Upper bound on neutrino mass based on T2K neutrino timing measurements

The Tokai to Kamioka (T2K) long-baseline neutrino experiment consists of a muon neutrino beam, produced at the J-PARC accelerator, a near detector complex and a large 295-km-distant far detector. The present work utilizes the T2K event timing measurements at the near and far detectors to study neutrino time of flight as a function of derived neutrino energy. Under the assumption of a relativistic relation between energy and time of flight, constraints on the neutrino rest mass can be derived. The sub-GeV neutrino beam in conjunction with timing precision of order tens of ns provide sensitivity to neutrino mass in the few $\mathrm{MeV}/c^2$ range. We study the distribution of relative arrival times of muon and electron neutrino candidate events at the T2K far detector as a function of neutrino energy. The 90% C.L. upper limit on the mixture of neutrino mass eigenstates represented in the data sample is found to be $m_{\nu }^2<5.6{\mathrm{MeV}}^2/c^4$.

Figure 1

Time distribution of the hits selected in SMRD (red: ${\mathrm{t}}_{\mathrm{SMRD}}$) and CT1 (black: ${\mathrm{t}}_{\mathrm{CT}1}$) integrated over run period III. The eight-bunch structure of the beam is clearly visible.

Figure 2

Zoomed in version of the fourth bunch of the hits selected in SMRD (red: ${\mathrm{t}}_{\mathrm{SMRD}}$) and CT1 (blue: ${\mathrm{t}}_{\mathrm{CT}1}$) for run period III.

Figure 3

Data for ${\mathrm{t}}_{\mathrm{CT}1}$ as a function of calendar time for run period III and bunch 4.

Figure 4

Data for ${\mathrm{t}}_{\mathrm{SMRD}}$ as a function of calendar time for run period III and bunch 4.

Figure 5

Data for $\mathrm{\Delta }T=t_{\mathrm{SMRD}}-t_{\mathrm{CT}1}$ as a function of calendar time for run period III and bunch 4.

Figure 6

Distribution of $t_{\mathrm{SMRD}}$ for run period III and bunch 4 with a superimposed Gaussian fit.

Figure 7

Timing distribution (top) and timing residuals (bottom) of fully contained (FC) events for run periods I–IV.

Figure 8

SK residual timing distributions for run period III for FC non-CCQE sample.

Figure 9

Energy scale stability of the SK detector [1]. The hatched grey regions correspond to run periods I through IV. The dotted horizontal lines correspond to the $\pm 1\%$ stability range.

Figure 10

Relative timing stability of the two official SK GPS systems at SK (top panel) and NU1 (middle panel). Black and red lines are daily averages and 95% C.L. ranges, respectively. The bottom panel shows the time residuals of the SK FC events.

Figure 11

Monte Carlo energy conversion matrix for muon neutrino candidate events. The vertical axis shows reconstructed neutrino energy and the horizontal axis shows true neutrino energy.

Figure 12

Central value of distributions of best-fit values of ${\mathrm{m}}_{\nu }^2$ versus ${\mathrm{m}}_{\nu -\text{true}}^2$ for 300 toy MC data sets for which event arrival times had been shifted based on the assumption of the shown ${\mathrm{m}}_{\nu -\text{true}}^2$ values. The error bars represent the RMS of the corresponding distributions.

Figure 13

Log-likelihood curve averaged over 1000 sampled true neutrino energies from the energy conversion matrix. The minimum value is found to be ${\mathrm{m}}_{\nu }^2=2.4{\mathrm{MeV}}^2/c^4$. The blue squares and red dots indicate the F-C 90% C.L. upper critical values without and with systematic uncertainties, respectively. The corresponding 90% C.L. upper limits on ${\mathrm{m}}_{\nu }^2=4.4{\mathrm{MeV}}^2/c^4$ and ${\mathrm{m}}_{\nu }^2=5.6{\mathrm{MeV}}^2/c^4$ are derived.

Figure 14

Timing residuals of T2K CCQE neutrino candidate events as a function of derived neutrino energy. Events have been grouped into energy bins of 100 MeV below 1 GeV and bin sizes of 1 GeV above 1 GeV.