MINERvA Review

Tomasz Golan

Neutrino Seminar, 14.11.2016

Postdoc summary

University of Rochester

Department of Physics and Astronomy

MINERvA (prof. Kevin McFarland)

  • flux systematic errors
  • physics reconstruction using machine learning
  • generator group coordinator
  • production group leader

Fermi National Accelerator Laboratory

Scientific Computing Division

GENIE (dr. Gabe Perdue)

  • nuclear effects in GENIE
  • automated validation system
  • "user support"

Main INjector ExpeRiment \(\nu\)-A

MINERvA Experiment

  • MINERvA is a neutrino-scattering experiment at Fermilab
  • Collaboration of about 50-100 physicist
  • NuMI beam is used to measure cross section for neutrino-nucleus interactions
  • The detector includes several different nuclear targets

Detector

Nuclear targets

NuMI Beamline

Low vs Medium Energy


  • by changing distance between horns one can change energy spectrum

  • by changing horns polarization one can switch between neutrino and anti-neutrino mode

Flux

NuMI Beam Simulation

  • Flux simulation starts with a Geant4 with the NuMI geometry

  • All of the information about interactions leading to neutrino are stored

  • The results of the simulation are corrected by external data

  • Similar approach to the one T2K used

Hadron Scattering Data

  • NA49 - charged hadron production in proton scattering off thin targets

  • FLUKA is used to scale proton energy from \(158\) to \(120\) GeV

  • MIPP - charged hadron production on thin target and NuMI target replica

NA49 Data

MIPP data

Reweighting

\[w_{HP} = \frac{f_{data}(x_F, p_T, E)}{f_{MC}(x_F, p_T, E)}, ~~~~~ f\equiv\frac{E}{\sigma}\frac{d^3\sigma}{dp^3}\]

  • Using external hadron production data events are weighted using the above formula

  • An event is reweighted on "interaction-by-interaction" basis

  • Whenever possible - "thick" target data is used

Uncertainties

  • Many-Universes method is used to propagate external data uncertainties to our flux

  • for each universe (u) data central value is shifted (respect to data uncertainties)

\[w_u \sim \prod_i w_{HP, u, i}\]

Flux history

Flux generations

  • Generation 0 -> no MIPP data

  • Generation 1 -> MIPP thin target data + other improvements

  • Generation 2 -> MIPP thick target data + other improvements

Flux vs publications

Flux Analysis Reference
Generation 0 \(\nu_\mu\) CCQE PRL 111 (2013) 022502
Generation 0 \(\bar{\nu_\mu}\) CCQE PRL 111 (2013) 022501
Generation 1 \(\nu_\mu\) muon + proton PRD 91 (2015) 071301
Generation 1 \(\nu_\mu\) CC \(\pi^\pm\) PRD 92 (2015) 092008
Generation 1 \(\bar{\nu_\mu}\) CC \(\pi^0\) PLB 749 (2015) 130
Generation 1 Coherent \(\pi\) PRL 113 (2014) 261802
Generation 1 CC target ratios PRL 112 (2014) 231801

Generation 0 vs Generation 2 (thick off)

Generation 1 vs Generation 2 (thick off)

Generation 2 (thin vs thick)

Generation 2: ratio

Flux constraints

\(\nu e\) constraint

  • Weighting up universes that agree better with data
  • Experimental signature is a very forward single electron in the finale state

Generation 1 + \(\nu e\)

low-\(\nu\) method

\[\frac{d\sigma}{d\nu} = A + B\frac{\nu}{E} - \frac{C}{2}\frac{\nu^2}{E^2}\]

  • Differential cross-section can be expressed by the above formula

  • It is a constant for \(\frac{\nu}{E} \rightarrow 0\)

  • It can be used to constraint the flux prediction (with high-energy normalization taken from other measurements, like NOMAD)

low-\(\nu\) vs generation 2

CCQE measurements

CCQE-"true" aka "1-track"

Flux Analysis Reference
Generation 0 \(\nu_\mu\) CCQE PRL 111 (2013) 022502
Generation 0 \(\bar{\nu_\mu}\) CCQE PRL 111 (2013) 022501


  • require only muon track

  • target -> scintillator (CH)

High-\(Q^2\) candidates

Low-\(Q^2\) candidates

Background

Background subtraction

\(\nu_\mu\) CCQE

\(\bar\nu_\mu\) CCQE

CCQE-"like" aka "muon+proton"

Flux Analysis Reference
Generation 1 \(\nu_\mu\) muon + proton PRD (2015) 071301


  • CC \(\nu_\mu\) on \(CH\)

  • require a muon, at least one proton, and no pions in the final state

  • based on hadronic kinematics

  • proton kinetic energy > 110 MeV

\(\nu_\mu\) CCQE-like

Pion production measurements

Charged pion production

CC \(1\pi^\pm\)

  • require a muon and exactly one charged pion

  • \(W < 1.4\) GeV

CC \(N\pi^\pm\)

  • require a muon and at least one charged pion

  • \(W < 1.8\) GeV

Invariant mass


\[E_\nu = E_\mu + E_{recoil}\] \[Q^2 = 2E_\nu(E_\mu - |\vec p_\mu|\cos\theta_\mu) - m_\mu^2\] \[W_{exp}^2 = M_p^2 - Q^2 + 2M_pE_{recoil}\]


  • \(E_{recoil}\) is measured colorimetrically

Backrounds vs invariant mass

CC \(1\pi^\pm\)

CC \(N\pi^\pm\)

\(\bar\nu_\mu\) CC \(1\pi^0\)

  • require a muon and a single neutral pion (visible as two photons)
  • about 70% of background - multipion events \(\pi^0 + \pi^\pm\)
    • \(pi^\pm\) is not tracked
    • \(\pi^-\rightarrow\pi^0\)
  • the rest of the background is mostly due to energy deposit by \(\pi^-\) and neutrons misidentified as photons

Invariant mass of \(\gamma\gamma\)



The \(\gamma\gamma\) invariant mass is reconstructed from the photon energies (\(E_1\), \(E_2\)):

\[m_{\gamma\gamma}^2 = 2E_1E_2(1 - \cos\theta_{\gamma\gamma})\]

Differential cross sections

Coherent pion production

  • require two final state particles: \(\mu^\pm\) and \(\pi^\mp\) and no extra visible recoil

Background

Low energy transfer requirement

Total cross section

Differential cross section (energy)

Differential cross section (angle)

Other measurements highlights

Inclusive \(\nu_\mu\) CC ratios

  • shadowing at low \(x\)

  • no MEC in simulations (high \(x\) dominated by QE)

DIS \(\nu_\mu\) CC ratios

  • \(W > 2\) GeV and \(Q^2 > 1\) GeV\(^2\)

  • \(E_\nu\) up to 50 GeV

Available energy vs momentum transfer


  • \(E_{avail}\) - sum of proton and charged pion kinetic energy and neutral pion, electron, and photon total energy

NC diffractive \(\pi^0\) production


  • The most plausible source of the excess seen in the data is diffractive NC \(\pi^0\) production from hydrogen in the scintillator target of MINERvA.

CC pion production (muon variables)


  • \(\nu\)-CC(\(\pi^+\)) and \(\bar\nu\)-CC(\(\pi^0\))

  • total cross section

  • differential cross sections:

    • muon momentum
    • muon angle
    • \(Q^2\)

Kaon production

Summary

  • MINERvA offers an unique opportunity to measure neutrino cross section on different nuclear targets

  • There is still a lot of collected data to study (e.g. in nuclear target region)

  • Medium energy data will allow to study more precisely DIS (and transition region?)