Sensitivity of Hyper-Kamiokande to sub-eV Sterile Neutrinos

TL;DR

This work evaluates Hyper-Kamiokande's sensitivity to light sterile neutrinos in a (3+1) framework, focusing on Δm^2_{41} up to 1 eV^2 and using both accelerator and atmospheric data to constrain active-sterile mixing angles θ_{14} and θ_{24}. The analysis solves neutrino propagation with matter effects, parameterizes the 4×4 mixing matrix U, and employs χ^2 fits with detailed systematics (flux, shape, and cross sections) across beam and atmospheric channels, using tools such as GLoBES and PREM-based Earth modeling. Accelerator data provide Δm^2_{41}-independent constraints in some regimes via ν_{e} survival and ν_{μ}→ν_{e} appearance, while atmospheric data exploit MSW resonances in ν_{e} and ν_{μ} channels to probe sub-eV splittings, particularly around Δm^2_{41} ~ 10^{-4}–10^{-3} eV^2. The combined beam plus atmospheric analysis yields stronger bounds than existing results in the targeted region and demonstrates Hyper-K's potential to explore sub-eV sterile neutrinos with competitive reach to future dedicated experiments. These results advance the experimental landscape for sterile neutrino searches by leveraging Hyper-K's large volume, improved performance, and broad energy–baseline coverage.

Abstract

In this work, we investigate the sensitivity of Hyper-Kamiokande (Hyper-K) to light sterile neutrinos within the $(3+1)$ framework, consisting of three active and one sterile neutrino state. We focus on the regime where the new mass-squared splitting satisfies $Δm_{41}^{2} \lesssim 1$ eV$^{2}$, a parameter space complementary to short-baseline sterile-neutrino searches. Using both accelerator and atmospheric neutrino samples, we evaluate the expected capability of Hyper-K to constrain active-sterile mixing. Our results show that Hyper-K can significantly improve current bounds on sterile-neutrino parameters and achieve sensitivity that is competitive with that of future dedicated experiments.

Figures (7)

• Figure 1: Top panels: The electron neutrino appearance (left) and the muon neutrino (right) survival probabilities as function of neutrino energy, setting $L$ = 295 km. Bottom panels: differences between the probabilities in $(3+1)$ and $3\nu$ scenarios, as function of neutrino energy. The black curves represents the standard $3\nu$ scheme, while the colored curves correspond to the $(3+1)$ scenario. The chosen values for the standard parameters are shown in Table \ref{['tab:std_osc']}. The sterile parameters not displayed in the figure are set to zero. Normal ordering among active neutrinos was assumed. We note that all the curves were obtained by numerical computations without using approximated probability formulas.
• Figure 2: Electron (top panels) and muon (bottom panels) neutrino survival probabilities, $P_{ee}$ and $P_{\mu\mu}$, as functions of neutrino energy $E_{\nu}$ for up-going trajectories ($\cos \theta_{z} = -1$). Left panels correspond to neutrinos, and right panels to antineutrinos. The dashed black line shows the standard $3\nu$ scheme, while solid colored curves correspond to the (3+1) sterile neutrino scenario. In the top part of each panel we show survival probabilities for two representative sterile mass-squared splittings, $\Delta m^2_{41} = 10^{-4}\,\mathrm{eV}^2$ and $10^{-3}\,\mathrm{eV}^2$, compared with the standard case, while the bottom part shows the differences between the sterile and standard probabilities, $P_{\alpha\beta}^{3+1}-P_{\alpha\beta}^{3\nu}$. Normal ordering among active neutrinos was assumed. We note that as done for Figure \ref{['fig:probBeam']}, all curves were obtained by numerical computations without using approximated probability formulas found in the previous subsection.
• Figure 3: Number of expected events in the appearance channel assuming ten years of collected data by Hyper-K experiment. The left (right) panel displays the event's energy distribution for neutrino (antineutrino) mode. We compare the $3\nu$, depicted by black solid line, with the $(3+1)$ scenario assuming $\Delta m_{41}^2 = 10^{-3}~{\rm eV}^2$ and $\sin^2\theta_{14}=0.3$ (red dashed line) and $\sin^2\theta_{14}=0.1$ (blue dot-dashed line). For standard oscillation parameters not labeled in the figure, we have assumed the values reported in Table \ref{['tab:std_osc']}. The mixing parameters involving the 4th state not displayed in the figure are set to zero.
• Figure 4: The same as Figure \ref{['fig:events_nue']}, but for disappearance channel, which can probe the $(\sin^2\theta_{24},\Delta m_{41}^2)$ parameters in $(3+1)$ scenario. $(\sin^2\theta_{24},\Delta m_{41}^2)=(0.03,10^{-3}~{\rm eV}^2)$ is depicted by red dashed line, and for $(\sin^2\theta_{24},\Delta m_{41}^2)=(0.01,10^{-2}~{\rm eV}^2)$ we use the blue dot-dashed line.
• Figure 5: Atmospheric neutrino event distributions as function of the zenith angle of lepton $\theta_l$. Left panel: Distributions of SubGeVlow($e$), SubGeVhigh($e$) and MultiGeV($e$) events, in the $3\nu$ and $(3+1)$ scenarios, whit active-sterile mixing values $(\sin^2\theta_{14},\Delta m_{41}^2)=(0.018,10^{-3}~{\rm eV}^2)$. Right panel: Distributions of SubGeVhigh($\mu$), MultiGeV(PC $\mu$) and MultiGeV(FC $\mu$) events, in the $3\nu$ and $(3+1)$ scenarios, for $(\sin^2\theta_{24},\Delta m_{41}^2)=(0.038,10^{-3}~{\rm eV}^2)$. In both panels, the color-shaded bands show the $\pm1\sigma$ statistical uncertainty of $3\nu$ scenario. In the $(3+1)$ scenario, the event distributions are shown after marginalization over nuisance parameters. For both panels we assume 6 years of data collection. All the mixing parameters of active neutrinos are fixed to the values reported in Table \ref{['tab:std_osc']}.