Improved Search for ${\nu }_{\mu }\to {\nu }_e$ Oscillation in a Long-Baseline Accelerator Experiment

We performed an improved search for ${\nu }_{\mu }\to {\nu }_e$ oscillation with the KEK to Kamioka (K2K) long-baseline neutrino oscillation experiment, using the full data sample of $9.2\times {10}^{19}$ protons on target. No evidence for a ${\nu }_e$ appearance signal was found, and we set bounds on the ${\nu }_{\mu }\to {\nu }_e$ oscillation parameters. At $\Delta m^2=2.8\times {10}^{-3}{\mathrm{eV}}^2$, the best-fit value of the K2K ${\nu }_{\mu }$ disappearance analysis, we set an upper limit of ${sin}^{2}2{\theta }_{\mu e}<0.13$ at a 90% confidence level.

Figure 1

Distribution of candidate events in the $M_{\mathrm{inv}}\mathrm{\text{−}}{E}_e$ space. The selection boundary is indicated by a line with a bend for (a) signal MC events and (b) ${\nu }_{\mu }$-originated background MC events. Remaining events before the ${\pi }^0$ rejection in data (circles) are also shown for K2K-I and K2K-II combined. Signal MC distributions are obtained under the assumption of ${sin}^{2}2{\theta }_{\mu e}=0.05$ and $\Delta m^2=2.8\times {10}^{-3}{\mathrm{eV}}^2$.

Figure 2

The upper bound on ${\nu }_{\mu }\to {\nu }_e$ oscillation parameters at 90% and 99% C.L. The sensitivities of the K2K experiment for each C.L. are also indicated with dashed lines.