First MiniBooNE Neutrino Oscillation Results

Prepared by J.M. Conrad and W.C. Louis of the MiniBooNE collaboration for the DNP webpage

The MiniBooNE Collaboration reports first results of a search for νe appearance in a νμ beam. Using a 40-foot spherical tank filled with 800 tons of mineral oil in the Booster neutrino beam at Fermilab, no signicant excess of events above background is observed for reconstructed neutrino energies above 475 MeV. The data are consistent with no oscillations within a two-neutrino appearance-only oscillation model. However, an unexplained excess of events is observed for reconstructed neutrino energies below 475 MeV.

MiniBooNE [1] was motivated by the result from the Liquid Scintillator Neutrino Detector (LSND) experiment [2], which has presented evidence for ¯νμ → ¯νe oscillations at the Δm² ∼ 1 eV² scale. The MiniBooNE experiment uses the Fermilab Booster neutrino beam, which is produced from 8 GeV protons incident on a 71-cm-long by 1-cm-diameter beryllium target that is located near the upstream end of a magnetic focusing horn. A schematic of the experiment is shown in Fig. 1. The detector is located 541 m from the front of the beryllium target and consists of a spherical tank of radius 610 cm that is covered on the inside by 1520 8-inch photomultiplier tubes and filled with 800 tons of pure mineral oil (CH2).

experiment_532

Figure 1: Schematic of the MiniBooNE experiment, not to scale.


experiment_532

Figure 2: The top plot shows the number of candidate νe events as a function of EνQE . The points represent the data with statistical error, while the histogram is the expected background with systematic errors from all sources. The vertical dashed line indicates the threshold used in the two-neutrino oscillation analysis. Also shown are the best-fit oscillation spectrum (dashed histogram) and the background contributions from νμ and νe events. The bottom plot shows the number of events with the predicted background subtracted as a function of EνQE, where the points represent the data with total errors and the two histograms correspond to LSND solutions at high and low Δm².


limitdata_stacked

Figure 3: The top plot shows the MiniBooNE 90% CL limit (thick solid curve) and sensitivity (dashed curve) for events with 475 < EνQE < 3000 MeV within a two neutrino oscillation model. Also shown is the limit from the boosted decision tree analysis (thin solid curve) for events with 300 < EνQE < 3000 MeV. The bottom plot shows the limits from the KARMEN [4] and Bugey [5] experiments. The MiniBooNE and Bugey curves are 1-sided upper limits on sin² 2θ corresponding to Δχ² = 1.64, while the KARMEN curve is a "unified approach" 2D contour. The shaded areas show the 90% and 99% CL allowed regions from the LSND experiment.

The MiniBooNE experiment took data for three and a half years with a νμ beam (the experiment is now taking data with a ¯νμ beam) and collected over a million clean, neutrino events. Detailed Monte Carlo simulations of the beam and detector were used to make estimates of the ux and detector efficiencies, while the v3 NUANCE [3] event generator simulated neutrino interactions in mineral oil. About 99.5% of the MiniBooNE neutrino events are estimated to be νμ-induced, while 0.5% are estimated to be νμ-induced. After the complete νe event selection is applied, the total background is estimated to be 358 ± 35 events, while 163 ± 21 signal events are expected for the LSND central expectation of 0.26% νμ → νe transmutation.

The top plot of Fig. 2 shows candidate νe events as a function of reconstructed neutrino energy (EνQE). The vertical dashed line indicates the minimum (EνQE) used in the two-neutrino oscillation analysis. There is no signicant excess of events (22±19±35 events) for 475 < EνQE < 1250 MeV; however, an excess of events (96±17±20 events) is observed below 475 MeV. This low-energy excess cannot be explained by a two-neutrino oscillation model, and its source is under investigation. The dashed histogram in Fig. 2 shows the predicted spectrum when the best-fit two-neutrino oscillation signal is added to the predicted background. The bottom panel of the figure shows background-subtracted data with the best-fit two-neutrino oscillation and two oscillation points from the favored LSND region.

A single-sided raster scan to a two neutrino appearance-only oscillation model is used in the energy range 475 <EνQE < 3000 MeV to find the 90% CL limit corresponding to Δχ² = Δχ²limit - Δχ²best fit = 1.64. As shown by the top plot in Fig. 3, the LSND 90% CL allowed region is excluded at the 90% CL. A joint analysis as a function of Δm², using a combined χ² of the best fit values and errors for LSND and MiniBooNE, excludes at 98% CL two-neutrino appearance-only oscillations as an explanation of the LSND anomaly. The bottom plot of Fig. 3 shows limits from the KARMEN [4] and Bugey [5] experiments.

In summary, while there is a presently unexplained discrepancy with data lying above background at low energy, there is excellent agreement between data and prediction in the oscillation analysis region. If the oscillations of neutrinos and antineutrinos are the same, this result excludes two neutrino appearance-only oscillations as an explanation of the LSND anomaly at 98% CL. First antineutrino oscillation results should be available within the next year.

We acknowledge the support of Fermilab, the Department of Energy, and the National Science Foundation. We thank Los Alamos National Laboratory for LDRD funding.

References

[1] The MiniBooNE webpage is located at www-boone.fnal.gov.
[2] C. Athanassopoulos et al., Phys. Rev. Lett. 75, 2650 (1995); C. Athanassopoulos et al., Phys. Rev. Lett. 77, 3082 (1996);
C. Athanassopoulos et al., Phys. Rev. Lett. 81, 1774 (1998); A. Aguilar et al., Phys. Rev. D 64, 112007 (2001).
[3] D. Casper, Nucl. Phys. Proc. Suppl. 112, 161 (2002).
[4] B. Armbruster et al., Phys. Rev. D 65, 112001 (2002).
[5] Y. Declais et al., Nucl. Phys. B434, 503 (1995).