searching for oscillating neutrinos at t2k: results and ... · neutrino interactions n p ν l-l w+...
TRANSCRIPT
Searching for Oscillating Neutrinos at T2K: Results and
Prospects
Mark Hartz
(University of Toronto, York University)
University of Toronto Seminar, March 30, 2012
2University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
OutlineOutline● Review of neutrino oscillations
● Measuring θ13
● Overview of the T2K experiment and data collected● The Optical Transition Radiation Monitor
● Oscillation Analysis Overview● The extrapolation method● Neutrino production and interaction modeling● Near detector measurement● Far detector selection
● T2K Oscillation Results
● Other measurements of θ13
and prospects for T2K
● Conclusions
3University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Neutrinos in the Standard ModelNeutrinos in the Standard Model
Neutrinos:
Carry no electrical or color charge
Only interaction via the weak force
Very small mass compared to quarks, charged leptons
4University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Neutrino InteractionsNeutrino Interactions
n p
l-ν
l
W+
n n
Z
νl
νl
Charged Current (CC) Neutral Current (NC)
t
Weak interactions via the W or Z boson
For neutrino energies in this talk, scattering off nucleons is important
Weak interactions means very small cross section: ~10-38 cm2
Need massive detectors and large neutrino fluxes
5University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Neutrino SourcesNeutrino Sources
AtmosphericAtmospheric
SolarSolar
p
π+
μ+
ReactorReactor
AcceleratorAccelerator
νμ
νμ
νe
e+
Cosmic ray showers produce pions and muons that decay to neutrinos
Produced in fusion reactions inside sun.
Energy thresholds matter for experiments
8 B 8 Be∗ee
p+e - p+ 2He
...
...
...
pπ+
μ+
νμ
νμ
νe
e+
Proton beam from accelerator collides with fixed target
Most muons absorbed before decaying
~1-10 GeV
~ 10 MeV
Where do the neutrinos that experiments measure come from?
νe From β decay
of isotopes in nuclear reactors
6University of Toronto Seminar, March 30, 2012
Brief History of Neutrino OscillationsBrief History of Neutrino Oscillations
● 1957: Bruno Pontecorvo proposes theory of neutrino mixing analagous to mixing seen in neutral kaons
● Late 1960s:
● John Bahcall calculates expected νe flux from fusion reactions in sun
● Raymond Davis Jr.'s Homestake experiment measures solar νe flux -
1/3 of what is expected.
● Could be explained by neutrinos changing flavor (oscillations)
● 1998: Super-Kamiokande data is consistent with νμ oscillations to ν
τ
● 2001: SNO measures total (νe+ν
μ+ν
τ) solar neutrino flux and ν
e flux.
Consistent with Bahcall's calculation and Davis's experiment respectively
● 2000's: Various reactor, accelerator, solar and atmospheric neutrino experiments confirm the phenomenon
7University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Mixing with 3 NeutrinosMixing with 3 Neutrinos
e
=100 0
cos23
−sin 23
0sin23
cos23× cos13
0−sin13e
iCP
010
sin13e−iCP
0cos13
× cos12
−sin 12
0
sin12
cos12
0
001×1
2
3
● The three neutrino PMNS mixing matrix Uli:
● Probability to oscillate depends on energy (E), distance traveled (L), the mixing matrix U and the differences in the squared neutrino masses :
P=−4∑i j
ℜU i* U iU jU j
* sin2mij
2 L
4 E
2∑i j
ℑU i* U iU jU j
* sin mij
2 L
2E
● Three mixing angles: θ12
, θ13
, θ23
● δCP
phase allows for CP violation – neutrinos and anti-neutrinos mix differently
Solar and reactorAtmospheric and accelerator
8University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Knowledge of Oscillation ParametersKnowledge of Oscillation Parameters
νe ν
μ ν
τ
m232
m221
m32
m22
m12
m32
m22
m12
m32−m2
2
InvertedNormal
m232
m221Ambiguity in sign of
Two possible mass hierarchies
m122 ≈7.6×10−5 eV 2
m322 ≈2.4×10−3 eV 223≈45°
SK, K2K, MINOS
12≈34°SNO, KAMLAND, SK
1310° 90 % C.L. MINOS(2010), CHOOZ
CP=?
As of 2010
9University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Measuring θ13
With AcceleratorsMeasuring θ13
With AcceleratorsAccelerator based experiments - access to θ
13 through oscillations of
muon neutrinos to electron neutrinos:
P e=sin223sin2 213sin2 m322 L
4 Esubleading terms
νe appearance probability for
L=295 km and sin2(2θ13
)=0.1
Design experiment with νμ beam
peaked at first oscillation maximum
Search for νe appearance
Measuring θ13
is an important first step to searching for CP violation in νe
appearance
10University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Measuring δCP
in νe AppearanceMeasuring δ
CP in ν
e Appearance
∝ ∓sin 212sin 213sin 223sinCP
Sign flip for neutrino vs. antineutrino
Full appearance probability includes term that goes as sinδcp
:
Need non-zero value for θ13
Two ways to measure δCP
:
1. neutrino vs. antineutrino appearance probability
2. 1st vs 2nd oscillation maximum
11University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Measuring θ13
: Reactor NeutrinosMeasuring θ13
: Reactor Neutrinos
Beam of reactor νe with short baseline (~1 km). Oscillation probability:
P e e≈1−sin2213sin2 1.27m31
2 L
E
CHOOZ Experiment:
1 km from 2 reactors in Chooz, France
5 ton liquid scintillator detector
Collected 3600 events
No disappearance observed
Set upper limit:
sin22130.17 90 % C.L.
Phys. Lett. B 466, 415 (1999)
12University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Current Reactor ExperimentsCurrent Reactor ExperimentsNew reactor experiments employ near and far detectors to reduce flux uncertainties
All are now running
Daya Bay (China)
Double CHOOZ (France)
RENO (S. Korea)
Daya Bay ultimately has the best sensitivity for θ
13
Disappearance probability not sensitive to δ
cp.
13University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Current Reactor ExperimentsCurrent Reactor ExperimentsNew reactor experiments employ near and far detectors to reduce flux uncertainties
All are now running
Daya Bay (China)
Double CHOOZ (France)
RENO (S. Korea)
Daya Bay ultimately has the best sensitivity for θ
13
Disappearance probability not sensitive to δ
cp.
14University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
OutlineOutline● Review of neutrino oscillations
● Measuring θ13
● Overview of the T2K experiment and data collected● The Optical Transition Radiation Monitor
● Oscillation Analysis Overview● The extrapolation method● Neutrino production and interaction modeling● Near detector measurement● Far detector selection
● T2K Oscillation Results
● Other measurements of θ13
and prospects
● Conclusions
15University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
T2K (Tokai to Kamioka) ExperimentT2K (Tokai to Kamioka) Experiment
TokaiTokaiKamiokaKamioka
295 km295 km
● Experiment's immediate goals:
● Search for νe appearance:
● Precision νμ disappearance
J-PARC: 30 GeV protonbeam, design power of
750 kW
J-PARC: 30 GeV protonbeam, design power of
750 kW
Super-Kamiokande22.5 kton (fiducial)
water cherenkovdetector at 295 km
Super-Kamiokande22.5 kton (fiducial)
water cherenkovdetector at 295 km
P e≈sin223 sin 2 213 sin2 m322 L
4 E
ννμμ
P ≈1−sin 2 223 sin2 m322 L
4 E
16University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
The T2K CollaborationThe T2K Collaboration
CanadaCanadaTRIUMFTRIUMFU. AlbertaU. AlbertaU.B. ColumbiaU.B. ColumbiaU. ReginaU. ReginaU. TorontoU. TorontoU. VictoriaU. VictoriaU. WinnipegU. WinnipegYork. U.York. U.
FranceFranceCEA SaclayCEA SaclayIPN LyonIPN LyonLLR E. Poly.LLR E. Poly.LPNHE ParisLPNHE Paris
GermanyGermanyU. AachenU. Aachen
ItalyItalyIPNF, U. RomaIPNF, U. RomaIPNF, U. NapoliIPNF, U. NapoliIPNF, U. PadovaIPNF, U. PadovaIPNF, U. BariIPNF, U. Bari
JapanJapanICRR KamiokaICRR KamiokaICRR RCCNICRR RCCNKEKKEKKobe U.Kobe U.Kyoto U.Kyoto U.Miyagi U. Edu.Miyagi U. Edu.Osaka City U.Osaka City U.U. TokyoU. Tokyo
PolandPolandA. Soltan, WarsawA. Soltan, WarsawH. Niewodnicsanki,H. Niewodnicsanki,
CracowCracowT.U. WarsawT.U. WarsawU. Silesia, KatowiceU. Silesia, KatowiceU. WarsawU. WarsawU. WroklawU. Wroklaw
RussiaRussiaINRINR
S. KoreaS. KoreaChonnam N.U.Chonnam N.U.Dongshin U.Dongshin U.Seoul N.USeoul N.U
SpainSpainIFIC, ValenciaIFIC, ValenciaIFAE(Barcelona)IFAE(Barcelona)
SwitzerlandSwitzerlandU. BernU. BernU. GenevaU. GenevaETH ZurichETH Zurich
United KingdomUnited KingdomImperial C. LondonImperial C. LondonQueen Mary U.L.Queen Mary U.L.Lancaster U.Lancaster U.Liverpool U.Liverpool U.Oxford U.Oxford U.Sheffield U.Sheffield U.Warwick U.Warwick U.
STFC/RALSTFC/RALSTFC/DaresburySTFC/Daresbury
USAUSABoston U.Boston U.B.N.L.B.N.L.Colorado S. U.Colorado S. U.Duke U.Duke U.Louisiana S. U.Louisiana S. U.Stony Brook U.Stony Brook U.U. C. IrvineU. C. IrvineU. ColoradoU. ColoradoU. PittsburghU. PittsburghU. RochesterU. RochesterU. WashingtonU. Washington
~500 collaborators, 60 institutes, 12 countries
17University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
T2K OverviewT2K Overview
p π,K ν
120m120m0m0m 280m280m 295 km295 km
off-axisμ-mon
Decay volume
MUMON measures muons from pion decay
Beam on graphite target
3 magnetic horns focus positively charged hadrons
Off-axis far detector at 295 km: SK water cherenkov detector measures oscillated flux
30 GeV proton beam
On-axis INGRID detector measures neutrino rate, beam profile
Off-axis near detector: ND280 detector measures spectra for various neutrino interactions
Off-axis = narrow band beam
18University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Neutrino Interactions at T2KNeutrino Interactions at T2K
T2K beam peak energy
In region of interest for T2K:
Large contribution from charge current quasi-elastic (CCQE)
W+
n p
e-,μ-νe,ν
μ
T2K signal at SK
Significant CCπ component with additional pion in final state
NCπ0 is significant background mode:
Z
nn
νe,μ,τ
νe,μ,τ
π0 γγ Photons from π0 can
fake an electron
νμ
19University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
J-PARC AcceleratorJ-PARC Accelerator
● Located in Tokai village
● Completed in 2009
● Accelerator Design/Performance
● Design goal of 750 kW
● 30 GeV protons to neutrino beam line
● Reached 145 kW before earthquake
● 150 kW post earthquake
181 MeV LINAC181 MeV LINAC
3 GeV RCS3 GeV RCS
30 GeV Main Ring30 GeV Main Ring
T2K Neutrino BeamlineT2K Neutrino Beamline
20University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
J-PARC Neutrino Beam LineJ-PARC Neutrino Beam Line
21University of Toronto Seminar, March 30, 2012
Optical Transition Radiation MonitorOptical Transition Radiation Monitor Proton beam monitor built and operated by UofT, York U. and TRIUMF
Measure proton beam profile just before it hits the target
Optical transition radiation produced when protons cross thin foil (change in dielectric constant)
Paper submitted to Nuclear Instruments and Methods arXiv:1201.1922
22University of Toronto Seminar, March 30, 2012
Imaging OTR LightImaging OTR Light
Beam into page
The radiation level near the target is too high for a camera: 0.8e4 Sv/hr at full beam power
Transport the light through the shielding with 4 parabolic mirrors
Camera sits above the shielding
Optical system distorts the image - correct using images of a back-lit calibration foil
23University of Toronto Seminar, March 30, 2012
Correcting the DistortionCorrecting the DistortionDistorted image of the hole pattern After correcting for the known hole
positions (black points)
After corrections, OTR monitor measures the mean beam position to better than 500 microns
Important to know precisely for the neutrino flux prediction
Example Beam Image
24University of Toronto Seminar, March 30, 2012
Real Time Beam MonitoringReal Time Beam MonitoringOTR online dislay – first image of beam to the T2K beam line after the earthquake
The OTR monitor measurements are used to “tune” the beam orbit to hit the center of the target
Target Edge
Target Center
25University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ND280 (Near) Off-axis DetectorND280 (Near) Off-axis Detector● 0.2 T UA1 magnet
● Fine Grained Detectors (FGD)● Scintillator bars and water targets
(FGD2)● Interaction mass and tracking
● Time Projection Chambers (TPC) – momentum and dE/dx measurements
● P0D π0 detector – measures NC π0 rates
● Electromagnetic calorimeters – measure EM showers from inner detectors
● SMRD muon detector installed in the magnet yoke – detect muons, cosmics trigger, side muon veto
Used in current analysis
Important for future analyses
26University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
SK (Far) DetectorSK (Far) Detector
● 50 kton (22.5 kton fiducial volume) water cherenkov detector
● ~11,000 20'' PMT for inner detector (ID) (40% photo coverage)
● ~2,000 outward facing 8'' PMT for outer detector (OD): veto cosmics, radioactivity, exiting events
● Good reconstruction for T2K energy range
IDIDODOD
Cherenkov light produces a ring detected by the PMTs
27University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Particle Identification at SKParticle Identification at SK
Muons
Muon follows a straighter trajectoryRing with sharper edges
Electrons
Electrons shower
Fuzzy ring edges
Neutral Pions
γs from π0 decays shower and look like electrons
Differentiate by finding second ring
MC MCMC
28University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Data CollectedData Collected
● 1.43 x 1020 POT for SK analysis
● Reached continuous running at 145 kW
● increase protons per pulse, increase rep rate
● 2% of the design goal for T2K
Run 1
Run 2
6 → 8 bunches per pulse
Rep period: 3.64 s → 3.2 s → 3.04 s
29University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
OutlineOutline● Review of neutrino oscillations
● Measuring θ13
● Overview of the T2K experiment and data collected● The Optical Transition Radiation Monitor
● Oscillation Analysis Overview● The extrapolation method● Neutrino production and interaction modeling● Near detector measurement● Far detector selection
● T2K Oscillation Results
● Other measurements of θ13
and prospects
● Conclusions
30University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Typical Long Baseline Osc. AnalysisTypical Long Baseline Osc. Analysis
Start with models of the neutrino flux production and interaction cross sections (may be tuned to external experiments)
The tuned flux and cross section modelsare used to predict the event rate at the
far detector under various oscillationhypothesis. Data constrains the
allowed parameter space for oscillations.
Tune the models with measurementsof the “not oscillated” event rate at
the near detector. Also reduces systematicuncertainties from the model
FLUKA and GEANT3 used for flux simulation
NEUT used to model neutrino interactions
T2K Analysis:
Measure the inclusive charge current interaction rate and take ratio with model based prediction:
RND , Data /RND
,MC
N SKexp=RND
, Data×N SKMC /RND
,MC
Expected number of events at SK is corrected with the ratio measurement
Model dependent errors enter hereare reduced in the ratio
31University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Neutrino Flux ModelingNeutrino Flux Modeling
νμ
μ
π, K
Focused by magnetic horns
Protons interact in T2K target, produce hadrons
Proton beam monitors measure beam properties
Decays produce neutrinos
Muon Monitor
ND280SK
Flux Simulation:● Proton beam monitor measurements as inputs● In target hadron production:
● NA61 experimental (at CERN) data to model π± production● Kaon production, other hadron interactions – model with FLUKA
● Out of target interactions, horn focusing, particle decays● GEANT3 simulation● Interaction cross sections are tuned to existing external data
32University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
NA61 ExperimentNA61 ExperimentLarge acceptance spectrometer and time-of-flight detectors
30 GeV proton beam to match T2K
Two target types:
1) 0.04 λ “thin target”
2) T2K replica “long target”
Pion production from thin target used in this analysis
Good TOF and dE/dx performance allows for particle separation
33University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Flux Tuning with NA61 ResultsFlux Tuning with NA61 ResultsNA61 Measure 30 GeV proton cross section on carbon:
prod=229.3±9.2 mb
Measure differential π± production multiplicity
Phys. Rev. C 84 034604 (2011)
p [GeV/c] p [GeV/c]
SK νμ Flux Tuning
The flux prediction is tuned to the production measured by NA61
~10% effect near the oscillation maximum
Also tuning for particle interaction lengths in GEANT3
34University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Neutrino Flux PredictionNeutrino Flux Predictionallkaon parentspion parentsmuon parents
allkaon parentspion parentsmuon parents
νe at SKν
e at SKν
μ at SKν
μ at SK
Region of oscillation maximum
● Muon neutrino flux around oscillation maximum predominantly from pion decays● Intrinsic electron neutrino flux in beam from muon and kaon decays ~1% of total flux below 1 GeV
● Dominant source around oscillation maximum is from muon decays
++
+e+e
Flux depends on pion production
35University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Flux UncertaintyFlux Uncertainty
Error Sources RND
μ,MC NSK
MC NSK
MC/RND
μ,MC
Pion Production 5.7% 6.2% 2.5%
Kaon Production 10.0% 11.1% 7.6%
Other Hadron Int. 9.7% 9.5% 1.5%
Beam Direction, Alignment, Horn Current 3.6% 2.2% 2.3%
Total 15.4% 16.1% 8.5%
Percent Errors from Flux Uncertainties (θ13
=0)
Precise measurement of the proton beam direction by the OTR monitor keeps the uncertainty on the flux small
Uncertainties on hadronic interactions dominate:
1. Pion production: systematic uncertainties from NA61
2. Kaon production: from comparison of FLUKA to external data
3. Secondary nucleon production: comparison of FLUKA to external data
4. Hadron interaction probabilities: from external measurements of π, p, K cross sections
Reduced in future analyses by using NA61 kaon production data
36University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ν Cross Section Uncertaintiesν Cross Section Uncertainties
NEUT comparison to MiniBooNE NCπ0 diff. cross section
Phys. Rev. D 81, 013005 (2010)
Nominal cross sections from NEUT model (used for SK atmospheric analyses)
Cross section uncertainties from:
1. Comparisons of models to data: MiniBooNE, SciBooNE, SK atmospheric
2. Variation of model parameters: Nucleon form factors, nuclear binding energy, etc.
3. Comparisons between models: relativistic fermi gas vs. spectral function for nuclear model
37University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ND280 Inclusive νμ AnalysisND280 Inclusive ν
μ Analysis
● Select negative μ-like negative charge tracks originating in FGDs and tracked by TPCs
● High Purity: 90% νμ CC and 50% CCQE
● For 2.88x1019 POT (Run 1): 1529 events
TPC1TPC1 TPC2TPC2 TPC3TPC3
FGD1FGD1 FGD2FGD2
νμ
No selection on additional particles
μ-
38University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ND280 Inclusive νμ SampleND280 Inclusive ν
μ Sample
Data and MC prediction for 2.88x1019 POT:
RND , Data /RND
,MC=1.036±0.028stat.−0.0370.044 det. sys.±0.038 phys.model
Comparison to POT normalized MC:
NA61+FLUKA flux model
NEUT neutrino interaction model
Dominant sources: dE/dx and FGD/TPC matching
39University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
νe Signal & Background at SKν
e Signal & Background at SK
νμ → ν
ee-
p (undetected)
Oscillation Signal:
νe
e-
p (undetected)
Beam νe Background:
MC
Identical for given neutrino energy.
Beam background has harder spectrum
νl
N+others (undetected)
γ
γπ0
MC
Can be removed by identifying second photon ring
Neutral Current π0:
40University of Toronto Seminar, March 30, 2012
SK Event SelectionSK Event Selection
νe Event Selection ν
μ Event Selection
Fully containted in inner detector fiducial volume
Visible energy > 100 MeV Visible energy > 30 MeV
Number of rings = 1
PID identifies ring as electron like PID identifies ring as muon like
No decay electrons detected -
π0 reconstructed mass < 105 MeV/c2 -
Reconstructed ν energy < 1250 MeV -
- Reconstructed μ momentum > 200 MeV/c
● Event selection criteria frozen before data collection to avoid bias
● νe: Light pattern fit of π0 mass assuming 2 γ rings, cut removes 75% NC
bgnd.● Decay electron cut removes electrons from muon decays (events with muons)
41University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
νe Selection at SK, Cont.ν
e Selection at SK, Cont.
Signal Efficiency = 66%Background Rejection: 77% for beam ν
e
99% for NC
N SKexp=RND
, Data×N SKMC /RND
,MC
Recall expected number of events is rescaled by Data/MC rate measured at ND280:
Sources NSK
exp
NC background 0.6
Beam νe background 0.8
Osc. through θ12 0.1
νμ CC background 0.03
Total 1.5±0.3
Breakdown of errors on next slide
42University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Systematic Uncertainty SummarySystematic Uncertainty Summary
Error Source sin2(2θ13
)=0 sin2(2θ13
)=0.1
Beam flux 8.5% 8.5%
ν cross sections 14.0% 10.5%
ND280 detector +5.6 -5.2
+5.6 -5.2
SK detector 14.7% 9.4%
ND280 statistics 2.7% 2.7%
Total +22.8 -22.7
+17.6 -17.5
% %
%%
Room for improvement in the flux, cross section and SK detector systematic errors
Flux and cross section errors can be improved by more ND280 measurements
43University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
OutlineOutline● Review of neutrino oscillations
● Measuring θ13
● Overview of the T2K experiment and data collected● The Optical Transition Radiation Monitor
● Oscillation Analysis Overview● The extrapolation method● Neutrino production and interaction modeling● Near detector measurement● Far detector selection
● T2K Oscillation Results
● Other measurements of θ13
and prospects
● Conclusions
44University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
SK νe Candidate SampleSK ν
e Candidate Sample
After νe selection is applied → 6 candidate events remain!
Recall, background expectation is 1.5 ± 0.3 events
6 Events
45University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Interpretation for sin2(2θ13
) Interpretation for sin2(2θ13
)
(fixing sin2(2θ23
) = 1.0, Δm2
23=2.4x10-3 eV2)
For sin2(2θ13
)=0 [sin2(2θ23
) = 1.0, Δm2
23=2.4x10-3 eV2 ],
probability to observe ≥6 events = 0.007
Published in Phys. Rev. Lett. 107, 041801 (2011)
46University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Measuring νμ Disappearance Measuring ν
μ Disappearance
T2K can measure θ23
and Δm2
32 through ν
μ disappearance:
P ≈1−sin 2 223 sin2 m322 L
4 E
Similar to νe appearance analysis, but now select muon like events at SK
31 candidate events
104 expected if no oscillations
T2K can make a strong measurement even with only 2% of design exposure
47University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Interpreting the Disappearance ResultInterpreting the Disappearance Result
Can produce an allowed region in the sin2(2θ23
) and |Δm2
32| plane
Results are consistent with those from MINOS and SK atmospheric
Systematics are already very important for this analysis (solid blue vs. dashed blue)
With just 2% of exposure, already becoming competitive with existing measurements
Phys. Rev. D85 (2012) 031103
T2K Stat+Syst 90% CLT2K Stat only 90% CLMINOS 2011 90%CLSuper-K 90% CLSuper-K L/E 90% CL
T2K Stat+Syst 90% CLT2K Stat only 90% CLMINOS 2011 90%CLSuper-K 90% CLSuper-K L/E 90% CL
48University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
OutlineOutline● Review of neutrino oscillations
● Measuring θ13
● Overview of the T2K experiment and data collected● The Optical Transition Radiation Monitor
● Oscillation Analysis Overview● The extrapolation method● Neutrino production and interaction modeling● Near detector measurement● Far detector selection
● T2K Oscillation Results
● Other measurements of θ13
and prospects
● Conclusions
49University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Other θ13
MeasurementsOther θ13
Measurements
● Since T2K released our result on June of 2011, other experiments have released measurements indicating non-zero θ
13
● MINOS – Also uses accelerator based muon neutrino beam, but L/E is not as well tuned as T2K
● Double CHOOZ – Reactor neutrino measurement using only their far detector
● Daya Bay – Reactor neutrino measurement using both near and far detectors
sin2 213=0.086±0.041stat ±0.030 syst arXiv:1112.6353
50University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Daya Bay MeasurementDaya Bay MeasurementRate measurements at both near and far halls
Measure non-zero θ13
at 5.2σ
sin2213=0.092±0.016 stat ±0.005syst
Daya Bay and T2K measurements are consistent
Remember, reactor measurements aren't sensitive to CP phase
Submitted to PRLarXiv:1112.6353
51University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
T2K and Large θ13
T2K and Large θ13
● A large value of θ13
is a good thing – can have CP violation in the lepton
mixing
● With 5 years of running at 750 kW, expect ~200 electron neutrino signal events
● Can we measure δCP
with T2K and Daya Bay (reactors)?
Illustrative, not a T2K plotAssume 10% T2K uncertainty and 3% Daya Bay uncertainty
If lucky, can start to disfavor some values of δ
CP
T2K has the capability to run in anti-neutrino mode – can look for CP violation independently of reactor measurements
3% Reactor Uncertainty
10% T2K Uncertainty
52University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
The Full Appearance ProbabilityThe Full Appearance Probability
Matter effect term that depends on the sign of Δm2
31
The full electron neutrino appearance has a matter effect term that flips sign for normal vs. inverted mass hierarchy
31=1.27m31
2 L
E
53University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
What About Mass HierarchyWhat About Mass Hierarchy
For large θ13
, T2K and NOvA may
resolve the mass hierarchy
Hierarchy resolved at 95% confidence for region to the right of the curves
Blue dotted line corresponds to T2K 6 years at 750 kW x2 and 1.2 MW for NOvA
Need to work hard to meet that exposure...
...but may be an argument for running longer in neutrino mode
NOvA plus a T2K successor (Hyper-K) should be able to determine the mass hierarchy
Courtesy NOvA
Daya Bay and 2sin2θ23
=1
54University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
T2K StatusT2K Status
● The J-PARC accelerators are now up and running, and the neutrino beam line is receiving beam
● So far beam power has reached 150 kW
● If all goes well, may double the data set in first half of 2012
● Analysis improvements are also ongoing
● Reduction of SK reconstruction uncertainties by improvements to studies on atmospheric data control samples
● Reduction of flux uncertainties with kaon production data from NA61
● Reduction of cross section uncertainties by tuning the NEUT models on existing MiniBooNE, SciBooNE and K2K measurements
● Measurement of the CC inclusive spectrum (muon p,θ bins) at ND280 to better constrain the flux and cross section models
● Studies of binned SK analyses (reconstructed neutrino energy, or electron p,θ bins)
Expect new and interesting results from T2K in 2012
55University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ConclusionConclusion● T2K has searched for ν
μ → ν
e oscillations based on 1.43 x 1020 p.o.t.
(2% exposure of T2K’s goal)
● Expected number of background events is 1.5 ± 0.3 (sin22θ13
= 0)
● 6 candidate events are observed
● Under θ13
=0 hypothesis, the probability to observe 6 or more
candidate events is 0.007 (equivalent to 2.5σ significance)
● Published in Phys. Rev. Lett. 107, 041801 (2011)
● T2K also measures the νμ disappearance and sees a deficit consistent with
SK atmospheric data and MINOS
● Non-zero value for θ13
confirmed by Daya Bay - CP violation measurement
is possible
● T2K has recovered from the earthquake and is collecting data – expect new results with improved analysis methods in 2012
57University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Example Candidate Event DisplayExample Candidate Event Display
58University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ν Cross Section Unc. Summaryν Cross Section Unc. Summary
Dominant source is uncertainty on pion final state interactions
Studied by adjusting NEUT microscopic pion cross section model and comparing to pion cross section data
Error Sources NSK
exp
CCQE low energy 3.1%
CC 1π 2.2%
CC Coherent π 3.1%
CC Other 4.4%
NC 1π0 5.3%
NC Coherent π 2.3%
NC Other 2.3%
σ(νe) 3.4%
FSI 10.1%
Total 14.0%
From studies of SciBooNE data
From studies of MiniBooNE data
Relativistic fermi gas model vs effective spectral function
59University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Hyper-KHyper-K990 kton water cherenkov detectorSite 8 km from SKSegmented design – 5x2 segments1750 m overburden
Call for international working group at Neutrino 2012?
Budget request to MEXT at the end of 2012
60University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Hyper-K SensitivityHyper-K Sensitivity
● Assuming 5% systematic error in event rate and negligible error in the spectrum shape
● Spectrum shape is important for differentiating δ near 0 and π
● Sensitivity is improved with known mass hierarchy
61University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
The Full Appearance ProbabilityThe Full Appearance Probability
Depends on the sign of the mass difference squared
The full electron neutrino appearance has a matter effect term that flips sign for normal vs. inverted mass hierarchy
31=1.27m31
2 L
E
62University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Measuring Mass HierarchyMeasuring Mass Hierarchy
NOvA: www-nova.fnal.gov/plots_and_figures/plots_and_figures.html
Like the CP phase, matter effects give different neutrino and antineutrino oscillation probabilities
For some regions, can determine the mass hierarchy without tightly constraining the CP phase
T2K can help NOvA differentiate in the overlapping region
63University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Comparison to MINOS MeasurementComparison to MINOS Measurement
T2K Allowed
MINOS Allowed
MINOS measures θ13
as well
Prefer a non-zero value and consistent with T2K
Phys.Rev.Lett. 107 (2011) 181802
64University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
SK Detector Systematic Unc.SK Detector Systematic Unc.
Evaluated on SK atmospheric sample
Control sample described on next slide
Uncertainty on signal
Uncertainty on background
Data driven evaluation of systematic uncertainties
65University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
SK Data SampleSK Data Sample
● SK synchronized to beam timing using GPS
● SK events fully contained in the ID show clear beam time structure
● In total, 121 FC events
Step 1:
66University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Vertex DistributionVertex DistributionVertex distribution of candidate events:
● KS test of R2 distribution yields p-value = 0.03 (~1-20% for other distributions)● Only one event seen outside fiducial volume that passes all other cuts
● If beam related background from outside FV, expect more events in this region
● OD event distributions show no indication of contamination from outside ID
67University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
More on Vertex DistributionsMore on Vertex DistributionsOD event vertex distributions:
OD events contained in OD
OD events entering ID
No significant data excess in OD samples
ID vertex distribution and MC with interactions simulated out to 550 cm from ID wall
If outside source, expect more data at large R2 outside FV
68University of Toronto Seminar, March 30, 2012 M. Hartz, UofT/YorkU
Effects of the EarthquakeEffects of the Earthquake
J-PARC
J-PARC, on the east coast of Japan, experienced a high seismic intensity
Luckily the tsunami did not reach the lab
No T2K or J-PARC personnel were injured
Earthquake damage to the roads and service pipes/cables
No significant damage to beam tunnels or detector halls
Recovery of the accelerator chain is underway
We may have beam to the neutrino beam line as early as the end of this year
Hope to start taking physics data at the beginning of 2012
69University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ND280 νe AnalysisND280 ν
e Analysis
● ND280 νe rate measurement using similar method as ν
μ measurement, same
data set (Run1, 2.88x1019 POT)
● Use TPC dE/dx measurement to select electrons instead of muons
● Reject events with second track that makes Minv
< 100 MeV/c2 with electron
candidate → reject photon conversions
Background sources:
1) Photons produced outside FGD that convert in the FGD (Out of FGD)
2) νμ interactions with γ from π0
that converts
3) νμ interactions where muon is
misidentified as an electron
200 600 1000 1400 1800
70University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ND280 νe Analysis ResultsND280 ν
e Analysis Results
Fit data with signal and background templates → extract ν
e signal events
Data based template for misid μ
MC based templates for Out of FGD and ν
μ FGD constrained by
control samples
N sel e=7.8±5.5 stat.±2.1 syst.
N eN
=N sel eeN sel
=1.0±0.7stat.±0.3 syst.%
[ N eN ]data÷[ N e
N ]MC=0.6±0.4 stat.±0.2 syst.
Dominated by uncertainty from fitting the photon conversion sample
νe Events from fit:
71University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Reconstructing the ν EnergyReconstructing the ν Energy
np
e-
νe
● Only lepton in final state is reconstructed● Can determine neutrino energy with assumptions:
● Neutrino direction is known (beam direction)● Recoil nucleon mass is known● Target nucleon is at rest – not exactly true, adds smearing to
energy
EQE=
2M nE e−M n2me
2−M p2
2 [M n−E eE e2−me2 cose]
θe
Fully reconstructed
Direction known (beam direction)
Assumed at rest in lab frame
72University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Proton Beam TargetingProton Beam TargetingProton beam orbit tuned with SSEM, ESM, OTR proton beam monitors and MUMON
Fit to SSEM and OTR monitors gives position at target
Uncertainty: 0.4 mm in x, 0.6 mm in y → ~0.3 mrad uncertainty in ν beam
1 mrad shift of neutrino beam = 2% shift in ν peak energy
MUMON measured beam profile center stable within ±0.3 mrad
73University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Next Steps for the νe AnalysisNext Steps for the ν
e Analysis
Aim to firmly establish νe appearance and better
determine the θ13
mixing angle
● Resume experiment and collect more data● Recovery works in progress
● No insurmountable problems found
● Resumption of J-PARC activity including accelerator complex and neutrino facility by December, 2011
● Neutrino facility ready by November
● Analysis improvements
● New analysis methods using νe shape information (e.g. recon. energy)
are under development
● ND280 measurements of νμ CCQE spectrum, improvements to beam ν
e,
NCπ0 measurements underway to better constrain flux and neutrino cross sections
74University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Vertex Distribution ProbabilitiesVertex Distribution Probabilities
● One must be careful when choosing distributions to test. If data guides choice, difficult to assign probability.
● Choose distributions that we would check regardless of what is seen in data
● Can use KS test for the probability of the cumulative distribution
● Concern about KS test for low statistics sample, so calculate probabilities from distributions of 100,000 toy MCs
● Probabilities for distributions of interest:
For distributions relativeto ID wall, it is more
natural to include all 7FC events
KS Toy MC Probabilities 6 FCFV Events 7 FC Events
Distance to nearest wall 3.7% 20.2%
From wall || to beam 0.1% 1.3%
To wall || to beam 1.2% 5.3%
R2 3.1% 9.4%
75University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Candidate Sample ChecksCandidate Sample Checks
Reconstructed cos(θbeam
):
θbeam
= Lepton angle relative to beam
direction
Reconstructed θbeam
vs. lepton
momentum:
76University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Contours in Δm2Contours in Δm2
sin2(2θ13
) allowed region as function of |Δm2
23|
(fixing sin2(2θ23
) = 1.0, δCP
=0)
77University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
KS + Toy MC ExampleKS + Toy MC Example
Cumulative distribution to extract maximum cumulative distance
p-value comes from toy MCs' distribution of maximum cumulative distance
78University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
MC Study of Sources Outside IDMC Study of Sources Outside ID
Generated MC with events produced in material up to 550 cm outside of ID wall
No significant contribution to FCFV sample simulated sources outside of ID
79University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Distance to ID WallDistance to ID Wall
Expected and observed distance to ID wall
Prediction includes events with true vertex outside of ID
80University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Measuring δ in νe AppearanceMeasuring δ in ν
e Appearance
∝ ±sin12 sin13sin 23sin
Sign flip for neutrino vs. antineutrino
Full appearance probability includes term that goes as sin(δ):
Need non-zero value for all three mixing angles including θ
13
Two ways to measure δ:
1. neutrino vs. antineutrino appearance probability
2. 1st vs 2nd oscillation maximum for neutrino mode
81University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Final State InteractionsFinal State Interactions
● Pions produced in the ν interactions can interact in the target nucleus:
● Absorption - no pion in final state
● Production - additional pions in the final state
● Charge exchange - change of pion charge
● Microscopic (internal to nucleus) pion interaction model employed in NEUT
● Tune microscopic model to reproduce macroscopic pion scattering data
Tuning: vary microscopic mean free path for different interaction types and vary models
Tuned (dotted lines) in much better agreement with data
82University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Decay Volume and MUMONDecay Volume and MUMONMuon production vertex simulation
● Pions (and other particles) decay in 100 m long decay volume:
● MUMON muon monitor
● Measures muons from pion decays
● Si PIN photodiodes sensitive at low intensity, but radiation damage
● Ionization chambers less suseptible to rad. damage
● Measure beam shape and direction
μ+
π+
νμ
83University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
ND280 TrackerND280 Tracker
● Neutrino target: 2.2 tonnes of material (including water targets)
● Tracking of particles
TPCTPC FGDFGD
● Track charged particles in magnetic field
● 10% momentum resolution at 1 GeV/c
84University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
SK νe Prediction BreakdownSK ν
e Prediction Breakdown
sin2(2θ13
) = 0
85University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Particle ID with TPC dE/dxParticle ID with TPC dE/dx
● Cut on TPC dE/dx with muon hypothesis to select muons
● Cut on TPC dE/dx with electron hypothesis to exclude electrons
Before Cut
After Cut
∣E ∣2.5, ∣E e∣2.0 E=dE /dx PullTwo dE/dx Cuts:
86University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Parent Pion Phase SpaceParent Pion Phase Space
Plots show the p-θ distribution of parent pions contributing to the ND280 (upper left), SK nu_e background (upper right) and SK nu_e signal (lower left) samples. Plots are normalized to by the sample size, so the z axis is sample fraction. θ is the polar angle relative to the beam direction
87University of Toronto Seminar, March 30, 2012 M. Hartz (UofT/YorkU)
Why Off-axis?Why Off-axis?
● Pion decay kinematics:
● In pion direction, neutrino energy proportional to pion momentum
● At non-zero angles, weak dependence on pion momentum
● 2.5° off-axis angle gives narrow band beam peaked at the first oscillation maximum
● More statistics in the oscillation region
● Less feed-down from backgrounds at higher energy
OA2.5°
Idea originally developed for long baseline proposal at BNL (E889)