Table Of ContentNew Results from the MINOS Experiment
2
1
0
2
n
a
J
7
1 AnnaHOLIN∗†
UniversityCollegeLondon
]
x E-mail: [email protected]
e
-
p
e The MINOS experiment is a long-baseline neutrino experiment designed to study neutrino be-
h
haviour,inparticularthephenomenonofneutrinooscillations.MINOSsendstheNuMIneutrino
[
beamthroughtwodetectors,aNearDetector1kmdownstreamfromthebeamsourceatFermilab,
1
v andaFarDetector735kmawayintheSoudanMineinMinnesota.MINOShasbeentakingbeam
5 datasince 2005. Thisdocumentsummarisesrecentneutrinooscillationsresults, withparticular
4
6 emphasisonelectronneutrinoappearance,whichprobestheangleq 13 oftheneutrinomassmix-
3 ingmatrix. Foranexposureof8.2×1020protonsontarget,MINOSfindsthatsin2(2q )<0.12
13
.
1 forthenormalmasshierarchy,and<0.20fortheinvertedmasshierarchyatthe90%C.L.,ifthe
0
2 CP-violatingphased =0.
1
:
v
i
X
r
a
XXIstInternationalEurophysicsConferenceonHighEnergyPhysics
21-27July2011
Grenoble,RhonesAlpesFrance
∗Speaker.
†OnbehalfoftheMINOSCollaboration
(cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommonsAttribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/
NewResultsfromtheMINOSExperiment AnnaHOLIN
1. Neutrino Oscillations
In the past 20 years, one of the most significant developments in particle physics was the
discoveryofneutrinooscillations [1]-[6],indicatingthatneutrinoshavemass. Therearethreegen-
erations (flavours) ofneutrino, theelectron, muon,andtauneutrino (n e,n m ,n t ). Eachneutrino also
has a corresponding anti-particle. Neutrino oscillations are parametrised using two mass squared
differences(D m2 -solarsector,andD m2 -atmosphericsector),threemixingangles(q ,q ,and
21 32 12 13
q ),andaCPviolating phase(d ).
23
2. The MINOSExperiment
The MINOS experiment [7] is designed to probe neutrino behaviour and the phenomenon of
neutrino oscillations by sending the NuMIneutrino beam through twodetectors [8], the Near De-
tector atFermilab, 1km downstream ofthebeam source, andtheFarDetector intheSoudanMine
in Minnesota, 735km away. Both detectors are magnetized iron scintillator tracking calorimeters
andaredesignedtobefunctionally identicaltoallowcancellation ofcertainsystematicerrors,like
for example any mismodelling of the neutrino flux or cross-section. The NuMI neutrino beam is
created by impacting 120 GeV protons onto a thin graphite target. The resulting hadrons (mostly
pionsandkaons)arecollimatedbytwomagnetichornsandthendecayproducingabeamofmostly
n m withasmall 7%component ofn m ,and a1.3% contamination ofbeam n e and n e. Itispossible
toreversethehorncurrentsoastoachieveabeamwithahigherproportionofmuonantineutrinos.
3. Muon Neutrino Disappearance
MINOS oscillation analyses use the Near Detector to measure the neutrino interaction rate
before any oscillations have occurred. These measured data are then extrapolated to the Far De-
tector to predict what would be seen in the absence of neutrino oscillations. This final data set is
then unblinded and compared to the prediction. In the case of the n m disappearance analysis, the
survival probability ofamuonneutrinoisgivenby:
P(n m →n m )≈1−sin2(2q 23)sin2(1.27D m232(L/E)) (3.1)
Figure 1 shows the results of the n m disappearance analysis for an exposure of 7.25×1020
protons-on-target (POT). For this data set, 2451 charged current muon neutrino events were pre-
dicted in the Far Detector fiducial volume, but 1986 events were observed [9]. Consequently, the
atmosphericmass-squared differencewasfoundtobeD m2=2.32+0.12×10−3eV2,andthemixing
−0.08
angle parameter sin2(2q ) >0.90 (90% C.L.). Exotic neutrino models like neutrino decoherence
andneutrino decaywereexcluded atthe9s and7s levelrespectively.
4. Muon Antineutrino Disappearance
MINOS has taken some data in antineutrino (reversed horn current) mode and carried out
an antineutrino disappearance analysis [10]. For an exposure of 1.71×1020 POT, MINOS pre-
dicted 156 charged current n m events in the absence of oscillations, however, 97 events were ob-
served, thusdisfavouring thenooscillations hypothesis atthe6.3s level. Thebestfitantineutrino
2
NewResultsfromtheMINOSExperiment AnnaHOLIN
· 10-3
3.5
MINOS best fit MINOS 2008 90%
MINOS 90% Super-K 90%
2)3.0 MINOS 68% Super-K L/E 90%
MINOS Far Detector V
e
GeV300 FNaor odsectiellcattoiorn dsata -3(10 2.5
s / 200 BNeCs bt aocskcgillraotuionnd fit 2| m
nt
e D|2.0
v100
E
0 1.5
0 5 10 15 20 30 50 0.80 0.85 0.90 0.95 1.00
Reconstructed neutrino energy (GeV) sin2(2q )
Figure1: The leftplotshowstheenergyspectrumoffullyreconstructedFarDetectoreventsclassified as
chargedcurrentinteractionsin black. Theredhistogramrepresentsthespectrumpredictedfrommeasure-
ments in the Near Detector assuming no oscillations, while the blue histogram reflects the best fit of the
oscillation hypothesis. The shaded area shows the predicted neutral current background. The right plot
shows the likelihood contoursof 68% and 90% C.L. around the best fit values for the mass splitting and
mixingangle.Alsoshownarecontoursfrompreviousmeasurements[2,3].
oscillation parameters were found to be |D m2| = (3.36+0.46(stat.)±0.06(syst.))×10−3 eV2 and
−0.40
sin2(2q )=0.86+0.11(stat.)±0.01(syst.).
−0.12
5. Electron Neutrino Appearance Analysis
Muon neutrinos may oscillate into electron neutrinos as they travel from the Near to the Far
Detector. Thecorresponding oscillation probability istofirstordergivenby:
P(n m →n e)≈sin2(2q 13)sin2q 23sin2(1.27D m2atm(L/E)) (5.1)
If the oscillation angle q is non-zero, then this should manifest itself as n appearance at
13 e
the Far Detector. In the case of MINOS, the search for n appearance [11] is however made very
e
difficult by low statistics and by a large background of neutral current events that mimic charged
current n events. To disentangle any potential signal from the large backgrounds, a number of
e
cuts are applied to the data. First, a number of data quality cuts like timing and fiducial cuts are
applied toselectgoodbeamevents. Then,sincen -likeeventsconsistofelectromagnetic showers,
e
any events with long muon tracks are removed from the sample. Only events with at least one
showerandawell-definedshowercore,andwithinanenergyrangeof1-8GeV,areselectedforthe
final sample. Finally a selection algorithm is used to obtain the final data sample. This selection
algorithm uses a MC library of 20 million signal and 30 million neutral current events to find the
50 best matches for each event and to construct three variables that are combined into a neural
networktoobtainafinaldiscriminant variable (LEM).
3
NewResultsfromtheMINOSExperiment AnnaHOLIN
In order to use the Near Detector data to predict the Far Detector data, the former is decom-
posedintoitscomponents: 60%neutralcurrentevents,29%short-trackchargedcurrentn m events,
and 11% intrinsic beam charged current n events after all selection cuts. Those components are
e
then extrapolated to the Far Detector using Monte Carlo Far/Near ratios accounting for various
systematic errors such as flux, cross-section, fiducial volume, energy smearing, detector effects,
and muon neutrino oscillation parameters. A n t appearance component resultant from oscillated
n m isalsoaddedtotheFarDetectorprediction.
For the final analysis, MINOS uses five bins in reconstructed energy and three bins for the
LEM discriminant variable (0.6-0.7, 0.7-0.8, and above 0.8). Systematic uncertainties taken into
account include the composition of the Near Detector spectrum, the calibration (relative energy
calibration, gains,absolute energycalibration), therelativeNear/Farnormalization, thehadroniza-
tionmodelandothersmalleruncertainties. InasignalenhancedregionwhereLEM>0.7,MINOS
predicts 49.6±7.0(stat.)±2.7(syst.) events in the Far Detector for an exposure of 8.2×1020, and
observes 62 events. The final Far Detector spectra can be seen in Figure 2. If the final data are
fittedtoaneutrino appearance oscillation hypothesis, foraCP-violating phased =0, MINOSfinds
sin22q <0.12forthenormalmasshierarchy,andsin22q <0.20fortheinvertedmasshierarchy
13 13
at the 90% C.L.. The corresponding best fit values are sin22q =0.041 for the normal mass hier-
13
archy, andsin22q =0.079fortheinverted masshierarchy. Figure3showsthefinaln appearance
13 e
limitsobtainedbyMINOS.ItcanbeseenthatMINOSwasabletoexcludeparameterspacebelow
thelimitsetbytheChoozexperiment[12]forallvaluesofd forthenormalmasshierarchy.
200
180 MINOS Far Detector MINOS Far Detector Best Fit (0.6 < LEM < 0.7)
20 PoT10111246000 Preselection RAenBgaailcoyksngirsound 2020202020 POT10 POT10 POT10 POT10 POT1012121212125050505050 FBSDiagc nDkaagltraound
·nts / 8.2 1068000 PDraetadiction ····· nts / 8.2 nts / 8.2 nts / 8.2 nts / 8.2 nts / 8.2 1111100000 sin2(2q13)=0.0M4e1r, gDemd23 2f>o0r, dFCiPt=0
e eeeee
v 40 vvvvv 55555
E EEEEE
20
0 00000
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 11111 22222 33333 44444 55555 66666 77777 88888
LEM PID RRRRReeeeecccccooooonnnnnssssstttttrrrrruuuuucccccttttteeeeeddddd EEEEEnnnnneeeeerrrrrgggggyyyyy (((((GGGGGeeeeeVVVVV)))))
MINOS Far Detector Best Fit (0.7 < LEM < 0.8) MINOS Far Detector Best Fit (LEM > 0.8)
TTTTT2222200000 FD Data TTTTT2222200000 FD Data
OOOOO OOOOO
PPPPP Background PPPPP Background
2020202020 0 0 0 0 01111155555 Signal 2020202020 0 0 0 0 01111155555 Signal
11111 11111
·····2 2 2 2 2 sin2(2q13)=0.041, D m232>0, dCP=0 ·····2 2 2 2 2 sin2(2q13)=0.041, D m232>0, dCP=0
s / 8.s / 8.s / 8.s / 8.s / 8.1111100000 Merged for Fit s / 8.s / 8.s / 8.s / 8.s / 8.1111100000 Merged for Fit
ntntntntnt ntntntntnt
eeeee eeeee
vvvvv 55555 vvvvv 55555
EEEEE EEEEE
00000 00000
11111 22222 33333 44444 55555 66666 77777 88888 11111 22222 33333 44444 55555 66666 77777 88888
RRRRReeeeecccccooooonnnnnssssstttttrrrrruuuuucccccttttteeeeeddddd EEEEEnnnnneeeeerrrrrgggggyyyyy (((((GGGGGeeeeeVVVVV))))) RRRRReeeeecccccooooonnnnnssssstttttrrrrruuuuucccccttttteeeeeddddd EEEEEnnnnneeeeerrrrrgggggyyyyy (((((GGGGGeeeeeVVVVV)))))
Figure 2: The top left plot shows the LEM discriminant variable distribution in the Far Detector. The
toprightplotandthebottomplotsshowthereconstructedenergyspectraforchargedcurrentn candidate
e
eventsforthethreeLEManalysisbins. Theblackpointsshowthedatawithstatisticalerrorbars. Thered
histogramsshowtheexpectedbackgroundtogetherwiththecontributionofn appearancesignal(hatched
e
area)forthebest-fitvalueofsin22q =0.041.
13
4
NewResultsfromtheMINOSExperiment AnnaHOLIN
22..00
D m2 > 0
6. Conclusions
11..55 MINOS Best Fit
68% C.L.
)) 90% C.L. The MINOS experiment has carried out a num-
ddpp ( ( 11..00 CHOOZ 90% C.L. ber of analyses that are helping to measure and hone
2sin2q = 1 for CHOOZ
23
in on the values of the neutrino oscillation parame-
00..55
ters. In the atmospheric sector, MINOS has been able
0022....000000 00..11 00..22 00..33 00..44 to carry out the world’s most precise measurement of
D m2 < 0 the oscillation parameter D m2 . For the electron neu-
32
11..55 trinoappearance analysis, MINOShasbeenabletoset
theworld’stightestlimiton2sin2(2q )sin2q forthe
13 23
pdpd) () ( 11..00 M8.I2N· O10S20 POT normalmasshierarchy.
ThisworkwassupportedbytheU.S.DOE;theU.K.STFC;
00..55
theU.S.NSF;theStateandUniversityofMinnesota; theUniver-
sityofAthens,Greece;andBrazil’sFAPESP,CNPq,andCAPES.
00..00
00 00..11 00..22 00..33 00..44 WearegratefultotheMinnesotaDepartmentofNaturalResources,
22ssiinn22((22qq ))ssiinn22qq thecrewof theSoudanUnderground Laboratory, andthestaffof
1133 2233
Fermilabfortheircontributionstothiseffort.
Figure3: Allowedrangesandbestfitsfor
2sin2(2q )sin2q as a function of CP- References
13 23
violating phase d . The upper panel as-
[1] B.Pontecorvo,
sumesthenormalneutrinomasshierarchy,
JETP34,172(1958);V.N.GribovandB.Pontecorvo,
and the lower panel assumes the inverted
Phys.Lett.B28,493(1969);Z.Maki,M.Nakagawa,
mass hierarchy. The vertical dashed line
andS.Sakata,Prog.Theor.Phys.28,870(1962).
showstheChooz90%C.L.upperlimitas-
sumingq =45◦andD m2 =2.32eV2. [2] Y.Ashieetal.,Phys.Rev.Lett.
23 32
93,101801(2004);Y.Ashieetal.,Phys.Rev.D71,112005
(2005);J.Hosakaetal.,Phys.Rev.D74,032002(2006).
[3] P.Adamsonetal.,Phys.Rev.Lett.101,131802(2008).
[4] B.Aharmimetal.,Phys.Rev.C72055502(2005).
[5] S.Abeetal.,Phys.Rev.Lett.100,221803(2008).
[6] M.H.Ahnetal.,Phys.Rev.D74,072003(2006).
[7] P.Adamsonetal.,Phys.Rev.D77(2008).
[8] D.G.Michaeletal.,Nucl.Inst.Meth.A596(2008).
[9] P.Adamsonetal.,Phys.Rev.Lett.106,181801(2011).
[10] P.Adamsonetal.,Phys.Rev.Lett.107,021801(2011).
[11] P.Adamsonetal.,Phys.Rev.Lett.107,181802(2011).
[12] M.Apollonioetal.,EuropeanPhys.Jour.C27,331(2003).
5
