Cross section measurements and capabilities at NOvA Leo Aliaga - - PowerPoint PPT Presentation

cross section measurements and capabilities at nova
SMART_READER_LITE
LIVE PREVIEW

Cross section measurements and capabilities at NOvA Leo Aliaga - - PowerPoint PPT Presentation

Cross section measurements and capabilities at NOvA Leo Aliaga March 12, 2018 Cross Section Measurement Strategy Workshop, Fermilab Outline Overview. Overview of the NOvA beam, detector and simulation. Strategy and challeges. Recent


slide-1
SLIDE 1

Leo Aliaga

Cross section measurements and capabilities at NOvA

March 12, 2018 Cross Section Measurement Strategy Workshop, Fermilab

slide-2
SLIDE 2

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 2

Outline

Overview.

  • Strategy and challeges.

Conclusions. Recent results.

  • Overview of the NOvA beam, detector and simulation.

Inclusive Analyses.

slide-3
SLIDE 3

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 3

Overview

slide-4
SLIDE 4

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 4

Introduction

NOvA is a long baseline oscillation experiment to measure: The ND provides an excellent opportunity to measure neutrino interaction cross sections with high statistics.

  • Constrain our cross section systematics.
  • Contribute to the current efforts of the neutrino community on understanding

neutrino interactions.

  • Collaborate with future experiments such as DUNE.
  • Mixing angle Θ23.
  • CP-violating phase.
  • Mass hierarchy determination.

With these measurements we can

slide-5
SLIDE 5

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 5

NuMI Beam at NOvA

NUMI FHC

NOvA detectors are off-axis, 14 mrad w.r.t NuMI beam axis.

  • It is a narrow-band beam centered

around 2GeV. In FHC: 96.2% νμ, 3.3% νμ, and 0.5% νe.

slide-6
SLIDE 6

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 6

NuMI Beam at NOvA

Even with a narrow band beam, NOvA is still sensitive to many different nu+A interaction channels.

  • 8.09 x 1020 in the FHC mode.
  • Currently 6.26 x 1020 in the RHC mode.

Protons on target:

NOvA

Muon Neutrino

J.A. Formaggio, G.P. Zeller

  • Rev. Mod. Phys. 84, 1307 (2012)

High data rate at the ND.

slide-7
SLIDE 7

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 7

NuMI Beam at NOvA

Even with a narrow band beam, NOvA is still sensitive to many different nu+A interaction channels.

Muon Neutrino

J.A. Formaggio, G.P. Zeller

  • Rev. Mod. Phys. 84, 1307 (2012)

T2K + MicroBooNE + NOvA + MINERvA

  • 8.09 x 1020 in the FHC mode.
  • Currently 6.26 x 1020 in the RHC mode.

Protons on target: High data rate at the ND.

slide-8
SLIDE 8

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 8

NuMI Beam at NOvA

Even with a narrow band beam, NOvA is still sensitive to many different nu+A interaction channels.

T2K + MicroBooNE + NOvA + MINERvA

Muon Antineutrino

J.A. Formaggio, G.P. Zeller

  • Rev. Mod. Phys. 84, 1307 (2012)
  • 8.09 x 1020 in the FHC mode.
  • Currently 6.26 x 1020 in the RHC mode.

Protons on target: High data rate at the ND.

slide-9
SLIDE 9

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 9

NOvA Near Detector

Wavelength- shifting fibers routed to a single cell on an Avalanche Photodiode (APD). Made of PVC with liquid scintillator, 3.9m x 3.9 m x 12.67

  • m. 193 ton, 192 planes and ~20k channels.
  • Fully active region: 77% hydrocarbon, 16% chlorine

and 6% TiO2.

  • Muon Catcher: steel + NOvA cell at downstream end

to range-out muons. O(10) ns single hit timing resolution

slide-10
SLIDE 10

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 10

NOvA Near Detector Events Display

Top View Side View

Muon Catcher

Muon Catcher

Hits associated in time and space are used to form a candidate interaction. Tracks and showers are reconstructed from these hits. Colors show time:

slide-11
SLIDE 11

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 11

Strategy

Simulation

  • We use G4NuMI for the beam simulation, GENIE (2.12.2) for the neutrino

interactions and Geant4 (4.10.1) for propagating the particles.

  • A correction to the central value is made coming from:
  • The beam: PPFX for the hadrons production in the beamline.
  • The cross section: a tuning is applied to account for FSI current knowledge

(see Aaron’s talk yesterday).

  • The beam and cross section systematics are determined by PPFX and the

GENIE knobs scheme.

  • The simulation of the intensity dependent of high rate of muons originating in the

surrounding rocks (rock muons) is integrated overlaying with the neutrino events .

  • The detector response is also simulated and the uncertainties on the calibration

parameters are dealt with systematic shifted MC.

slide-12
SLIDE 12

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 12

Strategy

  • Vertices should be inside a fully active (fiducial) region to cut rock muons.

Data and MC analysis Top View Side View

Muon Catcher

Muon Catcher

  • Tracks should be contained in the fiducial + Muon Catcher to avoid shower leaking.
slide-13
SLIDE 13

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 13

Strategy: main challenges

  • Cross section analysis requires refine existing or develop new tools.
  • 1. Most of the tools have been optimized for the oscillation analysis.
  • 2. We are more sensitive to some systematics than the oscillation analysis: beam

normalization or cross section mis-modeling goes directly in our uncertainty.

  • We use PPFX (Package to Predict the FluX) and the GENIE reweighing

scheme for systematics.

  • We implemented the multi-universe approach to handle beam and cross

section uncertainties.

  • We are working to a fully generated event by event MC with different

generators, such as GiBUU and NEUT.

slide-14
SLIDE 14

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 14

Strategy: main challenges

  • 3. Develop a PID algorithm for non-lepton final state particles.
  • Some analyses are using a convolutional visual network (CVN) trained
  • n topological features of individual prongs itself at the NOvA detectors.
  • We are moving to a final state particle identification for recognizing the

neutrino event.

  • NOvA uses a Convolutional Neural Network (CNN) where a series of image

filters are applied to hit map images to extract features associated with an interaction.

slide-15
SLIDE 15

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 15

Current Analyses

slide-16
SLIDE 16

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 16

Current Analyses

slide-17
SLIDE 17

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 17

Our Strategy

First results presented

  • n Dec 1, 2017
slide-18
SLIDE 18

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 18

Our Strategy

Top Priority

slide-19
SLIDE 19

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 19

Our Strategy

And then, we will make the ratios for the semi [ex,in]clusive channels.

slide-20
SLIDE 20

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 20

Our Strategy

Will constraint the flux

slide-21
SLIDE 21

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 21

Our Delivery

Measurements respect to different particle kinematics (momentum and angle) and neutrino energy.

  • Unfolded cross-sections.
  • Event rates at the detector and a folding matrix.
  • Correlation and covariance matrices.
slide-22
SLIDE 22

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 22

Beam Hadron Production Uncertainties

Muon neutrino flux:

  • ~ <1.5 hadronic interactions> contributes to the peak.
  • ~ 8% uncertainty at the peak.

Incident mesons, quasi-elastic and proton interacting in materials beyond carbon would reduce the beam uncertainty significantly. New data on HP experiments such as NA61 and EMPHATIC will help.

slide-23
SLIDE 23

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 23

Recent Results from NOvA

Preliminary results were presented in JETP, Dec 1st, 2017.

  • νμ CC Neutral Pion.
  • NC Coherent Neutral Pion.
slide-24
SLIDE 24

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 24

νμ CC Neutral Pion

Flux-average cross section of muon and neutral pion kinematics (angle respect to the beam and momentum), Q2 and W. It uses a data-driven technique for the signal and background fit: makes a template fit of the PID distribution of the signal and background per kinematic bin to match the MC to Data. The CC Nue Inclusive uses this procedure (see some slides ahead).

0.5 < pμ < 0.6 GeV/c

Before fit After fit The CCπ0ID separates signal from background.

Photon score and a CCπ0ID are developed based on dE/dx and “gappiness" of the tracks.

slide-25
SLIDE 25

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 25

νμ CC Neutral Pion: Some Results Shown in JETP, Dec 1st, 2017

Result consistent with GENIE FSI model. GENIE shape prediction lightly over-predicts around pμ ~ 0.3 GeV/c and Q2 ~ 0.6 GeV2.

slide-26
SLIDE 26

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 26

NC Coherent Neutral Pion

Cut on invariant mass reduces background. Two showers identified as photons by dE/dx-based likelihoods. Data-driven technique is used to constrained the photon simulation: we compare the isolated EM showers in data and MC from “golden” muons. One challenge: relative small number

  • f signal.
slide-27
SLIDE 27

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 27

Inclusive Analyses

slide-28
SLIDE 28

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 28

νμ CC Inclusive

Flux-averaged double differential cross section in the muon angle respect to the beam and the muon momentum, as well the neutrino energy. Statistical uncertainties are typically < 2%. Expected systematics are ~10% highly correlated (normalization) flux uncertainties, and all others systematics combined to be 5-8%. Restricted kinematic phase space due to limited statistics and low efficiency.

slide-29
SLIDE 29

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 29

νμ CC Inclusive: Efficiency and Background

Selection efficiency is dominated by containment cuts. Backgrounds are small near the peak, larger in the tail. Uncertainties are at the level of a few %.

slide-30
SLIDE 30

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 30

νμ CC Inclusive PID

We are moving to a Muon PID based only on the <dE/dx> and scattering likelihoods along the track to select a muon respect to a pion. A muon is select from a LLR: An optimum cut that minimizes the cross section uncertainty is found at Muon LLR ~ 0.8.

slide-31
SLIDE 31

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 31

νe CC Inclusive

Flux-averaged double differential cross section in the electron angle respect to the beam and the electron energy, as well the neutrino energy. Restricted kinematic phase space due to limited statistics and low efficiency. Challenging because there are <1% νe in the beam.

slide-32
SLIDE 32

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 32

νe CC Inclusive: Efficiency and Purity

Xsec, FSI and calibration systematics included in the error bands. Uncertainties on efficiency and backgrounds is between 5-10%.

Electron Energy

Preselection based on containment and rejecting muons.

slide-33
SLIDE 33

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 33

νe CC Inclusive: Efficiency and Purity

Electron Angle

CVN separates signal from backgrounds. We are using the “template fit” technique by per “cosθe vs electron energy bin” to match the MC to Data A data-driven technique is used to check the efficiency: we replace a “golden” muon with a simulated electron in data and MC to study the hadronic response.

slide-34
SLIDE 34

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 34

Other analyses

CC N π+- (N>=1)

  • Working on improving π+/π- reconstruction.
  • Using prong-level CVN classifier for charged pions.
  • The first goal is to report differential cross section in muon kinematics.

CC 0 π

  • The first goal is to report differential cross section in muon kinematics.

NC π0 (N>=1)

  • Important background to νe.
  • A event-level Boost Decision Tree (BDT) developed using shower variables as

inputs.

  • The first goal is to report differential cross section in π0 kinematics.
  • Working on improving the pion rejection.
  • Exploring proton reconstruction capabilities to study FSI.
slide-35
SLIDE 35

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 35

Conclusions

The NOvA experiment has and excellent opportunity to make a high precision neutrino-nucleus cross section measurements for both, FHC and RHC. The CC semi-inclusive neutral pion and the NC coherent neutral pion cross sections are in the last stages for publication. The CC inclusive channels have the highest priority for the next months. Ratio measurements of semi-[in,ex]clusive channels respect to the inclusive cross section are planed to be pushed forward soon. We already have presented results for the CC semi-inclusive neutral pion and the NC coherent neutral pion cross sections that will be published soon.

slide-36
SLIDE 36

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 36

Backup

slide-37
SLIDE 37

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 37

νμ CC Neutral Pion: Photon Selection

With this input, the neutral pion identifier is formed: CCπ0ID The photon identifier from 4 variables: two variables to describe dE/dx: the Brag peak and the energy per hit; and two to describe “gappiness”: distance from the vertex and the skipped planes along the prong.

slide-38
SLIDE 38

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 38

NC Coherent Neutral Pion: Background Constraint

Background constraint based on control and signal region Signal region: almost 90% of the signal Fit the backgrounds to control sample data in π0 energy vs angle 2D space

slide-39
SLIDE 39

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 39

NC Coherent Neutral Pion: MRBrem

Rock muons can induce EM shower via bremsstrahlung radiation

slide-40
SLIDE 40

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 40

NC Coherent Neutral Pion: MRBrem

Technique: isolate those EM showers in data and MC.

slide-41
SLIDE 41

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 41

NC Coherent Neutral Pion: MRBrem

This provides a method to constraint the photon simulation.

slide-42
SLIDE 42

3-12-2018 Leo Aliaga | Cross Section Measurement Strategy Workshop, Fermilab 42

NC Coherent Neutral Pion: MRBrem