Latest Results from the MINOS Experiment

Justin Evans, University College Londonfor the MINOS Collaboration

NOW 20089th September 2008

9th September 2008 2Justin Evans

The MINOS Experiment

Two detectors to mitigate systematics e.g. neutrino flux or cross section

mismodellings Use measured near detector data to

predict what should be observed at the far detector

An observed νμ deficit at the far detector tells us about the oscillation parameters

Measuring νμ disappearance

Near detector, 1.0 ktonne, 1km from sourceFar detector, 5.4 ktonne, 735 km from sourceTracking, sampling calorimeters

Alternate steel and scintillator planes Magnetised to 1.3 T

Near detector

Far detector

9th September 2008 3Justin Evans

Use the measured ND energy spectrum to predict the FD spectrum:

Spread of pion decay directions smears neutrino energies Different energy spectra at the two detectors

Encode the pion decay kinematics into a beam transfer matrix Convert ND to FD spectrum

FD

Decay Pipe

π+Target

NDp

MC MC

CC νμ Beam Extrapolation

9th September 2008 4Justin Evans

Effect of uncertainties estimated by fitting systematically shifted MC in place of data

Analysis is still statistically limitedThree largest uncertainties included as

penalty terms in fit to data Relative (ND to FD) normalisation (4%) Absolute hadronic energy scale (10%) NC background (50%)

Systematic Uncertainties

Relativenormalisation

Absolutehadronicenergy

NC background

9th September 2008 5Justin Evans

FD data not looked at until the analysis was finalised

Expected 1065 ± 60 with no oscillations

Observed 848 events

Energy spectrum fit with the oscillation hypothesis

Far Detector Data

Best Fit:|m2| = 2.43x10-3 eV2

sin2(2) =1.00

P( )sin2(2)sin2

1.27m2L

E

2/NDoF = 90/97

9th September 2008 6Justin Evans

Constrained fit |m2| =(2.43±0.13)x10-3

eV2 (68% C.L.) sin2(223) > 0.90 (90%

C.L.) 2/NDoF = 90/97

Unconstrained fit |m2| = 2.33 x 10-3 eV2 sin2(223) = 1.07

Δ2 = -0.6

Allowed Region

9th September 2008 7Justin Evans

Alternative ModelsTwo alternative disappearance

models are disfavoured

Decay:

V. Barger et al., PRL82:2640(1999)2/ndof = 104/972 = 14disfavored at 3.7

Decoherence:

G.L. Fogli et al., PRD67:093006 (2003)2/ndof = 123/972 = 33disfavored at 5.7

P 1sin2 22

1 exp 2L2E

P sin 2 cos 2 exp L E2

9th September 2008 8Justin Evans

NC Event Selection

Excluded

Excluded

Excluded

9th September 2008 9Justin Evans

NC Energy Spectrum

Far detector NC energy spectrum

Shown at θ13= 0 and θ13 at Chooz limit

Fraction of νμ which oscillate to

sterile neutrinos

No νe appearance

fs = 0.28+0.25-0.28 (stat.+sys.)

fs < 0.68 (90% c.l.)

With νe appearance

fs = 0.43+0.23-0.27 (stat.+sys.)

fs < 0.80 (90% c.l.)

9th September 2008 10Justin Evans

νe spectrum is heavily background dominated Dominant backgrounds: NC and CC νμ events

Near detector shows a large discrepancy between selected and expected νe energy spectrum

Using data-driven methods to correct the data/MC differences

νe Appearance Analysis

9th September 2008 11Justin Evans

With 3.25x1020 protons, at the Chooz limit, we expect 12 signal and 42 background events

At current exposure MINOS will reach the Chooz limitIn the future MINOS can improve on the Chooz limit by a factor of ~2

(normal hierarchy)

νe Appearance Sensitivity

9th September 2008 12Justin Evans

MINOS has made a new measurement of the atmospheric oscillation parameters |m2| =(2.43 ± 0.13) x 10-3 eV2 (68% C.L.) sin2(223) > 0.90 (90% c.l.)

Alternative models disfavoured Decay at 3.7σ, decoherence at 5.7σ

Measurement of the NC event spectrum places a limit on oscillations to sterile neutrinos fs < 0.68 (90% c.l.)

Work is ongoing towards a νe appearance measurement At the current exposure, MINOS’s sensitivity will rival the Chooz limit

Summary

9th September 2008 13Justin Evans

9th September 2008 14Justin Evans

Backup Slides

9th September 2008 15Justin Evans

CC νμ Event SelectionAim to separate charged and neutral current νμ interactions

Four variables combined using a k-nearest-neighbour algorithm Track length Mean signal in track planes Transverse track profile Signal fluctuation along the track

9th September 2008 16Justin Evans

NC Energy Spectrum

Far detector NC energy spectrum

Shown at θ13= 0 and θ13 at Chooz limitEnergy Range (GeV)

0—3 0—120

Data events 100 291

MC events

(θ13 = 0)

115.16 ± 7.61

292.63 ± 15.02

Significance 1.15σ 0.10σ

9th September 2008 17Justin Evans

νe spectrum is heavily background dominated Dominant backgrounds: NC and CC νμ events

Near detector shows a large discrepancy between selected and expected νe energy spectrum

Using data-driven methods to correct the data/MC differences

νe Appearance Analysis

9th September 2008 18Justin Evans

Event Topologies

νμ CC Event NC Event νe CC EventUZ

VZ

long μ track & hadronic activity at

vertex

short, with typical EM shower profile

short event, often diffuse

3.5m 1.8m 2.3m

Monte CarloEν = Eshower + pμ