SuperKamiokande Y Totsuka Kamioka n n n Introduction

Super-Kamiokande Y. Totsuka Kamioka n n n Introduction Contained events and upward muons Updated results l l l n Oscillation analysis with a 3 D flux Multi-ring events p 0/m ratio n 3 decay Search for t leptons nm ns Conclusion

Super-Kamiokande collaboration

Super-Kamiokande detector 50, 000 ton water Cherenkov detector (22. 5 kton fiducial volume) 1000 m underground (2700 m. w. e. ) 11, 146 20 -inch PMTs for inner detector 1, 885 8 -inch PMTs for outer detector

Atmospheric neutrinos p, He. . . L=10 -30 km p , K nm e nm ne L=up to 13000 km n m+ n m n e+ n e = ~ 2 @ low energy (En < 1 Ge. V) n m+ n m @ high energy n e+ n e Error in absolute flux~20%, but nm/ne ratio~5% Neutrino oscillations : nm+ nm ne+ ne M C data 1

Atmospheric neutrino spectrum (P. Lipari) (3 -D) Energy dependence of nm/ne ratio p m nm e n mn e <5% accuracy p m nm e n mn e

Primary cosmic ray flux protons He From P. Lipari

Bartol and Honda fluxes

Zenith angle distribution(1 D) Calculated zenith angle distribution En=0. 5 Ge. V En=3 Ge. V En=20 Ge. V For En > a few Ge. V, Upward / downward = 1 (within a few %) Up/Down asymmetry for neutrino oscillations

3 D neutrino flux calculation 1 D p 3 D p p p n n n 3 D calculation by G. Battistoni et al. nm (hep-ph/9907408) 10 -3– 0. 2 Ge. V 0. 5 -1 Ge. V horizontal vertical 0. 2 -0. 5 Ge. V 1 -5 Ge. V n

How to detect atmospheric neutrinos Contained events Upward throughgoing muons Upward stopping muons Interaction in the rock Initial neutrino energy spectrum contained stopping muons through-going muons

Contained event analysis Fully Contained (FC) Partially Contained (PC) m e or m No hit in Outer Detector One cluster in Outer Detector Reduction Automatic ring fitter Particle ID Energy reconstruction Fiducial volume (>2 m from wall, 22 ktons) Evis > 30 Me. V (FC), > 3000 p. e. (~350 Me. V) (PC) Final sample: FC: 8. 2 ev. /day, PC: 0. 58 ev. /day Evis < 1. 33 Ge. V : Sub-Ge. V Evis > 1. 33 Ge. V : Multi-Ge. V

Fully contained event summary (1289. 4 d (79. 3 kt. y)) Sub-Ge. V (Fully Contained) Evis < 1. 33 Ge. V, Pe > 100 Me. V, Pm > 200 Me. V Data MC(Honda flux) 1 ring e-like 2864 2667. 6 m-like 2788 4072. 8 Multi ring 2159 2585. 1 Total 7811 9325. 5 m /e Data = 0. 638 0. 017 0. 050 MC Multi-Ge. V Fully Contained (Evis > 1. 33 Ge. V) 1 ring e-like m-like Multi ring Total Data 626 558 1318 2502 MC(Honda flux) 612. 8 838. 3 1648. 1 3099. 1 Partially Contained (assigned as m-like) Total 754 1065. 0 m /e Data MC = 0. 675 +0. 034 -0. 032 0. 080

Zenith angle distribution 1289 days (79. 3 kt. yrs) No oscillation Best fit (Dm 2=2. 4 x 10 -3 e. V 2, sin 22 q=1. 00) (E<1. 33 Ge. V) (E>1. 33 Ge. V) c 2(best fit) = 132. 4/137 d. o. f. c 2(no osc. ) = 299. 3/139 d. o. f. Dc 2=167

Multi-ring event analysis 1289 days (79. 3 kt. yrs) Zenith angle distributions preliminary No oscillation Best fit (Dm 2=2. 0 x 10 -3 e. V 2, sin 22 q=1. 00) Sub-Ge. V muti-ring m-like sample 0. 6 Ge. V < E < 1. 33 Ge. V cosq Multi-Ge. V muti-ring m-like sample E > 1. 33 Ge. V cosq The zenith angle distortion is consistent with single-ring analysis.

Zenith angle distributions of upward-going muons Upward through-going muons 1416 events / 1268 days No oscillation: c 2(shape)=18. 7 / 10 d. o. f. (prob. =0. 044) Osc. best fit (Dm 2=5. 2 x 10 -3 e. V 2, sin 22 q=0. 86) vertical horizontal Upward stopping muons ( ( 345 events / 1247 days No oscillation: (Bartol, GRV 94) = ) ) stopping m through m Data stopping m through m MC 0. 241 0. 016 +0. 013 - 0. 011 0. 368 +0. 049 - 0. 044 = 0. 65 0. 04 0. 09 Oscillation (Dm 2=3. 2 x 10 -3 e. V 2, sin 22 q=1. 00) << 1

Allowed region (FC + PC + UP-thru + UP-stop) nm nt 79. 3 kt. yrs Best fit : Dm 2=2. 5 x 10 -3 e. V 2, sin 22 q=1. 00 (c 2=142. 1 / 152 d. o. f. ) 68% C. L. 90% C. L. 99% C. L. SK combined result Dm 2 = (1. 7~4)x 10 -3 e. V 2 sin 22 q > 0. 89 (90% C. L. )

Allowed region - II (FC + PC + UP-thru + UP-stop) nm nt 79. 3 kt. yrs Best fit : Dm 2=2. 5 x 10 -3 e. V 2, sin 22 q=1. 00 (c 2=142. 1 / 152 d. o. f. ) Dm 2 (e. V 2) unphysical region 68% C. L. 90% C. L. 99% C. L. sin 22 q SK combined result Dm 2 = (1. 7~4)x 10 -3 e. V 2 sin 22 q > 0. 89 (90% C. L. )

Zenith angle distributions for the best fit No oscillation Best fit (Dm 2=2. 5 x 10 -3 e. V 2, sin 22 q=1. 00)

Allowed region (grand global fit) (FC + PC + UP-thru + UP-stop + multi-rings) 79. 3 kt. yrs Within physical region; x 2 min = 157. 5/170 dof at sin 22 q = 1. 0, Dm 2 = 2. 5 10 -3 e. V 2 With unphysical region; x 2 min = 157. 4/170 dof at sin 22 q = 1. 01, Dm 2 = 2. 5 10 -3 e. V 2

Zenith angle distributions for the best fit (grand global fit) No oscillation Best fit (Dm 2=2. 5 x 10 -3 e. V 2, sin 22 q=1. 00)

Zenith angle distributions for the best fit (cont) (grand global fit) No oscillation Best fit (Dm 2=2. 5 x 10 -3 e. V 2, sin 22 q=1. 00)

Systematics in the 1 D fit

nm nsterile (p 0/m)Data (p 0/m)MC (p 0 method) { 1 for n n > 1 for nm nt m s Data 355. 2 events (BG subt. ) MC 323. 2 events (p 0/m)Data = 1. 49 0. 08(stat. ) 0. 11(sys. ) 0 (p /m)MC Experimental only

p 0 info from K 2 K-1 kt ( ( ) data p FC-m ) MC p 0 FC-m 0 = 0. 99 0. 03 0. 1 P RY A IN M LI E R

(p 0/m)data vs (p 0/m)MC-no-osc P Y R NA I IM L RE

nm nsterile (matter in earth) Using matter effect and enriched NC sample nm nt : No matter effect nm ns : With matter effect Neutrino oscillation in matter: ( ) ( nm ns = cosqm sinqm - sinqm cosqm )( ) n 1 n 2 22 q sin 22 qm = (z-cos 2 q)2+sin 22 q z = - 2 GFnn. En / Dm 2 sin 22 qm ~ For sin 22 q = ~ 1 And for En = 30~100 Ge. V Strategy: 1 z 2 + 1 z >>1 and sin 22 qm <<1 Suppression ! Obtain allowed region using lower energy events (Fully contained sample) Then, Test zenith angle of NC enriched events, high energy PC and throughgoing muon events.

Allowed region using only FC events

Zenith angle of high energy PC events > 45000 p. e. (E> ~ 5 Ge. V) <E>=~25 Ge. V nm ns nm nt Dm 2 = 3 x 10 -3 e. V 2 sin 22 q = 1 Zenith angle of upward-going muon nm ns nm nt Dm 2 = 3 x 10 -3 e. V 2 sin 22 q = 1

Zenith angle of NC enriched events Criteria > 400 Me. V visible energy Multi-ring event e-like ring is the most energetic ring Contents NC : 29 % ne CC : 46 % nm CC : 25 % nm nt nm ns Dm 2 = 3 x 10 -3 e. V 2 sin 22 q = 1

Ratios vs. Dm 2 sin 22 q = 1 nm nt Up/Down ( cos. Q <-0. 4 > 0. 4 ratio of NC enriched multi-ring Data ) nm ns Data nm nt 10 -3 nm nt 10 -2 e. V 2 Up/Down (cos. Q <-0. 4 > 0. 4 ) ratio of High Energy PC 10 -3 10 -2 e. V Vertical/Horizontal ratio (cos. Q > < -0. 4) of up muons

Allowed vs. excluded regions combine NC enriched, high E PC and up muons excluded nm ns is excluded with 99 % C. L.

Search for t leptons Neutrino CC cross sections Expected t events sin 22 q = 1 ~20 ev. /yr for 3 x 10 -3 e. V 2 nm CC All nt CC cosq<-0. 2 En(Ge. V) Signature of t appearance: Dm 2(e. V 2) nt + N t + N’ + p. . . mnn, enn, n+hadrons(p, p, . . ) cosq>0. 2 Higher multiplicity of Cherenkov rings · More m e decay signals · More spherical event pattern · Search for t appearance (3 methods) : (1) Energy flow and event shape analysis (2) Likelihood method using # of rings, m e, max p. e. ring and etc. (3) Neural network method Each method is optimized using only downward going events and then looks at upward going events. (I. e. blind method to disable systematic bias. )

Multi-ring samples : atm nm + ne w/o nt : nt CC

Zenith-angle distribution MC without t MC with t Dm 2=3 x 10 -3 e. V 2, sin 22 q=1. 00 (expected # of t : 74 events) Energy flow method Observed # of t : 25. 5 +14 -13 Efficiency for t: 32 # of t%production: 79 +44 -40 Likelihood method +17 +9 Observed # of t : 27 -16 -8 Efficiency for t : 43. 5 % +39 # of t production: 62 -27 +21 -18 Neural network method Observed # of t : 42 19 +13 -13 Efficiency for t : 45 % +14 # of t production: 92 35. 3 -0 cosq All methods show ~2 s excess of t-like events. The result is consistent with nm nt oscillations.

Probability of exotic oscillation models Test nm nt oscillation with : P(nm nt)=sin 22 q sin 2(b. L En) (q, b, n : parameters) n=-1 is the standard neutrino oscillation Use FC, PC, Up-through, and Up-stop data c 2 Magnified view -2 -1 0 1 index n n = -1. 06 0. 14

Neutrino decay Let neutrinos oscillate and decay n 3 P(nm nm) = sin 4 q + cos 4 q exp + sin 2 q exp ( ( m 3 L 2 t 3 E ) Consider two cases; dcy>> osc, and dcy<< osc, where t 3 E dcy = m 3 4 p. E , osc = Dm 2 m 3 L t 3 E X(invis); ) Dm 2 L cos 2 E ( )

dcy >> osc For Dm 2 , c 2 min = 221. 2/153 dof Bad fit !

dcy << osc For Dm 2 0, c 2 min = 147. 1/153 dof at (sin 2 q, m 3/t 3) = (0. 68, 0. 01 (Ge. V/km)) Good fit !

Up/down of NC enriched events (short dcy) FC, Nring>1, Evis>400 Me. V, Brightest ring = e-like Allowed from FC+PC+Upmu Excluded from NC The case of dcy<< osc is disfavored

Conclusions on atmospheric neutrinos n Oscillation parameters for nm nt : Dm 2 = 1. 7 ~ 4 x 10 -3 e. V 2, sin 22 q > 0. 89 (90%CL) n 3 D flux does not change the conclusion but more precise 3 D calculations are needed n nm ns is strongly disfavored n p 0/m ratio is consistent with nm nt n Excess from t leptons ~ 2 s n Decay senario is disfavored with > 2 s for dcy>> osc and dcy<< osc