2. 3/15/18 Moriond EW 2018 2
The T2K collaboration
Canada
TRIUMF
U. B. Columbia
U. Regina
U. Toronto
U. Victoria
U. Winnipeg
York U.
France
CEA Saclay
LLR E. Poly.
LPNHE Paris
Germany
Aachen U.
Italy
INFN, U. Bari
INFN, U. Napoli
INFN, U. Padova
INFN, U. Roma
Japan
ICRR Kamioka
ICRR RCCN
Kavli IPMU
KEK
Kobe U.
Kyoto U.
Miyagi U. Edu.
Okayama U.
Osaka City U.
Tokyo Institute Tech
Tokyo Metropolitan U.
U. Tokyo
Tokyo U of Science
Yokohama National U.
Poland
IFJ PAN, Cracow
NCBJ, Warsaw
U. Silesia, Katowice
U. Warsaw
Warsaw U. T.
Wroclaw U.
Russia
INR
Spain
IFAE, Barcelona
IFIC, Valencia
U. Autonoma Madrid
Switzerland
ETH Zurich
U. Bern
U. Geneva
United Kingdom
Imperial C. London
Lancaster U.
Oxford U.
Queen Mary U. L.
Royal Holloway U.L.
STFC/Daresbury
STFC/RAL
U. Glasgow
U. Liverpool
U. Sheffield
U. Warwick
USA
Boston U.
Colorado S. U.
Duke U.
Louisiana State U.
Michigan S.U.
Stony Brook U.
U. C. Irvine
U. Colorado
U. Pittsburgh
U. Rochester
U. Washington
Vietnam
IFIRSE
IOP, VAST
500 members, 66 Institutes, 12 countries
3. 3/15/18 Moriond EW 2018 3
Main subject: Neutrino oscillations
Neutrino oscillations, now well-
established phenomenon:
² Indicate massive neutrinos
² Mix flavor and mass eigenstates
² Beyond the Standard Model
(the only lab-based evidence)
After 90 yrs., neutrinos still keep surprising us!
0
@
⌫e
⌫µ
⌫⌧
1
A =
0
@
1 0 0
0 c23 s23
0 s23 c23
1
A
0
@
c13 0 s13e i CP
0 1 0
s13ei CP
0 c13
1
A
0
@
c12 s12 0
s12 c12 0
0 0 1
1
A
0
@
⌫1
⌫2
⌫3
1
A
Atmospherics / Accelerators Reactors/ Accelerators Solar/ Reactors
4. The T2K experiment
3/15/18 Moriond EW 2018 4
² Discovered appearance of (2013)
² Leading effort of CP violation search
² Vibrant programs of non-standard physics & neutrino interactions
⌫µ ! ⌫e Phys. Rev. Lett. 112, 061802 (2014)
Phys.Rev. D96, 9, 092006 (2017)
0
@
⌫e
⌫µ
⌫⌧
1
A =
0
@
1 0 0
0 c23 s23
0 s23 c23
1
A
0
@
c13 0 s13e i CP
0 1 0
s13ei CP
0 c13
1
A
0
@
c12 s12 0
s12 c12 0
0 0 1
1
A
0
@
⌫1
⌫2
⌫3
1
A
Neutrino produced at J-PARC &
detected at Super-Kamiokande
T2K sensitive parameters (with &m2
31)
6. Neutrino Oscillations at T2K (cont’d)
3/15/18 Moriond EW 2018 6
T2K is leading the efforsto measure !23, &m2
32 and (CP
Key ingredients for T2K
JHEP 01 (2017) 087
I Esteban et al, + Global Fit
Most intense and well-
monitored neutrino
beam, J-PARC
Large WC detector &
good flavor identification,
Super-Kamiokande
Support programs
(hadron production, neutrino
interaction models)
8. T2K neutrino beam
3/15/18 Moriond EW 2018 8
² Beam power steadily increased to 475 kW, high-quality data delivered
² 2.65x1021 Protons-on-target (POT) delivered. Data sample for results
presented today:
² Neutrino-mode: 1.47x1021 POT
² Antineutrino-mode: 0.76x1021 POT
Today result
Day
[events/1e14POT]
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Event rate Horn250kA
Horn205kA
Horn-250kA
[mrad]
1−
0.5−
0
0.5
Horizontal beam direction INGRID
MUMON
Day
[mrad]
1−
0.5−
0
0.5
1
Vertical beam direction INGRID
MUMON
T2K Run1
Jan.2010-Jun.2010
T2K Run2
Nov.2010-Mar.2011
T2K Run3
Mar.2012-Jun.2012
T2K Run4
Oct.2012-May.2013
T2K Run5
May.2014
-Jun.2014
T2K Run6
Oct.2014-June.2015
T2K Run7
Feb.2016-May.2016
T2K Run8
Oct.2016-Apr.2017
T2K Run9
Oct.2017-Dec.2017
Next results will be released
by this summer
High intensity, almost pure muon (anti) neutrino beam from J-PARC
9. T2K neutrino beam (cont’d)
3/15/18 Moriond EW 2018 9
High intensity, almost pure muon (anti) neutrino beam from J-PARC
,̅ −mode
(Anti-) neutrino flux prediction at T2K Far Detector (no oscillation)
² Hadron production at target
needed to infer , flux
² Constrained by external data
from NA61/SHINE
o Flux uncertainty ~ 10%
(absolute error)
o ~2-4% effect to analysis w/
Near Detector constraint
(Beam modes changed by switching horn polarity)
+ −mode
Errors of (anti-) neutrino flux prediction at T2K Far Detector
+ −mode ,̅ −mode
10. T2K far detector
3/15/18 Moriond EW 2018 10
Super-Kamiokande
(41.4 m tall x 39.3 m diameter)
1000 m underground
2.5
² Super-K is 2.50 off the beam’s axis to achieve narrow band beam peaked
at oscillation maximum (0.6 GeV)
(GeV)νE
0 1 2 3
(A.U.)295km
µνΦ
0
0.5
1 °OA 0.0
°OA 2.0
°OA 2.5
0 1 2 3
)eν→µνP(
0.05
0.1
= 0CP
δNH, = 0CP
δIH,
/2π=CP
δNH, /2π=CP
δIH,
0 1 2 3
)µν→µνP(
0.5
1
= 1.023θ22
sin
= 0.113θ22
sin
2
eV-3
10×= 2.432
2
m∆
Partice ID parameter
-10 -8 -6 -4 -2 0 2 4 6 8 10
0
50
100
150
200
250
300
350
Super Kamiokande IV 2166.5 days : Monitoring
e-like muon-like
Numberofevents
² Muon and electron are well-separated
à identify ,//,0 with high purity
Atmospheric , FCFV events
12. Strategy for oscillation analyses
3/15/18 Moriond EW 2018 12
Hadron prod. data
(NA61/SHINE)
Beam monitors
ND280 data
External cross-
section data
Neutrino
interaction model
Constrain flux
Constrain flux & interaction model simultaneously
(due to convolution of these factors on data )
13. Strategy for oscillation analyses
3/15/18 Moriond EW 2018 13
Hadron prod. data
(NA61/SHINE)
Beam monitors
ND280 data
Super-K data
External cross-
section data
Neutrino
interaction model
SK detector
model
Constrain flux
Constrain flux & interaction model simultaneously
(due to convolution of these factors on data )
Oscillation fit &
extract
parameters
Insensitive osc.
parameters
Data are used as much as possible to
reduce model dependence
14. Improvements in 2017 analysis
3/15/18 Moriond EW 2018 14
² Super-K event selection & new data sample
² Use reconstruction algorithm (fiTQun complete charge & time
information) à enables to extend detector fiducial volume, leads to 20%
effective statistic increase in selecting e-like events
² Add charged-current 11 e-like sample à increases 10% for neutrino-mode
e-like sample
² Usage of ND280 data to constrain flux & neutrino interaction model
² Incorporate FGD2 (water target) data to include interactions on water
(In previous analysis, only FGD1 (carbon target) data samples were used)
² Interaction models in neutrino event generator (NEUT)
² Improve pion production model by tuning to external data from Bubble
Chambers, MiniBooNE and MINERvA
² Include a model for multi-nucleon (2p-2h) scattering (~10-20% relative to
charged current quasi-elastic, main signal at T2K)
² Improve charged-current quasi-elastic model by including effect of long-
range correlations in nucleus
Phys. Rev. C83 (2011) 045501
15. Systematic errors
3/15/18 Moriond EW 2018 15
Error source
% errors on predicted event at SK
1 ring /-like 1 ring e-like
,-mode ,̅-mode ,-mode ,̅-mode ,-mode CC11 ,/ ,̅ [4]
SK detector 1.86 1.51 3.03 4.22 16.69 1.60
SK FSI+SI+PN 2.20 1.98 3.01 2.31 11.43 1.57
ND280-constrained
flux & cross section
3.22 2.72 3.22 2.88 4.05 2.50
2 34
2 35
,
2 364
2 365
[1]
0.00 0.0 2.63 1.46 2.62 3.03
NC 17 [2] 0.00 0.0 1.08 2.59 0.33 1.49
NC other [3] 0.25 0.25 0.14 0.33 0.98 0.18
Total error 4.40 3.76 6.10 6.51 20.94 4.77
[1] Theoretically motivated error based on
[2],[3] Not constrained by ND280, theoretical model & external data
[4] These errors are relevant for extracting ()* phase
Phys.Rev. D86 (2012) 053003
17. ∆#$%
%
, !%$ measurements
3/15/18 Moriond EW 2018 17
)23θ(2
sin
0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7
)-4c2
|(eV32
2
m∆|
2.3
2.35
2.4
2.45
2.5
2.55
2.6
2.65
2.7
3−
10×
Normal - 68CL
Normal - 90CL
Inverted - 68CL
Inverted - 90CL
Best fit
T2K Run1-8 PreliminaryFinal systematics pending
² Data are fitted separately for normal
and inverted hierarchy Data fit
Final
systematic
error
pending
(Study of ND data-driven
variation shows effect on
∆#$%
% , !%$ parameters)
² Pending final systematic error,
results will be updated in future
(Bayesian posterior probabilities)
Bias to lower value
Bias to
maximal mix.
[*] 2p-2h non-9 is a pure nucleon-nucleon correlation process
*
18. Mixing angle !13 & (CP
3/15/18 Moriond EW 2018 18
)13θ(2
sin
10 15 20 25 30 35 40 45 50
3−
10×
(Radians)CPδ
3−
2−
1−
0
1
2
3 Normal - 68CL
Normal - 90CL
Inverted - 68CL
Inverted - 90CL
Best fit
PDG 2016
T2K Run1-8 PreliminaryFinal systematics pending
)13θ(2
sin
15 20 25 30 35
3−
10×
(Radians)CPδ
3−
2−
1−
0
1
2
3 Normal - 68CL
Normal - 90CL
Inverted - 68CL
Inverted - 90CL
Best fit
T2K Run1-8 PreliminaryFinal systematics pending
² T2K sin2!13 measurement is
consistent with PDG 2016 average
T2K value:
PDG 2016:
² Including the reactor constraint on !13
improves constraint on (CP
sin2
✓13 = 0.0277+0.0054
0.0047 (NH)
sin2
✓13 = 0.0210 ± 0.0011
w/ reactor constraint
19. (CP measurement
3/15/18 Moriond EW 2018 19
(rad)CPδ
3− 2− 1− 0 1 2 3
ln(L)∆-2
0
5
10
15
20
25
30
Normal
Inverted
T2K Run1-8 PreliminaryFinal systematics pending
² 1: C.L. confidence interval
Normal Hierarchy [-2.49, -1.23] rad.
² 2 : C.L. confidence interval
Normal hierarchy [-2.91, -0.60] rad.
Inverted hierarchy [-1.54, -1.19] rad.
2: C.L. intervals
Reconstructed Energy (GeV)ν
0 0.2 0.4 0.6 0.8 1 1.2
NumberofEvents
4−
2−
0
2
4
6
8
10
12
14
16
T2K Run1-8 Preliminary
= 0CP
δ
Data
Frequentist
CP conserving values (0, 1) fall outside of
the 2: C.L. confidence/credible interval
Sample Prediction at true δCP Data
-;/2 0 +;/2
QE 73.5 61.5 49.9 74
11 6.9 6.0 4.9 15
QE 7.9 9.0 10.0 7
QE 267.8 267.4 267.7 240
QE 63.1 62.9 63.1 68
7 (pred. 9.0)⌫e
89 (pred. 67.4)⌫e
21. T2K future prospects
3/15/18 Moriond EW 2018 21
² Stay tuned for summer result w/ doubled data in anti-neutrino mode
² Approved T2K statistics, 7.8 x1021 POT, can be accumulated by 2021
² J-PARC beam aims for upgrade & operation at > 1MW from 2021
² Hyper-K and DUNE are expected to start around 2026
à T2K-II, if extend T2K operation until 2026, will collect 20x1021 POT.
Such amount of data along with neutrino beamline upgrade & analysis
improvements makes T2K(-II) physics potentials even more interesting!
JFY
2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027
MRBeamPower[kW]
0
200
400
600
800
1000
1200
1400
POT]21
IntegratedDeliveredProtons[10
0
5
10
15
20
25
30
35
40
45
POT]21
.[10July)−(Oct.DeliveredProtons/Period
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
MR Power Supply upgrade
T2K-II Protons-On-Target Request
22. T2K future prospects (cont’d)
3/15/18 Moriond EW 2018 22
)°(CP
δTrue
200− 100− 0 100 200
=0CPδtoexcludesin2
χ∆
0
5
10
15
20
POT w/ eff. stat. improvements (no sys. errors)21
20x10
POT w/ eff. stat. improvements & 2016 sys. errors21
20x10
POT (no sys. errors)21
7.8x10
POT w/ 2016 sys. errors.21
7.8x10
23
θ2
sin
0.4 0.5 0.6
32
2
m∆
2.2
2.4
2.6
2.8
3
3−
10×
POT by 2014 , 90% C.L
POT, 90% C.L21
7.8x10
POT w/improvement, 90% C.L21
20x10
Stat. only
Systematics
² 3< or higher significance sensitivity to CP
violation if (CP close to - 1/2
² Systematic error has large impact
à Motivate for ND280 upgrade
² 1% precision of &m2
32, 0.5o - 1.7o
precision of !23 ( depend on the truth)
(arXiv:1609.04111 [hep-ex])
(CERN-SPSC-P357)
SuperFGD
Exciting programs!
Welcome new collaborators.
23. Summary
3/15/18 Moriond EW 2018 23
² Stable operation at 470-475 kW beam power allows T2K to double
neutrino data in one year (also expectedly double anti-neutrino
data by this summer)
² Updates in T2K oscillation analyses:
² New reconstruction and event selections: statistically effective
improvement by 30%
² Improving neutrino interaction model
² CP conserving values (0, 1) fall outside of the 2: C.L.
confidence/credible interval
² T2K-II, an extended program to collect 20x1021 POT, has been
proposed in order to achieve 3: C.L. to exclude CP conserving
values for favorable true value of (CP
*A few anime drawings taken http://higgstan.com
25. Expansion of Super-K fiducial volume
3/15/18 Moriond EW 2018 25
Samples New selection Previous selection
Signal (MC) Purity Signal (MC) Purity
QE 69.5 81.2% 56.5 81.4%
11 6.9 78.8% 5.6 72.0%
QE 7.6 62.0% 6.1 63.7%
QE 261.6 79.7% 268.7 68.1%
QE 62.0 79.7% 65.4 70.5%
APFit
FiTQun
fiTQun, a maximum-likelihood approach for event
reconstruction at SK, offers significant improvement in
performance and allows us to re-optimize the fiducial cut
² APFit - based fiducial volume: requires to have
reconstructed vertex > 2 m from the detector wall
² fiTQun - based fiducial volume: 2-dimensional cut on
² “wall”: minimum distance from vertex to the wall
² “towall”: distance along the particle track to the wall
26. Near Detector Data Fit
3/15/18 Moriond EW 2018 26
One sample for illustration Flux parameters @SK
+ −mode
² 2p-2h for neutrinos is enhanced by 50%
² 2p-2h shape is shifted, tend to increase &–
enhanced component to maximum
² Fitted value of RPA parameter for low Q2
(<1GeV2) is increase, i.e CCQE enhancement
27. Impact of Fake data on (CP
3/15/18 Moriond EW 2018 27
² Maximum shift in the NH 2< confidence
interval mid-point is 1.7%
² Maximum change to the NH 2< confidence
interval is 2.3%
² Impact on (CP confidence intervals is small!
28. On (CP exclusion
3/15/18 Moriond EW 2018 28
² The exclusion of (CP conserving values is
stronger than our expected sensitivity. Is it
reasonable?
² We throw 104 toy experiments in which normal
mass hierarchy is assumed and (CP are fixed at -
1/2 , but other oscillation parameters,
statistics, systematic parameters are varied
² 30% of experiments exclude (CP =0 at >2: C.L.
² 25% of experiments exclude (CP = 1 at >2: C.L.
(rad)CPδ
3− 2− 1− 0 1 2 3
ln(L)∆-2
0
5
10
15
20
25
30
T2K Run1-8 PreliminaryFinal systematics pending
68.27% of toys MC
95.45% of toys MC
CLσ2
Data I
(rad)CPδ
3− 2− 1− 0 1 2 3
ln(L)∆-2
0
2
4
6
8
10
12
14
16
18
20
T2K Run1-8 PreliminaryFinal systematics pending
68.27% of toys MC
95.45% of toys MC
CLσ2
Data
29. Fitting methods
3/15/18 Moriond EW 2018 29
² Fit simultaneously 5 signal samples selected at Far Detector, Super-K
² A binned likelihood approach to fit data
² Perform both Frequentist approach (two analyses) and Bayesian approach
(one analysis)
ln(L) =
NSK bins
X
i
NSK
i (~o, ~p) MSK
i + MSK
i ln
⇥
MSK
i /NSK
(~o, ~p)
⇤
+
1
2
NoX
i
NoX
j
oi(V o
ij) 1
oj +
1
2
Np
X
i
Np
X
j
pi(V p
ij) 1
pj
NSK
i /MSK
i is the observed/predicted number of events in the ith
bin
~o/~p are the oscillation/systematics parameters
V o
/V p
is the oscillation/systematics covariance matrix
30. ND280 detector
3/15/18 Moriond EW 2018 30
Aim to understand unoscillated , beam: constrains flux
and cross-section parameters
² Tracker, composed of Fine-Grained Detector (FGD)
and Time Projection Chamber (TPC), is central part
o Two FGDs: active target w/ scintillator only
(FGD1) or scintillator-water interleaved (FGD2)
o Three TPCs: mainly Argon (95%) filled, for
momentum measurement and particle ID
² 10 detector (POD) for water-scintillator target and 10
tagging
² Electromagnetic calorimeters (ECal) to detect gamma
rays and reconstruct 10
² Side muon range detectors (SMRD) to tag entering
cosmic muons or side-exiting muons
Key features for cross-section:
o Narrow flux spectrum , mean ~ 0.85 GeV
o Multiple targets: scintillator, water, argon, lead
o High final state ID resolution, charge separation