1Caius Howcroft
Atmospheric Neutrinos at
MINOSCaius Howcroft
Caltech, 17/May/2005
✦Atmospheric Neutrinos
✦MINOS
✦MINOS Far Detector
✦Type of Atmospheric Neutrinos at MINOS
✦Two MINOS Atmospheric Analyses
2Caius Howcroft
Atmospheric Neutrinos
• Produced in the atmosphere from interactions of primary cosmic rays.
-( )!
µ-( )!
Down-!
Up-!
µ
"
p,He
#,$
µ
Detector
ZenithZenith
Detector
%
%
"
Earth
#&"
-( )!e
p + N → π± + X
π± → µ± + νµ(ν̄µ)
µ± → e± + νe(ν̄e) + νµ(ν̄µ).
1
νe + ν̄e
νµ + ν̄µ
= 1/2
νµ/ν̄µ ≈ 1
L
3Caius Howcroft
Atmospheric Neutrinos (2)
• First evidence of neutrino oscillations came from Super Kamiokande (1998).
0
100
200
300
400
500
600
-1 -0.5 0 0.5 1
Num
ber
of E
vents
Sub-GeV e-like
0
100
200
300
400
500
600
-1 -0.5 0 0.5 1
Sub-GeV µ-like
0
20
40
60
80
100
120
140
160
180
200
-1 -0.5 0 0.5 1
cos!
Num
ber
of E
vents
Multi-GeV e-like
0
50
100
150
200
250
300
350
-1 -0.5 0 0.5 1
cos!
Multi-GeV µ-like + PC
(νµ
ντ
)=
(cos θ sin θ
− sin θ cos θ
) (ν1
ν2
)
P (νµ → ντ ) = sin2 2θ sin2(δm2L/4E)
downup
νμ↔ντ OSCILLATIONS
4Caius Howcroft
MINOS Goals
• Demonstrate the oscillation behaviour
• Measure Δm
• Search for sub dominate ν appearance
223
e
Provide very high statistics discrimination against other models for neutrino disappearance , e.g. decoherence, neutrino decay, ...?
to better than 10%
θ13 3σ discovery limit is factor 2 improvement on CHOOZ current limit.
In addition: MINOS Far Detector capable of first direct observation of atmospheric ν and ν oscillations
5Caius Howcroft
MINOS: Basic Idea(Main Injector Neutrino Oscillation Search)
Position of minimum Δm2
Depth of minimum sin22θ
Near(unosc)
Far(oscillated)
735km
νμ
7Caius Howcroft
MINOS Far Detector Site
• Located in former iron mine in northern Mn, USA.
• Also the home of Soudan1&2 (ret.) and CDMS.
• MINOS is 2341 ft. (2070 mwe) below the surface.
MINOS
Soudan 2/CDMS IIshaft
8Caius Howcroft
The Far Detector
• 8m octagonal steel scintillator tracking calorimeter.
• Two 15m sections (supermodules)
• 5.23 kton total mass.
• Each supermodule has a 1.5T toroidal magnetic field
• Hadronic energy resolution ~55%/√E
• Completed in Aug 2003.A single supermodule
9Caius Howcroft
The Far Detector (2)
• Steel-Scintillator sandwich, each layer (plane) consists of a 2.54 cm steel +1 cm scintillator
• Each scintillator plane divided into 192 x 4cm wide strips
• Alternate planes have orthogonal strip orientations, U and V ⇒ 3D trackingU V U V U V U V
steel
scintillator
orthogonal orientations of strips
Scintillation light collected by fibres glued into each strip and read out by multi pixel PMTs
12Caius Howcroft
Live Time• Since completion in August 2003 the
Far Det has been taking cosmics data.
• Beam switched on 1st March 2005 at which point the Far Detector had collected 420 days of physics quality data, excellent for a detector still in commissioning stage.
• Total of 5.99 kton-years of data suitable for atmospheric neutrino studies, c.f. Soudan 2's 7.36 kton-years
• Will still take cosmics data during beam running.
Days since 2003-07-01
0 100 200 300 400 500
To
tal live t
ime
0
100
200
300
400
500
100% live time
live time
MINOS PRELIMINARY
13Caius Howcroft
What's in an Event?
DATA
UZ
VZ
Timing
Calorimetic
➪Two Views (U-Z and V-Z) are combined to give 3D tracks and showers.➪Event timing information ➪Calorimeteric information➪Event charge and muon momentum from curvature of tracks in B-field.
(2.4 ns)
14Caius Howcroft
Atmos ν Pros and Cons• MINOS Far Detector has a lot going for it:
• Deep (2070 mwe) 100,000 reduction in cosmic muon and large.
• Magnetic field, event charge and muon momentum, even muons that leave the detector.
• Potential problem, 80% of the detector surface is uninstrumented.
Event appears to start 1m from detector edge
Hit in VetoShield
15Caius Howcroft
Event Classesν
µ
ν
µ
ν
µ
ν
µ
Fully Contained
Partially ContainedDownward-going
Partially Containedupward-going
neutrino induced rock muon
FC
PCDN
PCUP
νinducedµ
Sneaky Stopping Cosmic Muons
Sneaky Thru-going Cosmic Muons
StoppingCosmic Muons (Direction Wrong)
Thru-going Cosmic Muons (Direction wrong)
Cut: Containment, topology
Cut: Containment, topology
Cut: timing
Cut: timing
16Caius Howcroft
Neutrino Induced Muons• Look for events coming from below the horizon. The flat
overburden at Soudan means can also look for events slightly above horizon.
• Event selection is based on event timing, timing resolution 2.4ns per channel. Require 20 planes and 2m tracks, look at event timing 1/β (c/v).
Downward going cosmic ray muons
Upward going events
Select 91 events.
B. Rebel (FNAL)
18Caius Howcroft
Neutrino Induced Muon(3)
NUANCE Monte Carlo: ✦Bartol'96 flux ✦MC normalised to data
Event Charge ID ✦Use the charge of the muon to separate ν and ν events ✦Efficiency is a function of, muon momentum, track length and vxB
Charge ID is a work in process
B. Rebel (FNAL)
19Caius Howcroft
FC PC-Down Selection
• FC and PC-Down have same background: sneaky cosmic muons. Therefore, use same cuts.
• Rejection factor of 1 in 107 required to achieve a signal to background ratio of 10:1.
• Use Monte Carlo signal (Barr et al.) and cosmic muon background to develop cuts based analysis to reduce signal:background to 1:1, then apply veto shield (~97% efficient).
• Use veto shield to measure remaining background in sample.
20Caius Howcroft
FC/PCDN:Containment
• Define a fiducial volume:
• 0.5m from detector edge
• 5 planes from each supermodule end
• Cuts:
• FC - require both ends be 'contained'
• PCDN - require upper end 'contained' and lower end non-contained.
Fiducial Distance / m 0 0.5 1 1.5 2 2.5 3 3.5 4
Eve
nts
10
210
310
410
510
610
Data (102 livedays)
Cosmic Muon MC
1000!CC MC µ"
Muons entering through coil hole. Removed with 40cm radial cut.
Signal:Cosmic ~ 1:300
MINOS PRELIMINARY
21Caius Howcroft
FC/PCDN: Δz• Background dominated by steep cosmic ray muons
• Cut out sneaky muons
Δz
/ m Z
! 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Even
ts
-110
1
10
210
310
data
µ MC cosmic
" MC atmos
Signal:Cosmic ~ 1:15
MINOS PRELIMINARY
22Caius Howcroft
FC/PCDN: Q/Direction• Remaining background is steep, highly curving cosmic ray muons
• Deposit lots of charge in first plane (Qvtx)
|z!|cos
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
/ P
Ev
txQ
0
50
100
150
200
250
300
"MC atmos
|z!|cos
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
/ P
Ev
txQ
0
50
100
150
200
250
300
µMC cosmic
zenith!cos
-1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1
/ P
Ev
txQ
0
50
100
150
200
250
300
"MC atmos
zenith!cos
-1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1
/ P
Ev
txQ
0
50
100
150
200
250
300
µMC cosmic
z
y (zenith)
23Caius Howcroft
FC/PCDN: Q/Direction(2)• Define 'steep' to be:
• Then require Qvtx<100
/PEvtxQ0 50 100 150 200 250 300
Even
ts
0
20
40
60
80
100
120
140
160
180
200
Data
µMC cosmic
10! "MC atmos
|<0.5z#>0.7, |coszenith#cos
cos θzenith > 0.7 or | cos θz| < 0.5
Signal:Cosmic ~ 1:1.5
MINOS PRELIMINARY
24Caius Howcroft
FC/PCDN:Veto Shield
• Added after detector construction, rather odd shape due to readout cables. Event so: 99.9% acceptance.
• Same scintillator strips used in detector, but parallel to Z axis. 4 sections (2 per supermodule) each 8 meters long.
X
Y
UV
Z out of page
WestEast
B-Field
Coil hole
Shield Layer
25Caius Howcroft
FC/PCDN: Veto Shield (2)
• An event is 'vetoed' if energy is deposited in the shield with 100ns of the event.
• Measure veto shield efficiency using steep stopping cosmic ray muons.
• The signal rejection is dominated by shield noise:
Veto Efficiency = 97.1±0.2(sys) %
Signal Inefficiency = 2.5±0.2(sys) %
26Caius Howcroft
FC/PCDN: Veto Shield• Applying veto shield:
• Know veto shield efficiency and number of vetoed events, therefore can estimate cosmic ray muon background in selected sample.
•
Measured Cosmic Muon BG
MC Cosmic Muon BG
Events 4.4±0.4(stat)±0.3(sys) 4.9±0.5(stat)±0.7(sys)
DataTotal MC
(no oscillations)Total MC
(Δm2=0.0025eV2)
Total 94 120.1±24.0 94.3±18.9
PCDN 25 22.6±4.3 20.5±4.1
FC 69 98.5±19.7 73.8±14.8MC predictions include all backgrounds
27Caius Howcroft
Partially Contained - UP• Very different backgrounds to FC/PCDN. Stopping cosmic
ray muons with incorrect direction from timing, therefore timing cuts do the work here.
• However still need a containment and a Δz cut to remove upgoing neutrino induced muons (reduced to 1% of signal)
Signal:BG ~ 1:15
28Caius Howcroft
PCUP:Timing • Direction of track is determined from timing of hits along
its length.
• The rms of the hits times about the track is calculated for the two hypotheses of +c or -c. The hypothesis with smallest rms is selected as track direction.
+c hypothesis+c hypothesis
29Caius Howcroft
PCUP:Timing• Use difference in RMS to selected events with good timing
• Remaining background is estimated from high statistics Monte Carlo studies <0.36 (68% c.f.). A x-check is made by extrapolating data distribution into signal region giving 0.5 events.
/ ns down
- RMSup RMS-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0
-110
1
10
210
data
µ MC cosmic
! MC atmos
DataTotal MC
(no oscillations)Total MC
(0.0025eV2)
Events 13 19.3±3.9 10.7±2.1
MINOS PRELIMINARY
30Caius Howcroft
Event Summary
DataMC νµ/νµCC
no oscillations
MC νµ/νµCC
0.0025eV2Cosmic Muon
MCOther
Backgrounds
Total 107 128.9±25.8 93.5±19 4.9±0.7 5.6±1.1
• Summary of selected events
Assumes no oscillations
Barr et al flux. Sys error dominated by (flux and
cross-section)
Other Backgrounds:ντCC
Rock muons νeCC
Neutral Current
32Caius Howcroft
Event Direction
• Define a 'high resolution' sample for FC events, by applying similar timing cuts to PCUP.
• Place 30 events into a 'low resolution' sample which have an ambiguous direction.
Correct Track DirectionBefore Cut = 96.0 %After Cut = 99.9 %
FC events
33Caius Howcroft
Up/Down Ratio
• For high-resolution events can calculate the up-down ratio (Robs) and compare it to MC expectation in the absence of oscillations (RMC).
zenith!cos
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Ev
en
ts
0
5
10
15
20
25
30DataMC null oscillations
2 = 0.0025 eV23
2 m"MC
MC Background
DataMC null oscillations
2 = 0.0025 eV23
2 m"MC
MC Background
Robs/RMC = 0.62±0.15(stat)±0.03(sys)
MINOS PRELIMINARY
34Caius Howcroft
Charge Separation• Measure charge from curvature of track in magnetic field.
• Require that |Q/P|/ σQ/P > 2.0
✦Track length✦Muon momentum✦vxB
Q/P!Q/P/
-20 -15 -10 -5 0 5 10 15 200
5000
10000
15000
20000
25000
Data
+µMC
-µMC
+µ+ -µMC
Stopping cosmic ray muons
μ- μ+Good data/MC
agreement.
MINOS PRELIMINARY
35Caius Howcroft
Charge Separation (2)
Q/P!Q/P/
-10 -8 -6 -4 -2 0 2 4 6 8 10
Events
0
2
4
6
8
10
12
14
16
18
20
22
24
Data
MC Expectation (no oscillations)
)2
=0.0025 eV2m"MC Expectation (
MC background
ν ν ν/ν
Data 34 18 55
MC (no oscillations)
45.6±8.9 24.9±4.7 68.5±14
MC
(0.0025 eV2)33.8±6.6 18.7±3.4 51.9±10
N+/N- = 0.53±0.15(stat)±0.02(sys)Expected N+/N- = 0.55
Charge Purity = 98.7%
MINOS PRELIMINARY
36Caius Howcroft
Q Separated Up-Down
zenith!cos
-1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1
Even
tsµ"
0
5
10
15
20
25
30
zenith!cos
-1 -0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 0.8 1
Even
tsµ"
0
5
10
15
20
25
30
Data
MC nill oscil
)2
=0.0025 eV2m#MC (
MC background
Robs/RMC
ν 0.72±0.35(stat)±0.03(sys)
ν 0.83±0.29(stat)±0.03(sys)
MINOS PRELIMINARY
37Caius Howcroft
Summary
• MINOS has collected 5.99 kton-years of cosmic data before beam switched on in March 2005.
• Have observed a total of 198 atmospheric neutrino events, 107 contained vertex and 91 upward-going muons.
• Contained event up-down ratio excludes no oscillations to 2.5σ.
• Contained events can be charged id'ed with 98% purity.
• Measured N+/N- = 0.53±0.15(stat)±0.02(sys) , c.f. expectation 0.55.
• No enough statistics yet to make any statement about charge separated oscillation parameters.
Acknowledgements: A. Blake, J. Chapman, B. Rebel and M. Thomson
`Thus I descended from the first circle down into the second, which girdles less space, and so much more woe that it goads to wailing. There abides Minos horribly, and snarls; he examines the sins at the entrance; he judges, and he sends according as he entwines himself.'
Dante Alighieri - Divina Commedia
Gustave Dore
39Caius Howcroft
backup1
2 / eV2
23 m
-510
-410
-310
-210
-110
2 /
eV
2 23
m
-510
-410
-310
-210
-110
! 2
= 0.0025 eV2
m ,2
= 0.0025 eV2 m !
!
!
90% C.L.95% C.L.Input Value
)2
(eV-!2m!
-410
-310
-210
-110
)2 (
eV
+!2
m!
-410
-310
-210
68%
90%
99%
FC Expected upmu Expected
MINOS PRELIMINARY MINOS PRELIMINARY