Flux Error Analysis

MINOS Collaboration Meeting, 10-15-05

Flux Error Analysis

Žarko Pavlović

For

Beam Systematics Group


Systematic Uncertainties

• Goal of the study was to produce an “error-band” for neutrino flux

• Effects considered– Uncertainty on number of protons on target– Misaligned angle of horn 1 & horn 2 – Misalignment offset of horn 1 & horn 2 – Uncertainty on the location of the chase shielding blocks– Incorrect calibration of horn current – Incorrect description of horn current in the inner conductor of the

horns (skin depth effect)– Proton beam scraping on the baffle– Hadron production model

• Used old MC studies and PBEAM to estimate how much of an effect these uncertainties cause


Protons on Target

• Old beam MC was off in scale by 4.8%. This had 2 contributions:– The toroids TOR101 and TORTGT were miscalibrated.

Disagreement with Main Injector Toroid (called I:BEAM) at the level of ~4% has been known for some time.

– The full NuMI beam does not hit the target. On average, 0.6% misses to the left and right and goes to the hadron absorber.

• The net result of these two effects is that the beam MC was too high by 4.8%.

• Since the toroids were calibrated net uncertainty on POT is ~1%. Will use this value for an error band.


Fraction of beam on Target

• Must look at this repeatedly throughout our analyses.

• Through July 1, ~0.6% of beam misses the target and goes down the aperture between the target and baffle.

p beam profile

target size

=1.2mm

PMTGT (1 spill)

• The proton beam size increases as we go to higher intensity. Also varies from day-to-day (“bad” quality beam).

• We directly measure the size at (near) the target, so can correct for any fraction missing the target.


Horn Offsets & Angles

• Bob Zwaska performed beam-based alignment.– www.hep.utexas.edu/numi/beamMC/

• Net outcome of his work:– Horns are not on the exact same line as the target by

(0.5-1.0) 0.2 mm– The horns make angles with respect to the primary

beam at the (0.1-0.2)0.1mrad level.

• To be conservative we take an uncertainty of 1mm (offset) and 0.2mrad (angle)


Incorrect Horn Current calibration

• GNuMI-v17 MC (used in R1.18) used nominal values for Horn Current

• Jim Hylen measured the current and found the true current to be 1.60.5% lower than the nominal values

• New GNuMI-V18 MC run with true values

• Assume ~1% uncertainty on Horn Current


Skin Depth Effect

• Field between conductors is known and can be measured but the field in the inner conductor is worrisome

– Can model if we know the skin depth of the current

– Difficult since the conductor will change as it heats and is irradiated.

• This plot (shown at Ely ’04) gives GNUMI calculation if =0 or if = (latter closer to truth)

– Variation shown here probably overstates the uncertainty on the flux from this effect.

• To get uncertainty, we modeled– Current uniformly distributed

throughout the IC (= )– Current exponentially falling as a

function of radius with =6mm (cf J. Hylen).

Current fills IC

Current on inner sheet of IC

M. Messier


Baffle Scraping

• Collimating baffle can also act as a target.• Halo of proton beam scrapes on the baffle.• We measure the halo using PMTGT (<1%, )

G3NuMI Monte Carlo

G3NuMI Monte Carlo

M. Messier


(Almost) Final Uncertainties: LE Beam

• All effects excluding Hadron Production!

Uncertainty on ND Spectrum Uncertainty on F/N ExtrapolationF

ocus

ing

Pea

k

Foc

usin

g P

eak


(Almost) Final Uncertainties: LE(-10) Beam


Uncertainty on ND Spectrum Uncertainty on F/N ExtrapolationF

ocus

ing

Pea

k

Foc

usin

g P

eak


(Almost) Final Uncertainties: pME Beam


Uncertainty on ND Spectrum Uncertainty on F/N Extrapolation

Foc

usin

g P

eak

Foc

usin

g P

eak


(Almost) Final Uncertainties: pHE Beam


Uncertainty on ND Spectrum Uncertainty on F/N Extrapolation

Foc

usin

g P

eak

Foc

usin

g P

eak


Hadron Production Uncertainties

• Spread of models is not the same as uncertainty– Correlated models

(eg. Kaon content).– Some models

known to have ‘flaws’ at certain kinematic regions

• Assumed 8-15% error on near spectrum coming from hadron production

See NuMI-B-768

Figure from F. Yumiceva


Hadron Production & F/N

• To estimate the error on F/N coming from hadron production used older cocktail of MARS, BMPT and Malensek

• We implement some ‘smoothing’ to the above curves:– Clearly statistical fluctuations in the MC. – Remove GFLUKA effect at high energies

Nominal LE Beam Semi-ME Beam Semi-HE Beam

See NuMI-B-768

Estimated Uncertainty

Estimated Uncertainty Estimated Uncertainty


Final Uncertainties:

ND Spectrum

• All effects including Hadron Production!

LE Beam LE(-10) Beam

pME BeampHE Beam

Box height indicates magnitude of systematic uncertainty


Final Uncertainties:

F/N Ratio

• All effects including Hadron Production!

LE Beam LE(-10) Beam

Box height indicates magnitude of systematic uncertainty

pME BeampHE Beam


NuMI-BooNE Flux

+ decays

K+ decays

LE-10

• Flux at MiniBooNE from pi/K decay-in-flight

• Working on the error band on flux at MiniBooNE coming from NuMI beam

• Expect small errors on flux

Figure from Alexis Aguilar Arevalo


LE & LE-10 Flux

• Biggest contribution to the error comes from:– Uncertainty of neutrino interaction location – Hadron production (10% for pi and 15% for K?)

LE Beam LE-10 Beam


pME & pHE Flux

pME Beam pHE Beam


Summary

• Uncertainties from a variety of effects have been investigated for LE, LE(-10), pME, and pHE beams.

• Vectors of uncertainties now are available for the G3NUMI beam MC:

http://www.hep.utexas.edu/numi/beamMC

• Flux at MiniBooNE is not sensitive to any misalignment or the uncertainties considered in this study gives opportunity to measure hadron production

Flux Error Analysis

Documents