Overview of MiniBooNE Ray Stefanski Fermilab August 29, 2010 An - - PowerPoint PPT Presentation
Overview of MiniBooNE Ray Stefanski Fermilab August 29, 2010 An overview of backgrounds & systematic effects; concentrating on interaction cross-section measurements Rough Outline a.) detector configuration b.)sources of systematic
Ray Stefanski Fermilab August 29, 2010
7th International Workshop on Neutrino Beams and Instrumentation
An overview of backgrounds & systematic effects; concentrating on interaction cross-section measurements Rough Outline a.) detector configuration b.)sources of systematic uncertainty b1.) flux b2.) x-sections b3.) reconstruction c.) MiniBooNE measurement d.) future
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 2
per POT per POT second per pulses 5 15Hz @ ppp ν 10 10 22 22 ν 10 10 121 10 10 4 ~
μ 17 17
17 17
12
n interactio
λ 1.7 totals long radius each slugs target Beryllium cm 10.2 cm 0.48 7
pulses/s. 5 for Designed . : horn magnetic Focussing Hz 15 @ kA 170 beamline GeV 8
M’BooNE Schematic Geometry
air. with filled radius long : region Decay cm 90 m 50
n Collimatio absorber m 25
region.
the in PMTs
array an and inward; facing PMTs 8" ,
array an with ed instrument is and sphere the with concentric is , radius
shell
inner An . CH Marcol7
with filled target; from located radius;
sphere a is detector The
2
240 1280 cm 575 T 800 541m cm 610.6 . .
threshold 3
g/cm 0.86 0.68 1.47
SciBooNE ~100m from target L/E ~1, similar to LSND
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 3
What constitutes an “event” in M’BooNE The ~1 GeV beam at M’BooNE results in interactions that are relatively low in outgoing multiplicity. The largest interaction channel is the l CCQE process l +n -> l + p, which accounts for ~40% of all the interactions in the M’BooNE detector. Since the recoil proton is typically below threshold, only the outgoing lepton, or
0 for NC interactions, produces significant light.
While the recoiling nucleon can produce significant scintillation light, this additional source of light is not considered in the reconstruction.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 4
What constitutes an “event” in M’BooNE
If we define a “hit” as a PMT with a signal above threshold, then we eliminate many backgrounds with a simple set of cuts.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 5
What constitutes an “event” in M’BooNE
An event display:
blue comes later. time cluster
event 1ST subevent 2nd subevent
e
e
e
CCQE events must contain 1 & only 1 subevent. 146,070 CCQE events with 5.58 X 1020 POT efficiency = 27% purity = 77 %
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 6
Understanding the detector: response of the oil to
Absorption B3. dependence c. response temperal b. isotropy a. nce Fluorecse B2. dependence c. cos 1 b. (prompt) response temperal a. (Rayleigh) Scattering B1.
Propagati
4 2
response temperal b. dE/dx a. ion Scintillat A2. dependence light Cherenkov A1. Creation
devices multiple B3 B2; B1; (Fermilab) n Attenuatio B2 (JHU) py spectrosco Temperal B2 (Fermilab) py spectrosco nce Fluoresce B4 B1; ) (Princeton Goniometry (IUCF) w/p repeated A2 w/ ) (Fermiloab ion Scintillat A2 w/p (IUCF) ion Scintillat
the
properties the
t Measuremen B2 1; laser B beam Pencil B2 ; light B1 laser Diffuse B2 A2; A1; muons Cosmic B2 A2; A1; electrons Michel ts measuremen time Run
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 7
Understanding the detector: Energy dependence
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 8 mode. in Horn mode. in Horn me. decay volu and area target in ns interactio tertiary includes Simulation 0.24 1.21 by prediction exceeds flux measured ns, interactio CCQE From . simulation flux neutrino based
mode. in Horn mode. in Horn
bins. between ns correlatio include and energy
bins in calculated is matrix error complete The energy. MeV 800 ~ at flux peak the at error 9% ~ a in result which data, HARP the fit to spline a from derived are ies uncertaint error flux practice, In effects. target for thick account to extented ts, measuremen HARP the formula to Wang
the
fit parameter 9 a
based ies uncertaint production published BooNE M' the are These Understanding the detector: Flux uncertainties – particle production
flux. predicted equals roughly flux measured ns, interactio CCQE From
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 9
MeV. 788
energy mean with GeV) 3 ( /POT/cm 10 5.15 is flux integrated The
contributi various into grouped ies uncertaint l Fractiona (b) detector. BooNE M' at the flux Predicted (a)
2 10
Change in flux from
+ due to
dominant sources of systematic uncertainty: horn current; n-N qe x-sect.;
+ -N qe x-sect..
increased decreased & skin effect
mode. in Horn mode. in Horn
Understanding the detector: Flux uncertainties – beam properties
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 10
Predictions from the NUANCE event generator for fractional
interactions in neutrino mode. Resonant and coherent processes are included.
photons. and electrons for used is relations
set similar A MeV. 9 34 set to is and carbon in energy separation the is E where , E M M masses. muon and proton neutron, the are m and , M , M where ), cos m E E ( E 2 _ m Q , ] cos m E E ) M [( 2 ) M m ) M (( E ) M ( 2 E energy muon total the m T E : are s
reported Additional angle. g scatter5in muon energy kinetic muon T . hypothesis a assumin tion reconstruc track the from extracted Variables
B B ' 2 2 QE 2 2 2 2 ' 2 2 2 ' ' QE n n p n QE n p n n
Understanding the detector: cross-sections – NUANCE
scattering for y uncertaint ion normalizat section
; 10 . 1.0 Nubar H
ns interactio EL NC and QE for mass axial ; 10 . 13 . 1 MA section cross resonant NC 0.947; pi0 RES sections.
production pion coherent CC NC, in mass axial 0.275; GeV 03 . 1 macoh factor scale blocking Pauli 0.012; 1.007 kappa energies. NE at MiniBoo known not well
DIS 0.25; 1 dis FSI. and BF ) ( radiative .1245; 1.022 delrad section
coherent NC 1; cog mode production pion
and N in mass axial GeV; 0.52 GeV 3 . 1 MA mode production 1 resonant in mass axial GeV; 0.275 GeV 1 . 1 MA spin proton to
contributi strange 0.10; model gas Fermi in momentum Fermi MeV; 30 MeV/c 220 p model gas Fermi in energy binding 9MeV; MeV 34 events EL NC and QE for mass axial ; 17 . 35 . 1 M : models NUANCE in used Constants
2 H QE N 1 F QE A
N s EB
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 11
In the track based reconstruction, four signal patterns are used: 1. single electron track, 2. single muon track, 3. two tracks, 4. and two tracks with a
0 invariant mass.
The CH2 has and extinction length of ~20m, the radiation length is ~50cm, and exhibits a wide range of optical phenomena near the peak of the PMT sensitivity: 400 nm. Cherenkov light and scintillation light are accompanied by
Also, photon reflection from the surface of the tubes, and the surface of the main detector region must be considered in the simulations. The electronics dead time is ~ 300 ns. A Geant3-based Monte Carlo simulation serves as the main tool for developing reconstruction algorithm’s predictive models.
Understanding the detector: event reconstruction
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 12
Understanding the detector: event reconstruction The quantities measured by the detector are: the number of PMTs that have recorded a light pulse – “hits” – the charge recorded on each PMT; and the time of the hit. . E : energy kinetic The ; , : direction the ; t : time starting The ; z , y , x : point starting The : produced is , , variables seven ith a vector w , quantities measured these From x In simulation, the flux, cross-section model (NUANCE), and detedtor characteristics are combined to convert an event type as input to generation
and data are passed through the same reconstruction routines to generate x, which is used to test our ability to reproduced the data in simulation.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 13
MiniBooNE event reconstruction – particle identification:
The process leads to the development of negative log-likelihoods for each of the fitted hypothesis. If Le, L , and L are the maximized likelihoods returned by the electron, muon, and (fixed-mass) two-track fits, respectively, then the ratios Re/ = log Le - log L and Re/ = log Le - log L , can be used as a test of the electron hypothesis, compared to a
0 .
The events undergo pre-selection based on:
decay.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 14
cross sections: Llewellyn Smith
s. experiment scattering neutrino from determined is M
g
value the fixes decay beta Neutron s. experiment scattering electron presision in measured are F and F factors, form vector the
dependence Q The . parameters emperical are M and , g , M and moments, magnetic anomalous neutron and proton the are and mass, pion the is m where , 2M F and ), M Q /(1 g F , ) / 1 )( 1 ( ) ( F and , ) / 1 )( 1 ( ) 1 ( 1 F by given are , and factors, form vector two The . 4 where ), ( 4 1 ) ( ]] ) 1 ( 4 ) 2 ( ) [( 4
) 1 ( ) 1 ( ) 1 [( ) ( parameters the in explicit is dependence Q The n. interactio the in produced lepton charged the
mass the with 4ME u)
and energy, neutrino incident the is E mass, nucleon the is M constant, Fermi the is G transfer, momentum
squared the is Q , scattering rino (anti)neut to refers
( where, , ) ( 8
A 2 1 2 A A V 2 2 2 P 2 A 2 A A 2 2 2 2 2 2 2 1 2 1 2 2 2 2 1 2 2 1 2 2 2 2 2 2 1 2 2 2 2 1 2 2 2 1 2 2 2 2 2 2 2 F 2 4 2 2 2 2 2 A n P l n P l n P A A P P A l l A l l l F
Q m m Q m Q F F M Q F F F C F F F M Q B F F F F F M m m F F F F F M Q m A A, B, C. m m Q C M u s M u s A E M G dQ d
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 15
Results: cross sections in NUANCE CCQE measurements (not M’BooNE )on a variety
Llewellyn Smith prediction from NUANCE with Mv = 0.84 GeV, MA = 1.0 GeV, and gA =-1.26(solid). Also shown is prediction on carbon from another model – the Smith-Moniz relativistic Fermi gas model.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 16
cross-sections: Pauli suppression In the RFG model, “Pauli-blocking” causes a suppression in the cross-section for low values of the momentum transfer, Q2. The struck nucleon is forbidden from entering a state already occupied by one
CCQE prediction on neutrons bound in carbon using this model. is a measure of the Pauli-blocking, and is normally set to 1. To get out of the nucleus, the final state lepton must have p > pF
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 17
. 38 / . 47 / ; 012 . 007 . 1 GeV/c 17 . 35 . 1 M : parameters model yields sample subevent 2 fit to
2 2 eff A
dof
. for M value average
the indicates circle work.The previous the from countour and point fit best the indicates star
The work. this from extracted countour and point fit best the is star filled .The vs. for M plot countour 1 A
A eff A
and
eff A
M
2 error ellipses are shown for the previous work – the larger is the total uncertainty included in the previous paper.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 18
errors. shape the represent bands lighter the and values, measured reresent bars Dark process. CCQE for the neutron per target section cross al differenti double integrated Flux Results: cross sections: Provide differential cross-sections , correctly normalized with a predicted flux (not normalized to a different reaction channel in the same data. Based on the world’s largest sample of CCQE events(~150,000) @ 1 GeV region.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 19
Results: cross sections
nucleons. free from scattering for and variations parameter different 2 with model RFG a for s prediction are shown Also NOMAD. and LSND, from results with along shown is range energy larger a (b), In bars. error total the with along boxes shaded as shown are errors shape (a), In Energy.
function a as neutron per CCQE unfolded Flux
The excess over predicted background for E QE <475 MeV cannot be explained as an oscillation
2 independent blind analyses were performed: Track Based Analysis (TBA); Boosted Decision Tree (BDT) Prior to box-opening, the collaboration decided to present TBA, which had somewhat better sensitivity. BDT analysis was used as a confirmation of the TBA.
. hypothesis n
2 al convention a fit not does excess This ) (3.0 3 . 38 20.4 128.8 exists events candidate
excess an MeV 475
e QE
E
) 4 , 10 2 (sin : fit best for the 99% is y probabilit The 5%. ~ is LSND ity with Compatibil . hypothesis null for the % 93 is y probabilit the in fit n
The 36 19 22 :
events candidate
excess subtracted background a is There MeV 3000 E 475 For E
2 2 3 2 2 2 e QE QE
eV m
Results: neutrino oscillation
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 21
The future: Short baseline experiment concentrate, to some extent
Many models have been proposed that profess the possible existence of this exotic form of matter. For example, the possible existence of a V+A world concurrent with
More sensitive experiments, that extend current measurements to smaller values of sin2(2 ), may be the only method by which we can probe this otherwise invisible world.
* See for example, Introduction to Sterile Neutrinos, Raymond R. Volkas – hep-ph/0111326 – 26 Nov 2001
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 22
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 23
MiniBooNE
. 10 3.39 1st the in than POT 10 2.27 last the in higher 1.9 was rate Event 0.6% is fit n
best the
likelihood the fit to
background the
likelihood the
ratio the
y probabilit The ) eV 64 . , 96 . ( ) , 2 (sin fit. best for the 8.7%
y probabilit a yields range energy MeV 1250 E 475 the in fit n
The 14.0. 20.9 :
is events
excess An : Fit Region
20 20 2 2 2 2 QE
m
The analysis was performed with the same tools as the TBA analysis, and the addition of a minimum likelihood fitting process. The Boosted Decision Tree (BDT) analysis is still being pursued for the anti-neutrino data.
. 3 . 4 1 18.5 : events
excess an is there MeV 475 For EQE
figures these in included not were excess energy low from events 12 78% 22% events CCQE 24,771 POT 10 33 . 5
20
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 25
Moment Magnetic Neutrino Muon the
Limit n Productio Coherent Ratio Section Cross /QE CC sections cross production single NC and section
Elastic NC beam axis
NuMI the with Analysis Supernovae Collapse
for Search SciBooNE with analysis combined analysis nce disappeara / : talk this in covered not MiniBooNE from More
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 26
Processes available in NUANCE: Particle lifetimes & producing decay modes with their branching rations
Understanding the detector: cross-sections – NUANCE
A little review:
3 2 1 13 23 13 23 12 23 12 13 23 12 23 12 13 23 13 23 12 23 12 13 23 12 23 12 13 13 12 13 12
c c e s c s s c e s c c s s c s e s s s c c s s c c s e s c s c c
i i i i e
) cos( ) sin(
ij ij ij ij
c s
2009 12, November Albright, Carl ) 010 . 016 . ( 0046 . sin 644 . 466 . 331 . sin 375 . 312 . 263 . sin 81 . 2 39 . 2 06 . 2 ) 10 ( 19 . 8 67 . 7 14 . 7 ) 10 ( ) ( limit upper 2 best value limit lower 2
13 2 23 2 12 2 2 3 2 2 5 . . 2
eV m eV m
atm A L sun
992 . cos 13 . sin 73 . cos 68 . sin 83 . cos 56 . sin : fit values Best
13 13 23 23 12 12
5 . 33 . 17 . 5 . 33 . 17 . 33 . 67 . ) | (|
2 3 2 1
e Ul
Tri-bimaximal mixing:
53 . 38 . 09 . 46 . 31 . 23 . 016 . 31 . 68 . ) | (|
2 3 2 1
e Ul
Plugging in the best fit values: Normal & inverted hierarchy Majorana condition? =
C
2- decay
m13
2 = m12 2 + m23 2
2 2
eV 2 / 1 additional for room No
LSND
m
e
2 3
m
2 2
m
2 1
m
atmospheric solar solar atmospheric 2
m
5
10 9 . 7 ~
5
10 9 . 7 ~
e 2 s 1 s e 2 s 1 s
solar
5
10 9 . 7 ~
2 3
m
2 3
m
2 2
m
2 2
m
2 1
m
2 1
m
atmospheric solar atmospheric 2
m
2
m
CPT conserving
LSND LSND 2 3
m
2 3
m
2 2
m
2 2
m
2 1
m
2 1
m
atmospheric solar atmospheric 2
m
2
m
5
10 9 . 7 ~
LSND LSND Kamland Kamland 2 3
m
2 4
m
2 4
m
CPT violating
Sterile neutrinos
Can also be part
hierarchy
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 29
MiniBooNE 90% CL limit (dark solid curve) compared to KARMEN and Bugey experiments. MiniBooNE 90% and 99% CL allowed regions compared to KARMEN and Bugey experiments.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 30
Savannah River Neutrino Detector, circa 1955
2
CdCl with doped targets, r liter wate 200 PMTs. 110 by viewed detectors,
scintillat liter 1,400
The detector weighted about 10 T. s. 3 ~ lifetime a has nucleus CdCl excited The ) 2 . 2 ( : shower EM : e e : pair Dalitz ) 2 . 2 ( : detector In .
beam intense an creates reactor in decay β
2
MeV MeV d p n n e p
e e
This technique is still used in reactor experiments, the observation of two distinct signatures separated in time by a known amount.
A.
discovery to prior decay inverse cm 10 6 / 10 flux MeV 3 E 4/1 ~ S/N : attributes
2 44
3
s cm
e
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 31
Cherenkov Emission Profiles generated in simulation. Scintillation Emission Profiles generated in simulation.
NC 0 NC 0
To these we must add the effects of indirect light from scattering, fluorescence, etc. MiniBooNE Sources of systematic uncertainties: charge likelihood
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 32
Distribution of corrected time for direct light, can be fit a with Gaussian mean and width. For 300 MeV muons, the mean and width parameters from fits like those above vs. direct Cherenkov predicted charge. Parameterized tc likelihood distributions as a function of predicted charge. Scintillation light E0 = 1500 MeV Cherenkov light Internal fit parameters for (a) a single track and (b) two photon tracks. Each photon track contains a conversion distance parameter s. MiniBooNE event reconstruction – time likelihood:
Using the known optical photon and Particle optical properties of the detector, one determines for a given particle type (e/ ) and a set of track parameters, the average number of pe’s that a particular PMT should observe. This quantity is referred to as the predicted charge.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 33
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 34
Birk’s Law
ns
as d interprete be might which ) (3.8 6.0 22.4 87.9 events
excess an
LSND
e e
Beam-related backgrounds Expectation for
( m2=0.24 eV2)
Liquid Sscintillator Neutrino Detector
‘s
e
) 2 . 2 ( MeV d p n n e p e
e e
decay inverse cm 10 MeV 50 E 4/1 ~ S/N : attributes
2 42
167 T mineral oil 14 lbs of scintillating material 1220 PMTs accidental are backgrounds. Similar experiments that fail to observe the LSND result: BNL E776(counter) Karmen(ISIS) Bugey(reactor)
i i
m . eV 2 . 1 17 . masses. the
sum the
limit upper an places data alpha Lyman ans SDSS, WMAP, combined from determined scale, mass absolute The
2
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 36
The smallness of the neutrino mass may be the result of mixing between Dirac and Majorana mass terms in the Electroweak Lagrangian. If we assume no Majorana neutrinos, the Lagrangian takes the form:
This is basically a talk about sterile neutrinos: Let’s look first at the See-Saw mechanism
; ; ) (
D R R L L R L L R D D R L R L
m m L
where , m i L
c R L R D D L R c L c L L L c L R c R R R c R R R L L R D R L L R D M R M L c R L c R R R c R R M R c L R c L L L c L R M L L T c R L T L c L R
N N m m m m N N m m m m m m C C 2 1 ) ( 2 1 ) ( 2 1 ) ( 2 1 ) ( 2 1 : four terms
sum a get we terms, Majorana and Dirac the combine we When ; ) ( 2 1 ; ) ( 2 1 : gin EW Lagran the into terms mass new introduces This les. antipartic ing correspond the as same the are particles Majorana is, That
n conjugatio charge the is C where , ; neutrino
its as particle same the is neutrino that the states condition Majorana The
D M D *
L L L L L L
0. assumed we where ; ; ; 2 2 tan
2 2 1 L R D R L R D
m m m m m m m m m
When we diagonalize the mass matrix we get a see-saw configuration:
. ph/0302238
Laveder and Giunti
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 38
Example: Sterile neutrinos can enter SUSY models through the See-Saw mechanism
Takahashi and Tanimototo [arXiv:07040186] The authors developed a model using the mass varying neutrinos of Fardon, Nelson & Weinner, JCAP, 0410, 005 (2004), in which the left handed Majorana is also massive, and a sterile neutrino plays a part. The authors assume a chiral super-field, A, in dark matter. The superfield A couples to both the left-handed lepton doublet super-field L, and the right-handed neutrino super-field R. The authors introduce a scalar potential of the form: 2 4 ) ( 2 m 2 4 ) ( 2 m : form the
s eigenvalue mass yields and n), ( basis the in , : becomes matrix mass The . . : an a Lagrangi to leads , | | | | | | 4 )
2 2 2 2 2 2 2 2 L 2 2 2 2 2 4 2
L
D A R D A R D n D A R D A R D A D D R D R R R R D L D L A a D A
m M M M M M M m M M M M M M M m m M M c h M M n m n nM n n m M V(φ M L is a coupling constant MA , MD , MR , and mD are mass parameters The scalar and spinor component of A are and n. The scalar component corresponds to the acceleron causing the present cosmic acceleration. The spinor component n is a sterile neutrino.
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 40
data. production fit to by the determined parameters are c and proton, incident the
momentum the is p muon, the
momentum total the is p section, cross al differenti double the is where , cos ( exp 1 ) , (
1,....,9 B 2 7 6 3 9 1 2
8 8 4 2
dpd d p c p c p p c c p p p c p dpd d
c B c B c B c
Understanding the detector: Sanford-Wang formula
August 29, 2010 7th International Workshop on Neutrino Beams and Instrumentation 42
Results: cross sections:
ion. reinteract
for corrected been not have data The . simulation with compared ies, uncertaint systmatic and l statistica both including , CH
ratio section
like
Observed
2