Hyper-Kamiokande (also called Hyper-K or HK) is a neutrino observatory and experiment under construction in Hida, Gifu and in Tokai, Ibaraki in Japan. It is conducted by the University of Tokyo and the High Energy Accelerator Research Organization (KEK), in collaboration with institutes from over 20 countries across six continents. As a successor of the Super-Kamiokande (also Super-K or SK) and T2K experiments, it is designed to search for proton decay and detect neutrinos from natural sources such as the Earth, the atmosphere, the Sun and the cosmos, as well as to study neutrino oscillations of the man-made accelerator neutrino beam. The beginning of data-taking is planned for 2027.

The Hyper-Kamiokande experiment facility will be located in two places:

• The neutrino beam will be produced in the accelerator complex J-PARC (36°26′42″N 140°36′22″E / 36.445°N 140.606°E / 36.445; 140.606) and studied by the set of near and intermediate detectors located in Tokai village, Ibaraki prefecture, on the east coast of Japan.
• The main detector, also called Hyper-Kamiokande (HK), is being constructed under the peak of Nijuugo Mountain in Hida city, Gifu Prefecture, in the Japanese Alps (36°21′20.105″N 137°18′49.137″E / 36.35558472°N 137.31364917°E / 36.35558472; 137.31364917). The HK detector will be used for proton decay searches, studies of neutrinos from natural sources and will serve as a far detector for the measurement of the oscillations of an accelerator neutrino beam at the distance corresponding to the first oscillation maximum.

Neutrino oscillations are a quantum mechanical phenomenon in which neutrinos change their flavour (neutrino flavours states:
ν
_e,
ν
_μ,
ν
_τ) while moving, caused by the fact that the neutrino flavour states are a mixture of the neutrino mass states (ν₁, ν₂, ν₃ mass states with masses m₁, m₂, m₃, respectively). The oscillation probabilities depend on the six theoretical parameters:

• three mixing angles (θ₁₂, θ₂₃ and θ₁₃) governing the mixing between mass and flavour states,
• two mass squared differences (∆m₂₁ and ∆m₃₂, where ∆m_ij = m_i – m_j)
• one phase (δ_CP) responsible for the matter-antimatter asymmetry (CP symmetry violation) in neutrino oscillations,

and two parameters which are chosen for a particular experiment:

• neutrino energy
• baseline – the distance travelled by neutrinos at which oscillations are measured.

Continuing studies done by the T2K experiment, the HK far detector will measure the energy spectra of electron and muon neutrinos in the beam (produced at J-PARC as an almost pure muon neutrino beam) and compare it with the expectation in case of no oscillations, which is initially calculated based on neutrino flux and interaction models and improved by measurements performed by the near and intermediate detectors. For the HK/T2K neutrino beam peak energy (600 MeV) and the J-PARC – HK/SK detector distance (295 km), this corresponds to the first oscillation maximum, for oscillations driven by ∆m₃₂. The J-PARC neutrino beam will run in both neutrino- and antineutrino-enhanced modes separately, meaning that neutrino measurements in each beam mode will provide information about muon (anti)neutrino survival probability P
_{ν
μ →
ν
μ}, P
_{ν
μ →
ν
μ}, and electron (anti)neutrino appearance probability P
_{ν
μ →
ν
e}, P
_{ν
μ →
ν
e} , where P_να → P_νβ is the probability that a neutrino originally of flavour α will be observed later as having flavour β.

Comparison of the appearance probabilities for neutrinos and antineutrinos (P
_{ν
μ →
ν
e} versus P
_{ν
μ →
ν
e}) allows measurement of the δ_CP phase. δ_CP ranges from −π to +π (from −180° to +180°), and 0 and ±π correspond to CP symmetry conservation. After 10 years of data taking, HK is expected to confirm at the 5σ confidence level or better if CP symmetry is violated in the neutrino oscillations for 57% of possible δ_CP values. CP violation is one of the conditions necessary to produce the excess of matter over antimatter at the early universe, which forms now our matter-built universe. Accelerator neutrinos will be used also to enhance the precision of the other oscillation parameters, |∆m₃₂|, θ₂₃ and θ₁₃, as well as for neutrino interaction studies.

In order to determine the neutrino mass ordering (whether the ν₃ mass eigenstate is lighter or heavier than both ν₁ and ν₂), or equivalently the unknown sign of the ∆m₃₂ parameter, neutrino oscillations must be observed in matter. With HK beam neutrinos (295 km, 600 MeV), the matter effect is small. In addition to beam neutrinos, the HK experiment studies atmospheric neutrinos, created by cosmic rays colliding with the Earth's atmosphere, producing neutrinos and other byproducts. These neutrinos are produced at all points on the globe, meaning that HK has access to neutrinos that have travelled through a wide range of distances through matter (from a few hundred metres to the Earth's diameter). These samples of neutrinos can be used to determine the neutrino mass ordering.

Ultimately, a combined beam neutrino and atmospheric neutrino analysis will provide the most sensitivity to the oscillation parameters δ_CP, |∆m₃₂|, sgn ∆m₃₂, θ₂₃ and θ₁₃.

Core-collapse supernova explosions produce great quantities of neutrinos. For a supernova in the Andromeda galaxy, 10 to 16 neutrino events are expected in the HK far detector. For a galactic supernova at a distance of 10 kpc about 50,000 to 94,000 neutrino interactions are expected during a few tens of seconds. For Betelgeuse at the distance 0.2 kpc, this rate could reach up to 10 interactions per second and such a high event rate was taken into account in the detector electronics and data acquisition (DAQ) system design, meaning that no data would be lost. Time profiles of the number of events registered in HK and their mean energy would enable testing models of the explosion. Neutrino directional information in the HK far detector can provide an early warning for the electromagnetic supernova observation, and can be used in other multi-messenger observations.

Neutrinos cumulatively produced by supernova explosions throughout the history of the universe are called supernova relic neutrinos (SRN) or diffuse supernova neutrino background (DSNB) and they carry information about star formation history. Because of a low flux (few tens/cm/sec.), they have not yet been discovered. With ten years of data taking, HK is expected to detect about 40 SRN events in the energy range 16–30 MeV.

For the solar
ν
_e's, the HK experiment goals are:

• Search for a day-night asymmetry in the neutrino flux – resulting from different distances travelled in matter (during the night neutrinos additionally cross the Earth before entering the detector) and thus the different oscillation probabilities caused by the matter effect.
• Measurement of the
  ν
  _e survival probability for neutrino energies between 2 and 7 MeV – i.e. between regions dominated by oscillations in vacuum and oscillations in matter, respectively – which is sensitive to new physics models, like sterile neutrinos or non-standard interactions.
• The first observation of neutrinos from the hep channel: ${\displaystyle {^{3}}{\text{He}}+{\text{p}}\to {^{4}}{\text{He}}+{\text{e}}^{+}+\operatorname {\nu } _{\text{e}}}$ predicted by the standard solar model.
• Comparison of the neutrino flux with the solar activity (e.g. the 11-year solar cycle).

Geoneutrinos are produced in decays of radionuclides inside the Earth. Hyper-Kamiokande geoneutrino studies will help assess the Earth's core chemical composition, which is connected with the generation of the geomagnetic field.

The decay of a free proton into lighter subatomic particles has never been observed, but it is predicted by some grand unified theories (GUT) and results from baryon number (B) violation. B violation is one of the conditions needed to explain the predominance of matter over antimatter in the universe. The main channels studied by HK are
p
→
e
+
π
which is favoured by many GUT models and
p
→
ν
+
K
predicted by theories including supersymmetry.

After ten years of data taking, (in case no decay will be observed) HK is expected to increase the lower limit of the proton mean lifetime from 1.6 · 10 to 6.3 · 10 years for its most sensitive decay channel (
p
→
e
+
π
) and from 0.7 · 10 to 2.0 · 10 years for the
p
→
ν
+
K
channel.

Dark matter is a hypothetical, non-luminous form of matter proposed to explain numerous astronomical observations suggesting the existence of additional invisible mass in galaxies. If the dark matter particles interact weakly, they may produce neutrinos through annihilation or decay. Those neutrinos could be visible in the HK detector as an excess of neutrinos from the direction of large gravitational potentials such as the galactic centre, the Sun or the Earth, over an isotropic atmospheric neutrino background.

The Hyper-Kamiokande experiment consists of an accelerator neutrino beamline, a set of near detectors, the intermediate detector and the far detector (also called Hyper-Kamiokande). The far detector by itself will be used for proton decay searches and studies of neutrinos from natural sources. All the above elements will serve for the accelerator neutrino oscillation studies. Before launching the HK experiment, the T2K experiment will finish data taking and HK will take over its neutrino beamline and set of near detectors, while the intermediate and the far detectors have to be constructed anew.