Super-Kamiokande Roger Wendell, Duke University NNN 2010 Toyama, - - PowerPoint PPT Presentation

Super-Kamiokande

Roger Wendell, Duke University NNN 2010 Toyama, Japan

1 Kamioka Observatory, ICRR, Univ. of Tokyo, Japan 2 RCCN, ICRR, Univ. of Tokyo, Japan 3 IPMU, Univ. of Tokyo, Japan 4 Boston University, USA 5 Brookhaven National Laboratory, USA 6 University of California, Irvine, USA 7 California State University, Dominguez Hills, USA 8 Chonnam National University, Korea 9 Duke University, USA 10 Gifu University, Japan 11 University of Hawaii, USA 12 Kanagawa, University, Japan 13 KEK, Japan 14 Kobe University, Japan 15 Kyoto University, Japan 16 Miyagi University of Education, Japan 17 STE, Nagoya University, Japan 18 SUNY, Stony Brook, USA 19 Niigata University, Japan 20 Okayama University, Japan 21 Osaka University, Japan 22 Seoul National University, Korea 23 Shizuoka University, Japan 24 Shizuoka University of Welfare, Japan 25 Sungkyunkwan University, Korea 26 Tokai University, Japan 27 University of Tokyo, Japan 28 Tsinghua University, China 29 Warsaw University, Poland 30 University of Washington, USA

From PRD81, 092004 (2010)

~120 collaborators 31 institutions, 6 countries

Autonomous University of Madrid, Spain (Nov.2008~)

The Super-Kamiokande Collaboration The Super-Kamiokande Collaboration

Inner detector (ID) ~11,146 50 cm PMTs ~ 2ns timing resolution Outer detector (OD) 1,885 20 cm PMTs 50 kton water Cherenkov detector 22.5 kton fiducial volume Depth of 2700 m.w.e cosmic ray background ~3 Hz Roughly ~10 Solar ν events ~10 Atmospheric ν per day Multi-purpose detector: (this talk) Solar neutrinos Supernova neutrinos ( Relic SN's ) Atmospheric neutrinos Nucleon decay Beam neutrinos: K2K, T2K ( R. Wilson, today) Exotic particles

Super-Kamiokande Super-Kamiokande

Super-K : Generations

SK-1/3/4 SK-2

Analyses Analyses Data Update

SK-IV Upgraded DAQ system SK-IV Upgraded DAQ system

New

Electronics

(QBEE) Readout (Ethernet) Periodic trigger

(17µ sec x 60kHz)

Clock Event build with variable time windows

SK-I,II,III: partial data above threshold were read (1.3µ sec window x3kHz) SK-IV: All hits are read, then apply complex triggers by software.

Collect all hits every 17µ sec .

PMT

signals

Typical event time windows: Super-Low-Energy (SLE) events (<~6.5MeV): -0.5/+1.0µ sec Normal events(>~6.5MeV): -5/+35µ sec Supernova Relic ν (SRN) candidates(>~10MeV, No OD): -5/+535µ sec T2K events: -512/+512µ sec at T2K beam spill timing

high rate (~3kHz) decay electrons neutrons

Wider dynamic range for charge measurement of each channel (>2000pC) No dead time up to ~6MHz/10sec for Supernova burst neutrinos Apply precise event reconstruction to remove more low-e BG events in real-time Precise analysis in parallel in real-time

x5 x100 IEEE Trans. Nucl. Sci. 57 (2010) 428

T2K GPS from J-PARC

Solar Neutrinos

Introduction to Recent Solar and SN Introduction to Recent Solar and SN ν ν Developments Developments

SK-IV solar ν data, new 8B flux measurement SK-I + II + III Oscillation Fits 2- and 3-Flavor Updated Supernova Relic Neutrino analysis

This talk

Neutron Tagging in SK-IV , Poster by H. Zhang Supernova burst DAQ system, Poster by T. Yokozawa Status of SK Gadolinium R&D project, Poster by A. Kibayashi Search for GUT Monopoles, Poster by K.Ueno

Ee = 9.1MeV cosθ sun = 0.95

Energy: 14% Vertex: 87cm Direction: 26o SK-I Energy: 14% Vertex: 55cm Direction: 23o SK-III

• Timing information

vertex position

• Ring pattern

direction

• Number of hit PMTs

energy ~6hit / MeV (SK-I, III, IV)

OD ID (color: time) (software improvement)

Typical Low Energy Event Typical Low Energy Event

Resolutions:

SK-IV Flux 2.28±0.04 (106/cm2/s)

SK-III official:1 2.32±0.04±0.05 (106/cm2/s)

Preliminary

SK-IV SK-IV 8

8B Flux

B Flux 567 days

BLACK: SK3 RED : SK4

Events/day/kton/bin

Fluxes in SK-IV are consistent with those from SK-III

• SK-IV data looks good so analysis under way
• Following oscillation analyses are for SK-I+II+III

SK SK-I 1496 days, spectrum 5.0-20MeV + D/N : E ≥ 5.0MeV SK-II 791 days, spectrum 7.0-20MeV + D/N : E ≥ 7.5MeV SK-III 548 days, spectrum 5.0-20.0MeV + D/N : E ≥ 5.0MeV SNO CC flux (Phase-I & II & III) NC flux (Phase-III & LETA combined) ( = 5.14±0.2 × 106cm-2s-1) Day/Night asymmetry (Phase-I & II) Radiochemical : Cl, Ga Ga rate: 66.1+/-3.1 SNU (All Ga global) , PRC80, 015807(2009) Cl rate: 2.56+/-0.23 , Astrophys. J. 496 (1998) 505 Borexino

7Be rate: 48 +/- 4 cpd/100tons, PRL101, 091302(2008)

KamLAND : 2008

8B spectrum : Winter(2006)

Oscillation Analysis Data Set and Inputs Oscillation Analysis Data Set and Inputs

Items in red are updates since the analysis presented in PRD78 , 032002 (2008)

Global Data

Two-Flavor analysis of SK-I+II+III w/ Flux Constraint Two-Flavor analysis of SK-I+II+III w/ Flux Constraint

Min χ2 = 48.8 ∆m2 = 6.1×10-5 eV2 tan2θ = 0.48 ΦB8 = 0.89× ΦB8,SSM

95% C.L.

* 8B rate is constrained by the SNO (NCD + LETA) Neutral Current Flux LMA

• nly

Add in global and KamLAND Solar global Solar+KamLAND

Min χ2 = 57.7 ∆m2 = 7.6×10-5 eV2 tan2θ = 0.44 ΦB8 = 0.89× ΦB8,SSM

Solar Global + KamLAND Preliminary

Solar global KamLAND Solar+KamLAND

Three-Flavor Analysis ( including SK-I+II+III ) Three-Flavor Analysis ( including SK-I+II+III )

68, 95, 99.7% C.L. Sin2θ 13=0.025+0.18

• 0.16 ( < 0.059 at 95%C.L.)

Solar Global + KamLAND

sin2θ 13=0.06 @95%C.L.

Solar Global

In both fits best θ13 is small but consistent with 0

• arXiv:1010.0118

Preliminary

Supernova Relic Neutrinos

Supernova Relic Neutrino Search Supernova Relic Neutrino Search

Look for these Ando , NJP 6 (2004) 170

Supernova explosions occurred commonly throughout the history of the universe

• Expect a diffuse ν flux

Measurement of diffuse flux ⇒ galactic evolution, matter distribution in the universe

• Current limit , 1.2 / cm2 / s at 90% C.L. SK-I Malek et al. PRL 90, 061101 (2003)

Solar neutrinos are a considerable background at low energies

Large background from solar Neutrinos Spallation, Atmosheric ν , also background

Search for inverse beta decay interactions 16 < Ee+ < 80 MeV

Improved SN Relic Search in Improved SN Relic Search in SK-I+II+III

ν

e

e +

p n

(invisible)

Signal Events

42 o

μ, π Low angle events

25-45o

Isotropic Events

ν N ν

reconstructed angle near 90o

Update of the Inverse-beta decay cross-section (overall decrease)

• Strumia-Vissani PLB 564 (2003) 42

Use Poisson probability based likelihoods during fitting Improvements to data selection

• Change in Cherenkov angle cut (next slide)
• Spallation and solar angle cuts

Neutral current Elastic Scattering

Each of these categories has a different Cherenkov angle distribution

Efficiency was 58% in SK-I now: 78% (SK-I ) 69% (SK-II) 77% (SK-III)

After the event selection there are 3 Types of events remaining

Preliminary

Cherenkov Angle ° low region (μ / π) signal region

isotropic region (NC)

Cherenkov Angle °

MC

• ATM. νe CC

μ/π NC elastic

SN Relic Fitting in SN Relic Fitting in SK-I+II+III

Remaining events populate different regions of the distribution

• Previous analysis selected 37° < θc < 50°
• Fit the backgrounds outside of the signal region simultaneously to better

constrain their contribution in the overlap (Three regions)

20-38 degrees (low region) 38-50 degrees 78-90 degrees (NC ES region) E (MeV)

data Relic* all BG νμ CC νe CC NC elastic μ/π > C. thr.

Supernova Relic Neutrino Fit , Supernova Relic Neutrino Fit , SK-I+II+III

E (MeV) E (MeV)

Combined fit shows good agreement in both the signal and non-signal regions

* Ando , NJP 6 (2004) 170 Preliminary

combined 90% c.l.: < 5.1 ev / yr / 22.5 ktons interacting < 2.7 /cm2/s (>16 MeV) < 1.9 /cm2/s (scaled to >18 MeV)

combined 90% c.l. ev/yr in 22.5 ktons l

• g

L i k e l i h

• d

SK-I+II+III combined likelihood

Supernova Relic Neutrino Fit Supernova Relic Neutrino Fit

SK-I SK-II SK-III combined

ev/yr in 22.5 ktons

SK-I alone fit prefers almost no signal SK-II and SK-III Fits allow more relic ν Slightly larger than published limit

Preliminary

/cm2/s >18 MeV

Published limit 1.2 cross section update to Strumia-Vissani 1.2 → 1.4 Gaussian statistics → Poissonian statistics in fit 1.4 → 1.9 New SK-I Analysis: ETHRESH 18 → 16 MeV ε = 52% →78 % (small statistical correlation in samples) improved fitting method takes into account NC 1.9→1.6 New SK-I/II/III combined fit 1.6 → 1.9

Comparison With Published Limit , Comparison With Published Limit , SK-I Analysis Change

Preliminary

Atmospheric ν

Roughly νe : νµ ~ 1:2

Introduction to Recent Atmospheric Neutrino Updates Introduction to Recent Atmospheric Neutrino Updates

» SK-IV Data » Oscillation Analyses Using SK-I+II+III » Update to the search oscillation induced τ-neutrinos » Searches for Rare particles and processes (nucleon decay)

SK-IV Atmospheric Neutrino Data SK-IV Atmospheric Neutrino Data

Sub-GeV e-like Multi-GeV e-like Multi-GeV µ-like Partially Contained

» SK-IV Data look good, consistent with SK-III » No oscillation result yet, but analyses are

coming

• Oscillations already appearing

Remainder of the Talk will concentrate on SK-I+II+III unless otherwise noted

Unoscillated SK-IV MC

Preliminary

Zenith angle & lepton momentum distributions : SK-I+II+III Zenith angle & lepton momentum distributions : SK-I+II+III

M-like samples show large deficits in the upward- going bins that are well described by oscillations Live time: SK-I 1489d (FCPC) 1646d (Upmu) SK-II 799d (FCPC) 827d (Upmu) SK-III 518d (FCPC) 636d (Upmu) µ -like e-like momentum ν µ–ν τ oscillation (best fit) null oscillation

Global Picture of Oscillations Agrees Global Picture of Oscillations Agrees

Experiments are in good agreement about these oscillations SK Data disfavor other types of disappearance strongly, sterile ν ~7σ We should look for oscillation induced ντ appearance!

SK Zenith Analysis (1σ) ∆m23

2 = 2.11 +0.11 -0.19 ×10-3 eV2

sin2 2θ23 > 0.96 ( 90% C.L.) SK L/E Analysis (1σ) ∆m23

2 = 2.19 +0.14 -0.13 ×10-3 eV2

sin2 2θ23 > 0.96 ( 90% C.L.)

Induced by θ13 + solar terms Induced by θ13

P( νµ → νe )

Three-Flavor Oscillations in Matter Three-Flavor Oscillations in Matter

Solar Terms Matter Driven Interference term sensitive to δ cp

» Presence of electrons in the Earth

alter the neutrino ineraction potential and induce additional νµ → νe

• scillations

» Higher energy , 2-10 GeV, (anti-)

neutrinos experience resonant enhanced transitions, for normal (inverted) hierarchy

» Lower energy oscillations, < 1GeV ,

are moderated by octant of θ23

Simultaneously considering all of these effects gives sensitivity to many of the remaining questions on oscillation physics...

Full Three-Flavor Oscillation Analysis, Full Three-Flavor Oscillation Analysis, Normal Hierarchy

SK-I+II+III 99% SK-I+II+III 90% SK-I+II+III 68%

» No Strong preference for either hierarchy (∆χ2 = 1.6) » θ13 is consistent with zero and the Chooz limit » No preference for θ23 octant or δcp

Chooz Exclusion region

Poster by M. Lee

Preliminary

e

• r



• r

h a d r

• n

s

Energy Threshold: 3.5 GeV  

 

Hadron

SK-I MC

ντ Events at Super-K

» Complicated event topology

complicate identification of the leading lepton

• Use a Neural Network procedure

» Many light producing

particles

» Most events are deep

inelastic scattering interactions GOAL : Observe ντ events in the atmospheric data How inconsistent is the “no appearance” hypothesis?

» Many light producing

particles

» Most events are deep

inelastic scattering interactions

» Negligible primary flux

• Observed tau events would be
• scillation induced

Update Fitting Technique

» Use an un-binned two-dimensional likelihood fit to extract the most from the data » Previous analysis ( PRD 2006 ) fit only in one dimension

• Events with NN output > 0.5 in this plot

» Tau and Background events appear in dramatically different regions of the plot

• The signal appears exclusively in the upward-going direction

BKG

τ-like BKG-like

θ13 BKG would be here

Neural Network and Fitting

NN

» 7 variables in NN : Fraction of energy in leading ring, number of decay electrons ,

number of ring fragments, visible energy, leading ring's PID, distance to decay-e

» Neural Network is good at separating signal from background » DIS events appear τ-like because of their many out-going particles

• To account for this we fit for the fraction of DIS events constrained to an

uncertainty of 10%

Signal Check output with downward-going data

downward-going

τ-like events (NN > 0.5) SK-I+II+II SK-I+II+II

Fit Results

Background Signal Result α β SK-I 0.96 SK-II 0.96 SK-III 0.94 SK-I+II+III 1.5 ± 0.48 1.67 ± 0.62 2.36 ± 0.77 0.94 ± 0.02 1.63 ± 0.35 DIS γ 1.10 ± 0.05

Fitting Error only

» Tau signal clearly appears in upward-

going region

» DIS fits to +1 σ » τ normalization fit is 1.63 × expectation

Fitted τ Excess Atm ν BKG MC

If no τ appearance , β = 0 Preliminary

Systematic Errors

E x p e c t ν ντ Observed ντ

» Flux : Up/down Ratio , Horizontal/Vertical

Ratio, K/p ratio

» X-Sec: NC / CC Ratio » Detector: Up/Down Energy Calib. Asymm » Oscillation Paramaters: Atm. 1σ region

0 < θ13 < Chooz

» Atmospheric oscillation ν analysis

sytematics ( 27 errors )

» Tau neutrino cross-section

.+6.1% .-5.0% .+1.5% .-0.5% .+0.0% .-13.0% .+12.5% .-14.2% .±25%

These affect the significance (next page)

Leading Systematic

This corresponds to 213.6 τ Events

Preliminary

SK data are inconsistent with no τ appearance at 3.8σ

Significance Calculation

» Prepare an asymmetric gaussian

centered about the best fit from the data with widths corresponding to the systematic errors (previous page)

» The integral of the PDF below

zero (corresponding to no τ appearance) is a measure of the significance

SK I+II+III : Integral below zero = 5.8 × 10-5 corresponds to 3.8 σ The Expected signficance : 2.6σ ( 2.9σ without systematics )

Central value of SK-I+II+III Fit Preliminary

Rare Particles and Proton Decay

EXAMPLE: illustration of 5.6 GeV WIMP annihila1tion signal in SK signal is before FIT Monoenergetic Eν = Mχ Isotropic assume 100% BR

» Search in atmospheric neutrino

data from SK-I, -II & –III

livetime: FC/PC 2806 days, UPMU 3109 days

» FIT: for each tested WIMP mass,

find the best configuration of ATM MC + DM signal that would match DATA the best

» Simulate signal in NUMU, NUE

and NUTAU

» Distinctive signatures: » using all SK samples:

e-like + mu-like FC+PC+UPMU (wide energy range)

Momentum [GeV/c]

cosθ

Mχ = 5.6 GeV Mχ = 5.6 GeV

Search for Diffuse Dark Matter Annihilation Search for Diffuse Dark Matter Annihilation

DM signal shape

enhanced for illustration

DATA

SK1,2,3

ATM MC with oscillations

SK two-flavor best fit

χ + χ → ν + ν

» Conservative upper limit on WIMP total

self-annihilation cross section <σV>

» No allowed excess of DM-induced ν’s

for Mχ in range 3GeV – 3 TeV

» FIT based on Evis & cosθ distr., systematics

included (120 sys. terms fitted) Limit on <σv>

J∆ Ω integrated intensity over all sky related

to DM halo density profile; includes information about DM density cusp in GC

Fit Results Fit Results

excluded above

M W H a l

• A

v e r a g e ( * )

» Focus on signal arising from Milky Way

halo (diffuse flux)

(*) H.Yuksel et al., Phys. Rev. D76, 123506 (2007), arxiv: 0707.0196 [astro-ph]

Nucleon Decay Limits, 2010

This talk

Super-K

7.5 × 1033 protons 6.0 × 1033 neutrons » Proton is predicted to be

stable in the standard model

» GUT model, Super-

Symetric Models predict various types (and lifetimes!) of proton decay

Search for p Search for p → → e e+

+ +

+ π π0

0 ,

, SK-I+II+III+IV

τ / B > 1.21 × 1034 yr (205.7 kton yr )

τ / B > 1.9 × 1033 yr (32.9 kton yr )

SK-IV Only SK-I-IV combined

Still no Candidates! Signal MC Background MC Data

SK4 MC Data Cut efficiency (%) 535.2d 2 or 3 Rings 73.6 1193 1117 All e-like 65.5 708.5 674 63.5 590.7 575 0 Decay electron 62.5 409.6 386 Total Mass Cut 45.0 ± 19.0 0.05 p → e+ + π0 85 < π0

m ass < 185 MeV

Into the next decade!

Preliminary

Summary Summary

SK-IV is underway , Solar and Atmospheric data look good New Solar results including global fits with SK-I+II+III

• Supernova Relic Neutrino Analysis has been updated ~ 1.9 ev/cm2/s

New Atmospheric results including oscillations and nucleon decay

• No oscillation induced tau events disfavored at 3.8σ

Tour of Kamioka Observatory will include Super-K, so come have a look!