Download - νμ Results from the MINOS experiment
νμ Results from the MINOS experiment
Nicholas DevenishRecontres de Blois 2010
The MINOS Experiment
Antineutrino Disappearance
Recent Results
➤
Summary
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Main Injector Neutrino Oscillation Search
735 km
The oscillated neutrino beam is measured by the Far Detector, at a mine in Minnesota.
Neutrinos are produced by the NuMI beam. The beams composition and energy are measured by the Near Detector
31 Institutions, 140 Physicists
Measurements and limits include:
Δm232, sin22θ32, Δm232, sin22θ32, θ13, sterile neutrinos, CPT
conservation, cross-sections...
3
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Physics Goals
‣ With three active neutrinos there are two independent mass splittings
‣ MINOS is most sensitive to the larger of these and θ23 through νμ disappearance
€
Δmatm2
€
Δmsol2€
Δmatm2 ≈ Δm32
2 ≈ 2.4 ×10−3 eV2
€
Δmsol2 ≈ Δm21
2 ≈ 8.0 ×10−5 eV2
€
P νµ →νµ( ) =1− sin2 2θ23( )sin2 1.27Δmatm2 LE
⎛
⎝ ⎜
⎞
⎠ ⎟
(Monte Carlosin22θ = 1.0,
Δm2 = 3.35x10-3 eV2)
Unoscillated
Oscillated
4
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Near Detector‣At Fermilab, IL‣282 Planes‣980 Ton
MINOS Detectors
B
Far Detector
Near DetectorPlanes of 2.54cm Steel and 1cm Scintillator
Alternating scintillator planes in perpendicular directions
for 3D event reconstruction
Toroidal magnetic field allows charge sign determination
Both detectors are functionally equivalent, in order to reduce systematics.
Far Detector‣At Soudan, MN‣486 Planes‣5400 Ton
5
Recent Results➤
Summary
Antineutrino Disappearance
The MINOS Experiment
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
νe Appearance Results
for δCP = 0, sin2 2θ23( ) = 1,
Δm322 = 2.43×10−3 eV2
sin2 (2θ13) < 0.12 normal hierarchysin2 (2θ13) < 0.20 inverted hierarchyat 90% C.L.
arXiv:1006.0996v1 [hep-ex]
‣ Looking for appearance of νe in the νμ beam
‣ Discriminate against NC events using 11 shape variables in an Artificial Neural Network
‣ Find Limits:
7
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Neutral Current Results‣ Neutral Current event rate
should not change in standard 3 flavour oscillations
‣ A deficit in the Far event rate could indicate mixing to sterile neutrinos
‣ νe CC events would be included with the selected NC sample, so result depends on the possibility of νe appearance
Expect:Observe:
757 Events802 Events
No deficit of NC Events
fs ≡Pνµ→νs
1− Pνµ→νµ
< 0.22 (0.40) at 90% C.L.no (with) νe appearance
8
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
νμ Disappearance Results
‣ Last published results 2008 (PRL 101:131802,2008)
‣ Additional data 3.4×1020→7.2×1020 POT
‣ Many analysis improvements including:
• Updated simulation and reconstruction• New selection with improved efficiency• Improved shower energy resolution• Separate data into bins of energy
resolution• Smaller systematic uncertainties
† Super-Kamiokande Collaboration (preliminary)
Δm2 = 2.35−0.08+0.11 ×10−3eV2
sin2 (2θ) > 0.91 (90%C.L.)
Best fit to oscillations at:
†
9
Antineutrino Disappearance➤
Summary
The MINOS Experiment
Recent Results
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Antineutrinos - Motivation
‣ Previously we presented an antineutrino analysis based on 3.2×1020 of neutrino mode data
‣World measurements of antineutrino disappearance are still limited
‣ MINOS magnetised detectors allow direct measurement of the charge sign of the muon
‣We recently took 1.71×1020 of data with NuMI in antineutrino mode
11
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Making Antineutrinos
120 GeV protons from MI
Focusing Horns2 m
675 m15 m 30 m
€
νµ = 91.7%ν µ = 7.0%
νe +ν e =1.3%
Target
Neutrino modeHorns focus π+, K+
Decay Pipe
π-
π+
νµνµ
Monte Carlo ‣With NuMI in neutrino mode only 7% of the beam is antineutrino
‣ These are of a higher average energy, making it harder to measure oscillations
12
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Antineutrino Mode
π+
π-
Target Focusing Horns Decay Pipe
2 m
675 m
νµ
15 m 30 m
νµ120 GeV protons from MI
Monte CarloNeutrino modeHorns focus π+, K+
€
νµ = 91.7%ν µ = 7.0%
νe +ν e =1.3%
Monte CarloAntineutrino modeHorns focus π-, K-
€
ν µ = 39.9%νµ = 58.1%
νe +ν e = 2.0%
13
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Track Charge Sign / Momentum (c/GeV)1 0.5 0 0.5 1
pot
1810×
Even
ts /
ton
/ 1
0
10
20
30
40
DataMC Expectation
Near Detector PoT19 10×8.65
Antineutrino Running
MINOS Preliminary
CC/NC Separation Parameter0 0.2 0.4 0.6 0.8 1
pot
1810×
Even
ts /
ton
/ 1
110
1
10
0 0.2 0.4 0.6 0.8 1
110
1
10 DataMC ExpectationTotal Background
Near Detector PoT19 10×8.65
Antineutrino Running
MINOS Preliminary
Selecting Antineutrinos
‣ Events in-time with the spill‣ In the fiducial volume‣ At least one reconstructed track‣ Positive reconstructed charge
Basic Selection
CC/NC SeparationUsing a k-Nearest-Neighbour algorithm - data is compared to monte-carlo events in four parameters:
‣ Track Length‣ Mean energy of track hits‣ Energy fluctuations along the track‣ Transverse track profile
Accept
Accept
14
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Reconstructed Energy (GeV)0 5 10 15 20 25Se
lect
or P
erfo
rman
ce (%
)
0
20
40
60
80
100
MINOS PreliminarySimulated Far Detector
NC Contamination Contaminationµ
Selection Efficiency
Selection Efficiency and Contamination
‣ High energy νμ contamination does not affect the oscillation result
‣ Highest purity is in the oscillation region
Signal Bkgd.
0-6 GeV 106 1.9
6-20 GeV 38 4.3
> 20 GeV 8 3.0Monte Carlo
15
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Results: Far Detector
‣ No Oscillation Prediction: 155 Events
‣Observe: 97 Events
‣ Disfavour no-oscillations at 6.3σ
‣ Best fit to oscillations:
Δm2 = 3.36−0.40+0.45 ×10−3eV2
sin2 (2θ) = 0.86 ± 0.11Reco. Energy (GeV)
FD E
vent
s/G
eV0
10
20
30
0 5 10 20 30 40 50
MINOS Preliminary
running, Far Detectorµ POT MINOS 20 10×1.71
MINOS dataNo oscillationsBest oscillation fitBackground
16
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Comparison with νμ Result
‣ Contours include the effects of systematic errors
)(22) and sin(22sin0.5 0.6 0.7 0.8 0.9 1
)2 e
V-3
| (10
2m
| and
|2
m|
2
3
4
5
6-310×
MINOS Preliminary
-modeµ POT 20 10×1.71
runningµMINOS 90%µMINOS 68%µMINOS
FitµBest POT20 10×1.71
90%µMINOS 68%µMINOS
FitµBest POT20 10×7.24
17
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
The Future‣With more antineutrino running, MINOS can quickly improve
the precision by a significant amount
‣ Doubling out current six-month data set would decrease the error by ~30%
18
)(22) and sin(22sin0.5 0.6 0.7 0.8 0.9 1
)2 e
V-3
| (10
2m
| and
|2
m|
234567
-310×
MINOS Preliminary90% MC Sensitivity
POT20 10×1.7 POT20 10×3.5
POT20 10×5 POT20 10×7
Data 90% C.L.µ
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Summary
‣With 7×1020 POT of neutrino beam, MINOS finds:
‣With 1.71×1020 POT of antineutrino beam, MINOS finds:
• Muon neutrinos disappear:
• NC event rate is not diminished:
• Electron-neutrino appearance is limited:
Δm2 = 2.35−0.08+0.11 ×10−3eV2,
sin2 (2θ) > 0.91 (90%C.L.)
fs < 0.22(0.40) at 90% C.L.
sin2 (2θ13) < 0.12 (0.20) at 90% C.L.
Δm2 = 3.36−0.40+0.45 ×10−3eV2,
sin2 (2θ) = 0.86 ± 0.11
• Muon antineutrinos also disappear with:
‣We look forward to taking more anti-neutrino data!
19
Backup Slides➤
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
The NuMI Beam‣ Neutrinos at the Main Injector
‣ 120 GeV Protons incident on a thick, segmented graphite target produces a spray of hadrons
‣ Magnetic horns can focus either sign by reversing the direction of current
‣ Can enhance the νμ flux by focusing π+, K+ (or vice versa)
‣ Adjustable energy by moving the target relative to the horns
π-
π+120 GeV p’s from MI
Target Focusing Horns
Evacuated Decay Pipe
2 m
675 m
νµνµ
1.5 m 3 m
Low EnergyMedium Energy
Monte Carlo
21
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Beam Performance
Run I Run II Run IIIHigh
Ene
rgy
Antin
eutrino
s
Run IV
€
3.21×1020 POT νµ modePrevious Analyses
22
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Beam Performance
Run I Run II Run IIIHigh
Ene
rgy
Antin
eutrino
s
Run IV
€
7.24 ×1020 POT νµ modeCurrent νµ Analysis
€
1.71×1020 POTν µ mode
23
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Previous Antineutrino Results
‣ Analysed 3.2×1020 POT of neutrino-mode running
‣ Antineutrinos are on average of higher energy, moving the spectrum peak out of the expected region of oscillations
‣ Very poor sensitivity to oscillations
24
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
νμ Near Detector Data Spectrum
‣ Selected Antineutrino data energy spectrum
‣ Good Data/MC agreement at the Near detector
25
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
‣ Flux and cross-section uncertainties cancel when we extrapolate from Near to Far detectors
‣ Near and Far detector energy spectra are not identical
• Due to π/K decay kinematics, neutrino energy varies with angle - the Near detector covers a wider solid angle
• Higher energy π travel further and decay closer to the Near detector
Extrapolation
Monte Carlo
‣ A beam matrix encapsulates our knowledge of decay kinematics, and transports the measured Near detector spectrum to the Far detector
26
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Systematics
‣ The antineutrino analysis is statistically limited
‣ Studies were done to assess the effect of systematics on the fitted result
‣ Results contours were produced with systematic uncertainties included
Monte Carlo
27
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Reco. Energy (GeV)
FD E
vent
s/G
eV
0
10
20
30
0 5 10 20 30 40 50
MINOS Preliminary
running, Far Detectorµ POT MINOS 20 10×1.71
MINOS dataNo oscillations
)=1(22, sin2eV-310×=2.352mBest oscillation fitBackground
Comparison with νμ Result
Reco. Energy (GeV)R
atio
to N
o O
scilla
tions
-0.5
0
0.5
1
1.5
0 5 10 20 30 40 50
MINOS Preliminary BackgroundSubtracted
running, Far Detectorµ POT MINOS 20 10×1.71
MINOS dataBest oscillation fit
)=1(22, sin2eV-310×=2.352m
28
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
The Future - At νμ Best Fit
‣ This shows the sensitivity to future antineutrino running, in the scenario where the new data matches the νμ disappearance best fit value
)(22sin0.5 0.6 0.7 0.8 0.9 1
)2 e
V-3
| (10
2m
|
2
3
4
5
6-310×
MINOS Preliminary
90% MC Sensitivity
POT20 10×1.7 POT20 10×3.5
POT20 10×5 POT20 10×7
Dataµ POT MINOS 20 10×1.7 with additional MC oscillated at
)=1(22, sin2 eV-3 10×=2.35 2m
29
Nicholas Devenish, Recontres de Blois 2010, 17th July 2010, for the MINOS collaboration / 19
Q/P1 0.5 0 0.5 1
Even
ts
0
20
40
60
80
DataOscillated MCNo Oscillations
Far Detector POT20 10×1.71
Antineutrino Running
MINOS Preliminary
Data/MC Agreement - Far Detector
Reconstructed Shower Energy (GeV)0 5 10 15 20
Even
ts
0
20
40
60
80DataOscillated MCNo Oscillations
Far Detector POT20 10×1.71
Antineutrino Running
MINOS Preliminary
CC/NC Separation Parameter0 0.2 0.4 0.6 0.8 1
Even
ts
0
10
20
30
40
50
60
DataOscillated MCNo Oscillations
Far Detector POT20 10×1.71
Antineutrino Running
MINOS Preliminary
Energy (GeV)+µReconstructed 0 5 10 15 20
Even
ts
0
10
20
30
40
DataOscillated MCNo Oscillations
Far Detector POT20 10×1.71
Antineutrino Running
MINOS Preliminary
30