1 / 61

Teppei Katori for the MiniBooNE collaboration Massachusetts Institute of Technology

La Primera Medida de la Sección Eficaz Doble Diferencial en la Dispersión Quasi-elastica de la Corriente Cargada del Neutrino Muónico. Work based on PhD thesis at Indiana University . Teppei Katori for the MiniBooNE collaboration Massachusetts Institute of Technology

lobo
Download Presentation

Teppei Katori for the MiniBooNE collaboration Massachusetts Institute of Technology

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. La PrimeraMedida de la SecciónEficazDobleDiferencial en la Dispersión Quasi-elastica de la CorrienteCargada del Neutrino Muónico Work based on PhD thesis at Indiana University Teppei Katori for the MiniBooNE collaboration Massachusetts Institute of Technology NuInt 09, Sitges, May, 19, 09 TeppeiKatori, MIT

  2. First Measurement of Muon Neutrino Charged Current Quasielastic (CCQE) Double Differential Cross Section outline • Booster neutrino beamline • CCQE eventsin MiniBooNE • CC1pbackground constraint • CCQE MAeff-k shape-only fit • CCQE absolute cross section • Conclusion Work based on PhD thesis at Indiana University Teppei Katori for the MiniBooNE collaboration Massachusetts Institute of Technology NuInt 09, Sitges, May, 19, 09 Teppei Katori, MIT

  3. 1. Booster neutrino beamline 2. CCQE events in MiniBooNE 3. CC1p background constraint 4. CCQE MAeff-k shape-only fit 5. CCQE absolute cross section 6. Conclusion Teppei Katori, MIT

  4. 1. Booster Neutrino Beamline Booster Target Hall decay region absorber target and horn FNAL Booster detector dirt nm K+ p+ Booster primary beam secondary beam tertiary beam (protons) (mesons) (neutrinos) MiniBooNE extracts 8.9 GeV/cmomentum proton beam from the Booster Teppei Katori, MIT

  5. 1. Booster Neutrino Beamline Magnetic focusing horn nm target and horn FNAL Booster decay region absorber detector K+ p+ Booster dirt primary beam secondary beam tertiary beam (protons) (mesons) (neutrinos) Protons are delivered toa beryllium target in a magnetic horn (flux increase ~6 times) p- p+ p+ p- Teppei Katori, MIT

  6. 1. Booster Neutrino Beamline HARP experiment (CERN) Modeling of meson production is based on the measurement done by HARP collaboration - Identical, but 5% l Beryllium target - 8.9 GeV/cproton beam momentum HARP collaboration, Eur.Phys.J.C52(2007)29 Booster neutrino beamlinepion kinematic space HARP kinematic coverage Majority of pions create neutrinos in MiniBooNE are directly measured by HARP (>80%) Teppei Katori, MIT

  7. 1. Booster Neutrino Beamline HARP experiment (CERN) Modeling of meson production is based on the measurement done by HARP collaboration - Identical, but 5% l Beryllium target - 8.9 GeV/cproton beam momentum HARP collaboration, Eur.Phys.J.C52(2007)29 HARP data with 8.9 GeV/c proton beam momentum The error on the HARP data (~7%) directly propagates. The neutrinoflux error is the dominant source of normalization error for an absolute cross section in MiniBooNE. Teppei Katori, MIT

  8. 1. Booster Neutrino Beamline nm target and horn detector FNAL Booster decay region absorber Booster dirt primary beam secondary beam tertiary beam (protons) (mesons) (neutrinos) Predicted nm-flux in MiniBooNE The decay of mesons make the neutrino beam. The neutrino beam is dominated by nm (93.6%), of this, 96.7% is made by p+-decay MiniBooNEcollaboration, PRD79(2009)072002 p- p+ K+ p+ p+ p- Teppei Katori, MIT

  9. 1. Booster neutrino beamline 2. CCQE events in MiniBooNE 3. CC1p background constraint 4. CCQE MAeff-k shape-only fit 5. CCQE absolute cross section 6. Conclusion Teppei Katori, MIT

  10. 2. CCQE event measurement in MiniBooNE nmcharged current quasi-elastic (nmCCQE) interaction is an important channel for the neutrino oscillation physics and the most abundant (~40%) interaction type in MiniBooNE detector MiniBooNE detector (spherical Cherenkov detector) MiniBooNEcollaboration, NIM.A599(2009)28 12C n Teppei Katori, MIT

  11. 2. CCQE event measurement in MiniBooNE nmcharged current quasi-elastic (nmCCQE) interaction is an important channel for the neutrino oscillation physics and the most abundant (~40%) interaction type in MiniBooNE detector MiniBooNE detector (spherical Cherenkov detector) muon like Cherenkov light and subsequent decayed electron (Michel electron) like Cherenkov light are the signal of CCQE event MiniBooNEcollaboration, NIM.A599(2009)28 Cherenkov 1 e m n-beam 12C Cherenkov 2 n p proton measurement in neutral current elastic, see D. Perevalov and R. Tayloe’s talk, May 20 (Wed.) (Scintillation) Teppei Katori, MIT

  12. 2. CCQE event measurement in MiniBooNE nm CCQE interactions (n+nm+p) has characteristic two “subevent” structure from muondecay  nm + ne+ e- + p 1 2  m-+ p nm+ n muon high hits Michel electron low hits 26.5% efficiency 75.8% purity 146,070 events with 5.58E20POT Teppei Katori, MIT

  13. 2. CCQE event measurement in MiniBooNE Em m 12C n-beam cosq All kinematics are specified from 2 observables, muon energy Em and muon scattering angle qm Energy of the neutrino EnQEand 4-momentum transfer Q2QEcan be reconstructed by these 2 observables, under the assumption of CCQE interaction with bound neutron at rest (“QE assumption”) Teppei Katori, MIT

  14. 1. Booster neutrino beamline 2. CCQE events in MiniBooNE 3. CC1p background constraint 4. CCQE MA-k shape-only fit 5. CCQE absolute cross section 6. Conclusion Teppei Katori, MIT

  15. 3. CC1p background constraint, introduction data-MC comparison, in 2 subevent sample (absolute scale) Problem 1 CCQE sample shows good agreement in shape, because we tuned relativistic Fermi gas (RFG) parameters. However absolute normalization does not agree. The background is dominated with CC1p without pion (CCQE-like). We need a background prediction with an absolute scale.  nm + ne+ e- + p 1 1 2 2 CCQE nm+ n CC1p nm+ N  m-+ p MiniBooNE collaboration, PRL100(2008)032301  m-+ p+ + N (p-absorption)  nm +ne+ e- + N Teppei Katori, MIT

  16. 3. CC1p background constraint, introduction data-MC comparison, in 3 subevent sample (absolute scale) Problem 2 CC1p sample is worse situation, data and MC do not agree in shape nor normalization. Under this situation, we cannot use CC1p prediction for background subtraction for CCQE absolute cross section measurement. nm + ne + e+ + N  nm +ne+ e-+ N 1 3 2 CC1p nm+ N  m-+ p+ + N recent development of prediction in CC1p, see J. Novak’s talk, May 22 (Fri.), pion measurement in CC1p, see M. Wilking’s talk, May 22 (Fri.), nm + m+ Teppei Katori, MIT

  17. 3. CC1p background constraint data-MC comparison, before CC1p constraint (absolute scale) Solution Use data-MC Q2 ratio in CC1p sample to correct all CC1p events in MC. Then, this “new” MC is used to predicts CC1p background in CCQE sample This correction gives both CC1p background normalization and shape in CCQE sample Teppei Katori, MIT

  18. 3. CC1p background constraint data-MC comparison, after CC1p constraint (absolute scale) Now we have an absolute prediction of CC1p background in CCQE sample. We are ready to measure the absolute CCQE cross section! Teppei Katori, MIT

  19. 1. Booster neutrino beamline 2. CCQE events in MiniBooNE 3. CC1p background constraint 4. CCQE MAeff-k shape-only fit 5. CCQE absolute cross section 6. Conclusion Teppei Katori, MIT

  20. 4. Pauli blocking parameter “kappa”, k We performed shape-only fit for Q2 distribution to fix CCQE shape within RFG model, by tuning MAeff (effective axial mass) and k Pauli blocking parameter "kappa”, k To enhance the Pauli blocking at low Q2, we introduced a new parameter k, which is the energy scale factor of lower bound of nucleon sea in RFG model in Smith-Moniz formalism, and controls the size of nucleon phase space Smith and Moniz, Nucl.,Phys.,B43(1972)605 final nucleon phase space Initial nucleon phase space k+q k+q k k PF Pauli blocked phase space Pauli blocking is enhanced Teppei Katori, MIT

  21. 4. MAeff-k shape-only fit • MAeff goes even up, this is related to our new background subtraction. • goes down due to the shape change of the background. Now k is consistent with 1. • doesn’t affects cross section below ~0.995. • MAeff - k shape-only fit result • MAeff = 1.35 ± 0.17 GeV (stat+sys) • = 1.007 + 0.007- ∞ (stat+sys) • c2/ndf = 47.0/38 MAeff only fit (MAeff = 1.37 ± 0.12 GeV, c2/ndf = 48.6/39) data-MC Q2 comparison before and after fit Fit parameter space Teppei Katori, MIT

  22. 4. MAeff-k shape-only fit MiniBooNE anti-neutrino CCQE data J. Grange poster, May19 (Tue.) • MAeff - kshape-only fit result • MAeff= 1.35 ± 0.17 GeV (stat+sys) • = 1.007 + 0.007- ∞ (stat+sys) This new CCQE model doesn’t affect our cross section result. data-MC ratio in Tm-cosq kinematic plane after fit Data-MC agreement in Tm-cosq kinematic plane is good. World averaged RFG model MAeff = 1.03, k = 1.000 Teppei Katori, MIT

  23. 1. Booster neutrino beamline 2. CCQE events in MiniBooNE 3. CC1p background constraint 4. CCQE MA-k shape-only fit 5. CCQE absolute cross section 6. Conclusion Teppei Katori, MIT

  24. 5. CCQE absolute cross section Flux-averaged single differential cross section (Q2QE) The data is compared with various RFG model with neutrino flux averaged. Compared to the world averaged CCQE model (red), our CCQE data is 35% high Our model extracted from shape-only fit has better agreement (within our total normalization error). Teppei Katori, MIT

  25. 5. CCQE absolute cross section Flux-unfolded total cross section (EnRFG) New CCQE model is tuned from shape-only fit in Q2, and it also describes total cross section well. Teppei Katori, MIT

  26. 5. CCQE errors Error summary (systematic error dominant) Flux error dominates the total normalization error. Cross section error is small because of high purity and in situ background measurement. Detector error dominates shape error, because this is related with energy scale. Unfolding error is the systematic error associated to unfolding. Teppei Katori, MIT

  27. 5. QE cross section comparison with NOMAD Flux-unfolded total cross section (EnRFG) New CCQE model is tuned from shape-only fit in Q2, and it also describes total cross section well. Comparing with NOMAD, MiniBooNE cross section is 35% higher, but these 2 experiments leave a gap in energy to allow some interesting physics. Teppei Katori, MIT

  28. 5. CCQE total cross section model dependence Flux-unfolded total cross section (EnRFG) Unfortunately, flux unfolded cross section is model dependent. Reconstruction bias due to QE assumption is corrected under “RFG” model assumption. One should be careful when comparing flux-unfolded data from different experiments. Teppei Katori, MIT

  29. 5. CCQE total cross section model dependence Flux-unfolded total cross section (EnRFG) Unfortunately, flux unfolded cross section is model dependent. Reconstruction bias due to QE assumption is corrected under “RFG” model assumption. One should be careful when comparing flux-unfolded data from different experiments. Teppei Katori, MIT

  30. 5. CCQE double differential cross section Flux-averaged double differential cross section (Tm-cosq) This is the most complete information about neutrino cross section based on muon kinematic measurement. The error shown here is shape error, a total normalization error (dNT=10.8%) is separated. Teppei Katori, MIT

  31. 5. CCQE double differential cross section Flux-averaged double differential cross section (Tm-cosq) This is the most complete information about neutrino cross section based on muon kinematic measurement. The error shown here is shape error, a total normalization error (dNT=10.8%) is separated. cross section value shape error fractional shape error Teppei Katori, MIT

  32. 6. Conclusions Using the high statistics and high purity MiniBooNEnm CCQE data sample (146,070 events, 26.5% efficiency, and 75.8% purity), the absolute cross section is measured. We especially emphasize the measurement of flux-averaged double differential cross section, because this is the most complete set of information for muon kinematics based neutrino interaction measurement. The double differential cross section is the model independent result. We measured 35% higher cross section than RFG model with the world averaged nuclear parameter. Interesting to note, our total cross section is consistent with RFG model with nuclear parameters extracted from shape-only fit in our Q2 data. Teppei Katori, MIT

  33. BooNE collaboration University of Alabama Bucknell University University of Cincinnati University of Colorado Columbia University Embry Riddle Aeronautical University Fermi National Accelerator LaboratoryIndiana University University of Florida Los Alamos National Laboratory Louisiana State University Massachusetts Institute of Technology University of Michigan Princeton University Saint Mary's University of Minnesota Virginia Polytechnic Institute Yale University MoltesGrácies! (¡Muchas Gracias!) Teppei Katori, MIT

  34. Back up Teppei Katori, MIT

  35. 1. CCQE event measurement in MiniBooNE CC inclusive cut veto hits <6 for all subevents 1st subevent is within beam window, 4400<T(ns)<6400 fiducial cut, muon vertex <500cm from tank center visible energy cut, muon kinetic energy >200MeV m to e log likelihood cut 2 and only 2 subevent m-e vertex distance cut This cut is not designed to remove CC1p events, but trying to remove “others”. This is an important step for CC1p background fit. Teppei Katori, MIT

  36. 1. CCQE event measurement in MiniBooNE CC inclusive cut  CCQE cut veto hits <6 for all subevents 1st subevent is within beam window, 4400<T(ns)<6400 fiducial cut, muon vertex <500cm from tank center visible energy cut, muon kinetic energy >200MeV m to e log likelihood cut 2 and only 2 subevent m-e vertex distance cut nm CCQE interactions (n+n m+p) has characteristic two “subevent” structure from muon decay nm+ n m+ pmnm +ne+ e muon >200 hits Michel electron <200 hits Teppei Katori, MIT

  37. 1. CCQE event measurement in MiniBooNE This cut is not designed to remove CC1p, but trying to remove “mis-reconstructed CC1p” and “others”. This is an important step for CC1p background fit. CC inclusive cut  CCQE cut veto hits <6 for all subevents 1st subevent is within beam window, 4400<T(ns)<6400 fiducial cut, muon vertex <500cm from tank center visible energy cut, muon kinetic energy >200MeV m to e log likelihood cut 2 and only 2 subevent m-e vertex distance cut Teppei Katori, MIT

  38. 1. CCQE event measurement in MiniBooNE 26.5% cut efficiency 75.8% purity 146,070 events with 5.58E20POT Teppei Katori, MIT

  39. 2. CC1p background fit data-MC Q2 ratio in 3subevent after fit with various assumption Since we can fit with any assumptions, Q2 ratio is always flat. Teppei Katori, MIT

  40. 2. CC1p background fit data-MC Tm-cosq plane ratio in 3subevent after fit with various assumption However, we can differentiate them by 2 dimensional kinematic plane. 15% increase of piabs and 0% of coherent fraction gives the best fit. We chose 15% for piabs, and 50% for cohfrac as new cv MC which will be used to estimate background from all kinematic distribution. The rest of models go to make a new error matrix Teppei Katori, MIT

  41. 2. CC1p background fit CCQE events with 2 subevent MC Tm-cosq plane CC1p kinematics has different shape from CCQE kinematics. The background cross section error is maximum at the bins where CC1p has larger number of event comparing with CCQE. CC1p events with 2 subevent Teppei Katori, MIT

  42. 2. Energy scale of MiniBooNE Mis-calibration of the detector can mimic large MA value. Roughly, 2% of energy shift correspond to 0.1GeV change of MA. MA-k fit for 2% muon energy shifted data To bring MA=1.0GeV, 7% energy shift is required, but this is highly disfavored from the data. Question is what is the possible maximum mis-calibration? (without using muon tracker data) Teppei Katori, MIT

  43. 2. Energy scale of MiniBooNE Energy resolution is very good. Typical resolution is <10%, and the error is 20-80MeV. Tm resolution is various bins of Tm Teppei Katori, MIT

  44. 2. Energy scale of MiniBooNE Range is the independent measure of muon energy. So range-Tm difference for data and MC can be used to measure the possible mis-calibration. Range - Tm X 0.5+100 This variable agrees in all energy regions within 1.5%. Teppei Katori, MIT

  45. 4. Kappa and (e,e’) experiments In low |q|, The RFG model systematically over predicts cross section for electron scattering experiments at low |q| (~low Q2) Data and predicted xs difference for 12C triangle: RFG model circle: DWIA model Butkevich and Mikheyev Phys.Rev.C72:025501,2005 Teppei Katori, MIT

  46. 4. Kappa and (e,e’) experiments • In low |q|, The RFG model systematically over predicts cross section for electron scattering experiments at low |q| (~low Q2) • We had investigated the effect of Pauli blocking parameter “k” in (e,e’) data. • cannot fix the shape mismatching of (e,e’) data for each angle and energy, but it can fix integral of each cross section data, which is the observables for neutrino experiments. We conclude k is consistent with (e,e’) data. E=730MeV q=37.1 degree Q2=0.182GeV2 E=240MeV q=60 degree Q2=0.102GeV2 black: (e,e’) energy transfer data red: RFG model with kappa (=1.019) blue: RFG model without kappa w (MeV) w (MeV) 05/17/2009 Teppei Katori, MIT Teppei Katori, MIT, NuInt '09 46

  47. 4. Kappa and (e,e’) experiments In low |q|, The RFG model systematically over predicts cross section for electron scattering experiments at low |q| (~low Q2) We had investigated the effect of Pauli blocking parameter “k” in (e,e’) data. k cannot fix the shape mismatching of (e,e’) data for each angle and energy, but it can fix integral of each cross section data, which is the observables for neutrino experiments. We conclude k is consistent with (e,e’) data. RFG prediction-(e,e’) data ratio in Q2 (GeV2) red: RFG prediction with kappa (=0.019) blue: RFG prediction without kappa prediction / data Q2 (GeV2) 05/17/2009 Teppei Katori, MIT Teppei Katori, MIT, NuInt '09 47

  48. 4. CCQE normalization fit data-MC comparison, after CCQE normalization fit fit region After the CC1p correction, normalization of CCQE is also found from CCQE sample. We use limited Q2 region to find CCQE normalization, so that this fit is insensitive with CCQE shape very much. Now, CCQE normalization and CC1p normalization and CC1pshape looks good, except CCQE shape. Butkevich arXiv:0904.1472 Teppei Katori, MIT

  49. 4. MA-k fit Least c2 fit for Q2 distribution c2 = (data - MC)T (Mtotal)-1 (data - MC) c2 minimum is found by global scan of shape only fit with 0.0<Q2(GeV2)<1.0 The total output error matrix keep the correlation of Q2 bins Mtotal = M(p+ production) + M(p- production) + M(K+ production) + M(K0 production) + M(beam model) + M(cross section model) + M(detector model) + M(data statistics) Input error matrices keep the correlation of systematics dependent p+ production (8 parameters) p- production (8 parameters) K+ production (7 parameters) K0 production (9 parameters) beam model (8 parameters) cross section (20 parameters) detector model (39 parameters) independent Teppei Katori, MIT

  50. 4. CCQE absolute cross section Absolute flux-averaged differential cross section formula Uij :unsmearing matrix i :true index j : reconstructed index The cross section is function of true value, for example, ds2/Tm/cosqm, ds/dQ2QE, etc Integrated flux is removed, so it is called flux-averaged cross section dj :data vector bj :predicted background T :integrated target number ei :efficiency F :integrated n-flux Teppei Katori, MIT

More Related