Experimental study of the atmospheric neutrino backgrounds for $p\to e^{+}{\pi }^0$ searches in water Cherenkov detectors

The atmospheric neutrino background for proton decay via $p\to e^{+}{\pi }^0$ in ring imaging water Cherenkov detectors is studied with an artificial accelerator neutrino beam for the first time. In total, $3.14\times {10}^5$ neutrino events corresponding to about 10 megaton-years of atmospheric neutrino interactions were collected by a 1000 ton water Cherenkov detector (KT). The KT charged-current single ${\pi }^0$ production data are well reproduced by simulation programs of neutrino and secondary hadronic interactions used in the Super-Kamiokande (SK) proton decay search. The obtained $p\to e^{+}{\pi }^0$ background rate by the KT data for SK from the atmospheric neutrinos whose energies are below 3 GeV is ${1.63}_{-0.33}^{+0.42}(\mathrm{stat}{)}_{-0.51}^{+0.45}(\mathrm{syst})(\mathrm{megaton}\mathrm{\text{−}}\mathrm{year}{)}^{-1}$. This result is also relevant to possible future, megaton-scale water Cherenkov detectors.

Figure 1

The top figure compares the number of total ${\nu }_{\mu }$ interactions in the KT 50 ton fiducial volume for $7.4\times {10}^{19}$ protons on target (red solid) and the number of total neutrino interactions from atmospheric neutrinos in SK [24, 25] for 1 Mtyr exposure (blue dashed for ${\nu }_e+{\overline{\nu }}_e$ and blue dotted for ${\nu }_{\mathrm{all}}+{\overline{\nu }}_{\mathrm{all}}$, where “all” stands for all the neutrino flavors). Disappearance of atmospheric ${\nu }_{\mu }$’s CC interactions due to neutrino oscillation is taken into account. The NEUT simulation program [31] is used for calculation of total neutrino interaction for each neutrino flavor. The bottom figure shows the ratio of neutrino energy spectra for the atmospheric neutrino interactions in SK and the K2K beam ${\nu }_{\mu }$ interactions in KT. The meanings of the dashed and dotted lines are the same as above.

Figure 2

Reconstructed—true vertex position of “cosmic ray pipe” muons entering at the center of KT. Red crosses (blue boxes) show the data (MC).

Figure 3

PID likelihood for cosmic ray muons (shaded histogram) and the decay electrons (open histogram). The red (blue) histogram [top (bottom) figure] shows data (MC). Positive (negative) likelihood corresponds to muon(electron)-like.

Figure 4

Reconstructed momentum divided by range for vertical (top figure) and horizontal (bottom) cosmic ray muons. Red crosses (blue boxes) show the data (MC).

Figure 5

Reconstructed mass of ${\pi }^0$ candidates produced by NC interactions in the K2K beam. Red crosses (blue boxes) show the data (MC).

Figure 6

Event display of “$p\to \mu {\pi }^0$” sample in the KT real data satisfying selection criteria (A)–(E). Each hit PMT is represented by a small circle whose area is proportional to the PMTs measured charge. The small blue triangles show the projection of the reconstructed vertex onto the walls. Cyan (red) thick circles show reconstructed rings identified by the PID as muon(electron)-like (two overlapped thick circles are identified as electronlike rings). The reconstructed $M_{{\pi }^0}$, $M_{\mathrm{tot}}$, and $P_{\mathrm{tot}}$ for this event are $164\mathrm{MeV}/c^2$, $851\mathrm{MeV}/c^2$, and $226\mathrm{MeV}/c$, respectively.

Figure 7

Event rate after application of each selection criterion. The event selections (A)–(E) are explained in the text. The event rates at D and E correspond to “$\mu {\pi }^0$” and “$p\to \mu {\pi }^0$” samples, respectively. The data, NEUT, and NUANCE are shown as red circles, blue squares, and blue triangles, respectively. The simulated data are normalized by the total number of neutrino interactions in the 25 ton fiducial volume.

Figure 8

Reconstructed hadronic invariant mass $W$ for three-ring “$\mu {\pi }^0$” events, assuming CC single ${\pi }^0$ production via resonance. Red crosses show the data with statistical and total measurement errors. The relatively large systematic errors in this plot are correlated between bins, and arise from the absolute energy scale uncertainty. The NEUT (NUANCE) prediction is shown by the solid (dashed) blue histogram. The hatched histogram shows the distribution of NEUT events that originate from resonant CC single-pion production. Both MC samples are normalized to the data by number of entries.

Figure 9

Reconstructed momentum transfer $Q^2$ for three-ring “$\mu {\pi }^0$” events, assuming CC single ${\pi }^0$ production via resonance. Red crosses show the data with statistical and total measurement errors. The NEUT (NUANCE) prediction is shown by the solid (dashed) blue histogram. The hatched histogram shows the distribution of NEUT events that originate from resonant CC single-pion production. Both MC samples are normalized to the data by number of entries.

Figure 10

Total momentum $P_{\mathrm{tot}}$ vs total invariant mass $M_{\mathrm{tot}}$ for “$\mu {\pi }^0$” events from the KT data ($7.4\times {10}^{19}$ pot), with the proton decay signal box superimposed.

Figure 11

Total momentum $P_{\mathrm{tot}}$ vs total invariant mass $M_{\mathrm{tot}}$ for “$\mu {\pi }^0$” events from NEUT ($1.1\times {10}^{20}$ pot), with the proton decay signal box superimposed.

Figure 12

Total momentum $P_{\mathrm{tot}}$ vs total invariant mass $M_{\mathrm{tot}}$ for “$\mu {\pi }^0$” events from NUANCE ($5.4\times {10}^{19}$ pot), with the proton decay signal box superimposed.

Figure 13

Total invariant mass $M_{\mathrm{tot}}$ for “$\mu {\pi }^0$” events with total momentum $P_{\mathrm{tot}}\le 250\mathrm{MeV}/c$. Red crosses show the data with statistical and total measurement errors. The NEUT (NUANCE) predictions are shown by a solid (dashed) blue histogram. Both MC samples are normalized to the data by number of entries. The green arrow shows the range accepted by the proton decay signal box.

Figure 14

Total momentum $P_{\mathrm{tot}}$ for “$\mu {\pi }^0$” events with total invariant mass $800\le M_{\mathrm{tot}}\le 1050\mathrm{MeV}/c^2$. Red crosses show the data with statistical and total measurement errors. The NEUT (NUANCE) predictions are shown by a solid (dashed) blue histogram. Both MC samples are normalized to the data by number of entries. The green arrow shows the range accepted by the proton decay signal box.

Figure 15

$L$ (as defined in the text) for “$\mu {\pi }^0$” events. Red crosses show the data with statistical and total measurement errors. The NEUT (NUANCE) predictions are shown by a solid (dashed) blue histogram. Both MC samples are normalized to the data by number of entries. The green arrow indicates the range accepted by the proton decay signal box.

Figure 16

Predicted number of CC [$m^{\mathrm{CC}}(i)$ shown in top] and NC [$m^{\mathrm{NC}}(i)$ shown in bottom] “$p\to \mu (e){\pi }^0$” events from NEUT in the KT for $7.4\times {10}^{19}$ pot, in bins of true parent neutrino energy $i$. The errors shown are MC statistical.