Hongyue Duyang University of South Carolina on Behalf of the Nova Collaboration Quy Nhon, Vietnam NUFACT 2016, 08/26/2016

Home , Gargamelle

Coherent π0 Measurement in the NOvA Near Detector Hongyue Duyang University of South Carolina On Behalf of the NOvA Collaboration

Quy Nhon, Vietnam NUFACT 2016, 08/26/2016 What is Coherent?

• Neutrinos can coherently scatter off target nucleus via charge/ neutral current interaction and produce pions. • The target nucleus stays in ground state. • Small momentum transfer. No quantum number (charge, spin, isospin) exchange. • Single forward-going pion in the final state, no vertex activity.

2 Why Coherent?

• Coherent π0 is background to νe appearance measurement. • Coherent meson (π+/π-) production can be useful to the next generation oscillation experiment (DUNE) • Identical topology for neutrino and antineutrino. • Small nuclear effect. • Constraint on (anti)neutrino energy scale. • Physics in its own right: Partially Conserved Axial Current (PCAC) hypothesis, used in Rein-Seghal model and in most neutrino event generators such as GENIE.

3 Figure 1: Feynman Diagram of the neutrino induced coherent ⇡0 production.

Coherent π0: World Measurement

There are relatively few coherent π0 measurement, most suffer from large

Results scaled to uncertainty. Carbon (A=12) Target

Table 1: Summary of world coherent ⇡0 measurement.

40 2 Experiments A (GeV) (10 cm /N ) /(⌫µ-CC) /(RS) Aachen-Padova 27 2 29 10 ± Gargamelle 31 3.5 31 20 ± CHARM 20 30 96 42 ± SKAT 30 7 79 28 4.3 1.5 ± ± 15’ BC 20 20 0.20 0.04 ± NOMAD 12.8 24.8 72.6 10.6 3.21 0.46 ± ± MiniBooNE 12 0.8 0.65 0.14 ± SciBooNE 12 0.8 0.9 0.20 +15.8 ± MINOS 48 4.9 77.6 17.5 4

2 The NOvA Near Detector NO�A NO�A Far Detector (Ash River, MN) MINOS Far Detector (Soudan, MN) A broad physics scope • 0.3 kton, 4.2mX4.2mX15.8m,Using ��→�e , � ͞ �→� ͞ e … Determine the � mass hierarchy Determine the � octant • 1 km from source, underground at Fermilab.23 Constrain �CP

• PVC cells filled with liquid scintillatorUsing ��→�� , � ͞.� →� ͞ � … Precision measurements of 2 2 sin 2�23 and m 32. • Alternating planes of orthogonal (Exclude view. �23=�/4?) Over-constrain the atmos. sector (four oscillation channels) Also … Neutrino cross sections at the NO�A Near Detector Sterile neutrinos Supernova neutrinos Fermilab Other exotica

Ryan Patterson, Caltech

4 cm ⨯ 6 cm 5 The NOvA Near Detector NOνA Near Detector Construction NO�A NO�A Far Detector (Ash River, MN) MINOS Far Detector (Soudan, MN) A broad physics scope • • Detector construction and instrumentation0.3 kton, completed4.2mX4.2mX15.8m, Aug. Using ��→�e , � ͞ �→� ͞ e … Determine the � mass hierarchy Determine the � octant 2014 • 1 km from source, underground at Fermilab.23 Constrain �CP

• PVC cells filled with liquid scintillatorUsing ��→�� , � ͞.� →� ͞ � … • Neutrinos observed within seconds of turning on! Precision measurements of 2 2 sin 2�23 and m 32. • Alternating planes of orthogonal (Exclude view. �23=�/4?) Over-constrain the atmos. sector (four oscillation channels) Also … Neutrino cross sections at the NO�A Near Detector Sterile neutrinos Supernova neutrinos Fermilab Other exotica

Ryan Patterson, Caltech

11 Jonathan M. Paley 4 cm ⨯ 6 cm 6 The NOvA Near Detector NOνA Near Detector Construction NO�A NO�A Far Detector (Ash River, MN) MINOS Far Detector (Soudan, MN) A broad physics scope • • Detector construction and instrumentation0.3 kton, completed4.2mX4.2mX15.8m, Aug. Using ��→�e , � ͞ �→� ͞ e … Determine the � mass hierarchy Determine the � octant 2014 • 1 km from source, underground at Fermilab.23 Constrain �CP

Ryan Patterson, Caltech

11 Jonathan M. Paley 4 cm ⨯ 6 cm 7 The NOvA Near Detector NOνA Near Detector Construction NO�A NO�A Far Detector (Ash River, MN) MINOS Far Detector (Soudan, MN) A broad physics scope • • Detector construction and instrumentation0.3 kton, completed4.2mX4.2mX15.8m, Aug. Using ��→�e , � ͞ �→� ͞ e … Determine the � mass hierarchy Determine the � octant 2014 • 1 km from source, underground at Fermilab.23 Constrain �CP

• PVC cells filled with liquid scintillatorUsing ��→�� , � ͞.� →� ͞ � … • Neutrinos observed within seconds of turning on! Precision measurements of 2 2 sin 2�23 and m 32. • Alternating planes of orthogonal (Exclude view. �23=�/4?) Over-constrain the atmos. sector Results (four oscillation channels) Also … Neutrino cross sections at the NO�A Near Detector Bin to bin correlation matrix: Sterile neutrinos Supernova neutrinos Fermilab Other exotica

Ryan Patterson, Caltech

Mass weight of detector component: C12 Cl35 H1 Near DetectorTi48 O16 Others 0.3 kton 11 Jonathan M. Paley 66.8% 16.4% 10.5% 206 3.3%layers 2.6% 0.4% 4 cm ⨯ 6 cm 8 The measured inclusive cross section from Gargamelle, T2k, and NOvA as shown. There is also shown the predicted cross section for nue on carbon from GENIE. There is large correlation between the energy bins for NOvA results (see Top table). Our detector material is dominant by the carbon, chlorine, and hydrogen.

11/17 NuInt 2015 Xuebing Bu (Fermilab) 28 The NOvA Near Detector NOνA Near Detector Construction NO�A NO�A Far Detector (Ash River, MN) MINOS Far Detector (Soudan, MN) A broad physics scope • • Detector construction and instrumentation0.3 kton, completed4.2mX4.2mX15.8m, Aug. Using ��→�e , � ͞ �→� ͞ e … Determine the � mass hierarchy Determine the � octant 2014 • 1 km from source, underground at Fermilab.23 Constrain �CP

• PVC cells filled with liquid scintillatorUsing ��→�� , � ͞.� →� ͞ � … • Neutrinos observed within seconds of turning on! Precision measurements of 2 2 sin 2�23 and m 32. • Alternating planes of orthogonal (Exclude view. �23=�/4?) Over-constrain the atmos. sector Results (four oscillation channels) Also … Neutrino cross sections at the NO�A Near Detector Bin to bin correlation matrix: Sterile neutrinos Supernova neutrinos Fermilab Other exotica

Ryan Patterson, Caltech

• Low-Z, fine-grained (1 plane ~ 0.15X0), highly-active tracking calorimeter, optimized for EM shower Mass weight of detector component: measurement. C12 Cl35 H1 Near DetectorTi48 O16 Others 0.3 kton 11 Jonathan M. Paley 66.8% 16.4% 10.5% 206 3.3%layers 2.6% 0.4% 4 cm ⨯ 6 cm 9 The measured inclusive cross section from Gargamelle, T2k, and NOvA as shown. There is also shown the predicted cross section for nue on carbon from GENIE. There is large correlation between the energy bins for NOvA results (see Top table). Our detector material is dominant by the carbon, chlorine, and hydrogen.

11/17 NuInt 2015 Xuebing Bu (Fermilab) 28 NuMI off-axis beam

NO�A detectors are sited NuMI NuMI Beam Beam 14 mrad off the NuMI The NuMI Beam beam axis

With the medium-energy NuMI tune, yields a narrow 2-GeV spectrum at the NO�A detectors ➔ Detectors are installed by being ➔ Detectorsoff beam are installed axis by being off beam axis on axis → Reduces NC and �e CC ➔ Narrow band beam peaked at 2 GeV

backgrounds in the ➔ 14 mrad Narrow➔ bandNear maximumbeam peaked oscillation at 2 GeV oscillation analyses (NO�A) ➔ Near ➔maximumReduced oscillation NC background while maintaining high � flux at 2 GeV. ➔ Reduced➔ Electron NC background neutrino flux counts ~1% � of total flux. ➔ Electron neutrino flux counts ~1% of total flux.

Ryan Patterson, Caltech 7 Fermilab JETP, August 6, 2015 • Narrow11/17 NuInt 2015band neutrino beam 1~3GeVXuebing Bu (Fermilab) peak at ~2GeV. 5

11/17 •NuIntDominated 2015 by νμ (94%), withXuebing small Bu (Fermilab) contribution from νe (1%). 5

10 Flux uncertainty NOvA Flux #Strategy of numu as a function of true neutrino energy at NOvA Near Detector. NOvA Simulation ➔ Two major uncertainties 1012 • νμ flux comes from pion and kaon decay. From all parents

➔ Beam transport (5%)• Uncertainties come from hadron 2 From π+ production and beam transport simulation. 1011 ➔ horn current, horn positions, beam direction, POT/cm 20 From K+

• Use external thick target (MIPP) and thin 10 beam spot size, and magnetic field ×

/6 10

target (NA49) data to constrain the hadron 10Φ Flux uncertainty➔ Hadron productionproduction uncertainty to ~10%. ➔ • ~5% uncertainty from beam transport. Using external data (see below table) 2 4 6 8 10 12 14 16 18 20 E (GeV) νµ ➔ Conservative systematic uncertainty is assigned ➔ Two major uncertainties for the region not covered by data. 2 Data p range p range Carbon Proton energy T Z Target (GeV) ➔ Beam transport (5%) (GeV) (GeV) NA49 pion 0 - 2 0 - 60 thin 158 ➔ horn current, horn positions, NA49beam kaon direction,0 - 1 0 - 27 thin 158 beam spot size, and magneticMIPP kaon/pion field ratio 0 - 2 27 - 60 thin 120 ➔ Hadron production MIPP pion 0 - 2 0 - 60 thick 120 NA49 pion cross section: Eur. Phys. J. C49 (2007) ➔ Using external dataNA49 (see kaon below cross section: table) G. Tinti Ph.D. Thesis MIPP kaon/pion ratio: A. Lebedev Ph.D. Thesis 11 ➔ Conservative systematicMIPP pion uncertaintyyield: Phys. Rev. is D assigned90, 032001 (2014) for the region not coveredFermilab by JETP data. seminar, 02/26/16 Xuebing Bu (Fermilab) 46 Data p range p range Carbon Proton energy T Z (GeV) (GeV) Target (GeV) NA49 pion 0 - 2 0 - 60 thin 158 NA49 kaon 0 - 1 0 - 27 thin 158 MIPP kaon/pion ratio 0 - 2 27 - 60 thin 120 MIPP pion 0 - 2 0 - 60 thick 120 NA49 pion cross section: Eur. Phys. J. C49 (2007) NA49 kaon cross section: G. Tinti Ph.D. Thesis MIPP kaon/pion ratio: A. Lebedev Ph.D. Thesis MIPP pion yield: Phys. Rev. D 90, 032001 (2014) Fermilab JETP seminar, 02/26/16 Xuebing Bu (Fermilab) 46 Flux uncertainty NOvA Flux #Strategy of numu as a function of true neutrino energy at NOvA Near Detector. NOvA Simulation ➔ Two major uncertainties 1012 • νμ flux comes from pion and kaon decay. From all parents

➔ Beam transport (5%)• Uncertainties come from hadron 2 From π+ production and beam transport simulation. 1011 ➔ horn current, horn positions, beam direction, POT/cm 20 From K+

• Use external thick target (MIPP) and thin 10 beam spot size, and magnetic field ×

/6 10

target (NA49) data to constrain the hadron 10Φ Flux uncertainty➔ Hadron productionproduction uncertainty to ~10%. ➔ • ~5% uncertainty from beam transport. Using external data (see below table) 2 4 6 8 10 12 14 16 18 20 E (GeV) νµ ➔ Conservative systematic uncertainty is assigned ➔ Two major uncertainties for the region not covered by data. 2 Data p range p range Carbon Proton energy T Z Target (GeV) ➔ Beam transport (5%) (GeV) (GeV) NA49 pion 0 - 2 0 - 60 thin 158 ➔ horn current, horn positions, NA49beam kaon direction,0 - 1 0 - 27 thin 158 beam spot size, and magneticMIPP kaon/pion field ratio 0 - 2 27 - 60 thin 120 ➔ Hadron production MIPP pion 0 - 2 0 - 60 thick 120 NA49 pion cross section: Eur. Phys. J. C49 (2007) ➔ Using external dataNA49 (see kaon below cross section: table) G. Tinti Ph.D. Thesis MIPP kaon/pion ratio: A. Lebedev Ph.D. Thesis 12 ➔ Conservative systematicMIPP pion uncertaintyyield: Phys. Rev. is D assigned90, 032001 (2014) for the region not coveredFermilab by JETP data. seminar, 02/26/16 Xuebing Bu (Fermilab) 46 Data p range p range Carbon Proton energy T Z (GeV) (GeV) Target (GeV) NA49 pion 0 - 2 0 - 60 thin 158 NA49 kaon 0 - 1 0 - 27 thin 158 MIPP kaon/pion ratio 0 - 2 27 - 60 thin 120 MIPP pion 0 - 2 0 - 60 thick 120 NA49 pion cross section: Eur. Phys. J. C49 (2007) NA49 kaon cross section: G. Tinti Ph.D. Thesis MIPP kaon/pion ratio: A. Lebedev Ph.D. Thesis MIPP pion yield: Phys. Rev. D 90, 032001 (2014) Fermilab JETP seminar, 02/26/16 Xuebing Bu (Fermilab) 46 Flux uncertainty NOvA Flux #Strategy of numu as a function of true neutrino energy at NOvA Near Detector. NOvA Simulation ➔ Two major uncertainties 1012 • νμ flux comes from pion and kaon decay. From all parents

➔ Beam transport (5%)• Uncertainties come from hadron 2 From π+ production and beam transport simulation. 1011 ➔ horn current, horn positions, beam direction, POT/cm 20 From K+

• Use external thick target (MIPP) and thin 10 beam spot size, and magnetic field ×

/6 10

target (NA49) data to constrain the hadron 10Φ Flux uncertainty➔ Hadron productionproduction uncertainty to ~10%. ➔ • ~5% uncertainty from beam transport. Using external data (see below table) 2 4 6 8 10 12 14 16 18 20 E (GeV) νµ ➔ Conservative systematic uncertaintyMore is aboutassigned NOvA flux ➔ Two major uncertainties for the region not covered by data. 2 Data p range p range Carbon Protonsee energy the poster T Z Target (GeV) ➔ Beam transport (5%) (GeV) (GeV) NA49 pion 0 - 2 0 - 60 thin 158 ➔ horn current, horn positions, NA49beam kaon direction,0 - 1 0 - 27 thin 158 beam spot size, and magneticMIPP kaon/pion field ratio 0 - 2 27 - 60 thin 120 ➔ Hadron production MIPP pion 0 - 2 0 - 60 thick 120 NA49 pion cross section: Eur. Phys. J. C49 (2007) ➔ Using external dataNA49 (see kaon below cross section: table) G. Tinti Ph.D. Thesis MIPP kaon/pion ratio: A. Lebedev Ph.D. Thesis 13 ➔ Conservative systematicMIPP pion uncertaintyyield: Phys. Rev. is D assigned90, 032001 (2014) for the region not coveredFermilab by JETP data. seminar, 02/26/16 Xuebing Bu (Fermilab) 46 Data p range p range Carbon Proton energy T Z (GeV) (GeV) Target (GeV) NA49 pion 0 - 2 0 - 60 thin 158 NA49 kaon 0 - 1 0 - 27 thin 158 MIPP kaon/pion ratio 0 - 2 27 - 60 thin 120 MIPP pion 0 - 2 0 - 60 thick 120 NA49 pion cross section: Eur. Phys. J. C49 (2007) NA49 kaon cross section: G. Tinti Ph.D. Thesis MIPP kaon/pion ratio: A. Lebedev Ph.D. Thesis MIPP pion yield: Phys. Rev. D 90, 032001 (2014) Fermilab JETP seminar, 02/26/16 Xuebing Bu (Fermilab) 46 Coherent π0 Candidate in the NOvA ND

A coherent π0 candidate events with 2 photons from π0 decay.

14 Reconstruction: Slicing

Group hits together in time and space for each neutrino interaction.

15 Reconstruction: Vertexing

Find particle paths, and use the intersection to form vertex

16 Reconstruction: Clustering

Group hits from each shower together using clustering algorithm.

17 Calibration and energy scale

I Detector response varies substantially over length of cell due to attenuation in ﬁber

I Use cosmic ray muons as a standard candle I Calibrate every channel (344,064) individually Calibration

• Use stopping muons as standard candles to correct for attenuation, threshold and shadowing effect. I Use dE/dx near the end of• stoppingCross-checked muons by Michael to electron, set absolute beam muons, scale hadronic energy/hit, and π0 mass peak (see plot later). FD cosmic data - plane 84 (horizontal), cell 12 • 5% uncertainty on energy calibration. 25 NOνA Preliminary

10 Mean PE / cm

0 -500 0 500 Distance from center (cm)

18 C. Backhouse (Caltech) NOvA 39 / 27 EM Shower Identification

• Identify EM showers by likelihoods build upon shower dE/dx information.

19 Rock-Muon Induced EM Showers

• Rock muons induce EM showers in the detector via bremsstrahlung radiation. • A muon-removal (MR) technique is developed to isolate those EM showers . • Provide a data-driven method to check detector performance and benchmark EM shower modeling and likelihoods. 20 Rock-Muon Induced EM Showers

θshw - θμ (rad)

• A “measured”36 Jonathan M.angular Paley resolution in data by comparing the reconstructed EM shower direction to the muon direction. • The NOvA ND has good angular resolution (~0.02rad) for EM shower measurement. • Important to the coherent π0 cross-section measurement.

23 EM Shower Selection Efficiency

• Use MR Brem showers to benchmark the MC modeling and likelihood selection efficiency of EM showers. • Very good agreement between data and MC. • 1% difference in selection efficiency taken into systematic uncertainty.

24 Analysis Strategy

• Select NC π0 sample: no muon track, two EM showers, reconstruct the invariant mass. • Using kinematics, further select a signal sample with most of the coherent signal. • Define a control sample (sideband), dominated by non-coherent π0 s, to constrain background modeling in normalization and shape. • Apply the normalization and kinematic shape tuning to the signal sample. • Get a measurement from the signal sample as the data event excess over background prediction in the coherent region.

25 Neutral Current π0 Selection

• NC π0 events with no muon track and two EM shower prongs. • Invariant mass plot show good data/MC agreement: also serve as a calibration check. • Non-coherent background comes from Resonance (RES), Deep- Inelastic Scattering (DIS) and charge-current events.

26 Coherent Signal Sample Selection

• Define the signal sample by events with most of their energy on the 2 showers and low vertex energy: it has >90% of the signal. • Define the events left as a control sample (sideband), dominated by non-coherent π0 s. • The control sample will be used for background tuning (shape and normalization).

27 Control sample

• The control sample is dominant by non-coherent π0 s (RES, DIS). • Select the events in the π0 peak region for further analysis.

28 Control sample

• Energy and angle shapes agree well between data and MC. • It is possible to tune the background in 1D ζ = E*(1-cosθ), or 2D (Eγγ vs Angle) space. • The final background normalization method is under work. 29 Signal Sample

• Signal region data is blinded for now. • Select the events in the π0 peak region for further analysis.

30 Signal Sample

• It is possible to do the measurement (data - background) in 1D ζ = E*(1- cosθ), or 2D (Eγγ vs Angle) space.

31 Figure 2: Left: number of 3D fussyK prongs per interaction from the coherent ⇡0 MC. Right: data and MC comparison of the number of prongs in the neutral current events.

64 in the low-⇣ region. It is therefore important to use a data-driven method to constrain both 65 the normalization and shape of the background. 0 0 66 The strategy of the coherent ⇡ analysis is as follows. First, we select single ⇡ events in 67 the NC sample defined by the absence of a reconstructed muon in the final state. Both photons 0 68 from ⇡ decay should be reconstructed as 3D prongs. The sample is composed of both coherent 69 and non-coherent (Resonance and DIS) interactions. Next, using kinematics, we define a control 0 70 sample, entirely dominated by non-coherent ⇡ , and a signal sample containing coherent and 71 non-coherent events. The control sample is used to tune the normalization and shape of the Cross-Section Measurement 72 non-coherent ⇣, which is then applied to the non-coherent background in the signal sample. 73 Finally, the coherent signal is measured in the low-⇣ region of the coherent signal sample as the 74 excess over non-coherent prediction. 0 Normalized Background 75 The cross-section of coherentSelected⇡ production data is calculated as: N N = Data,selected Bkg,norm (1) ✏ N ⇥ Target ⇥ Signal efficiency 76 where NData,selected and NBkg,norm are the number of data andFlux normalized MC background in 77 the selected coherent (low-⇣) region of the signal sample, ✏ is the eciency of coherent signal Number of target nucleus 78 selection calculated by MC, NTarget is the number of target nucleus in the fiducial volume, and 79 is the muon neutrino flux. 80 Uncertainty to this analysis comes from both statistics and systematics. To reduce the statis- 81 tic uncertainty, we want to reduce the number of background (NBkg) while keeping relatively 32 82 high signal eciency (✏). Systematic uncertainty mainly comes from the measured number of 83 background (NBkg) and flux. Coherent modeling and detector simulation also contribute to 84 the uncertainty through ✏. The uncertainty to NBkg is constrained by the control sample as 85 described above. External data (MIPP/NA49) are used to constrain the flux uncertainty from 86 hadron production. 87 The neutrino flux, data and MC used in this analysis will be discussed in section 2 and 3. 0 88 Section 4 focuses on the selection of NC ⇡ sample, including both coherent signal and non- 89 coherent background. Section 5 present the selection of coherent signal sample and non-coherent 90 control sample, and the data-driven method of background constraint. Systematic uncertainties 91 will be discussed in section 6.

3 Figure 2: Left: number of 3D fussyK prongs per interaction from the coherent ⇡0 MC. Right: data and MC comparison of the number of prongs in the neutral current events.

64 in the low-⇣ region. It is therefore important to use a data-driven method to constrain both 65 the normalization and shape of the background. 0 0 66 The strategy of the coherent ⇡ analysis is as follows. First, we select single ⇡ events in 67 the NC sample defined by the absence of a reconstructed muon in the final state. Both photons 0 68 from ⇡ decay should be reconstructed as 3D prongs. The sample is composed of both coherent 69 and non-coherent (Resonance and DIS) interactions. Next, using kinematics, we define a control 0 70 sample, entirely dominated by non-coherent ⇡ , and a signal sample containing coherent and 71 non-coherent events. The control sample is used to tune the normalization and shape of the Cross-Section Measurement 72 non-coherent ⇣, which is then applied to the non-coherent background in the signal sample. 73 Finally, the coherent signal~8% is statistical measured uncertainty in the low-⇣ region of the coherent signal sample as the with 3.7E20POT data 74 excess over non-coherent prediction. 0 Normalized Background 75 The cross-section of coherentSelected⇡ production data is calculated as: N N = Data,selected Bkg,norm (1) ✏ N ⇥ Target ⇥ Signal efficiency 76 where NData,selected and NBkg,norm are the number of data andFlux normalized MC background in 77 the selected coherent (low-⇣) region of the signal sample, ✏ is the eciency of coherent signal Number of target nucleus 78 selection calculated by MC, NTarget is the number of target nucleus in the fiducial volume, and 79 is the muon neutrino flux. 80 Uncertainty to this analysis comes from both statistics and systematics. To reduce the statis- 81 tic uncertainty, we want to reduce the number of background (NBkg) while keeping relatively 33 82 high signal eciency (✏). Systematic uncertainty mainly comes from the measured number of 83 background (NBkg) and flux. Coherent modeling and detector simulation also contribute to 84 the uncertainty through ✏. The uncertainty to NBkg is constrained by the control sample as 85 described above. External data (MIPP/NA49) are used to constrain the flux uncertainty from 86 hadron production. 87 The neutrino flux, data and MC used in this analysis will be discussed in section 2 and 3. 0 88 Section 4 focuses on the selection of NC ⇡ sample, including both coherent signal and non- 89 coherent background. Section 5 present the selection of coherent signal sample and non-coherent 90 control sample, and the data-driven method of background constraint. Systematic uncertainties 91 will be discussed in section 6.

3 Figure 2: Left: number of 3D fussyK prongs per interaction from the coherent ⇡0 MC. Right: data and MC comparison of the number of prongs in the neutral current events.

64 in the low-⇣ region. It is therefore important to use a data-driven method to constrain both 65 the normalization and shape of the background. 0 0 66 The strategy of the coherent ⇡ analysis is as follows. First, we select single ⇡ events in 67 the NC sample defined by the absence of a reconstructed muon in the final state. Both photons 0 68 from ⇡ decay should be reconstructed as 3D prongs. The sample is composed of both coherent 69 and non-coherent (Resonance and DIS) interactions. Next, using kinematics, we define a control 0 70 sample, entirely dominated by non-coherent ⇡ , and a signal sample containing coherent and 71 non-coherent events. The control sample is used to tune the normalization and shape of the Cross-Section Measurement 72 non-coherent ⇣, which is then applied to the non-coherent background in the signal sample. ~5% systematic uncertainty 73 Finally, the coherent signal~8% is statistical measured uncertainty in the low-by ⇣changingregion control of the sample coherent definition, signal sample as the with 3.7E20POT data normalization method, etc. 74 excess over non-coherent prediction. 0 Normalized Background 75 The cross-section of coherentSelected⇡ production data is calculated as: N N = Data,selected Bkg,norm (1) ✏ N ⇥ Target ⇥ Signal efficiency 76 where NData,selected and NBkg,norm are the number of data andFlux normalized MC background in 77 the selected coherent (low-⇣) region of the signal sample, ✏ is the eciency of coherent signal Number of target nucleus 78 selection calculated by MC, NTarget is the number of target nucleus in the fiducial volume, and 79 is the muon neutrino flux. 80 Uncertainty to this analysis comes from both statistics and systematics. To reduce the statis- 81 tic uncertainty, we want to reduce the number of background (NBkg) while keeping relatively 34 82 high signal eciency (✏). Systematic uncertainty mainly comes from the measured number of 83 background (NBkg) and flux. Coherent modeling and detector simulation also contribute to 84 the uncertainty through ✏. The uncertainty to NBkg is constrained by the control sample as 85 described above. External data (MIPP/NA49) are used to constrain the flux uncertainty from 86 hadron production. 87 The neutrino flux, data and MC used in this analysis will be discussed in section 2 and 3. 0 88 Section 4 focuses on the selection of NC ⇡ sample, including both coherent signal and non- 89 coherent background. Section 5 present the selection of coherent signal sample and non-coherent 90 control sample, and the data-driven method of background constraint. Systematic uncertainties 91 will be discussed in section 6.

3 Figure 2: Left: number of 3D fussyK prongs per interaction from the coherent ⇡0 MC. Right: data and MC comparison of the number of prongs in the neutral current events.

64 in the low-⇣ region. It is therefore important to use a data-driven method to constrain both 65 the normalization and shape of the background. 0 0 66 The strategy of the coherent ⇡ analysis is as follows. First, we select single ⇡ events in 67 the NC sample defined by the absence of a reconstructed muon in the final state. Both photons 0 68 from ⇡ decay should be reconstructed as 3D prongs. The sample is composed of both coherent 69 and non-coherent (Resonance and DIS) interactions. Next, using kinematics, we define a control 0 70 sample, entirely dominated by non-coherent ⇡ , and a signal sample containing coherent and 71 non-coherent events. The control sample is used to tune the normalization and shape of the Cross-Section Measurement 72 non-coherent ⇣, which is then applied to the non-coherent background in the signal sample. ~5% systematic uncertainty 73 Finally, the coherent signal~8% is statistical measured uncertainty in the low-by ⇣changingregion control of the sample coherent definition, signal sample as the with 3.7E20POT data normalization method, etc. 74 excess over non-coherent prediction. 0 Normalized Background 75 The cross-section of coherentSelected⇡ production data is calculated as: N N = Data,selected Bkg,norm (1) ✏ N ⇥ Target ⇥ Signal efficiency 76 where NData,selected and~4% NUncertaintyBkg,norm fromare the number of data andFlux normalized MC background in 77 the selected coherent (low-signal modeling⇣) region by of the signal sample, ✏ is the eciency of coherent signal comparing BS vs RS model. Number of target nucleus 78 selection calculated byEM MC, ShowerN selectionTarget is the number of target nucleus in the fiducial volume, and 79 is the muon neutrinoefficiency flux. constrained by MR Brem 80 Uncertainty to this analysis comes from both statistics and systematics. To reduce the statis- 81 tic uncertainty, we want to reduce the number of background (NBkg) while keeping relatively 35 82 high signal eciency (✏). Systematic uncertainty mainly comes from the measured number of 83 background (NBkg) and flux. Coherent modeling and detector simulation also contribute to 84 the uncertainty through ✏. The uncertainty to NBkg is constrained by the control sample as 85 described above. External data (MIPP/NA49) are used to constrain the flux uncertainty from 86 hadron production. 87 The neutrino flux, data and MC used in this analysis will be discussed in section 2 and 3. 0 88 Section 4 focuses on the selection of NC ⇡ sample, including both coherent signal and non- 89 coherent background. Section 5 present the selection of coherent signal sample and non-coherent 90 control sample, and the data-driven method of background constraint. Systematic uncertainties 91 will be discussed in section 6.

3 Figure 2: Left: number of 3D fussyK prongs per interaction from the coherent ⇡0 MC. Right: data and MC comparison of the number of prongs in the neutral current events.

64 in the low-⇣ region. It is therefore important to use a data-driven method to constrain both 65 the normalization and shape of the background. 0 0 66 The strategy of the coherent ⇡ analysis is as follows. First, we select single ⇡ events in 67 the NC sample defined by the absence of a reconstructed muon in the final state. Both photons 0 68 from ⇡ decay should be reconstructed as 3D prongs. The sample is composed of both coherent 69 and non-coherent (Resonance and DIS) interactions. Next, using kinematics, we define a control 0 70 sample, entirely dominated by non-coherent ⇡ , and a signal sample containing coherent and 71 non-coherent events. The control sample is used to tune the normalization and shape of the Cross-Section Measurement 72 non-coherent ⇣, which is then applied to the non-coherent background in the signal sample. ~5% systematic uncertainty 73 Finally, the coherent signal~8% is statistical measured uncertainty in the low-by ⇣changingregion control of the sample coherent definition, signal sample as the with 3.7E20POT data normalization method, etc. 74 excess over non-coherent prediction. 0 Normalized Background 75 The cross-section of coherentSelected⇡ production data is calculated as: N N = Data,selected Bkg,norm (1) ✏ N ⇥ Target ⇥ Signal efficiency 76 where NData,selected and~4% NUncertaintyBkg,norm fromare the number of data andFlux normalized MC background in 77 the selected coherent (low-signal modeling⇣) region by of the signal sample, ✏ is the eciency of coherent signal comparing BS vs RS model. Number of target nucleus 78 selection calculated byEM MC, ShowerN selectionTarget is the number of target nucleus in the fiducial volume, and ~1% Uncertainty from 79 is the muon neutrinoefficiency flux. constrained by MR Brem detector mass simulation 80 Uncertainty to this analysis comes from both statistics and systematics. To reduce the statis- 81 tic uncertainty, we want to reduce the number of background (NBkg) while keeping relatively 36 82 high signal eciency (✏). Systematic uncertainty mainly comes from the measured number of 83 background (NBkg) and flux. Coherent modeling and detector simulation also contribute to 84 the uncertainty through ✏. The uncertainty to NBkg is constrained by the control sample as 85 described above. External data (MIPP/NA49) are used to constrain the flux uncertainty from 86 hadron production. 87 The neutrino flux, data and MC used in this analysis will be discussed in section 2 and 3. 0 88 Section 4 focuses on the selection of NC ⇡ sample, including both coherent signal and non- 89 coherent background. Section 5 present the selection of coherent signal sample and non-coherent 90 control sample, and the data-driven method of background constraint. Systematic uncertainties 91 will be discussed in section 6.

3 Figure 2: Left: number of 3D fussyK prongs per interaction from the coherent ⇡0 MC. Right: data and MC comparison of the number of prongs in the neutral current events.

64 in the low-⇣ region. It is therefore important to use a data-driven method to constrain both 65 the normalization and shape of the background. 0 0 66 The strategy of the coherent ⇡ analysis is as follows. First, we select single ⇡ events in 67 the NC sample defined by the absence of a reconstructed muon in the final state. Both photons 0 68 from ⇡ decay should be reconstructed as 3D prongs. The sample is composed of both coherent 69 and non-coherent (Resonance and DIS) interactions. Next, using kinematics, we define a control 0 70 sample, entirely dominated by non-coherent ⇡ , and a signal sample containing coherent and 71 non-coherent events. The control sample is used to tune the normalization and shape of the Cross-Section Measurement 72 non-coherent ⇣, which is then applied to the non-coherent background in the signal sample. ~5% systematic uncertainty 73 Finally, the coherent signal~8% is statistical measured uncertainty in the low-by ⇣changingregion control of the sample coherent definition, signal sample as the with 3.7E20POT data normalization method, etc. 74 excess over non-coherent prediction. 0 Normalized Background 75 The cross-section of coherentSelected⇡ production data is calculated as: N N = Data,selected Bkg,norm (1) ✏ N ⇥ Target ⇥ Signal efficiency 76 where NData,selected and~4% NUncertaintyBkg,norm fromare the number of data andFlux normalized MC background in 77 the selected coherent (low-signal modeling⇣) region by of the signal sample, ✏ is~11% the uncertainty eciency of coherent signal comparing BS vs RS model. Number of target nucleus Constrained by 78 selection calculated byEM MC, ShowerN selectionTarget is the number of target nucleusMIPP/NA49 in data the fiducial volume, and ~1% Uncertainty from 79 is the muon neutrinoefficiency flux. constrained by MR Brem detector mass simulation 80 Uncertainty to this analysis comes from both statistics and systematics. To reduce the statis- 81 tic uncertainty, we want to reduce the number of background (NBkg) while keeping relatively 37 82 high signal eciency (✏). Systematic uncertainty mainly comes from the measured number of 83 background (NBkg) and flux. Coherent modeling and detector simulation also contribute to 84 the uncertainty through ✏. The uncertainty to NBkg is constrained by the control sample as 85 described above. External data (MIPP/NA49) are used to constrain the flux uncertainty from 86 hadron production. 87 The neutrino flux, data and MC used in this analysis will be discussed in section 2 and 3. 0 88 Section 4 focuses on the selection of NC ⇡ sample, including both coherent signal and non- 89 coherent background. Section 5 present the selection of coherent signal sample and non-coherent 90 control sample, and the data-driven method of background constraint. Systematic uncertainties 91 will be discussed in section 6.

3 Figure 2: Left: number of 3D fussyK prongs per interaction from the coherent ⇡0 MC. Right: data and MC comparison of the number of prongs in the neutral current events.

64 in the low-⇣ region. It is therefore important to use a data-driven method to constrain both 65 the normalization and shape of the background. 0 0 66 The strategy of the coherent ⇡ analysis is as follows. First, we select single ⇡ events in 67 the NC sample defined by the absence of a reconstructed muon in the final state. Both photons 0 68 from ⇡ decay should be reconstructed as 3D prongs. The sample is composed of both coherent 69 and non-coherent (Resonance and DIS) interactions. Next, using kinematics, we define a control 0 70 sample, entirely dominated by non-coherent ⇡ , and a signal sample containing coherent and 71 non-coherent events. The control sample is used to tune the normalization and shape of the Cross-Section Measurement 72 non-coherent ⇣, which is then applied to the non-coherent background in the signal sample. ~5% systematic uncertainty 73 Finally, the coherent signal~8% is statistical measured uncertainty in the low-by ⇣changingregion control of the sample coherent definition, signal sample as the with 3.7E20POT data normalization method, etc. 74 excess over non-coherent prediction. 0 Normalized Background 75 The cross-section of coherentSelected⇡ production data is calculated as: N N = Data,selected Bkg,norm (1) ✏ N ⇥ Target ⇥ Signal efficiency 76 where NData,selected and~4% NUncertaintyBkg,norm fromare the number of data andFlux normalized MC background in 77 the selected coherent (low-signal modeling⇣) region by of the signal sample, ✏ is~11% the uncertainty eciency of coherent signal comparing BS vs RS model. Number of target nucleus Constrained by 78 selection calculated byEM MC, ShowerN selectionTarget is the number of target nucleusMIPP/NA49 in data the fiducial volume, and ~1% Uncertainty from 79 is the muon neutrinoefficiency flux. constrained by MR Brem detector mass simulation 80 Uncertainty to this• Expect analysis ~15% comes uncertainty from (stat both + syst) statistics using current and systematics.data (3.7E20 To reduce the statis- POT): a very competitive result. 81 tic uncertainty, we want to reduce the number of background (NBkg) while keeping relatively 38 82 high signal eciency (✏). Systematic uncertainty mainly comes from the measured number of 83 background (NBkg) and flux. Coherent modeling and detector simulation also contribute to 84 the uncertainty through ✏. The uncertainty to NBkg is constrained by the control sample as 85 described above. External data (MIPP/NA49) are used to constrain the flux uncertainty from 86 hadron production. 87 The neutrino flux, data and MC used in this analysis will be discussed in section 2 and 3. 0 88 Section 4 focuses on the selection of NC ⇡ sample, including both coherent signal and non- 89 coherent background. Section 5 present the selection of coherent signal sample and non-coherent 90 control sample, and the data-driven method of background constraint. Systematic uncertainties 91 will be discussed in section 6.

3 Summary

• Coherent is an important interaction mode for neutrino oscillation measurement, and also has its own physics interest. • NOvA near detector is good for measurement of EM particles including coherent induced π0. • We have enough data to do a cross-section measurement of coherent π0. • We have • A data-driven technique to constrain the background. • MR Brem showers to constrain signal efficiency. • External thin and thick target data to constrain flux uncertainty. • Expect to report a result soon. Stay tuned!

39 Back Up Slides Coherent π0 in the NOvA ND

• The signal we are looking for is one π0 decay into 2 photons, both reconstructed as 3D showers.

41 NOvA Flux #Strategy of numu as a function of true neutrino energy at NOvA Near Detector. NOvA Simulation

1012 • νμ flux comes from pion and kaon decay. From all parents

• Uncertainties come from hadron 2 From π+ production and beam transport simulation. 1011 POT/cm

20 From K+

• Use external thick target (MIPP) and thin 10 ×

/6 10

target (NA49) data to constrain the hadron 10Φ production uncertainty to ~10%. • ~5% uncertainty from beam transport. 2 4 6 8 10 12 14 16 18 20 E (GeV) νµ π+ phase space K+ phase space

42 Coherent π0: World Measurement

100 Results scaled to Carbon Target

80 /Nucleus)

2 There are relatively few

cm 0

-40 60 coherent π measurement,

GENIE

) (10 most suffer from large 0 40 NOMAD Worldπ Data Aachen-Padova uncertainty. Gargamelle 20

(COH CHARM

σ SKAT 0 0 10 20 30 Neutrino Energy (GeV)

40 2 Experiments A < E⌫ > (10 cm /N) /(⌫µ-CC) /(RS) Aachen-Padova 27 2 29 10 ± Gargamelle 31 2 31 20 ± CHARM 20 30 96 42 ± SKAT 30 7 79 28 4.3 1.5 ± ± 15’ BC 20 20 0.20 0.04s ± NOMAD 12.8 24.8 72.6 10.6 3.21 0.46 ± ± MiniBooNE 12 0.8 0.65 0.14 ± SciBooNE 12 0.8 0.9 0.20 ± NO⌫A 10.6 2.7 13.7 1.9 43 ±

Coherent ⇡0 Hongyue Duyang 44 Calibration and energy scale NOνA Preliminary Calibration and energy scale Calibration Data I Detector response varies substantially over length of cell due to 300 MC π0 signal attenuation in ﬁber I Response varies substantially MC bkg • Use stoppingalong cell muons due to as light standard atten. candles to correct for attenuation, I Use cosmic ray muons as a standard candle 200 Data µ: 134.2 ± 2.9 MeV threshold and shadowing effect. Data σ: 50.9 ± 2.1 MeV Events I Calibrate every channel (344,064) individuallyI Use cosmic ray muons as a MC µ: 136.3 ± 0.6 MeV • MC σ: 47.0 ± 0.7 MeV Cross-checkedstandard candle by Michael to calibrate electron, beam100 muons, hadronic energy/hit, and πevery0 mass channel peak (see individually plot later). I Use dE/dx near the end of stopping muons to set absolute scale 0 • 5% uncertainty on energy calibration. 0 100 200 300 400 500 M γγ (MeV) FD cosmic data - plane 84 (horizontal), cell 12 I Use dE/dx near the end of ND, 1.66×1020 POT NOνA Preliminary ×103 25 NOνA Preliminary stopping muon to set abs. scale 60 Simulated νµ CC Simulated Background 20 Data I Multiple calibration x-checks 15 40 I Beam muon dE/dx

10 I Michel energy spectrum Events Mean PE / cm 0 I 20 5 ⇡ mass peak I Hadronic energy/hit 0 -500 0 500 Distance from center (cm) 0 I Take 5% abs. and rel. errors on 0 0.01 0.02 0.03 0.04 Average Hadronic Energy Per Hit (GeV) C. Backhouse (Caltech) NOvA energy scale 39 / 27

C. Backhouse (Caltech) NOvA 38 / 27

44 What is Coherent?

Fig. 1. Diagram of the Cohπ0 process, ν + ν + + π0. A → A • Neutrinositself can as acoherently forward electromagnetic scatter shower off posing target a backgro nucleusund for the νviae- charge/ neutral currentinduced signal. interaction This is relevant and to the produce long baseline experi pions.ments searching for νe appearance with the purpose of measuring the mixing angle Θ13.Aprecise 0 • measurement of Cohπ ,althoughconductedatenergieshigherthanthose The targetof nucleus the long baseline stays projects in at Fermilabground (MINOS/NO state.νA), will constrain the • error on a model-prediction of this background to the νe appearance. Finally, Small momentumthe study of coherent transfer. pion production No quantum provides an insight number into the structure(charge, spin, isospin) exchange.of the weak hadronic current [1,8], and oﬀers a test of the partially conserved axial-vector current hypothesis (PCAC) [9]. Ref. [10] presents an excellent • Single forward-goingreview of these topics. pion in the final state, no vertex activity. Acoherentinteraction,Figure1,wherenochargeorisospinis exchanged between the ν and the target nucleus ( )whichrecoilswithoutbreakup,leads A 0 to an enhancement in the cross-section.45 In the Cohπ process the interaction is mediated by a pomeron-like particle bearing the quantum number of the vacuum. The cross section is dominated by the axial vector current. The contribution of the isovector current to the Cohπ0 process is minimal where Z0 can be viewed as a ρ meson which produces a π0 exchanging an isoscalar ω with .ThisminimalcontributionoftheisovectorcurrenttotheCohπ0 arises A from two reasons: (a) the cross section of the isovector ρ- interaction is zero A in the forward direction, a direction preferred by the nuclear form factor; and (b) the vector component has a contribution proportional to (1 2sin2 θ )2 − W reducing the isovector contribution further, the net reduction with respect to the axial part being a factor of 3.5. The PCAC hypothesis stipulates that for zero-momentum transfer (Q2 =0,whereQ2 is the negative of the square of the four-momentum transfer from the incident neutrino to the target), the ν- cross section can be related to the π- cross section. The ν- cross A A A section in the forward direction is related to the strong π- interaction as A follows: