NuMI Neutrino Flux Predictions Alexander Radovic College of - - PowerPoint PPT Presentation

▶
numi neutrino flux predictions
SMART_READER_LITE
LIVE PREVIEW

NuMI Neutrino Flux Predictions Alexander Radovic College of - - PowerPoint PPT Presentation

Jun 08, 2023 419 likes •813 views

NuMI Neutrino Flux Predictions Alexander Radovic College of William and Mary Alexander Radovic NuMI neutrino flux predictions 1 The NuMI Beam NuMI = Neutrinos at the Main Injector Target: two distinct styles (MINOS-era [le],

slide-1
SLIDE 1

NuMI Neutrino Flux Predictions

Alexander Radovic College of William and Mary

1

Alexander Radovic NuMI neutrino flux predictions

slide-2
SLIDE 2

The NuMI Beam

2

Alexander Radovic NuMI neutrino flux predictions

NuMI = Neutrinos at the Main Injector

  • Target: two distinct styles (MINOS-era [le], NOvA-era [me])
  • both are graphite, but internal structure is different
  • MINOS-era target could be repositioned relative to horn 1

Two “horns” produce magnetic fields that focus secondaries

slide-3
SLIDE 3

The NuMI Beam

3

Alexander Radovic NuMI neutrino flux predictions

NuMI = Neutrinos at the Main Injector

slide-4
SLIDE 4

The NuMI Targets

4

Alexander Radovic NuMI neutrino flux predictions

MINOS (LE): NOvA (ME):

slide-5
SLIDE 5

5

Alexander Radovic NuMI neutrino flux predictions

MINOS (LE): NOvA (ME):

The NuMI Targets

slide-6
SLIDE 6

Comparing LE and ME Hadron Production

6

Alexander Radovic NuMI neutrino flux predictions

slide-7
SLIDE 7

NuMI-X

7

Alexander Radovic NuMI neutrino flux predictions

Who uses NuMI? One Beamline - Many Experiments:

  • MINOS (+): steel & plastic scintillator sandwich (Near + Far ), on axis
  • MiniBooNE: liquid scintillator, off-axis 121mrad
  • ArgoNeuT: small liquid Ar TPC, on axis
  • Minerva: fine grained calorimeter w/ variety of nuclear targets, on axis
  • NOvA: large segmented liquid scintillator (Near + Far), off-axis 14mrad
  • ther soon to exist experiments (e.g. off-axis microBooNE, far future

LBNF) The NuMI-X mission statement: Get all the NuMI experiments “on board” to work

  • n Beam Simulation and Beam Analysis (including hadron production studies).

To produce a reference flux that all NuMI experiments can use.

slide-8
SLIDE 8

The NuMI Beam Seen at the NuMI Experiments

8

Alexander Radovic NuMI neutrino flux predictions

Relative positioning affects focusing, and thus the spectrum: (on-axis) (14 mrad) (7.34 mrad)

slide-9
SLIDE 9

A Priori Predictions and Uncertainties

9

Alexander Radovic NuMI neutrino flux predictions

What do we know a priori?

slide-10
SLIDE 10

Recent NuMIX Efforts

10 10

Alexander Radovic NuMI neutrino flux predictions

Unifying disparate code bases:

  • steps taken, still work to be done
  • g4numi and g4numi_flugg should use the same geometry
  • — but, sadly, they currently don’t

Unifying the output format — Dk2Nu:

  • more structured; standardized naming conventions; flexibility for storing

pre-calculated detector location energies and weights

  • carry on Minerva’s addition of recording full ancestry
  • non-NuMI specific; LBNE & Booster adoption in progress
slide-11
SLIDE 11

The MC

11 11

Alexander Radovic NuMI neutrino flux predictions

g4numi_flugg = fluka physics + G4 geometry + flugg “glue”:

  • all (recent) MINOS and NOvA analyses to date have used flugg
  • historically it seemed to better represent what was seen in data when

used “out of the box” g4numi = pure Geant4:

  • primarily used by Minerva
  • interesting new development work being done here
  • more choices of physics models
  • local experts/development of G4
slide-12
SLIDE 12

ME at MINOS & NOvA Near Detectors

12 12

Alexander Radovic NuMI neutrino flux predictions

slide-13
SLIDE 13

13 13

Alexander Radovic

Neutrino flux prediction is notoriously difficult, relying on the extrapolation of sparse fixed target data to the energies seen on the NuMI target. Estimates of focusing and geometric uncertainties in the final flux prediction are estimated by producing alternative flux MC shifted to their 1sigma uncertainties.

Focusing Uncertainties

True Neutrino Energy (GeV) 5 10 15 20 25 30

nominal
  • /
shifted
  • 0.96

0.98 1.00 1.02 1.04 MINOS+ PRELIMINARY

Horn Current Miscalibration Systematic Beam Flux Simulation 200 kAmp Nominal vs. 199 kAmp Distribution Far Detector Ratio Near Detector Ratio Far/Near Double Ratio

True Neutrino Energy (GeV) 5 10 15 20 25 30

nominal
  • /
shifted
  • 0.96

0.98 1.00 1.02 1.04 MINOS+ PRELIMINARY

Horn 1 Alignment Systematic Beam Flux Simulation - Near Detector Only Nominal Position vs. 0.5mm Shifts 0.5mm Horizontal Shift + 0.5mm Horizontal Shift

  • 0.5mm Vertical Shift

+ 0.5mm Vertical Shift

  • NuMI neutrino flux predictions
slide-14
SLIDE 14

Beam Constraints

14

NuMI neutrino flux predictions

What can we learn about the beam from our data and external constraints?

Alexander Radovic

slide-15
SLIDE 15

NuMI Constraints

Alexander Radovic

15

NuMI neutrino flux predictions

Multiple detector locations:

  • Angle relative to beam direction
  • on-axis yields a broad spectrum beam
  • off-axis sees a more narrow spectrum
  • Different positions are sensitive to π production in different regions of pt & pz

Multiple target designs:

  • MINOS-era target could be repositioned relative to the horn

Horn current affects focussing:

  • “horn off” is valuable running condition
slide-16
SLIDE 16

Tuning to Measured Spectra

One approach is to use a physically motivated hadron production parameterization and focusing uncertainties to create a fit which uses all our available beam modes to constrain our flux prediction. At MINOS the hadron production parameterization is a slightly altered version of the BMPT parameterization where we use linear warpings of some

  • f it’s key variables to tune hadron

production in the fit.

Alexander Radovic

16

NuMI neutrino flux predictions

Muon-Neutrino CC selected sample

Reconstructed Energy (GeV)

slide-17
SLIDE 17

Different Beam Modes

Data/MC disagreement varies as a function of energy in different beam modes, suggesting that flux uncertainties rather than detector or cross-section uncertainties are dominating the Near Detector Discrepancy. Each beam mode also gives us access to a different region of Pion and Kaon production phase space so that we can better constrain our parameterization of the raw yield of hadron production coming off of the target.

NuMI neutrino flux predictions Alexander Radovic

17

slide-18
SLIDE 18

Fit Result: Final Tuned Flux

NuMI neutrino flux predictions Alexander Radovic

18

slide-19
SLIDE 19

Fit Result: Final Tuned Flux

NuMI neutrino flux predictions Alexander Radovic

19

slide-20
SLIDE 20

Fit Result: Hadron Production Weights

NuMI neutrino flux predictions Alexander Radovic

20

slide-21
SLIDE 21

Tuning in the ME Era

Historically we made use of the power of the old beam to run in different beam modes to access a wide range of Pion/Kaon kinematics and deconvolve cross section effects. Can we do something similar in the ME beam by looking at different spectra in the same beam?

Alexander Radovic

21

NuMI neutrino flux predictions

π+ MINOS π+ NOvA K+ MINOS K+ NOvA

slide-22
SLIDE 22

The Muon Monitors

Yet another approach is to attempt to measure the flux by measuring the rate and energy of muons produced in pion and kaon decays in the NuMI decay pipe. Laura Loiacono performed that analysis* using the Muon Monitors just after the decay pipe. Whilst the fit has a large uncertainty the final result is largely consistent with that of the MINOS beam fitting. With work they could be a powerful constraint on the new beam.

Alexander Radovic

22

P+ Weights

*Laura Jean Loiacono, University of Texas at Austin, May 2010 “Measurement of the muon neutrino inclusive charged current cross section on iron using the MINOS detector” Fermilab-Thesis-2011-06

slide-23
SLIDE 23

The Muon Monitors

Alexander Radovic

23

NuMI neutrino flux predictions

Monitor 1: Monitor 2: Monitor 3:

slide-24
SLIDE 24

The Low 𝝃 Method

Another approach is to attempt to measure the flux by selecting events with a well understood cross section. One approach is to select for CC events with a low inelasticity*neutrino energy or “𝝃”. Used in a preliminary MINOS cross section analysis* this study showed that data/MC discrepancy at the MINOS ND was indeed largely driven by the difference between the measured and predicted flux.

  • Alexander Radovic

24

NuMI neutrino flux predictions

*Debdatta Bhattacharya, March 2009, “Neutrino and antineutrino inclusive charged-current cross section measurement with the MINOS near detector”, Fermilab-Thesis-2009-11

slide-25
SLIDE 25

Using External Data

Alternatively we can reweight our MC our yield of cross-sections to information from fixed target experiments. MINERvA will cover this topic in detail in the next talk, but broadly we can use thin target data (NA49 etc.) and reweight each interaction or use thick target data (MIPP, USNA61) and reweight the yield. New thick target data is needed to characterize the new ME beam.

Alexander Radovic

25

NuMI neutrino flux predictions

slide-26
SLIDE 26

Summary

  • Working together the NuMI experiments can more

thoroughly & efficiently test and tune their MC geometries.

  • Thin and Thick target experiments like NA49/MIPP

give us strong constraints on our hadron yield but we need USNA61 to really understand the new beam.

  • NuMI-X has inherited some powerful tools for

constraining and understanding the new NuMI beam but will have to work harder than ever before now that we no longer have access to a plethora of alternative beam modes.

26 26

Alexander Radovic NuMI neutrino flux predictions

slide-27
SLIDE 27

The NuMI Beam Seen at the NuMI Experiments

27 27

Alexander Radovic NuMI neutrino flux predictions

Relative positioning affects focusing, and thus the spectrum:

E (GeV)

5 10 15 CC / 6E20 POT / kTON / 50 MeV ν

6

10

  • 3

10

  • 2

10

  • 1

10 1 10 Total

µ

ν

µ

ν MINOS+ Preliminary

98.3% νμ (1-3GeV) 1.7% ν̅μ 97.3% (1-3GeV) 2.7% ν̅μ

slide-28
SLIDE 28

28 28

Alexander Radovic

Those alternative parameterizations can then be propagated to our final MC neutrino reconstructed energy and used to generate a covariance matrix describing both the uncertainty on individual bins and how we would expect that any shifts to be correlated with neighboring energy bins.

A Priori Flux Prediction and Uncertainties

NuMI neutrino flux predictions

slide-29
SLIDE 29

29 29

Alexander Radovic

Hadron Production Uncertainty

Ultimately these large FD uncertainties cancel in a Far/Near ratio

A Priori Flux Prediction and Uncertainties

Those alternative parameterizations can then be propagated to our final MC neutrino reconstructed energy and used to generate a covariance matrix describing both the uncertainty on individual bins and how we would expect that any shifts to be correlated with neighboring energy bins.

Hadron Production Uncertainty

NuMI neutrino flux predictions

slide-30
SLIDE 30

The Branching MC

30 30

Alexander Radovic

MINOS numisoft NOvA fork Minerva fork nova branch minerva branch proto2-numix branch proto1-numix branch numi-beam-sim

flugg g 4 n u m i

2009-09 2010-08 1999-07 2013-05 2014-09

unified trunk

slide-31
SLIDE 31

The MINOS Parameterization

31 31

Alexander Radovic NuMI neutrino flux predictions

Original Fit to Hadron yield:

slide-32
SLIDE 32

The MINOS Parameterization

32 32

Alexander Radovic NuMI neutrino flux predictions

What we fit for:

slide-33
SLIDE 33

Fit Result: Final Tuned Flux

NuMI neutrino flux predictions Alexander Radovic

33

slide-34
SLIDE 34

Fit Result: Hadron Production Weights

NuMI neutrino flux predictions Alexander Radovic

34

slide-35
SLIDE 35

Components of the MINOS+ Beam

35 35

Alexander Radovic NuMI neutrino flux predictions

slide-36
SLIDE 36

Region of Interest and NA49 Coverage

36 36

Alexander Radovic

NA49 P+

NuMI neutrino flux predictions

slide-37
SLIDE 37

Region of Interest and NA49 Coverage

37 37

Alexander Radovic

NA49 K+

NuMI neutrino flux predictions

slide-38
SLIDE 38

38 38

Alexander Radovic

Example Alternative Parameterizations

NuMI neutrino flux predictions