Super-Kamiokande
Atmospheric Neutrino Results
Kimihiro Okumura
ICRR Univ. of Tokyo
([email protected])
11-September-2005
C2CR2005, Prague
11-September-2005
C2CR2005, Prague
1
Outline
Super-Kamiokande-I Results




new nm→nt oscillation analysis
search for tau neutrinos
search for q13
neutrino flux study
Status of Super-Kamiokande-II Analysis
11-September-2005
C2CR2005, Prague
2
Super-Kamiokande
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
SK-I
SK-II
accident
reconstruction
SK-III
• 50kton cylindrical water
Cherenkov detector
(22.5kt fiducial vol.)
• 1000m underground
(2700m water equiv.)
SK-II
• optically separated into
ID and OD
SK-I
11146
40%
Num. of inner detector PMTs
Photocathod coverage
PMT enclosure :
Acrylic (front) and
Fiberglass (back)
5182
19%
SK-III will start on June 2006.
11-September-2005
C2CR2005, Prague
3
Observation of Atmospheric Neutrino
Event classification
Fully Contained
(En ~1GeV, ne nm)
Partially Contained
(En ~10GeV, nm)
sensitive energy: 0.1GeV ~ 10TeV
11-September-2005
Stopping m (En ~10GeV, nm)
Through-going m (En ~100GeV, nm)
C2CR2005, Prague
4
Zenith Angle Distributions (SK-I)
e-like
m-like
<
<
>
>
11-September-2005
FC and PC : 1489 days
Up-going m : 1646 days
C2CR2005, Prague
nm–nt oscillation (best fit)
null oscillation
5
Number of events
Previous oscillation analysis
Zenith angle analysis :
UP/DOWN ~ sin2(2q)
good for sin2(2q)
Allowed region of
previous zenith and L/E
cosq
Data / MC
L/E analysis :
dip ~ l/2
good for Dm2
Question: Can single analysis give the best sin2(2q) and Dm2 ?
11-September-2005
C2CR2005, Prague
6
Oscillation analysis with finer binning
1/2 osci.
Important energy range to see the
oscillation dip = multi-GeV
Finer energy bins for multi-GeV events
previous zenith (180bins)
Expected sensitivity with 5yr MC
(sin22q,
Dm2)truth
= (1.00,
L/E
finer binning (370bins)
2.510-3 eV2)
zenith angle analysis with finer energy binning
gives better sensitivity for both sin2(2q) and Dm2
11-September-2005
C2CR2005, Prague
7
Allowed region with finer binning
Previous zenith-angle analysis
L/E analysis
90% CL allowed region:
sin22q > 0.93
2.0 < Dm2 < 3.0×10-3eV2
11-September-2005
C2CR2005, Prague
8
Search for CC nt events
CC nt events
nt
nt
t
hadrons
CC nt
MC
hadrons
nt events are accompanied with many hadrons
event topology similar to other hadronic events
 likelihood (or NN) analysis
nt appears only for up-going events
 use down-going events for
control sample
11-September-2005
C2CR2005, Prague
Only ～ 1.0 CC nt FC
events/kton・yr
9
Tau-like events selection criteria
Visible energy > 1.33 GeV (multi-GeV energy)
Multi-ring events (event w/ hadron particles)
Most energetic ring is e-like (showering event)
log(likelihood) > 0
or
NN > 0.5
11-September-2005
C2CR2005, Prague
10
Likelihood variables
Down-going data
visible energy
Preliminary
t MC
Max distance
from primary and
decay-e vertex
Down-going nm, ne MC
Max. momentum
for m-like ring
Sphericity in the
lab fame
Clustered
sphericity in the
center of mass
frame
Number of
candidate rings
11-September-2005
C2CR2005, Prague
11
Tau likelihood distributions
downward
upward
Cut
Cut
Number of events
DATA
BG-MC
tau-MC
DATA
BG-MC
tau-MC
(×1.82t MC)
t-likelihood
t-likelihood
Tau-like events : Likelihood > 0
11-September-2005
C2CR2005, Prague
12
Zenith angle dist. and fit results
NN analysis
Likelihood analysis
Data
Number of events
×1.82 t-MC
×1.93 t-MC
Data
nm, ne, & NC
background
nm, ne, & NC
background
cosqzenith
cosqzenith
Fitted # of t
events
145±48(stat)
+9 / -36 (osc. para. uncertainty)
152±47(stat)
+12 / -27 (osc. para. uncertainty)
Expected # of
t events
79±31(syst)
79±31 (syst)
11-September-2005
C2CR2005, Prague
13
Search for non-zero q13
Three flavor oscillation analysis with Dm2solar=0 assumption :
 1.27Dm2 L 

P(n m n e )  sin q 23  sin 2q13  sin 
E


2
2
2
1+multi-ring, e-like, 2.5 - 5 GeV
P(nm →ne)
cosqzenith
Electron appearance
s213=0.05
s213=0.00
null oscillation
cosqzenith
Neutrino Energy (GeV)
Non-zero q13 will enhance nm →ne oscillation in multi-GeV energy
due to matter effect in Earth
Electron appearance is expected in upward-going events
11-September-2005
C2CR2005, Prague
14
Multi-GeV electrons
Zenith angle
UP/DOWN asymmetry
Data
No oscillation
single-Ring
electron
Three flavor best fit
(normal hierarchy)
multi-Ring
electron
No significant excess due to matter effect in upward-going
multi-GeV electron sample
11-September-2005
C2CR2005, Prague
15
Allowed region for q13
Preliminary
excluded by CHOOZ
Dm2 (eV2)
90%CL
sin2q13
Normal hierarchy
(Dm2>0)
99%CL
c2min/ndf = 376.82/368
@(2.5x10-3, 0.5, 0.0)
sin2q23
sin2q13
Distributions are best fitted when sin2(q13)=0
Consistent with two flavor nm  nt oscillation
sin2(q13)<0.14 allowed in 90% C.L. from SK-I data
11-September-2005
C2CR2005, Prague
16
Atmospheric n flux measurement
Neutrino flux calculation
cosmic-ray proton flux
nm
Recent neutrino flux calculations have improved due to accurate
measurement of comic-ray protons / muons
Approximately 20% difference still exists among flux models in
higher energies
11-September-2005
C2CR2005, Prague
17
ne and nm energy spectrum
Events / yr / GeV
Events / yr / GeV
Preliminary
data
expectation
sub-GeV
multi-GeV
multi-GeV
sub-GeV
ne
nm
PC
upmu-stop
upmu-thru
SK-I ne and nm event rates
are compared with
expectation in each
energy range
Neutrino energies are
estimated from simulation
Neutrino oscillation
included for nm expected

nm→nt (Dm2, sin22q) =
(2.5x10-3 eV2,1.0) assumed
Neutrino energy (GeV)
11-September-2005
C2CR2005, Prague
18
Comparison with flux models
Preliminary
Neutrino flux models:
ne

Data / Expectation


HKKM04
Bartol
Fluka
Phys. Rev. D70 043008 (2004)
Phys. Rev. D70 023006 (2004)
hep-ph/0305208
Systematic uncertainties:



nm
detector efficiency
n cross section ~10%
n oscillation
~4% for nm
Results:

absolute flux and spectrum
are consistent within
systematic errors

HKKM04 and Fluka underestimated high energy nm flux
Error bar : (stat.)2+(sys.)2
11-September-2005
C2CR2005, Prague
19
New HKKM flux calculation
Preliminary
ne
Data / Expectation
HKKM flux revised:


nm
modified hadron interaction
model
see Dr. Honda’s talk
Agreement in high energy
nm becomes better
Error bar : (stat.)2+(sys.)2
11-September-2005
C2CR2005, Prague
20
Zenith Angle Distributions (SK-II)
e-like
m-like
<
<
>
>
11-September-2005
FC and PC : 627 days
Up-going m : 609 days
C2CR2005, Prague
nm–nt oscillation (best fit)
null oscillation
21
Contours for SK-II
SK-II
SK-I
SK-II data are consistent with SK-I data
SK-I + SK-II combined analysis : next step
11-September-2005
C2CR2005, Prague
22
SK-II L/E analysis
L/E analysis was carried out for the SK-II data with the identical
selection criteria as those in SK-I.
SK-II
SK-I
Decoh.
Decay
Osc.
Consistent with SK-I.
Oscillation still gives the best fit to the data.
11-September-2005
C2CR2005, Prague
23
Summary
SK-I analysis results:

New nm → nt oscillation analysis with finer binning
gives more stringent constraint :
• sin22q > 0.93, 2.0<Dm2<3.0x10-3 eV2 (90%C.L.)

Data are consistent with nt appearance

No evidence for multi-GeV electron appearance due
to non-zero q13

Neutrino flux calculation models are compared and
consistent with data
SK-II data was analyzed independently and
consistent with SK-I
11-September-2005
C2CR2005, Prague
24
End