FIRST OSCILLATION RESULTS
FROM THE NOVA
EXPERIMENT
Bruno Zamorano
Oxford - 17th November 2015
Why study neutrino oscillations?
•Second most abundant particle in the
Universe and yet the worst understood
•Dark Matter aside, the only measured
confirmation of Physics beyond the
Standard Model
•~20 000 neutrino papers since the
discovery of neutrino oscillations
•Nobel prize 2015 and Breakthrough
prize 2016
?
•Many open questions: CP violation
(matter-antimatter asymmetry), mass
ordering and mass scale, Dirac or
Majorana…
•Oscillation parameters are, to our best
knowledge, fundamental constants of
Nature
B. Zamorano - First oscillation results from the NOvA experiment
2
Neutrino oscillations overview
⎛ν e ⎞ ⎛ U e1 U e 2 U e 3 ⎞⎛ν 1 ⎞
⎟
⎟
⎟
⎜
⎜
⎜
PMNS
ν µ ⎟ = ⎜U µ 1 U µ 2 U µ 3 ⎟⎜ν 2 ⎟
⎜
matrix
⎜ν ⎟ ⎜ U
⎟⎜ν ⎟
U
U
τ2
τ 3 ⎠⎝ 3 ⎠
⎝ τ ⎠ ⎝ τ1
Atmospheric
Pµe =
!
"
∗
∗
U
U
U
µj
µk Uek exp −i
j,k ej
∆m2jk L
2E
#
Oscillations
Reactor
Solar
Osc.max
∆m232 ≃ ∆m231 −−−−→ L/E ≈ 500km/GeV
Osc.max
∆m221 = m22−m21 −−−−→ L/E ≈ 15000 km/GeV
B. Zamorano - First oscillation results from the NOvA experiment
3
Importance of reactor result
θ13: from unknown to
best measured
sin2 θ13 = 0.0219 ± 0.0012
θ13 ~ 8.5°
CP violation ⇐⇒ θ13 ̸= 0
A new door to probing CP violation, the mass ordering and the octant of θ23
B. Zamorano - First oscillation results from the NOvA experiment
4
𝜃13>0 ⇒ 𝜈𝜇→𝜈e
increasing mass
Importance
of
reactor
result
Makes feasible long-baseline
measurements of…
sin2 θ13 = 0.0219 ± 0.0012
θ13: from unknown to
neutrino mass hierarchy
via matter effects thatbest
modify
P(𝜈𝜇→𝜈e)
measured
13 to m ;
Implications for: 0𝜈𝛽𝛽 data and Majorana nature of 𝜈; approach
𝛽
theoretical frameworks for mass generation,
θ13cosmology;
̸quark/lepton
= 0 astrophysics;
unification; Is the lightest charged lepton associated with
the heaviest light neutrino?
θ ~ 8.5°
CP violation ⇐⇒
A new door to probing
CP violation,
the mass ordering and the octant of θ23
CP violation
Long-baseline
neutrino
oscillations
𝜈𝜇 disappearance:
via dependence of P(𝜈𝜇→𝜈e) on CP phase 𝛿. Amplified by 𝜈/𝜈̅ comparisons.
baryon asymmetry through see-saw/leptogenesis; fundamental
θ23question
octant
in the Standard Model (is CP respected by leptons?)
𝜈
3
𝜈e
𝜈3 flavor mixing
…to leading order
via leading-order factor sin2(𝜃23)
experimental data are consistent with unity
Is 𝜈 more strongly coupled to 𝜇 or 𝜏 flavor?;
(“maximal 3mixing”)
frameworks for mass generation, unification
𝜈𝜏
𝜈e
?
𝜈𝜇
𝜈𝜏
𝜈𝜇
2 )
Need a leap in precision on 𝜃23 (and m32
3
𝜈e appearance:
𝜃13>0 ⇒ 𝜈𝜇→𝜈e
…plus potentially
Makes feasible long-baseline
large CPv and
Daya Bay reactor experiment:
matter effect
of…
sin2measurements
(2𝜃 ) = 0.084 ± 0.005
13
modifications!
neutrino mass
hierarchychannel, and…
Non-zero 𝜃13 opens the long-baseline
appearance
Fermilab JETP, August 6, 2015
Mass hierarchy
increasing mass
Ryan Patterson, Caltech
via matter effects that modify P(𝜈𝜇→𝜈e)
B. Zamorano
experiment
2 the NOvAfor:
Ryan Patterson,
Caltech - First oscillation results from
Fermilab
JETP,
August 6, 2015
Implications
0𝜈𝛽𝛽 data
and
Majorana
nature of 𝜈; approach to m ;
5
SUMMARY
The NOvA
experiment
Muon neutrino
disappearance
Electron neutrino
appearance
Future
sensitivities
The NOvA experiment
About 150 physicists from 6 countries (list growing…)
2 UK institutions: Sussex (8 people, since 2012) and UCL (6 people, just joined)
New collaborators and institutions welcome!
B. Zamorano - First oscillation results from the NOvA experiment
7
The NOvA experiment
RAMPING UP
RAMPING UP
NuMI beam
• Capable of 700 kW
Joao Coelho
4 November 2015
Neutrinos
to Far Det.
• Reliably running ~470 kW
• Capable ofNOvA
700 kW
• Record 521 kW for 1 hour
Near Det.
• Reliably running
~470 kW
Hadron
• Full power by mid 2016
• 6.4 x 1020 PoT delivered
Decay
• Record 521 kW for 1 hour
NuMI
Target
Proton2016
• Full power by mid
•Capable of running at 700 kW
• Reliably running at ~500 kW
Accel.
• 6.4 x 1020 PoT delivered
• World record 521 kW for 1 hour
4
• Full power by mid 2016
•6.4 x 1020 POT delivered
B. Zamorano - First oscillation results from the NOvA experiment
8
The NOvA experiment
• NuMI Off-Axis νe Appearance, the
leading neutrino oscillation experiment
in the NuMI beam
Ash
River
• Two totally active scintillator detectors:
m
• 14 mrad off-axis narrowly peaked muon
neutrino flux at 2 GeV, L/E ~ 405 km/
GeV
0k
- Near Detector: 300 T, 105 m
underground
81
- Far Detector: 14 kT, on surface
Fermilab
• νµ disappearance channel: θ23, Δm232
• νe appearance channel: mass hierarchy,
δCP, θ13, θ23 and octant degeneracy
B. Zamorano - First oscillation results from the NOvA experiment
9
The NOvA experiment
• NuMI Off-Axis νe Appearance, the
leading neutrino oscillation experiment
in the NuMI beam
• Two highly active scintillator detectors:
15.
6m
- Far Detector: 14 kT, on surface
- Near Detector: 300 T, 105 m
underground
• νµ disappearance channel: θ23, Δm232
• νe appearance channel: mass hierarchy,
δCP, θ13, θ23 and octant degeneracy
m
To APD
Readout
Scintillation Light
15.5m
• 14 mrad off-axis narrowly peaked muon
neutrino flux at 2 GeV, L/E ~ 405 km/
GeV
4.5
Particle Trajectory
Waveshifting
Fiber Loop
3.9cm 6.6cm
B. Zamorano - First oscillation results from the NOvA experiment
10
The NOvA experiment
• NuMI Off-Axis νe Appearance, the
leading neutrino oscillation experiment
in the NuMI beam
• Two highly active scintillator detectors:
- Far Detector: 14 kT, on surface
- Near Detector: 300 T, 105 m
underground
• 14 mrad off-axis narrowly peaked muon
neutrino flux at 2 GeV, L/E ~ 405 km/
GeV
• νµ disappearance channel: θ23, Δm232
• νe appearance channel: mass hierarchy,
δCP, θ13, θ23 and octant degeneracy
B. Zamorano - First oscillation results from the NOvA experiment
Also: neutrino cross sections at the
ND, sterile neutrinos, supernovae…
11
The NOvA experiment
Far detector dataset
Far
Detector
data
set
• During the construction era, we began collecting Physics data with each far
detectorthe
“diblock”
(64 detector
soon as it was
fully data
commissioned
During
construction
era, weplanes)
beganascollecting
physics
with each
Far
Detector
data
set
andDetector
Physics-ready
Far
“diblock” (64 detector planes) as soon as it was fully
During the construction era, we began collecting physics data with each
and
physics-ready
Far size
Detector
“diblock”
(64throughout
detector planes)our
as soon
as itset
was fully
•commissioned
Thus, the FD
is not
static
data
commissioned and physics-ready
Thus, FD size is not static throughout our data set
Thus, FD size is not static throughout our data set
20 POT 20
Protons-on-target
data set:
Protons-on-target
inindata
set:3.45×10
3.45×10
POT
average)
Fraction of detector operational: 79.4% (POT-weighted
(POT-weighted
average)
Fraction of detector operational: 79.4%
Full-detector-equivalent exposure: 2.74×1020 POT-equiv
Full-detector-equivalent exposure: 2.74×1020 POT-equiv
××××××××
Partial Far Detector
during construction
(6 diblock example)
××××××××
22
Ryan-Partial
Patterson,
Caltech
B. Zamorano
First oscillation
from the NOvA experiment
Farresults
Detector
Full Far Detector
(14 diblocks)
Fermilab JETP, August 6, 2015
Full Far Detector
12
The NOvA experiment
1m
Proton
1m
νμ + n → μ + p
• Superb
granularity for a
detector this
scale
• Outstanding
event
identification
capability
Muon
Michel e-
νμ Charged Current
Proton
νe + n → e + p
Electron
νe Charged Current
Proton
ν + X → ν + X'
π0 (→γγ)
Neutral Current
1 radiation length = 38 cm
(6 cell depths, 10 cell widths)
B. Zamorano - First oscillation results from the NOvA experiment
13
The NOvA
experiment
Near
Detector: 10
𝜇s of readout during NuMI beam pulse
(color ⇒ time of hit)
Top view
Near detector event display
Side view
Ryan Patterson, Caltech
16
10 µs of readout during NuMI beam pulse
B. Zamorano - First oscillation results from the NOvA experiment
Fermilab JETP, August 6, 2015
14
The NOvA
experiment
Near
Detector: 10
𝜇s of readout during NuMI beam pulse
(color ⇒ time of hit)
Top view
Near detector event display
Side view
9.6 µs NuMI pulse
Ryan Patterson, Caltech
16
10 µs of readout during NuMI beam pulse
B. Zamorano - First oscillation results from the NOvA experiment
Fermilab JETP, August 6, 2015
15
The NOvA experiment
Far Detector Event Display
Far detector event display
Top view
Side view
Full 550
usec
readout
(colors
show
charge.)
Full 550 µs readout (colours show charge)
B. Zamorano - First oscillation results from the NOvA experiment
13(
16
The NOvA experiment
Far Detector Event Display
Top view
Side view
Zoomed
on
the
10
usec
beam
spill
window.
Zoomed on the 10 µs beam spill window
B. Zamorano - First oscillation results from the NOvA experiment
14(
17
The NOvA experiment
νµ
Colour denotes
deposited charge
Zoomed on the time slice
B. Zamorano - First oscillation results from the NOvA experiment
18
The NOvA experiment
NO ν A Preliminary
FD data
Data (basic
monitoring selection)
All current data
Events / (12 µ s)
Background (fit)
Beam window
100
150
200
250
300
350
∆ t from t (µ s)
0
Zoomed on the time slice
B. Zamorano - First oscillation results from the NOvA experiment
19
The NOvA experiment
νe
Colour denotes
deposited charge
Zoomed on the time slice
B. Zamorano - First oscillation results from the NOvA experiment
20
Muon neutrino
disappearance
Disappearance analysis in a nutshell…
Identify contained
νµ CC events in
both detectors
•
Measure both
energy spectra
B. Zamorano - First oscillation results from the NOvA experiment
Infer on the
oscillation
mechanism from
differences
between near and
far energy spectra
21
Muon neutrino
disappearance
Disappearance analysis in a nutshell…
Identify contained
νµ CC events in
both detectors
•
Measure both
energy spectra
Infer on the
oscillation
mechanism from
differences
between near and
far energy spectra
• Extrapolation
•Containment
•PID
• Calibration
• Far / near ratio
•NC rejection
• Energy scale
• Best fit
•Cosmic
rejection
• Hadronic energy
• Contours
B. Zamorano - First oscillation results from the NOvA experiment
• Systematics
22
Muon neutrino
disappearance
Signal selection: νµ CC
Simulated ν µ CC
Simulated Background
6
10
• Muon ID: 4-variable k-nearest
neighbours algorithm to identify muons
- Track length
- dE/dx along track
- Scattering along track
- Track-only plane fraction
ND, 1.66 × 1020 POT
Events
• First: basic containment cuts: require a
buffer of no activity around the event
Data
105
104
103
0
0.2
0.4
0.6
0.8
1
Muon ID
• Keep events with muon-ID >Figure:
0.75 Plot from ⌫µ manuscript. Distribution of the greatest
muon identication variable of all reconstructed tracks in selec
ND. The shaded region of the simulated distributions represe
bin-to-bin fluctuations that are introduced by the systematic
(shape-dependent e↵ects). Pure normalization uncertainties h
B. Zamorano - First oscillation results fromremoved
the NOvA experiment
23
from the error bands shown.
Muon neutrino
disappearance
Good spills
Cosmic rejection
Data quality
Cosmic rej.
• From beam timing: a factor 105
reduction
Containment
Cosmic background
NC rejection
CC ν µ prediction (max. mixing)
• From event topology: 107
E < 5 GeV
• Keep events with BDT > 0.535
• Final rate is measured directly using outof-time data
1
10
102
10
3
104
10
5
10
6
107
Number of events in the spill window
NOν A Preliminary
20
NOν A 2.74 ×10 POT-equiv.
60
# Events in Final Sample
• For the remaining, a boosted decision
tree is trained on data without beam
- Direction
- Position and length
- Energy and number of hits in event
Out of time data
ν µ CC MC
All cuts except BDT applied
20
0
0.3
B. Zamorano - First oscillation results from the NOvA experiment
Cut value
40
0.4
0.5
Cosmic Rejection BDT
0.6
0.7
24
Muon neutrino
disappearance
muon
Energy estimation
• Hadronic system:
Σ Total visible E
Eµ
Ehad
• Reconstructed neutrino energy is the
sum of them: Eν = Eµ + Ehad
• Energy resolution at beam peak: ~7%
Fractional energy resolution
• Muon track: length
hadronic system
NOνA Preliminary
0.15
FD
ND
Neutrino spectrum
0.10
0.05
0.00
1
2
3
4
5
True Neutrino Energy (GeV)
B. Zamorano - First oscillation results from the NOvA experiment
25
Muon neutrino
disappearance
×106
Simulated ν µ CC
0.12
Simulated Background
Data
Kinematic variables in the ND after all cuts
Events
0.10
(Sample purity in ND = 98%)
ND, 1.66 × 1020 POT
0.08
This ND
distribution is used
to create the
FD prediction
0.06
0.04
0.02
0.00
0
1
2
3
4
5
Reconstructed Neutrino Energy (GeV)
3
Figure: Plot from40⌫×10
µ manuscript. Reconstructed neutrino ene
Simulated ν CC
the ND. The shaded regions of the simulated
distributions rep
Simulated Background
shape-dependent systematic uncertainties.Data
Both data and MC
30
to 1.66 ⇥ 1020 POT
exposure. The plot ND,
includes
the
1.66 × 1020
POTcorrectio
energy scale.
Simulated ν µ CC
107
µ
Simulated Background
6
10
Data
ND, 1.66 × 1020 POT
5
Events
Events
10
104
3
10
102
20
10
10
1
0
0.2
0.4
0.6
Muon Track cos(θZ)
0.8
1
0
0
B. Zamorano
- First oscillation
results from of
the the
NOvA
experiment
Plot from
⌫µ manuscript.
Distributions
reconstructed
muon
5
10
15
Length of Primary Track (m)
26
Muon neutrino
disappearance
Far detector prediction
• Estimate the underlying true energy distribution of selected ND events
• Multiply by expected Far/Near ratio and νµ → νµ oscillation probability
• Convert FD true energy distribution into predicted FD reconstructed energy distribution
• Systematic uncertainties assessed by varying all MC-based steps
B. Zamorano - First oscillation results from the NOvA experiment
27
Muon neutrino
disappearance
Systematics
×103
Hadronic energy systematic is one
with a noticeable effect
(although the impact is reduced by
ND-to-FD prediction procedure)
100
ND, 1.66 × 1020 POT
80
Events
Most of the systematic uncertainties
have relatively little influence on
the result
Simulated ν µ CC
Simulated Background
Uncalibrated Data
Data
Tune hadronic energy in ND
simulation to achieve better
agreement in Eν and Ehad
60
40
Take the full size of this tuning
as a systematic uncertainty
20
0
0
0.5
1
1.5
2
2.5
3
Hadronic Energy (GeV)
Figure: Plot from ⌫µ manuscript. Distribution of the reconstructed hadron
• Hadronic energy (21%, equiv. to 6% in neutrino
energy, Ehad, of selected
⌫µand
CC tau
interactions
in(100%
the ND
in data and simula
NC
neutrinos
each)
•
energy)
The shaded regions of the simulated distributions represent the
calibration The
and light
(Hitrepresen
• Multiple
uncertainties.
solid level
blacksysts.
circles
transportation systematic
• Neutrino flux (NA49 + beam shape-dependent
fibrecorrection
attenuation,
effects)
data after a 21% Ehad energy,
calibration
is threshold
applied, and
the solid gray
model)
squares represent the data before the correction.
• Absolute, relative normalisation (1%, 2%)
•Neutrino interactions (GENIE / Intranuke model)
• Oscillation parameter uncertainties (current world
knowledge)
B. Zamorano - First oscillation results from the NOvA experiment
28
Muon neutrino
disappearance
Some of the selected candidates
(zoomed views, not showing the
whole detector)
Ryan Patterson, Caltech
31
Fermilab JETP, August 6, 2015
32
Fermilab JETP, August 6, 2015
Ryan Patterson, Caltech
B. Zamorano - First oscillation
results from the
NOvA experiment
33
Fermilab JETP, August 6, 2015
Ryan Patterson, Caltech
29
Muon neutrino
disappearance
50
In the absence of oscillation, the expectation is
200 events (2 beam bkg., 1.4 cosmic bkg.)
of
uld
nts
(bckgrnd subtracted)
ted
or
Ratio with unoscillated
spectrum
bkgnd
gnd)
2 and 𝜃
s well matched by oscillation fit for m32
23
Unoscillated pred.
Best fit pred. (systs)
40
Events / 0.25 GeV
33 selected events in the FD
Data
Backgrounds
Best fit pred. (no systs)
1-σ syst. range
30
Normal Hierarchy
20
2.74×10 POT-equiv.
Best fit χ2=12.1/16 dof
20
10
0
0
1
2
3
4
Reconstructed Neutrino Energy (GeV)
5
Figure:
Plot from
manuscript. The reconstructed ener
Clear
observation
of ⌫νµµ disappearance
selected events. The data points show the statistical unce
r observation of 𝜈𝜇 disappearance
histogram
corresponds to the predicted spectrum in the a
37
Fermilab JETP, August 6, 2015
B. Zamorano - First oscillation results
the NOvA
experiment
30
Thefromgreen
histogram
corre- sponds to the best fit predicti
uncertainties included in fit via nuisance parameters)
Muon neutrino
disappearance
Oscillation parameters
3.5
Normal Hierarchy, 90% CL
Normal Hierarchy
Normal hierarchy
NOν A 2015
∆m232 (10-3 ev2)
T2K 2014
3
−3 eV2
∆m232 = 2.35+0.19
×
10
−0.12
MINOS 2014
sin2 θ23 = 0.513 ± 0.106
2.5
Inverted hierarchy
−3 eV2
∆m232 = −2.40+0.12
×
10
−0.18
2
0.3
0.4
0.5
2
0.6
sin θ23
0.7
sin2 θ23 = 0.509+0.103
−0.106
2
from ⌫µ manuscript. The allowed values of sin2 (✓32 ) and m32
normal
mass
hierarchy. Will
slightly
changewith
withless
FCthan
correction
NOvA
measurement
already
compelling
8% of nominal exposure!
B. Zamorano - First oscillation results from the NOvA experiment
31
Electron neutrino
appearance
Appearance analysis in a nutshell…
Identify contained
νe CC events in
both detectors
Use ND candidates
to predict beam
backgrounds in the
FD
B. Zamorano - First oscillation results from the NOvA experiment
Interpret any FD
excess over
predicted
backgrounds as νe
appearance
32
Electron neutrino
appearance
Appearance analysis in a nutshell…
Identify contained
νe CC events in
both detectors
Use ND candidates
to predict beam
backgrounds in the
FD
Interpret any FD
excess over
predicted
backgrounds as νe
appearance
•Containment
•PID
• Extrapolation
•NC rejection
• Far / near ratio
•Cosmic
rejection
• Systematics
B. Zamorano - First oscillation results from the NOvA experiment
• Exclusions
• Significance
33
Pre-selection
Electron neutrino
appearance
First, basic containment cuts require sufficient distance from the
largest reconstructed shower to the walls.
Signal pre-selection: νe CC
Then, cuts applied to:
- shower length
• First: basic containment cuts: require
- number of hits in event
sufficient distance from the largest
- calorimetric energy
reconstructed shower to the edges
applied
to
• Then, cuts
All three related to the “size”
length
- Shower
of the
event
of know
hits inwell
event
- NumberWe
the range of
energy
- Calorimetric
energies
any appearing 𝜈e
might to
have
the “size” of the
• All three related
event (we know well the range of
energies any appearing νe might have)
Ryan Patterson, Caltech
B. Zamorano - First oscillation results from the NOvA experiment
41
Fermilab JETP, August 6, 2015
34
The 𝜈e neutrino
selectors themselves
Electron
appearance
provide
a lot of cosmic rejection
A
Cosmic rejection
• Cut events with large transverse
momentum (Rejects downward-going
cosmic showers)
Cosmic
rejection
• The νe selectors
themselves
are very
efficient atCut
rejecting
cosmic
background
events with large
Ryan Patterson, Caltech
42
pT/p
1 in 108 rejection
• Achieve areconstructed
Rejectsbackground;
downward-directed
0.06
• Expected cosmic
cosmic directly
shower using out-ofevents, measured
time dataThe 𝜈 selectors themselves
e
provide a lot of cosmic rejection
B. Zamorano - First oscillation results from the NOvA experiment
35
Electron neutrino
appearance
43
𝜈e CC event identification
νe CC event identification
We have developed two independent
𝜈e CC selection algorithms
→ Very different
designs
Two independent
selection algorithms with very different designs
LID: Likelihood identification
Color: p.d.f. for dE/dx in each plane (e– assumption)
LID: Likelihood Identification
dE/dX likelihoods calculated for longitudinal
dE/dx likelihoods calculated for longitudinal
and transverse slices of leading shower under
and transverse slices of leading shower
multiple
particle hypotheses
under multiple
particle hypotheses
Likelihoods
feed
an an
artificial
neural
network
Likelihoods
feed
artificial
neutral
network
along
and
topological
info:
e.g.,
alongwith
withkinematic
kinematic
and
topological
info:
energy
near
vertex,
shower
angle,
vertex-toe.g.,
energy
near
vertex,
shower
angle,
shower gap,
…
vertex-to-shower
gap
Likelihoods calculated for each red and yellow region
B. Zamorano
- First oscillation results from the NOvA experiment
43
Ryan Patterson,
Caltech
36
Fermilab JETP, August 6, 2015
Electron neutrino
appearance
LEM: Library event matching
Library
Event Matching
Spatial pattern ofLEM:
energy
depositions
is
Spatial
pattern of events
energy deposition
compared to that of
108 simulated
is compared directly to that of ~108
(“library”)
Left panels: candidate event, both views
Right panels: best-matched library event, both views
Middle panels: an intermediate step in calculating the match quality
simulated events (“library”)
Key properties of the
matched
library
Keybest
properties
of the best-matched
event (e.g., fractionlibrary
that are
signal
events)
events
(e.g.,
fractionare
that
aretree
signal
are input into a
input into a decision
toevents)
form discriminant
decision tree to form discriminant
•Identical performance as measured
with efficiency, S/√B and sensitivity to
oscillation parameters
Prior to unblinding, decided to show both results and use LID as primary
Ryanoscillation
Patterson, results
Caltechfrom the NOvA experiment
B. Zamorano - First
44
Fermilab JETP, August 6, 2015
37
Electron neutrino
appearance
LID and LEM distributions for ND data
and simulation
•Good agreement over full range
Total
BKG
0.94
± 0.09
νe - CC LID and LEM distributions
NC νµ - CC ντ - CC Cosmic
and simulation
(beam) for ND data
all preselection cuts applied
0.46
0.35
0.05
0.02
0.06
± few % depending on osc. param.
Good agreement over full range
Range of signal predictions
•
NH, δCP = 3π/2, θ23 = π/4 → 5.6 ± 0.7
•
IH, δCP = π/2, θ23 = π/4 → 2.2 ± 0.3
Ryan Patterson, Caltech
B. Zamorano - First oscillation results from the NOvA experiment
46
38
Electron neutrino
appearance
Far detector prediction
• ND data are Far
translated
to FD prediction
Detector
•
background expectation in each
NDfar/near
data is translated
to FD bckgnd
energy bin, using
ratios
from
expectation in each energy bin, using
simulation
Far/Near ratios from simulation
(LID selection)
FD signal expectation
• FD signal expectation
is pinnedistopinned
the to
𝜈𝜇 CC spectrum
ND-selected vµthe- ND-selected
CC spectrum
Most systematics are assessed via
systematics
are assessed
via ratios
variations
in the Far/Near
• Most
variations in the far/near ratios
Some FD sample stats:
Signal efficiency relative toSignal
containment
cuts:
35%
efficiency
relative
to containment cuts: 35%
Expected overlap in LID/LEM
samples of 62%
Expected overlap in
(Expected differences in LID/LEM
which events
each 62%
samples:
technique selects)
→ Differences in which events
each technique selects
Ryan Patterson, Caltech
B. Zamorano - First oscillation results from the NOvA experiment
48
After all selection,
After all 0.7%
selection,
of NC 0.7%
events of NC
remain,relative
relative toto those
events remain,
those after
containment
after
containment
Fermilab JETP, August 6, 2015
39
Electron neutrino
appearance
Far detector selected events
Far Detector selected events
LID: 6 νe candidates
6 𝜈e candidates
3.3σ significance forLID:
νe appearance
3.3𝜎 significance for 𝜈e appearance
At right:
Calorimetric energy
LEM: 11 νe candidates
5.5σ significance for νe appearance
(All 6 LID events
present
LEM:
11 𝜈e candidates
in LEM set)
5.5𝜎 significance for 𝜈e appearance
(All 6 LID events present in LEM set)
Patterson,
Caltech
B. Zamorano - FirstRyan
oscillation
results
from the NOvA experiment
55
Fermilab JETP, August 6,40201
Electron neutrino
appearance
LEM selected
νe candidates
Here comes the sun
LID, LEM selected
Because
B. Zamorano - First oscillation results from the NOvA experiment
LEM selected
You never give me your money
41
Electron neutrino
appearance
LID, LEM selected
LEM selected
Sun King
Mean Mr. Mustard
LEM selected
LID, LEM selected
Polyethylene Pam
She came in through
the bathroom window
B. Zamorano - First oscillation results from the NOvA experiment
42
Electron neutrino
appearance
LID, LEM selected
Golden Slumbers
LID, LEM selected
Carry that weight
LEM selected
LID, LEM selected
The End
Her Majesty
B. Zamorano - First oscillation results from the NOvA experiment
43
Electron neutrino
appearance
FD selection: 6 𝜈e candidates
For allowed regions of δCP and θ13
NO𝜈A Preliminary
Result using
LID selector
Result
using
LID selector
For (𝛿CP , sin22𝜃13) allowed regions
•Feldman & Cousins procedure applied
• Feldman-Cousins procedure applied
•Solar oscillation
parameters
• solar osc.
parameters varied
varied
2 varied by new NOvA measurement
• m32
2
•Δm 32 varied2 by new NOvA result
• sin 𝜃23=0.5
• sin2θ23 = 0.5
58
Ryan Patterson, Caltech
B. Zamorano - First oscillation results from the NOvA experiment
Fermilab JETP, August 6, 2015
44
Electron neutrino
appearance
Result using LID selector
Applying global reactor constraint of
With θ23 uncertainty
LID 2.74×1020 POT equiv.
Normal hierarchy
13 = 0.0219 ± 0.0012
• Again, apply Feldman-Cousins to
interpret -2ΔlnL
• Converted into significance (steps
due to discrete nature of counting
experiment)
Inverted hierarchy
Significance
2
sin θ
2σ
90%
NOνA Preliminary
3σ
1σ
0
• Using νµ constraint for sin2θ23
0
π/2
π
3π/2
δCP
2π
5σ
B. Zamorano - First oscillation results from the NOvA experiment
nificance
Inverted hierarchy for δCP ∈[0, 0.9π] is mildly disfavoured4σ (>1σ)
3σ
45
Electron neutrino
appearance
LEM 2.74×1020 POT equiv.
Normal hierarchy
FD selection: 11
𝜈e candidates
Inverted
hierarchy
4σ
Significance
NOνA Preliminary
5σ
3σ
NO𝜈A Preliminary
Result
using
LEM LEM
selectorselector
Result
using
Below: With reactor constraint applied
2σ
90%
(significance on next page)
1σ
0
π/2
π
δCP
3π/2
2π
•Inverted
hierarchy is disfavoured at 2.2σ
NH LID
NH LEM
for
all δCPIH LEM
IH LID
NO𝜈A Preliminary
0
• NH for δCP ∈[0,π] is mildly disfavoured
(>1σ)
61
Ryan Patterson, Caltech
B. Zamorano - First oscillation results from the NOvA experiment
Fermilab JETP, August 6, 2015
46
Electron neutrino
appearance
0
0
π/2
π
0
2π
3π/2
δCP
1σ
5σ
LID/LEM consistency
Both prefer normal hierarchy
•
Both prefer δCP near 3π/2
•
Given the expected correlations, the
observed event counts yield a reasonable
mutual p-Value of 10%
C. Backhouse (Caltech)
Significance
•
4σ
0
π/2
NH LID
NH LEM
IH LID
IH LEM
3σ
2σ
90%
1σ
0
0
π/2
π
δCP
Throwing θ23
3π/2
2π
The specific point IH, δCP = 3π/2 is disfavoured at 1.6σ (LID), 3.2σ (LEM) for
sin2θ23 ∈[0.4, 0.6]
B. Zamorano - First oscillation results from the NOvA experiment
47
Future sensitivities
3
20
•Measurements in the anti-neutrino
channel: CPT tests
MINOS result
θ = 0.5)
23
Non-maximal mixing (sin
T2K latest ν result
2
µ
θ = 0.4)
23
2.6
(10
-3
2
eV )
2.8
2
32
2.4
∆m
• Leading measurement in both Δm232
and sin2θ23 for nominal sensitivity
Maximal mixing (sin
2
•Potential to exclude maximal mixing,
depending on Nature’s choice
20
NO ν A 18 × 10 FHC PoT + 18 × 10 RHC PoT
B. Zamorano - First oscillation results from the NOvA experiment
2.2
2
0.3
0.4
0.5
2
sin θ
0.6
0.7
23
48
Future sensitivities
COMBINING MUON AND ELECTRON NEUTRINO ANALYSES
Combining$the$vμ$and$ve$Measurements:$$
36×1020 POT
2
2
sin
2
θ
=0.095,
sin
2θ =0.95, Normal, θ23>π/4, δCP=3π/2
13 A contours, 3+3 yr23
Example NO
Example NOνA contours
0.7
0.7
sin22
13=0.095,
sin22
23=0.95,
m2>0,
23>
/4, =3 /2
0.65
0.65
0.6
0.6
2
sin
2
23
sin θ23
0.55
0.55
Inverted
0.5
0.5
0.45
0.45
m2<0, 1 CL
0.4
Normal
2
m2<0, 2 CL
2 CL
Inverted,m1>0,
σ CL
0.350.4
0.3
0.35
0
0.2
m2>0, 1 CL
True parameters
Inverted, 2σ CL
0.4
0.6
0.8
1
1.2
1.4
/
0.3
0
0.2
0.4
0.6
0.8
1
Normal, 1σ CL
Normal, 2σ CL
1.6
1.8
2
True parameters
δCP / π
1.2
1.4
1.6
1.8
2
Best case scenario: NOvA simultaneously measures the mass ordering, CP
NOvA(obtains(simultaneous(hierarchy,(δ(CP,(and(Θ
$
23
violation and octant information!
octant(informa<on!(
M.Baird(Y(NOvA(numu(Disappearance(
B. Zamorano - First oscillation results from the NOvA experiment
41(
49
Future sensitivities
COMBINING MUON AND ELECTRON NEUTRINO ANALYSES
Combining$the$vμ$and$ve$Measurements:$$
36×1020 POT
sin22θ13=0.095, sin22θ23=0.95, Normal, θ23>π/4, δCP=π/2
Example NOνA contours
Example
NO A contours, 3+3 yr
0.7
0.7
sin22
13=0.095,
sin22
23=0.95,
m2>0,
23>
/4, = /2
0.65
0.65
0.6
0.6
23
sin θ23
0.55 0.55
Inverted
Normal
sin2
2
0.5
0.5
0.45
0.4
0.35
0.45
m2<0, 1 CL
m2>0, 1 CL
m2<0, 2 CL Inverted,
m2>0,1σ
2 CL
CL
0.4
Normal, 1σ CL
True parameters
0.3
00.350.2
0.4
0.6
0.8
Inverted, 2σ CL
1
1.2
1.4
Normal, 2σ CL
1.6
1.8
2
True parameters
/
0.3
0
0.2
0.4
0.6
0.8
1
δCP / π
1.2
1.4
1.6
1.8
2
Degenerate(case:((hierarchy(and(δ(informa<on(is(
Degenerate case: mass ordering and CP violation are coupled, but the octant
coupled,(BUT(Θ23$octant(can(be(resolved(AND(we(can(
information is not
claim(that(CP(viola<on(is(present.(
M.Baird(Y(NOvA(numu(Disappearance(
40(
B. Zamorano - First oscillation results from the NOvA experiment
50
Future sensitivities
MASS HIERARCHY AND CP-VIOLATION
4
NOνA CPV determination
sin22θ13=0.095, sin22θ23=1.00
36×1020 POT
Inverted
3.5
Normal
3
2.5
2
1.5
1
0.5
0
0
0.2
0.4
0.6
0.8
1
δCP / π
3.5
significance of CP violation (σ)
significance of hierarchy resolution (σ)
NOνA hierarchy resolution
1.2
1.4
1.6
1.8
2
3+3 years (νµ +anti-νµ): 2 sigma in
about 30% of the δCP range
B. Zamorano - First oscillation results from the NOvA experiment
sin22θ13=0.095, sin22θ23=1.00
36×1020 POT
Inverted
3
Normal
2.5
2
1.5
1
0.5
0
0
0.2
0.4
0.6
0.8
1
δCP / π
1.2
1.4
1.6
1.8
2
Only 1.5 sigma in 10% of the range
51
Future sensitivities
COMBINATION WITH T2K
36×1020 POT NOνA
4
sin22θ13=0.095, sin22θ23=1.00
+ 7×1021 POT T2K
Inverted
3.5
Normal
3
2.5
2
1.5
1
0.5
0
0
0.2
0.4
0.6
0.8
1
δCP / π
1.2
1.4
1.6
1.8
36×1020 POT NOνA
NOνA+T2K CPV determination
3.5
significance of CP violation (σ)
significance of hierarchy resolution (σ)
NOνA+T2K hierarchy resolution
2
Combining with T2K: At least 1 sigma
for the whole δCP range
B. Zamorano - First oscillation results from the NOvA experiment
sin22θ13=0.095, sin22θ23=1.00
+ 7×1021 POT T2K
Inverted
3
Normal
2.5
2
1.5
1
0.5
0
0
0.2
0.4
0.6
0.8
1
δCP / π
1.2
1.4
1.6
1.8
2
With T2K: 1.5 sigma in 25%
52
Future sensitivities
• Far and near detector completed in
August 2014
• The NuMI beam intensity is steadily
increasing, setting new records on a
weekly basis
• First analysis dataset included 7.8% of
full exposure (40% of a standard
running year)
• Other NOvA physics: programme
underway
- Neutrino cross-sections
- Sterile neutrinos, non-standard
interactions, CPT tests
- Supernova neutrinos
- Dark matter and monopole searches
- And more…
B. Zamorano - First oscillation results from the NOvA experiment
53
Future sensitivities
• Far and near detector completed in
August 2014
• The NuMI beam intensity is steadily
increasing, setting new records on a
weekly basis
• First analysis dataset included 7.8% of
full exposure (40% of a standard
running year)
With 2.74 x 1020 POT exposure…
• Unambiguous νµ disappearance signature
• 6.5% measurement of atmospheric mass
splitting
• Consistent with maximal mixing
• νe appearance signal at 3.3σ for primary
selector, 5.5σ for secondary selector
• Other NOvA physics: programme
underway
• At max mixing, disfavour IH for δCP ∈
- Neutrino cross-sections
[0,0.6π] at 90% with primary selector.
- Sterile neutrinos, non-standard
Further preference for NH with secondary
interactions, CPT tests
selector
- Supernova neutrinos
- Dark matter and monopole searches
Much more to come!
- And more…
B. Zamorano - First oscillation results from the NOvA experiment
54
SUMMARY
•
Both NOvA detectors are operational and taking high quality data in the
NuMI beam
SUMMARY
•
Both NOvA detectors are operational and taking high quality data in the
NuMI beam
•
Compelling measurement of muon neutrino disappearance with 8% of
nominal exposure
SUMMARY
•
Both NOvA detectors are operational and taking high quality data in the
NuMI beam
•
Compelling measurement of muon neutrino disappearance with 8% of
nominal exposure
•
NOvA will use both appearance and disappearance channels to provide
constraints on δCP, and leading measurements on the mass ordering and the
θ23 octant
SUMMARY
•
Both NOvA detectors are operational and taking high quality data in the
NuMI beam
•
Compelling measurement of muon neutrino disappearance with 8% of
nominal exposure
•
NOvA will use both appearance and disappearance channels to provide
constraints on δCP, and leading measurements on the mass ordering and the
θ23 octant
•
Full FD with higher beam power: at least 2 x data next summer
THANK YOU FOR
YOUR ATTENTION
Public live event display:
http://nusoft.fnal.gov/nova/public/