SN Physics Workshop September 17 th 2009 Michael Smy UC Irvine Super-Kamiokande Results

SN Physics Workshop September 17th 2009

Michael SmyUC Irvine

Super-Kamiokande ResultsSuper-Kamiokande Results

http://antwrp.gsfc.nasa.gov/apod/image/9710/skam_icrr_big.jpg

Super-KamiokandeSuper-Kamiokande• 50,000 tons of

ultra-pure Water• 11,129 20” PMTs

covering 40% of the inner 32,000 tons: ~six photo-electrons per MeV

• 1,885 8” PMTs with wavelength shifter plates view the outer 18,000 tonsMichael Smy, UC Irvine

Super-Kamiokande HistorySuper-Kamiokande History

11146 ID PMTs(40% coverage)



EnergyThreshold(total electron energy)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

SK-I SK-II SK-III SK-IV

Acrylic (front)+ FRP (back)

ElectronicsUpgrade

SK-I SK-II SK-III SK-IV

5.0 MeV 7.0 MeV 4.5 MeVwork in progress

< 4.0 MeVtarget

inner detector mass: 32kton fiducial mass: 22.5kton

Michael Smy, UC Irvine

SK New Front-End Electronics: QBEESK New Front-End Electronics: QBEE

QTC TDC FPGA

Network Interface Card

PMTsignal

EthernetReadout

60MHz ClockTDC Trigger

QTC-Based Electronics with

Ethernet (QBEE)

• 24 channel input• QTC (custom ASIC)

– three gain stages– wider (5x!) dynamic range

• Pipe line processing– multi-hit TDC (AMT3)– FPGA

• Ethernet Readout• 60MHz common clock• Internal calibration pulser• Low (<1W/ch!) power

Calibration Pulser


Difference in Readout SystemDifference in Readout System

FormerElectronics

(ATM)

Readout (backplane, SCH, SMP)

Trigger (1.3sec x 3kHz)

HITSUMTrigger

logic

NewElectronics

(QBEE)

Readout (Ethernet)

Periodic trigger(17sec x 60kHz)

Clock

Hardware Triggerusing number of hit

(HITSUM)

1.3secevent window

Variableevent window

by software trigger

No hardware trigger. All hits are read out. Apply software trigger.No hardware trigger. All hits are read out. Apply software trigger.

12PMTsignals

permodule

24PMTsignals

permodule

Collect ALL hits; trigger every 17sec with a 60kHz clock without “gaps”

Former readout system

New readout system


SuperSupernova nova Neu-Neu-

trinos trinos


Supernova Supernova Burst: Expected # of Events Burst: Expected # of Events

~7,300 e+p events~300 +e events~360 16O NC events ~100 16O CC events (with 5MeV thr.)

for 10 kpc supernova

Neutrino flux and energy spectrum from Livermore simulation (T.Totani, K.Sato, H.E.Dalhed and J.R.Wilson, ApJ.496,216(1998))

Courtesy M. Nakahata, ICRR

Time Variation Measurement withTime Variation Measurement with ee+p+pAssuming a supernova at 10kpc.

Time variation of event rate Time variation of mean energy

Enough statistics to discuss model predictions

ep e+n events give direct energy information (Ee = E – 1.3MeV).


SN at 10kpc

e+p

e+p

e+p e+p

+e +e

+e +e

+e+e Scattering EventsScattering Events

Spectrum of +e events can be statistically extracted using the direction to supernova.

Direction of supernova can be determined with an accuracy of ~5 degree.

Neutrino flux and spectrum from Livermore simulation


SN at 2kpcTime variation Visible energy spectrum

~240,000 events are expected for supernova at 2kpc.

~10,000 events are e scattering events.

Total number of events in parentheses

200 log bins from 20msec to 18sec

Close SupernovaeClose Supernovae


SN at 2kpc

Spectrum measurement up to ~40MeV.

ee++x x Energy Spectrum MeasurementEnergy Spectrum Measurement


SN at 2kpc

Number of events from 20msec to 0.1 sec (1bin=10msec)

Neutronization burst could be observed even with neutrino oscillations.

No oscillation

Normal PH=1 orInverted hierarchy

ν ＋ e －

Normal hierarchy PH=0

Neutronization Burst (eNeutronization Burst (e--+p+pn+n+ee))


Search for Neutronization Burst in SK-I/IISearch for Neutronization Burst in SK-I/II• use magnitude of average

direction of events within 1, 10, and 100ms: sumdir

• 84% efficiency if require sumdir>0.75

• also cut on mean distance between event vertices: >94% efficiency

• no cluster found with more than two events in SK-I or II

• found 194/19/1 doublets within 1/10/100ms in SK-I data while expecting 194/19/2.1

• found 0/0/10 doublets in SK-II data while expecting 0.125/1.25/12.5


expect between 1 and six events at 10kpc(depending on oscillation)

a diffuse neutrino signal from all past supernovae

Motivation SRN measurement enables us to investigate the history of past Supernovae. The SRN flux determines the star formation rate and supernova rate in galaxies.

Predicted Predicted SRN fluxSRN flux

Expected # SRN evts in SK10-30MeV: 0.8 -5.0 evts/22.5kt·y

16-30MeV: 0.5 -2.5 evts/22.5kt·y

18-30MeV: 0.3 -1.9 evts/22.5kt·y

Supernova Relic Neutrinos (SRN)Supernova Relic Neutrinos (SRN)

Courtesy Iida, ICRR

Many backgrounds

in SN relic energy

window:• electronic noise• solar ’s • reactor ’s• atmospheric ’s• cosmic ray ’s• spallation from ’s

(~600/day)• radioactive backgrounds

spallation is worst; products decay with energies up to 20.8 MeV

and lifetimes up to 13.8 s (practically forever): spallation limits the

energy threshold & cuts to reduce it causes greatest signal loss

SN Relic SN Relic ’s: ’s: Backgrounds in HBackgrounds in H22OO

atm. → stealth ±→e±

relic ’s

spallation productsfrom cosmic ’s


Visible energy [MeV]

SK-ISK-I

Visible energy [MeV]

SK-IISK-II

DATDATAA

Atmospheric e Atmospheric e

Invisible -e decay

Invisible -e decay

Spallation BGSpallation BG

DATDATAA

(1496day) (791day)

preliminarypreliminary preliminarypreliminary

90% C.L. Flux limit:90% C.L. Flux limit:SK-I : < 1.25 /cm2 /sec

SK-I + SK-II : < 1.08 /cm2 /secSK-II : < 3.68 /cm2/sec

Irreducible backgroundsIrreducible backgrounds:Atmospheric νe cc interactions

Decay of sub-Cherenkov ‘invisible μ’s’ from atmospheric νμ interactions

SK-I result:M. Malek, et al, Phys. Rev. Lett. 90, 061101 (2003)

Courtesy Iida, ICRR

0

0.5

1

1.5

2

2.5

3

3.5

4

Constant SN rate

(Totani et al. 1996)

Totani et al. 1997

Malaney et al. 1997)

Hartmann et al. 1997)

Kaplinghat et al. 2004

Ando et al. 2005

Fukugita et al. 2003

Lunardini et al. 2006

SK-II limit = 3.68 /cm2/sec

SK-I limit = 1.25 /cm2/sec Combined limit = 1.08 /cm2/sec

preliminary(E>18MeV)Flux limit VS Predicted FluxFlux limit VS Predicted Flux

Courtesy Iida, ICRR

Solar Solar ‘s‘s


Solar Neutrino Future Prospects in SKSolar Neutrino Future Prospects in SK

Vacuum osc. dominant

transition from vacuum to matter osc.“upturn” in 8B relative spectrum.

matter dominant

e survival probability(at best fit parameter)

Expected spectrum distortion with 5 years low BG SK data

BG is 70% reduced compared to SK-I below 5.5 MeVEnergy-cor. Syst. uncertainty is half compared to SK-Ifive years

SK-I

0.8

0.

6

0.

4

0.

2

0.

0

P(

e

e)

Courtesy L. OberauerCourtesy L. OberauerTU MTU Münchennchen(BOREXINO)(BOREXINO)

Neutrino Energy in MeV Michael Smy, UC Irvine

SK-III: Less Radioactive BackgroundSK-III: Less Radioactive Background

r2 [m2]

z [m

]

clean central

13.3kton

5.0-5.5MeV

5.5-6.0MeV

6.0-6.5MeV

SK-ISK-III

SK-ISK-III

SK-ISK-III

Courtesy Y. Takeuchi, ICRR

consistent with SK-I within statistical uncertainty!

Observed Observed 88B Flux in SK-IIIB Flux in SK-III

SK-I 8B flux: 2.35±0.02(stat)±0.08(sys) x106/cm2s(PRD73: 112001, 2006)

DataBest-fitBackground


Hint of Signal between 4.5-5.0MeV (recoil electron total energy)

Fiducial volume is central 9.0kton

Solar Peak at 4.5 MeVSolar Peak at 4.5 MeV

DataBest-fitBackground


88B B Flux Flux

SK-III 298day5.0-20MeV

(Preliminary)


Recoil Electron SpectrumRecoil Electron Spectrum

8B=2.36x106/cm2s

hep=15x103/cm2s(best-fit for SK-I)


Day/Night AsymmetryDay/Night Asymmetry• only direct test of matter effects on solar neutrino

oscillations• SK-I measured ADN=2(D-N)/(D+N)=-2.1±2.0%(stat)• SK-I also fit LMA day/night variations; expressed as

ADN the result is ADN=-1.8±1.6%(stat) • SK-II measured ADN=-6.3±4.2%(stat)• SK-III can measure ADN to ±4.3%(stat) with the shown

298 days of data; maybe to ±3.7%(stat) using the entire SK-III data set (including periods w/o SLE or high very low energy background runs)

• SK-I-III can determine ADN to ±1.6%(stat)• SK-I-III can fit LMA D/N variations to ±1.3%(stat) Michael Smy, UC Irvine

Solar Solar Oscillation Constraints Oscillation Constraints

Courtesy Ikeda, ICRR

excluded from spectrum& d/n variation

allowedusing 8B totalflux by SNO

SK combinedVery Preliminary

SK-IIIVery Preliminary

excluded byspectrum

global solarVery Preliminary

WWideband ideband IIntelligent ntelligent TTriggerrigger• have 2 modules:

32 cores • plan to buy four

more modules: 96 cores

• sufficient CPU for 3MeV thresholdI. convert inner detector hit ADC/TDC counts to real times/charges

II. sort hits by time

III. pre-filter based on N230 (# of hits within 230ns)

IV. Software Triggered Online Reconstruction of Events: coincidence after time-of-flight subtraction (vertex from selected four-hit combin.)

V. fast vertex fit

VI. if fiducial, precision vertex fit

VII. if fiducial, save event Michael Smy, UC Irvine

ProCurve Switch

WIT Machine IDual Quad-Core3GHz CPU

WIT Machine IIDual Quad-Core3GHzCPU

10Gbit 10Gbit

1Gbitmany “slow”ethernet lines two fast

ethernet lines

AtmosphericAtmospheric NeutrinosNeutrinos


100

0

0

0

010

0

0

0

001

1212

1212

1313

1313

2323

2323 CS

SC

CeS

eSC

CS

SCUi

i

Fanny Dufour WIN09 September 2009

Atmospheric Atmospheric ’s: It’s not just for ’s: It’s not just for atmospheric mixing any moreatmospheric mixing any more

Cij=cosθij

Sij=sinθij

Cij=cosθij

Sij=sinθij

SolarAtmospheric Accelerator / reactor

“2-3 sector” “1-3 sector” “1-2 sector”

Atmospheric mixing parameters:

• Zenith angle analysis → mainly sin2(2θ23)• L/E analysis → mainly Δm2

• Solar term analysis → octant degeneracy

θ13 and mass hierarchy:

• 3 flavors zenith angle analysis

Non-standard interactions are not covered in

this talk

Non-standard interactions are not covered in

this talk

Courtesy F. Dufour, Boston University

Two-Flavor: Zenith & L/E AnalysisTwo-Flavor: Zenith & L/E Analysis

L/E analysisGoal is to actually see the first oscillation dip.

Need events with good path-length (L) and energy (E) resolution.

Uses a subsample of events with good resolution.

cos θzenith

DatasetsSK-I FC/PC: 1489 daysSK-I Upmu: 1646 daysSK-II FC/PC: 798 daysSK-II Upmu: 828 daysSK-III FC/PC: 518 daysSK-III Upmu: 635 days

DatasetsSK-I FC/PC: 1489 daysSK-I Upmu: 1646 daysSK-II FC/PC: 798 daysSK-II Upmu: 828 daysSK-III FC/PC: 518 daysSK-III Upmu: 635 days

420 bins each for SK-I, II, and III; 122 syst. terms describe neutrino flux, cross section, reconstruction,

and data reduction uncertainties

where

Zenith angle analysisGoal is to observe a deficit of upward going neutrinos.



Zenith Analysis ResultsZenith Analysis ResultsData

MC (no oscillations)

MC (best fit oscillations)

New:Sub-GeV samples subdivided to improve sensitivity to low energy oscillation effects

16 sub-samples are used for the

oscillation analysis



L/E analysis resultsL/E analysis resultsDatasetsSK-I FC/PC μ-like: 1489 daysSK-II FC/PC μ-like: 798 daysSK-III FC/PC μ-like: 518 days

DatasetsSK-I FC/PC μ-like: 1489 daysSK-II FC/PC μ-like: 798 daysSK-III FC/PC μ-like: 518 days

We do see oscillation and not just disappearance and we compare against:

Neutrino decay (disfavored @ 4.4σ)Neutrino decoherence (5.4σ)

Grossman and Worah: hep-ph/9807511Lisi et al.: PRL85 (2000) 1166

Barger et al.: PRD54 (1996) 1, PLB462 (1999) 462

Δm2 = 2.2 * 10-3 eV2Δm2 = 2.2 * 10-3 eV2

sin2(2θ23

)=1.0sin2(2θ23

)=1.0

E

LmP

4sin2sin1)(

22322



Two-Flavor Results (SK I+II+III)Two-Flavor Results (SK I+II+III)Zenith angle analysis best fit

L/E analysis best fit

These two analyses are complementary:L/E has stronger Δm2 constraintEqually strong sin22θ23 constraint

These two analyses are complementary:L/E has stronger Δm2 constraintEqually strong sin22θ23 constraint

SK-1+2+3, Preliminary



Comparing with MINOS and K2KComparing with MINOS and K2KZenith angle analysis best fit

L/E analysis best fit

SK-1+2+3, Preliminary

The results agree well with other experimentsLong baseline constrains Δm2 betterAtmospheric still has stronger sin2 2θ constraint

The results agree well with other experimentsLong baseline constrains Δm2 betterAtmospheric still has stronger sin2 2θ constraint



Solar Term & Octant degeneracySolar Term & Octant degeneracyC

osin

e Z

enit

h A

ngle

Energy (GeV)

νe flux reduction

νe flux enhancement

(In constant density matter)

Driven by Δm212 and θ12.

Addition of solar terms shows no significant deviation of θ23 from π/4.

Addition of solar terms shows no significant deviation of θ23 from π/4.



θθ1313 with atmospheric neutrinoswith atmospheric neutrinos

MSW effect gives rise to additional scattering amplitudes in matter (for νe only). The clearest indication of non-zero θ13 at Super-K is a resonance @ ~2-10 GeV for up-going e-like eventsNormal hierarchy neutrino enhancement⇒Inverted hierarchy anti-neutrino enhancement⇒

Analysis uses 3 parameters (sin2θ13, sin2θ23, Δm223)

assuming a single “dominant mass scale” (Δm223 Δm≫ 2

12).

MSW effect gives rise to additional scattering amplitudes in matter (for νe only). The clearest indication of non-zero θ13 at Super-K is a resonance @ ~2-10 GeV for up-going e-like eventsNormal hierarchy neutrino enhancement⇒Inverted hierarchy anti-neutrino enhancement⇒

Analysis uses 3 parameters (sin2θ13, sin2θ23, Δm223)

assuming a single “dominant mass scale” (Δm223 Δm≫ 2

12).

sin2 θ13 = 0.005

Cos

ine

Zen

ith

Ang

le

sin2 θ13 = 0.015 sin2 θ13 = 0.04

Energy (GeV) Energy (GeV) Energy (GeV)



Three flavor Effects: Zenith Angle DataThree flavor Effects: Zenith Angle Data

Clear distortion of muon-like zenith distribution, well-described by 2-flavorνμ → ντ disappearance...

Allow also νμ → νe appearance in 3-flavor analysis, look for enhancement of high-energy upward-going e-like events.

No distortion in electron-like samples... No distortion in electron-like samples... no evidence for matter-enhanced no evidence for matter-enhanced ννee appearance.appearance.

Preliminary

Preliminary

Preliminary

Data

MC(no oscillations)

MC (best fit oscillations)



Three Flavor ResultsThree Flavor Results

Normal Hierarchy

Inverted Hierarchy

Data consistent with both hierarchies; no electron-like excess observed.Analysis assumes Δm2

12 = 0, next update will include solar terms.

Data consistent with both hierarchies; no electron-like excess observed.Analysis assumes Δm2

12 = 0, next update will include solar terms.

χ2/dof Δm223 sin2θ23 sin2θ13

Normal 469/417 2.1x10-3 0.50 0

Inverted 468/417 2.1x10-3 0.55 0.01



By combining the solar term analysis and the three flavor analysis we can get the global pictures: Analysis underway, no results yet.By combining the solar term analysis and the three flavor analysis we can get the global pictures: Analysis underway, no results yet.

Future:Future: The Global PictureThe Global Picture


Nucleon DecayNucleon Decay


SK-I SK-II

eff.(xBr.) (%)

atm. BG

candi-date

pe+ 44.6 43.5 0.20 0.11 0 0

p+ 35.5 34.7 0.23 0.11 0 0

pe+ 18.8 18.2 0.19 0.09 0 0

p+ 12.4 11.7 0.03 0.01 0 0

pe+ 8.1 7.6 0.08 0.08 0 0

p+ 6.1 5.4 0.30 0.15 0 2

pe+ 4.9 4.2 0.23 0.12 0 0

p+ 1.8 1.5 0.30 0.12 1 0

pe+ 2.4 2.2 0.10 0.04 0 0

p+ 2.8 2.8 0.24 0.07 0 0

pe+ 2.5 2.3 0.26 0.13 1 0

p+ 2.7 2.4 0.10 0.07 0 0

ne+ 19.4 19.3 0.16 0.11 0 0

n+ 16.7 15.6 0.30 0.13 1 0

ne+ 1.8 1.6 0.25 0.13 1 0

n+ 1.1 0.94 0.19 0.10 0 0

Charged lepton + meson modes

SK-I+II

IMB-3

KAM-I+II

exposure(kt ・ yr)

141 7.6 3.8

Total BG 4.7 47.9 11.5

candidates

6 32 9

6 candidates are observed while 4.7 events are expected from atmosphericB.G.For each mode,•p→+(30) : P(≥2)= 7.5%•p→+ : P (≥1)=34.9%•p →+(3) : P (≥1)=32.3%•n →+ : P (≥1)=34.3%•n →e+ : P (≥1)=31.6%no evidence of nucleon decay.

For all modes, efficiency and expected BG for SK-II is almost similar with SK-I,BG expectation is less than 0.5events.Courtesy Kaneyuki, ICRR

K-> Eff (%) BKG Obs.

+ P SK-1 37.0±0.4 188.9±5.7 198±14.1

SK-2 35.7±0.4 95.5±2.0 85± 9.2

Prompt tag

SK-1 7.2±1.6 0.16±0.05 0

SK-2 5.8±1.3 0.08±0.03 0

+0 SK1 6.2±0.5 0.43±0.13 0

SK2 4.8±0.4 0.31±0.10 0

No evidence of p→K+

Merged lifetime limit: 2.8 x1033 years @141 kton ・year(2.3x1033 years@92kton ・ year Phys.Rev.D 72, (2007) 052007)

Efficiency for SK-II are about 80% of SK-I.Expected backgrounds for K+→++prompt , K+→+0

are small.

summary of each analysis

p→K+ mode

Courtesy Kaneyuki, ICRR

Summary of nucleon decay search results

SK-I+II

SK-I+II

SK-I+II

10 34Courtesy Kaneyuki, ICRR

ConclusionsConclusions• still waiting for a galactic core-collapse supernova• still waiting for SN relic neutrinos to show up• SK electronics was upgraded successfully• now read out every hit• SK solar analysis has lower backgrounds <5.5 MeV in the

center of SK, start search for “upturn”• SK-III solar results consistent with SK-I and SK-II results

within statistical uncertainties• SK atmospheric neutrinos still dominate atmospheric

mixing angle constraints and contribute to mass splitting• SK atmospheric neutrinos start to constrain 1-3 mixing• full three-flavor analysis in preparation• SK has not yet found proton decay; sets the best limits

Documents

SN Physics Workshop September 17 th 2009 Michael Smy UC Irvine Super-Kamiokande Results