Atmospheric Neutrinos and the Quest for Neutrino Oscillations

1
(No Transcript)
2
(No Transcript)
3
Neutrino Physics at the Crossroads
Hugh Gallagher Tufts University Feb. 28, 2003

• Neutrino History
• Neutrino Oscillations
• Experimental Evidence
• Solar
• Atmospheric
• New Long-Baseline Experiments
• Future Experiments

4
50s 60s The Age of Discovery
Neutrinos have proved to be a valuable probe of
the structure of matter and the nature of
electroweak interactions.
Observation confirmed (2000) DONUT Experiment
From LEP we know that Nn 3
Reines and Cowan First n detection
1952-3 (1995 Nobel Prize)
Brookhaven 2 n Experiment (Lederman, Schwartz,
Steinberger 1988 Nobel Prize)
5
70s 80s Age of Exploration
Bubble chambers BNL, ANL, FNAL, CERN, Serpukhov

• hadronic weak currents
• observation of neutral currents
• cross section measurements

SN9187A n detection confirmed astrophysical
predictions!
1950
1990
1960
1970
1980
2000
counter experiments CDHS, CHARM, CHARM
II, CCFR, NuTEV

• high statistics (100k events)
• structure functions (F2, F3)
• parton universality quarks are quarks!
• Electroweak studies sin2(qw)
• strange sea studies
• cross sections

6
80s-90s The Age of Serendipity
First evidence for neutrino oscillations (and
neutrino mass) came from experiments designed to
measure something else entirely!
the solar neutrino problem
the atmospheric neutrino anomaly
Kamioka experiment Nobel prize 2002 Masatoshi
Koshiba
set out to confirm the standard solar model
Large underground experiments designed to detect
proton decay originally worried about atmospheric
neutrinos as a potential background
Homestake experiment Nobel prize 2002 Ray Davis
7
Neutrino Properties Mass
Do neutrinos have mass?? What kind of mass?
Dirac (4-component) DL 0 bb0n Majorana
(2-component) DL2 bb0n
8
The MNS Matrix
The Maki-Nakagawa-Sakata matrix
Oscillations between all 3 flavors at each Dm2
3 sets of 3 equations like these
9
Searching for Oscillations
For two generations, a good approximation is
Dm2 and sin2(2q) are the physics we are after
10
The Evidence for Oscillations
P(n?n)
Dm2 3 X 10-3 eV2 sin2(2q) 1.0 L 730 km
seeing oscillations
En (GeV)
11
Solar Neutrinos
12
Sudbury Neutrino Observatory
Elastic Events ne 6 x(nm,nt)
3 events / day good directionality
CC Events ne only
30 events / day spectral information
NC Events ne, nm, nt

• Results on data Nov. 1999 May 2002
• 1 kton ultra-pure D2O
• 12 m diameter acrylic vessel
• 9456 PMTs surrounded by light water
• PMT times and hit patterns are used to
• determine location, direction, and energy

30 events / day flavor independent!
phase II
Images courtesy of the SNO collaboration,
http//www.sno.phy.queensu.ca
13
SNO
The total flux is measured with the NC reaction
(106 cm-2 s-1)
in good agreement with the prediction
(106 cm-2 s-1)
14
KamLand
KamLand is a long-baseline reactor
neutrino experiment sensitive to the large mixing
angle solution of the solar neutrino problem.
First Results from KamLand Evidence
for Reactor Anti-Neutrino Disappearance PRL 90
021802 (2003) shows the first evidence for
neutrino disappearance from a reactor!
15
Kamland
Neutrino Oscillations enhanced by the MSW effect
in the sun. day/night differences SK and
SNO spectral distortions (KamLand)
16
Atmospheric Neutrinos
Originally studied as a potential background for
underground proton decay experiments. First hint
of oscillations... 1986!
IMB collab.
17
Oscillations with Atmospheric ns
18
Flux Calculations

• Ingredients
• Primary cosmic ray flux
• (time-dependent)
• 2. Geomagnetic model
• 3. Hadronic interactions
• in the atmosphere

Flux of primary cosmic rays (from Gaisser and
Honda, Ann. Rev. Nuc. Part. Sci (2002))
19
Flux Calculations (2)
overall uncertainty in the flux normalization is
20-25
The most robust prediction is that of the flavor
ratio nm / ne.
20
Soudan 2
Constructed 1988-1993.
21
The Detector
224 1m x 1m x 2.7 m modules
963 ton total mass
5.90 fiducial-kton yr exposure
The detector is surrounded by a 1700 m2 veto
shield which provides nearly 4p coverage for the
identification of charged particles entering /
exiting the detector cavern.
22
Event Pictures
contained events
nm
ne
23
Data vs. No Oscillations
Events/0.5
Events/0.4
24
Feldman-Cousins Analysis
One of the most controversial statistical
questions in physics is how to report a
measurement which is close to the edge or
even outside of the allowed physical
region.---Review of Particle Properties
Recommended approach is presented in G.J.
Feldman and R.D. Cousins, Phys Rev. D57, 3873
(1998).
25
Results
Best Fit ?m2 0.007 eV2 sin22?
0.98 fn(data/mc) 0.90 MC ? Bartol '96
Dm2 eV2
sin2(2?)
26
SuperK
50 kton (22.5 fiducial volume) 11,146 PMTs in the
inner tank 1885 PMTs in outer region 79.4
kton-yr data sample from SuperK I Measure fully
contained n, partially contained n, and
throughgoing muons.
SuperK I 4/1996 7/2001 (Accident in Nov. 2001
destroyed 6777 inner PMTs). Rebuilt 11/2001
10/2002 SuperK II 10/2002 ? now reduced
PMT coverage SuperK III before JHF
27
SuperK Zenith Angles
ne data are consistent with expectation. nm
are disappearing (in a way consistent with
oscillations) nm ? what??
28
SuperK nm ? ?
nm ? ne ruled out by ne data nm ? nt or nm ?
ns?
t are hard to identify. Ethreshold 3.5
GeV decays in 3 x 10-13 sec primarily
to hadrons (65)
Look for t-like events (neural networks
2s) NC events (up vs. down) matter effects
(up going)
nt and ns propagate differently through matter
29
All Experiments
Serendipity!
That factors completed unrelated to neutrino
physics allowed us to observe effects in solar
and atmospheric neutrinos that were nearly
maximal.
size of earth ? Dm2 10-3 eV2 solar density,
earth-sun distance ? Dm2 10-5 eV2.
For 3 massive neutrinos there are two
independent Dm2.
There is still one loose end ...
30
LSND
A similar beam dump experiment, KARMEN, failed to
see a signal.
31
miniBoone at Fermilab
8 GeV booster protons ? 1 GeV n 12 m sphere
filled with mineral oil 0.5 km from source Expect
huge signal (1000s of events) Now running
If miniBoone confims LSND things will get very
interesting!
32
All Experiment Allowed Region
If we ignore LSND, what have we learned, and what
questions remain?
Dm12 10-5 eV2
nm
ne
Dm23 10-3 eV2
nt

• Better precision on parameters
• Strictly speaking, no one has yet observed
  Oscillations
• What are the mixing modes nm ? nt or nm ? ns?
• q13 and dCP in the MNS matrix
• Normal or Inverted mass hierarchy
• Are neutrinos degenerate? measuring mn not Dm2.
• Majorana or Dirac particles?

33
The Age of Precision
Neutrino Oscillation experiments to this point
have largely relied on natural neutrino sources
(if you consider nuclear reactors natural).
Limited ability to perform cross-checks. Low
statistics Results ultimately limited by
knowledge of production mechanisms L, E not
optimal for observation of oscillation signatures
The new era relies on accelerators (Fermilab,
KEK, CERN, ???) to create high-intensity, well
understood neutrino beams aimed at distant
detectors optimized for particular oscillation
signatures.
34
MINOS (Main Injector Neutrino Oscillation Search)
MINOS (Fermilab to Minnesota) L 730 km 2004
near detector
far detector
35
NuMI Beam
zoom lens Vary the relative distances of the
source and focusing elements

• Expected CC Events Rates in Minos 5kt detector
• High 16,000 ev/yr
• Medium 7,000 ev/yr
• Low 2,500 ev/yr

36
MINOS
408 / 484 planes currently installed
37
MUX boxes

• Strips assembled into 20- or 28-strip modules
• Fire resistant aluminum light cases
• 2-ended WLS fiber readout
• WLS to clear fiber cables at module connectors
• MUX boxes route 8 (1 in Near Detector) fibers to
  one MAPMT pixel
• Custom front-end electronics
• Commercial processors used for Trigger and DAQ
  system

38
MUX boxes
Sheila Dupuis and Lisa Guittarr
½ of the MUX boxes used in the experiment were
built and tested here at Tufts.
39
MINOS Construction

• MINOS construction will complete by summer
• Outfitting at FNAL and beam installation this
  year and next.
• Beam Nov 2004.

40
MINOS Physics Tests
41
MINOS Status
42
The MINOS Mural
The Soudan Underground Laboratory is operated
partly under the auspices of the Minnesota
Department of Natural Resources. The MINOS
hall will be part of a special underground
scientific tour open to the public.
43
Outreach
44
K2K (KeK to Kamiokande)
First long baseline experiment to run! KEK 12 GeV
PS ? 1 GeV n 1.1 ms beam spill every 2.2 s
monitor pions and muons to predict beam near
detector similar to far (SuperK) Future 50 GeV
JHF ? 2 orders of magnitude higher flux!
45
K2K
P(rateshape) 99.3 CL
from J. Wilkes, hep-ex/0212035
46
CERN OPERA
nm ? nt appearance at atmospheric Dm2

• Emulsion bricks and electronic tracking
• Automated scanning finds events in emulsion
• Search for t decay (similar to DONUT)
• Very low backgrounds
• Signal strongly depends on true value of
• Dm2. (20 nt events measured at
• Dm2 3.5 10-3 eV2 in five years)

47
The Future Measuring q13
Look for nm ? ne at the atmospheric neutrino Dm2.
(Current limits are around 1) Dm210-3 means we
want L/E 103 (km/GeV) Backgrounds come
from ne in the beam (0.7) NC events t ? e
decay (17)
For the NuMI beam to Soudan 1.27Dm2L/E p/2
?E 1.5 GeV below tau threshold!
NC

• Requires
• A big detector
• good pattern recognition (e/p0)
• more neutrinos in the signal region!
• less neutrinos out of the signal region!

ne
A. Para
signal
48
NuMI Off-Axis
On axis En0.43Ep
49
NuMI Off-Axis
A. Para, M. Szleper, hep-ex/0110032
Minnesota 730 km Canada 990 km
50
Nu Factories

• High intensity ne, nm from muon storage rings
• highly collimated beams
• well understood beams
• both nm and ne
• ne and ne to study matter effects

51
Conclusions
We are entering a new era of experimentation
where neutrino beams will be used to probe the
physics of the lepton sector with high precision.

• Observing Oscillations
• Demonstrating oscillation modes (appearance
  experiments)
• Resolving mass hierarchy
• q13
• CP violation in the lepton sector

Theories beyond the Standard Model will have to
explain the smallness of the neutrino mass.
Precision measurements of the lepton sector of
the Standard Model provide valuable
constraints on the form these new theories can
take!
52
The Neutrino World
53
A Module
54
PCE Picture
55
Flavor Ratio

• We measure via the flavor ratio of ratios

MC null osc. R 1.0 if no osc

• Use all contained events having identified flavor
  (multiprongs events now included).
• Rraw 0.73 ? 0.10

Flavor tag

• De-convolve using Flavor Identification Matrix

True flavor
56
Monte Carlo
57
Production Heights
58
Azimuthal Angles

• Observe azimuthal symmetry for both ?e and ??
  fluxes
• - as expected.

• The ?? is uniformly depleted relative to null
  oscillation expectation.

59
Feldman-Cousins
We assume nm ? nt ne is unaffected. For
points ( i, j ) over the physical region of the
( sin2(2Q)i , log10Dm2j ) plane, we do a fit of
the oscillation expectation to our data at each
point using Maximum-Likelihood (Ldata)ij
L(sin2(2Q)i , log10 Dm2j fn ) where fn ?
flux normalization (Bartol '96 atm. flux). We
find (Ldata)min , and plot (DLdata)ij (Ldata)ij
- (Ldata)min
From an analysis carried out by Mayly Sanchez.
60
Feldman-Cousins
At each of 2000 points ( i, j ) ( sin2(2Q)i ,
log10Dm2j ) in the physical region, we run 1000
simulated experiments. Find (DL90)ij such that
(DLsim)ijlt(DL90)ij for 90 of the experiments at
( i, j ). The surface defined by local DL90 over
the oscillation plane
DL90
Sin2(2Q)
log10Dm2
61
All Data Plot
Hi-Res nm CC
PCE nm CC
Lo-Res nm CC
Contained ne CC
Lo-Res Multiprongs
NC / Flavor Ambiguous
62
Battistoni 3D flux
Best Fit ?m2 0.007 eV2 sin22?
0.98 fn(data/mc) 1.05 MC ? Battistoni 02
Dm2 eV2
?2/dof34.7/28
sin2(2?)
63
Misc
The Sun
Supernovae
Cosmic Rays
Particle
Power
Reactors
Accelerators

others
64
Rock events
Rock event sample- identified by presence of
shield hits.
The residual zero-hit background is calculated by
fitting contained event vertex-depth
distributions to a combination of rock and
neutrino Monte Carlo distributions.
Rock-muon associated background correction in the
hires sample is about 6.
65
Event Categories
66
Data vs. Oscillations
sin2(2?) 1.0
Events/0.4
Events/0.4
67
Zenith Angle
68
SuperK Flavor Ratios