Dual baseline search for muon antineutrino disappearance at $0.1{\mathrm{eV}}^2<\Delta m^2<100{\mathrm{eV}}^2$

The MiniBooNE and SciBooNE collaborations report the results of a joint search for short baseline disappearance of ${\overline{\nu }}_{\mu }$ at Fermilab’s Booster Neutrino Beamline. The MiniBooNE Cherenkov detector and the SciBooNE tracking detector observe antineutrinos from the same beam, therefore the combined analysis of their data sets serves to partially constrain some of the flux and cross section uncertainties. Uncertainties in the ${\nu }_{\mu }$ background were constrained by neutrino flux and cross section measurements performed in both detectors. A likelihood ratio method was used to set a 90% confidence level upper limit on ${\overline{\nu }}_{\mu }$ disappearance that dramatically improves upon prior limits in the $\Delta m^2=0.1–100{\mathrm{eV}}^2$ region.

Figure 1

Schematic view of the BNB from the beryllium target and magnetic horn to the SciBooNE and MiniBooNE detectors.

Figure 2

The neutrino and antineutrino flux prediction as a function of true neutrino (antineutrino) energy, in antineutrino mode at the MiniBooNE and SciBooNE detectors. The ${\overline{\nu }}_{\mu }$ flux is represented by the solid line, the ${\nu }_{\mu }$ flux is represented by the dashed line, the ${\overline{\nu }}_e$ flux is represented by the dot-dashed line, and the ${\nu }_e$ flux is represented by the dotted line.

Figure 3

Schematic view of the MiniBooNE detector.

Figure 4

Schematic view of the SciBooNE detector.

Figure 5

Reconstructed antineutrino and neutrino energy ($E_{\nu }^{\mathrm{QE}}$) distributions for selected RS and WS MiniBooNE events on different target types (hydrogen or carbon) from MiniBooNE MC. Total events are represented by the solid line, events with interaction on carbon are represented by the dashed line, and events with interaction on hydrogen are represented by the dot-dashed line.

Figure 6

Reconstructed antineutrino and neutrino energy ($E_{\nu }^{\mathrm{QE}}$) distributions for selected RS and WS MiniBooNE events for different interaction types (CCQE, $\mathrm{CC}1\pi$, other) from MiniBooNE MC. Total events are represented by the solid line, CCQE interaction events are represented by the dashed line, $\mathrm{CC}1\pi$ interaction events are represented by the dot-dashed line, and all other interaction (CC multi-$\pi$ or NC) events are represented by the short-dashed line.

Figure 7

Reconstructed antineutrino and neutrino energy ($E_{\nu }^{\mathrm{QE}}$) distributions for selected RS and WS SciBooNE events on different target types (hydrogen or carbon) from SciBooNE MC. Total events are represented by the solid line, events with interaction on carbon are represented by the dashed line, and events with interaction on hydrogen are represented by the dot-dashed line.

Figure 8

Reconstructed antineutrino and neutrino energy ($E_{\nu }^{\mathrm{QE}}$) distributions for selected RS and WS SciBooNE events for different interaction types (CCQE, $\mathrm{CC}1\pi$, other) from SciBooNE MC. Total events are represented by the solid line, CCQE interaction events are represented by the dashed line, $\mathrm{CC}1\pi$ interaction events are represented by the dot-dashed line, and all other interaction (CC multi-$\pi$ or NC) events are represented by the short-dashed line.

Figure 9

Antineutrino and neutrino path lengths for MiniBooNE and SciBooNE events from point of production to interaction in detector, as predicted by the MC simulation.

Figure 10

Bin-wise square root of the total (statistical and systematic errors combined) fractional error matrix $\sqrt{{\widehat{M}}_{ij}}=\sqrt{M_{ij}}/\sqrt{N_i{N}_j}$, where $M_{ij}$ is the total error matrix and $N_i$ ($N_j$) is the MC prediction for reconstructed antineutrino energy bin $i$ ($j$). Bins 1 through 21 are MiniBooNE, bins 22 through 42 are SciBooNE.

Figure 11

Correlation coefficients of the total (statistical and systematic errors combined) error matrix (${\rho }_{ij}=M_{ij}/({\sigma }_{ii}{\sigma }_{jj})$). Bins 1 through 21 are MiniBooNE, bins 22 through 42 are SciBooNE. No bins are anticorrelated.

Figure 12

RS (solid line) and WS (dashed line) contributions to the default MC prediction for MiniBooNE and SciBooNE reconstructed antineutrino and neutrino energy ($E_{\nu }^{\mathrm{QE}}$) distributions. Error bars are the systematic uncertainties from the diagonals of the error matrix $({\sigma }_{ii}=\sqrt{M_{ii}})$ and do not account for correlations.

Figure 13

Reconstructed antineutrino energy ($E_{\nu }^{\mathrm{QE}}$) distribution for data events, compared to Monte Carlo predictions, for MiniBooNE and SciBooNE. Vertical error bars on data are statistical uncertainty. Shaded error band around simulation is the systematic uncertainty computed as the square roots of the diagonals of the total error matrix.

Figure 14

90% C.L. exclusion region (solid line) and best fit point for the MiniBooNE-only ${\overline{\nu }}_{\mu }$ disappearance analysis. Also shown is the 90% C.L. result from the 2009 MiniBooNE disappearance analysis [8] (dashed line) and the Chicago-Columbia-Fermilab-Rochester (CCFR) experiment [41] (dot-dashed line). The expected 90% C.L. sensitivity band from fake data studies is also shown (shaded region); $1\sigma$ (68%) of fake data tests, where the fake data had statistical and systematic fluctuations but no oscillation signal, had 90% C.L. limit curves in this shaded region.

Figure 15

90% C.L. exclusion region (solid line) and best fit point for the joint MiniBooNE, SciBooNE ${\overline{\nu }}_{\mu }$ disappearance analysis. Also shown is the 90% C.L. result from the 2009 MiniBooNE disappearance analysis [8] (dashed line) and the CCFR experiment [41] (dot-dashed line). The expected 90% C.L. sensitivity band from fake data studies is also shown (shaded region); $1\sigma$ (68%) of fake data tests, where the fake data had statistical and systematic fluctuations but no oscillation signal, had 90% C.L. limit curves in this shaded region.

Figure 16

The ratio, with error bars, between data and default MC as a function of reconstructed antineutrino energy ($E_{\nu }^{\mathrm{QE}}$). The ratio of best fit signal MC to default MC is also shown (dashed line). The best fit results from the joint analysis were used to generate the signal MC. The shaded regions are the $1\sigma$ band from fake data with statistical and systematic fluctuations but no oscillation signal.

Figure 17

The double ratio [Eq. (11)], with error bars, as a function of reconstructed antineutrino energy ($E_{\nu }^{\mathrm{QE}}$). Some of the MiniBooNE and SciBooNE uncertainties cancel in this double ratio. The double ratio where the MiniBooNE and SciBooNE signal MC based on the best fit results from the joint analysis are used in placed of data is also shown (dashed line). The shaded region is the $1\sigma$ band from fake data with statistical and systematic fluctuations but no oscillation signal.