � like BG Tau

BG like

�� = �� + �×�� + ∑ ��×(�� + ��) � is the normalization of tau events. (��, ��)�� is the PDF of ith systematic error of shifting it by 1σ, � is the magnitude of the systematic error. a=1.47±0.32 preliminary 0.33 Mtyr compared to simulation (4.6s from 0)assuming NH tau-like events after the fit Sensitivity at α=1: 3.3s ICHEP 2016 9 SK1-4 0.33 Mtyr data and MC

3 flavor oscillation at best fit and reactor θ13 for MC.

ICHEP 2016 10 Neutrino oscillations at Super-K ~10 km P(�� → ��) P(�� → ��)

~13,000 km

Leading effect is νμ disappearance. Sub-leading effects help to resolve θ23 octant, δCP and mass hierarchy. ICHEP 2016 11 Neutrino oscillations at Super-K

P(�� → ��) P(�� → ��) δcp = 3π/2

sin2θ = 0.6 2 23 sin θ = 0.5 δcp = π/2 2 23 sin θ23= 0.4

More νe at π<δCP<2π.

Oscillation probability has 2 dependence on sin θ23

Matter effect and large Ratio to two-flavor θ , ν enhanced at NH. oscillations shown. 13 e sin2θ = 0.6 sin2θ = 0.6 sin2θ23 = 0.5 Multiple samples 2θ23 2 23 sin2 23 = 0.5 sin θ23 = 0.4 used in the analysis. sin θ23 = 0.4 ICHEP 2016 12 SK oscillation analysis-θ13 constrained preliminary Inverted Normal

2 2 2 2 2 Fit (517 dof) c sin q13 dCP sin q23 |Dm 32|eV SK (IH) 576.0 0.0219 4.2 0.58 2.5x10-3 SK (NH) 571.7 0.0219 4.2 0.59 2.5x10-3

• θ13 is constrained at PDG average, uncertainty is included as a systematic error. 2 2 2 • Dc = c NH-c IH= -4.3 (-3.1 of sensitivity) 2 2 • The p-value of obtaining Dc of -4.3 or less is 0.031 (sin q23=0.6) and 0.007 2 2 (sin q23=0.4) in IH hypothesis. Under NH hypothesis, the p-value is 0.45 (sin q23=0.6). ICHEP 2016 13 Oscillation analysis with constraint from published T2K data preliminary Inverted Normal

2 2 2 2 2 Fit (517 dof) c sin q13 dCP sin q23 |Dm 32|eV SK (IH) 644.8 0.0219 4.5 0.55 2.5x10-3 SK (NH) 639.6 0.0219 4.9 0.55 2.4x10-3 • Dc2 = -5.2 (-3.8 of sensitivity for SK best, -3.1 for combined best) 2 2 2 • The p-value of obtaining Dc of -5.2 is 0.024 (sin q23=0.6) and 0.001 (sin q23=0.4). 2 Under NH hypothesis, the p-value is 0.43 (sin q23=0.6).

ICHEP 2016 14 Summary

• Measurement of atmospheric neutrino flux of energies from sub-GeV to 10 Te V. • Tau neutrino appearance with significance of 4.6σ. • Normal hierarchy preferred by Dc2 = -5.2 , 2 p-value is between 0.024 (sin q23=0.6) and 0.001 2 (sin q23=0.4) in IH hypothesis.

• Weak preference of second octant and δCP near 3/2π. • More analyses in SK, indirect WIMP search in poster session.

ICHEP 2016 15 Super-Kamiokande Collaboration

• a

1 Kamioka Observatory, ICRR, Univ. of Tokyo, Japan 15 Imperial College London, UK 29 Shizuoka University of Welfare, Japan 2 RCCN, ICRResearch, Univ. of Tokyo, Japan 16 KEK, Japan 30 STE, Nagoya University, Japan 3 University Autonoma Madrid, Spain 17 Kobe University, Japan 31 Sungkyunkwan University, Korea 4 University of British Columbia, Canada 18 Kyoto University, Japan 32 SUNY, Stony Brook, USA 5 Boston University, USA 19 University of Liverpool, UK 33 Tokai University, Japan 6 Brookhaven National Laboratory, USA 20 Miyagi University of Education, Japan 34 University of Tokyo, Japan 7 University of California, Irvine, USA 21 National Centre For Nuclear Research, Poland35 Kavli IPMU (WPI), University of Tokyo, Japan 8 California State University, USA 22 Okayama University, Japan 36 Dep. of Phys., University of Toronto, Canada 9 Chonnam National University, Korea 23 Osaka University, Japan 37 TRIUMF, Canada 10 Duke University, USA 24 University of Oxford, UK 38 Tsinghua University, China 11 Fukuoka Institute of Technology, Japan 25 Queen Mary University of London, UK 39 University of Washington, USA 12 Gifu University, Japan 26 University of Regina, Canada 13 GIST College, Korea 27 Seoul National University, Korea ~150 collaborators 14 University of Hawaii, USA 28 University of Sheffield, UK 39 institutions

2016/7/4 8 countries 16 Backups

ICHEP 2016 17 Seasonal variation

] 2.8 -1 Multi-GeV 2.6

2.4

2.2

Neutrino events [counts day [counts events Neutrino Sub-GeV 4.6

4.4

4.2

4 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month No seasonable variation is seen as expected.

ICHEP 2016 18 Tau analysis

Seven input variables Fit result for different samples, to neural network. signal/BG MC scaled by fit result.

ICHEP 2016 19 Fitted systematic errors in tau analysis

Fitted deviation in unit of σ of systematic errors

Mean: 0.2 RMS: 0.8

Fitted value of systematic terms, distributed around 0. �� − ��

ICHEP 2016 �� − �� 20 Paper fit update of tau analysis

. Tau fraction:1.37±0.23., the significance is 5.1σ.

ICHEP 2016 21 Improved MH sensitivity with T2K constraint

NH IH

SK+T2K SK+T2K

SK only SK only

ICHEP 2016 22 p-value for MH preference

SK analysis with T2K constraint.

SK only analysis ICHEP 2016 23 MH preference in sub-samples

4 χ2 - χ2 IH NH

2 2 χ 0 ∆

-2

-4 Sub-GeV Sub-GeVe-like 0-dcy Sub-GeVe-like e 1-dcy Sub-GeV e Sub-GeV Sub-GeV Sub-GeV Multi-GeVMulti-GeV e-like Multi-GeV e-like Multi-Ring Multi-Ring e-like Multi-Ring e-like Multi-Ring PC Other Stop PC Thru Up Stop Non-showeringShowering µ µ-like 0-dcyµ-like e 1-dcyµ-like e 2-dcyπ 0 -likee 1-Rπ 0-like M-R µ-like µ-like µ ν ν ν ν µ e e e e

ICHEP 2016 24 Multi-RingMulti- e-likeRing e νe -like ne Multi-GeVMulti -e-likeGeV νe e-like ne 0.40.4 0.40.4 Normal2 2 2 Normal DcInverted=c NH-c IH: +0.35 Inverted +0.42

0.20.2 0.20.2 Hierarchy preference in 0 0 -0.2-0.2 -0.2-0.2 Normal sub-samples ( Up - Down ) / + Normal ( Up - Down ) / + Inverted Inverted -0.4-0.4 -0.4-0.4 3 3.2 2×1033.43×103 3.6 3.8 1044 3 2×103 3.5 1044 2×104 1 2.5 Lepton5 P [ MeV10 ] 1 2.5 5 Lepton10 20 P [ MeV ] Momentum (GeV) Momemtum (GeV)

ν Multi-RingMulti- OtherRing others Multi-GeVMulti - e-likeGeV ee-like ne Multi-RingMulti- e-likeRing e νe -like ne 0.40.4 0.40.4 0.40.4 Normal Normal Normal Inverted -2.97 Inverted -2.18 Inverted -0.92 0.20.2 0.20.2 0.20.2

0 0 0

-0.2-0.2 -0.2-0.2 -0.2-0.2 ( Up - Down ) / + ( Up - Down ) / + Normal Normal ( Up - Down ) / + Normal Inverted Inverted Inverted -0.4-0.4 -0.4-0.4 -0.4-0.4 11 3 3 3.5 44 4 3 3 3.5 44 4 3 3.2 33.4 3 3.6 3.8 44 1 2×102.5 5 10 2010 2×10 1 2×102.5 5 10 2010 2×10 1 2×10 2.53×10 5 10 Lepton P [ MeV ] Lepton P [ MeV ] Lepton P [ MeV 12] Momemtum (GeV) Momemtum (GeV) Momemtum (GeV) Indirect WIMP search preliminary

Data ATM MC

DM signal

Cosθzenith Search for an excess of neutrinos from galaxy, sun or earth(shown in the plot). Preliminary sensitivity result of 90% CL Indirect Dark Matter Searched with Super-Kamiokande limits for background only scenario for Poster by K. Frankiewicz WIMP decay in earth core.

ICHEP 2016 26