1. Neutrino Oscillation Results from MINOS Alexandre Sousa Oxford University (for the MINOS Collaboration) 30 th International Cosmic Ray Conference - ICRC 2007 Mérida, México 07/07/07

2. 2 Alex Sousa - ICRC2007 Outline Argonne – Athens – Benedictine – Brookhaven – Caltech – Cambridge – Campinas – Fermilab – College de France – Harvard – IIT – Indiana – Minnesota, Twin Cities – Minnesota, Duluth – Oxford – Pittsburgh – Rutherford Lab – Sao Paulo – South Carolina – Stanford – Sussex – Texas A&M – Texas-Austin – Tufts – UCL – Western Washington – William & Mary - Wisconsin Introduction to the MINOS experiment Data selection and oscillation analysis Results from the first year of NuMI beam running (exposure of 1.27×10 20 protons-on-target) Summary and Outlook The MINOS Collaboration

3. 07/07/07 3 Alex Sousa - ICRC2007 735 km The MINOS Experiment MINOS (Main Injector Neutrino Oscillation Search) – Long-baseline neutrino oscillation experiment – Neutrino beam provided by 120 GeV protons from the Fermilab Main Injector – Near Detector at Fermilab to measure the beam composition – Far Detector deep underground in the Soudan Mine, Minnesota, 735 km away, to search for evidence of neutrino oscillations

4. 07/07/07 4 Alex Sousa - ICRC2007 The NuMI Neutrino Beam LE pME pHE 120 GeV protons strike water cooled graphite target, producing kaons, pions 2 parabolic magnetic horns focus secondaries, which decay into muons, neutrinos Beam energy spectrum can be tuned by varying the relative positions of target, horns First year performance – 10μs spill of 120 GeV protons every 2s – Intensity: 2.3  10 13 POT/spill – 0.2 MW average beam power

5. 07/07/07 5 Alex Sousa - ICRC2007 Scintillator strip M64 M16 MINOS Detector Technology UVUVUVUV Steel Scintillator Orthogonal orientations of strips MINOS Near and Far Detectors are functionally identical: – 2.54cm thick magnetized steel plates – co-extruded polystyrene scintillator strips – orthogonal orientation on alternate planes – U,V – optical fiber readout to multi-anode PMTs

6. 07/07/07 6 Alex Sousa - ICRC2007 Beam spot Coil 3.8m 4.8m Located 1km downstream of the target ~1kt (980t) total mass Shaped as squashed octagon (4.8  3.8  15m 3 ) Partially instrumented (282 steel, 153 scintillator planes) Fast QIE readout electronics, continuous sampling during beam spill 8m Coil Veto shield Located 735km away in Soudan mine, MN 5.4kt, 2 supermodules Shaped as octagonal prism (8  8  30m 3 ) 486 steel planes, 484 scintillator planes Veto shield (scintillator modules) Spill times from Fermilab for beam trigger The MINOS Detectors Near Detector Far Detector

7. 07/07/07 7 Alex Sousa - ICRC2007 μ CC Event e CC Event UZ VZ Long μ track + hadronic activity at vertex Short event, with EM shower profile Short showering event, often diffuse 3.5m 1.8m 2.3m Monte Carlo E = E shower + p μ 55%/  E /GeV 6% range, 13% curvature Event Topologies NC Event

8. 07/07/07 8 Alex Sousa - ICRC2007 Event Classification Parameter Near Detector CC Event Selection Events are classified into charged current and neutral current samples Likelihood based procedure using three event variables – Event length – Fraction of event pulse height in the track – Track pulse height per plane CC events are selected with high efficiency (87%) and NC contamination is low (< 3%). Monte Carlo Visible Energy (GeV) Efficiency/Purity (87%) (97%)

9. 07/07/07 9 Alex Sousa - ICRC2007 Predicting Unoscillated FD Spectrum FD Decay Pipe π+π+ Target ND p × = Start with Near Detector data and perform extrapolation to the Far Detector Use knowledge of pion decay kinematics and beamline geometry to construct a beam transport matrix and predict FD spectrum from measured ND spectrum The “Beam Matrix” method is the primary method used in the MINOS CC analysis

10. 07/07/07 10 Alex Sousa - ICRC2007 Systematic Errors Magnitude of systematic shifts of the oscillation parameters is computed using MC fake data samples at the best fit values of  m 2, sin 2 (2  shifted by the values shown in the table Total systematic error is ~40% of statistical error in  m 2 3 largest systematic uncertainties are included in the oscillation fit as nuisance parameters Uncertainty Shift in  m 2 (10 -3 eV 2 ) Shift in sin 2 (2  Near/Far normalization  4% 0.0500.005 Absolute hadronic energy scale  11% 0.0600.048 NC contamination  50% 0.0900.050 All other systematic uncertainties0.0440.011 Total systematic (summed in quadrature)0.130.07 Statistical error (data)0.360.12

11. 07/07/07 11 Alex Sousa - ICRC2007 MINOS Best Fit Spectrum Best fit energy spectrum and Data/MC Ratio for 1.27  10 20 POT SampleObserved Expected (Unoscillated)  (<30 GeV) 215 336.0  14.4  (<10 GeV) 122 238.7  10.7  (<5 GeV) 76 168.4  8.8

12. 07/07/07 12 Alex Sousa - ICRC2007 MINOS Allowed Regions Contour obtained using the Beam Matrix extrapolation method Fit includes nuisance parameters corresponding to largest systematic uncertainties

13. 07/07/07 13 Alex Sousa - ICRC2007 Summary MINOS has completed an oscillation analysis of the first year of NuMI beam exposure, corresponding to 1.27  10 20 POT – Consistent with neutrino oscillations with parameters: – Published in PRL 97, 191801 (2006) The systematic uncertainties are under control and substantial improvements on understanding them should be made as MINOS collects more data Analysis of second year of beam data is underway, improved results available very soon!

14. 07/07/07 14 Alex Sousa - ICRC2007 Outlook MINOS will significantly improve the current result as the number of POT increases MCMC MINOS MC e Appearance  Disappearance MINOS could make the first ever measurement of  13 – Can improve on current best limit set by CHOOZ More to come soon!! MC Includes 10% syst. err. on background 90% C.L. stat. err. only sin 2 2   =1.00  m 2 32 =2.74x10 -3 eV 2

15. 07/07/07 15 Alex Sousa - ICRC2007 Backup Slides

16. 07/07/07 16 Alex Sousa - ICRC2007 First year of NuMI Running Spill Intensity (1e12 POT) Dataset used for the oscillation analysis (1.27×10 20 POT ) Observation of neutrinos in Near Detector! Start of LE running 10 20 POT! Main Injector Shut-down

17. 07/07/07 17 Alex Sousa - ICRC2007 Beam Target z position (cm) FD Events per 1  10 20 POT LE-10-10390 pME-100970 pHE-2501340 LE pME pHE Events expected in fiducial volume NuMI Beam Composition Movable target relative to horns allows beam energy tuning Currently running in the LE-10 configuration ~1.5  10 19 POT in pME and pHE configurations early in the run for commissioning and systematics studies

18. 07/07/07 18 Alex Sousa - ICRC2007 Help understand energy response to reconstruct E ν E ν = p µ + E had Measured in a CERN test beam with a “mini-Minos” operated in both Near and Far configurations Study e/µ/hadron response of detector Test MC simulation of low energy interactions Provides absolute energy scale for calibration Single particle energy resolution beam The MINOS Calibration Detector

19. 07/07/07 19 Alex Sousa - ICRC2007 Parameterize Fluka2005 prediction as a function of x F and p T Perform fit which reweights neutrino parent pion x F and p T to improve data/MC agreement Horn focusing, beam misalignments included as nuisance parameters in fits Small changes in x-section, neutrino energy scale, NC background also allowed LE-10/185kA Distribution of pions producing MINOS neutrinos MC LE-10/170kALE-10/185kA pME/200kA Horn off LE-10/200kA pHE/200kA Weights applied vs p z & p T Hadron Production Tuning

20. 07/07/07 20 Alex Sousa - ICRC2007 Beam monitoring and detector quality cuts  CC pre-selection cuts: – At least one good reconstructed track – The fitted track should have negative charge (selects  ) – The reconstructed track vertex should be within the fiducial volume of the detector Event Pre-Selection Cuts ν CalorimeterSpectrometer NEAR: 1m < z < 5m (from detector front), R < 1m from beam center. FAR: z > 50cm from front face, z > 2m from rear face, R < 3.7m from detector center.

21. 07/07/07 21 Alex Sousa - ICRC2007 CC Event Selection Event length – Related to p  Fraction of event PH in track – Related to inelasticity of CC events Track pulse height per plane – Related to track dE/dx Probability P  (P NC ) is the product of the three CC (NC) PDFs at the value of these variables taken by the event Near Detector Final selection employs a likelihood based procedure with probability density functions (PDFs) for three low level variables

22. 07/07/07 22 Alex Sousa - ICRC2007 Particle ID (PID) parameter obtained from the likelihood procedure is defined as: CC-like events are defined by PID > −0.2 in the FD (> −0.1 in the ND) – NC contamination limited to low energy bins (below 1.5 GeV) – High purity selection. Efficiency mostly flat as a function of visible energy CC Selection Parameter Monte Carlo Event Classification Parameter Visible Energy (GeV) Efficiency/Purity Near Detector

23. 07/07/07 23 Alex Sousa - ICRC2007 Beam Matrix encapsulates the knowledge of pion 2-body decay kinematics & geometry. Beam Matrix provides a very good representation of how the Far Detector spectrum relates to the near one. Near to Far extrapolation

24. 07/07/07 24 Alex Sousa - ICRC2007 FD Observed vs. Expected A 49% deficit with respect to the no disappearance hypothesis is observed for energies between 0-10GeV Significance of the effect is 6.2σ (stat+syst) Data Sample FD Data Expected (Matrix Method; Unoscillated) Data/MC (Matrix Method)  (<30 GeV) 215 336.0  14.40.64  0.05  (<10 GeV) 122 238.7  10.70.51  0.05  (<5 GeV) 76 168.4  8.80.45  0.06

25. 07/07/07 25 Alex Sousa - ICRC2007 Breakdown of FD Selection CutEventsEfficiency All Events in fiducial volume438- Events with a track38487.7% Track quality cuts36595.1% PID cut (CC-like)26773.2% Track charge sign cut (negative  only) 24491.4% Reconstructed energy <30 GeV21588.1%

26. 07/07/07 26 Alex Sousa - ICRC2007 Preliminary Uncertainty Shift in  m 2 (10 -3 eV 2 ) Shift in sin 2 2  Near/Far normalization  4% 0.0500.005 Muon energy scale  2% 0.0330.003 Relative shower energy scale  2% 0.0500.005 Absolute shower energy scale  6% 0.0330.005 Intranuclear re-scattering  10% 0.0500.048 NC contamination  50% 0.0900.050 Total (sum in quadrature)0.130.07 Statistical error (data)0.360.12 Systematic Shifts

27. 07/07/07 27 Alex Sousa - ICRC2007 3 other extrapolation methods were applied to the data: NDfit: 1D fit to E distribution in ND data used to determine systematic parameters and reweight FD MC accordingly 2D Grid Fit: 2D fit to (E, y) grid in ND data used to determine systematic parameters and reweight FD MC accordingly F/N Ratio: Uses Far/Near spectrum ratio from MC to scale the events in each ND energy bin into a number of FD events in that same bin Other Extrapolation Methods The 4 extrapolation methods investigated give consistent predictions and are robust to different types of systematics. 1.27×10 20 POT MINOS Size of statistical error

28. 07/07/07 28 Alex Sousa - ICRC2007 NC Events NC event selection no track or no track beyond shower short

29. 07/07/07 29 Alex Sousa - ICRC2007 NC Energy Spectrum Efficiency ~ 76% Purity ~ 52%

30. 07/07/07 30 Alex Sousa - ICRC2007 Multiple events in ND per Main Injector spill – Over 1x10 7 fiducial events collected Events are separated using topology and timing – Color in display indicates time Blue hits are early, red are late Near Detector Spill

31. 07/07/07 31 Alex Sousa - ICRC2007 Typical FD  CC event – Spatial information – used for track reconstruction – Timing information – used in atmospheric analysis to distinguish up/down events – Charge information – used for calorimetry U-Z View V-Z View X-Y View T-Z View PH-Z View Far Detector Event