{
  "id": "super-kamiokande",
  "title": "Super-Kamiokande",
  "lang": "en",
  "source_lang": "en",
  "summary": "Super-Kamiokande is a Japanese underground water Cherenkov detector that established atmospheric neutrino oscillations and studies neutrinos, stellar collapse, and proton decay.",
  "category": "science",
  "subcategory": "physics",
  "type": "place",
  "aliases": [
    "Super-K",
    "SK",
    "Super-Kamioka Neutrino Detection Experiment"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/super-kamiokande/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/super-kamiokande/",
      "json": "https://aiwiki.page/zh/super-kamiokande.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Super-Kamiokande is a large underground neutrino observatory in Kamioka, Hida, Gifu Prefecture, Japan. Its principal instrument is a cylindrical tank containing approximately 50,000 metric tonnes of highly purified water, viewed by thousands of sensitive light detectors. It records the faint flashes produced when charged particles generated by neutrino interactions travel through water. The experiment also searches for hypothetical processes, especially proton decay, that would reveal physics beyond currently established theories.\n\nOperating since April 1996, Super-Kamiokande supplied decisive evidence in 1998 that atmospheric neutrinos change type during flight. This phenomenon, called neutrino oscillation, demonstrates that neutrinos have nonzero masses and requires an extension of the original Standard Model of particle physics. The observatory also measures neutrinos from the Sun, serves as the far detector for accelerator neutrino experiments, and searches for neutrinos from stellar explosions.\n\nSuper-Kamiokande combines a very large target mass with the ability to reconstruct particle directions, energies, and interaction patterns. Its scientific program spans laboratory particle physics, solar physics, and astrophysics. The addition of dissolved gadolinium beginning in 2020 introduced a new capability: more efficient identification of neutrons, especially those accompanying interactions of electron antineutrinos."
    },
    {
      "heading": "Origins, underground setting, and detector construction",
      "text": "Super-Kamiokande developed from Kamiokande, an earlier water Cherenkov experiment built in the Kamioka mine. Kamiokande originally emphasized searches for nucleon decay but became an important neutrino observatory after improvements to its low-energy sensitivity. It detected neutrinos from Supernova 1987A and measured solar neutrinos. These achievements demonstrated that a large underground water detector could function both as a particle-physics experiment and as an astronomical instrument.\n\nThe larger successor was constructed during the first half of the 1990s and began observations on 1 April 1996. Its principal Japanese institutional base is the University of Tokyo’s Institute for Cosmic Ray Research, with participation from an international collaboration. Masatoshi Koshiba was central to the development of the Kamioka experimental program, while Yoji Totsuka led the construction and early scientific operation of Super-Kamiokande. Takaaki Kajita played a leading role in the atmospheric-neutrino analysis that established oscillations.\n\nThe detector stands approximately 1,000 metres beneath the mountain surface. The rock overburden, corresponding to roughly 2,700 metres of water-equivalent shielding, greatly reduces the flux of penetrating cosmic-ray muons. These particles would otherwise overwhelm the much rarer neutrino signals. Underground operation does not eliminate cosmic-ray backgrounds entirely: several muons still cross the detector each second, and their secondary products require careful identification.\n\nThe stainless-steel tank is approximately 39 metres in diameter and 41 metres high. An optically separated inner detector contains about 32,000 tonnes of water, while an outer detector acts primarily as an active veto and containment monitor. Many analyses use a central fiducial mass of 22,500 tonnes, conventionally defined by excluding a two-metre layer adjacent to the inner detector walls. This restriction reduces backgrounds entering from detector materials and provides enough surrounding water to reconstruct interactions reliably.\n\nIn its original configuration, the inner detector carried 11,146 photomultiplier tubes with photocathodes approximately 50 centimetres in diameter, covering about 40 percent of its inner surface. The outer detector used 1,885 smaller tubes. These instruments detect individual or small numbers of photons and convert their arrival times and light intensities into electronic signals.\n\nThe large water volume is simultaneously the neutrino interaction target, the optical medium, and a shield against external radiation. Although neutrinos interact extremely rarely, the tank contains enough electrons and atomic nuclei for measurable event samples to accumulate. Its scale is therefore not merely an engineering feature: it directly determines the experiment’s sensitivity to rare interactions and long hypothetical particle lifetimes."
    },
    {
      "heading": "Cherenkov radiation and the visible signatures of particles",
      "text": "Super-Kamiokande observes particles through Cherenkov radiation. A charged particle emits this light when its speed exceeds the phase velocity of light in the surrounding medium. This does not violate special relativity: the particle remains slower than the vacuum speed of light, \\(c\\), while moving faster than light propagates through water.\n\nFor a particle of speed \\(v=\\beta c\\), Cherenkov emission occurs when \\(\\beta n>1\\), where \\(n\\) is the refractive index. The emission angle satisfies\n\n\\[\\cos\\theta_{\\mathrm C}=\\frac{1}{\\beta n}.\\]\n\nFor a highly relativistic particle in water, the angle is about \\(42^\\circ\\). The expanding cone of light produces a roughly ring-shaped pattern on the detector walls. The geometry is more complicated for particles that scatter, shower, stop, or decay, but the ring remains the fundamental visual signature.\n\nNeutrinos themselves emit no Cherenkov light because they carry no electric charge. They are observed indirectly through interactions mediated by the weak interaction. A muon neutrino can undergo a charged-current interaction that produces a muon, while an electron neutrino can produce an electron. Neutral-current interactions, in which the outgoing particle remains a neutrino, can also yield visible secondary particles.\n\nMuon and electron events have characteristically different appearances. A muon generally follows a relatively straight track and produces a sharp-edged ring. An energetic electron repeatedly emits bremsstrahlung photons and initiates an electromagnetic shower; the many secondary tracks and their scattering produce a more diffuse ring. Statistical classification of these patterns allows Super-Kamiokande to distinguish muon-like from electron-like events without directly observing the incident neutrino.\n\nSome interactions generate several rings. A neutral pion, for example, usually decays into two gamma rays, each of which can initiate an electromagnetic shower. Recognizing this topology is important because an unresolved two-photon event can imitate a single electron. Charged pions, secondary muons, and delayed electrons from muon decay provide additional information about the interaction.\n\nThe method also has intrinsic limitations. Slow particles below the Cherenkov threshold are not directly visible, including many recoil protons and nuclear fragments. Neutrons are observed through secondary interactions or capture products rather than through their own light. Super-Kamiokande consequently reconstructs the visible portion of an event, not every outgoing particle. The relationship between visible energy and incident neutrino energy depends on the interaction channel, nuclear effects, and whether particles escape or remain below threshold."
    },
    {
      "heading": "Event reconstruction, calibration, and background control",
      "text": "Each recorded event consists principally of photomultiplier hit times and charges. Reconstruction algorithms compare these measurements with expectations for candidate vertices, particle directions, ring patterns, and energies. After correcting for the travel time of light through water, the distribution of hit times helps locate the interaction vertex. The spatial distribution of collected light then constrains the directions and identities of the visible particles.\n\nEnergy estimation relies on the number and distribution of detected photoelectrons, corrected for optical attenuation, detector geometry, tube response, and other effects. For a low-energy electron, collected light provides a measure of its energy. For a long muon track, track length and light yield supply complementary information. At higher energies, simultaneous fitting of several rings helps distinguish complicated final states and improves estimates of particle momenta.\n\nCalibration must connect these electronic signals to physical quantities across a broad energy range. Super-Kamiokande uses light sources to characterize timing, gain, and water transparency. An electron linear accelerator provides controlled low-energy electrons for solar-neutrino calibration. Other checks use radioactive sources, stopping cosmic-ray muons, electrons from muon decay, and reconstructed particle masses. Agreement between independent calibration samples limits biases that could otherwise imitate a change in neutrino rate or spectrum.\n\nBackground rejection depends strongly on the physics analysis. Low-energy events face backgrounds from natural radioactivity, especially radon and radioactive impurities in surrounding materials. Muons crossing the water can fragment oxygen nuclei and create unstable isotopes whose later decays resemble neutrino interactions. These “spallation” backgrounds are reduced using the position, time, and energy-loss characteristics of preceding muons.\n\nThe outer detector provides a separate means of identifying particles entering or leaving the inner volume. A fully contained atmospheric-neutrino event has no substantial exiting track, whereas a partially contained event includes a particle that escapes. Upward-going muons are another important class: their parent neutrinos interact in the rock below or around the detector, and the resulting muons enter the tank. The Earth suppresses the corresponding upward-going cosmic-ray muon background.\n\nInterpretation requires detailed simulations of neutrino fluxes, interactions with water, secondary particle propagation, light production, and detector response. Particularly important uncertainties arise from interactions inside oxygen nuclei, where nucleon motion, correlations, pion absorption, and secondary scattering modify the visible final state. Analyses therefore combine multiple event categories and explicitly account for measurement uncertainties. A convincing oscillation result depends not simply on counting fewer events than expected, but on reproducing their energy, direction, and flavor-dependent distributions under a coherent physical model."
    },
    {
      "heading": "Atmospheric neutrinos and the discovery of oscillations",
      "text": "Atmospheric neutrinos originate when cosmic rays strike nuclei in the Earth’s atmosphere. The resulting showers contain pions and kaons, whose decays produce neutrinos and muons; the muons subsequently decay into electrons and additional neutrinos. In a simplified low-energy decay chain, approximately two muon-flavor neutrinos or antineutrinos accompany each electron-flavor neutrino or antineutrino. Precise predicted ratios depend on energy, direction, and the atmospheric shower model.\n\nBefore Super-Kamiokande, several experiments had observed unexpectedly few muon-like events relative to electron-like events. Super-Kamiokande’s large sample and directional information enabled a much stronger test. Atmospheric neutrinos arriving from directly overhead typically travel tens of kilometres, whereas those arriving from below may cross nearly the Earth’s diameter. The detector could therefore compare neutrinos with very different travel distances within the same apparatus.\n\nIn 1998 the collaboration reported a pronounced zenith-angle dependence in the muon-neutrino sample. Upward-going muon-like events were depleted relative to expectations without oscillations, while downward-going events and the electron-like sample behaved differently. Uncertainties in the overall atmospheric flux could not naturally explain this combination. The observations were consistent with muon neutrinos changing predominantly into tau neutrinos during propagation.\n\nA two-flavor approximation expresses the muon-neutrino survival probability as\n\n\\[P(\\nu_\\mu\\rightarrow\\nu_\\mu) \\simeq 1-\\sin^2(2\\theta_{23}) \\sin^2\\!\\left( 1.267\\, \\frac{\\Delta m^2 L}{E} \\right),\\]\n\nwhere \\(\\Delta m^2\\) is measured in \\(\\mathrm{eV}^2\\), \\(L\\) in kilometres, and \\(E\\) in GeV. The relevant squared-mass splitting is of order \\(2.5\\times10^{-3}\\,\\mathrm{eV}^2\\), and the mixing is large. The full description involves three neutrino flavors and additional parameters, but the approximation captures the principal atmospheric signal.\n\nA later analysis, published in 2004, identified a dip in the distribution of events as a function of reconstructed \\(L/E\\), near the first oscillation minimum. This provided a more specific signature of oscillatory behavior than an overall deficit alone. Super-Kamiokande also developed statistical evidence for tau-neutrino appearance, using the complex interaction patterns expected when sufficiently energetic tau neutrinos produce tau leptons.\n\nThe 2015 Nobel Prize in Physics was awarded jointly to Takaaki Kajita and Arthur B. McDonald for the discovery of neutrino oscillations, which shows that neutrinos have mass. Super-Kamiokande’s atmospheric result was complementary to the solar-neutrino measurements of the Sudbury Neutrino Observatory. Neither result measured the absolute neutrino masses; oscillations determine differences between squared masses and the mixing of flavor states."
    },
    {
      "heading": "Solar neutrinos and propagation through matter",
      "text": "The Sun produces electron neutrinos through nuclear fusion reactions in its interior. Super-Kamiokande principally observes the relatively high-energy neutrinos produced in the beta decay of boron-8. Although these constitute only a small fraction of the total solar-neutrino flux, their energies are sufficient to produce detectable recoil electrons in water. The experiment’s practical threshold varies with analysis and operating period and lies in the range of a few MeV.\n\nThe dominant detection reaction is elastic scattering,\n\n\\[\\nu+e^-\\rightarrow\\nu+e^-.\\]\n\nThe recoil electron tends to point approximately away from the Sun. This directional correlation is crucial: most radioactive backgrounds have no corresponding solar alignment. By fitting the angle between reconstructed electron directions and the instantaneous solar direction, the collaboration extracts a solar-neutrino signal statistically even when individual events cannot be identified with certainty.\n\nElastic scattering is sensitive to all three active neutrino flavors, but electron neutrinos interact more strongly than muon or tau neutrinos at these energies. Super-Kamiokande therefore measures a flavor-weighted flux rather than a completely flavor-independent count. Its observed rate was substantially below the rate predicted if all neutrinos retained their original electron flavor, confirming the solar-neutrino deficit already established by earlier experiments.\n\nThe interpretation became especially clear through comparison with the Sudbury Neutrino Observatory, which used heavy water to distinguish reactions sensitive mainly to electron neutrinos from reactions sensitive to all active flavors. Together, these measurements showed that the Sun’s total active-neutrino output was broadly consistent with solar models while its electron-neutrino component was reduced. The deficit resulted from flavor transformation rather than an inadequate nuclear energy source in the Sun.\n\nSolar oscillations are strongly influenced by the Mikheyev–Smirnov–Wolfenstein effect. Coherent forward scattering in matter changes the effective propagation of electron-flavor neutrinos relative to the other flavors. Because the solar density decreases from the core outward, a neutrino’s evolution can differ markedly from simple vacuum oscillation. At boron-8 energies, the resulting electron-neutrino survival probability is approximately one-third, with an energy dependence set by the oscillation parameters and matter profile.\n\nSuper-Kamiokande tests this framework through the recoil-electron energy spectrum, time dependence, and day–night comparison. Neutrinos observed at night have crossed part of the Earth, where additional matter effects can slightly modify their flavor composition. The detector also measures the seasonal variation caused by the changing Earth–Sun distance. Searches for the expected low-energy spectral transition and precise measurements of day–night asymmetry probe details of the standard oscillation picture rather than merely repeating the original flux-deficit measurement."
    },
    {
      "heading": "Accelerator neutrinos: K2K and T2K",
      "text": "Atmospheric neutrinos provide a wide range of energies and baselines, but their production cannot be controlled experimentally. Accelerator beams offer a complementary approach: the source location, timing, beam direction, and approximate energy spectrum are known. Super-Kamiokande has served as the far detector for two major long-baseline neutrino experiments, allowing controlled tests of the oscillation interpretation.\n\nThe KEK-to-Kamioka experiment, K2K, sent a predominantly muon-neutrino beam approximately 250 kilometres from the KEK laboratory in Tsukuba to Super-Kamiokande. Operating from 1999 to 2004, it used near-detector measurements to constrain the beam before oscillation effects became substantial. The far detector observed both a reduction in the event rate and a distortion of the reconstructed energy distribution consistent with the atmospheric-neutrino parameters.\n\nT2K, the Tokai-to-Kamioka experiment, uses a beam produced at the Japan Proton Accelerator Research Complex, J-PARC, approximately 295 kilometres from Super-Kamiokande. Its first beam events at the far detector were observed in 2010. The beam axis passes slightly away from the detector: this off-axis configuration produces a comparatively narrow neutrino energy distribution concentrated near 0.6 GeV, close to the first atmospheric-scale oscillation maximum at that baseline.\n\nA central T2K measurement is electron-neutrino appearance in an initially muon-neutrino beam. Super-Kamiokande identifies candidate interactions through electron-like Cherenkov rings and uses additional topology and decay information to reject backgrounds. Important backgrounds include the small electron-neutrino component already present in the beam and neutral-current production of neutral pions. T2K reported initial indications of appearance in 2011 and established the process more firmly with subsequent data.\n\nBy changing the magnetic focusing polarity of the beamline, T2K can enhance either neutrinos or antineutrinos. Comparing their appearance probabilities provides sensitivity to the phase associated with charge–parity violation in the neutrino mixing matrix. Such a difference, if established, would show that neutrino and antineutrino oscillations are not exact mirror processes. Its extraction requires simultaneous control of backgrounds, interaction models, detector efficiencies, and parameter degeneracies.\n\nSuper-Kamiokande is not magnetized and generally cannot determine the charge of an individual electron or muon track. Neutrino–antineutrino separation therefore relies mainly on the beam operating mode and statistical information, rather than direct curvature measurements. Near-detector data and external interaction measurements are essential because water-nucleus effects influence how the incident energy is inferred. Accelerator results and atmospheric results are complementary: the former offer controlled source conditions, while the latter sample much broader ranges of distance, energy, and passage through the Earth."
    },
    {
      "heading": "Stellar collapse, diffuse supernova neutrinos, and gadolinium",
      "text": "A collapsing massive star releases most of its gravitational binding energy in neutrinos rather than visible light. During the formation of a neutron star, an energy of order \\(10^{53}\\) ergs is emitted in neutrinos and antineutrinos of all flavors over roughly several to tens of seconds. Detecting such a burst provides direct information about the dense stellar core, which electromagnetic observations cannot reveal.\n\nFor a core-collapse supernova within the Milky Way, Super-Kamiokande could record thousands of interactions, depending on distance, explosion properties, flavor conversion, and event-selection conditions. The dominant channel is inverse beta decay,\n\n\\[\\bar{\\nu}_e+p\\rightarrow e^++n.\\]\n\nThe positron produces prompt Cherenkov light, while the neutron thermalizes and is captured later. Elastic scattering on electrons contributes fewer events but is more strongly directional, helping estimate the supernova’s position on the sky. Interactions with oxygen provide further, more model-dependent information.\n\nThe time structure of a burst can constrain the accretion phase, cooling of the newly formed compact object, and possible abrupt termination associated with black-hole formation. Energy spectra and flavor-sensitive channels test models of neutrino transport under extreme density and temperature. The detector participates in the SuperNova Early Warning System, which combines alerts from multiple observatories. A neutrino signal can precede visible brightening because neutrinos escape while the explosion shock is still propagating through the stellar envelope.\n\nBeyond individual nearby explosions, every past core-collapse supernova contributes to a faint accumulated background. This diffuse supernova neutrino background is especially difficult to detect because its expected event rate is small and overlaps with reactor, atmospheric, and detector backgrounds. Identifying the delayed neutron accompanying inverse beta decay is a major means of improving discrimination.\n\nIn pure water, neutrons usually capture on hydrogen and emit a 2.2 MeV gamma ray, which produces a weak detectable signal. Gadolinium has an exceptionally large thermal-neutron capture cross section. Its capture typically releases a cascade of gamma rays with a combined energy of about 8 MeV, producing a much more recognizable delayed signal. The prompt positron and delayed neutron-capture signal together form a coincidence signature.\n\nThe SK-Gd program began adding gadolinium sulfate in 2020. The initial elemental gadolinium concentration was approximately 0.01 percent by mass, giving a neutron-capture fraction on gadolinium of about one-half; a subsequent loading in 2022 increased the concentration to roughly 0.03 percent and the capture fraction to about three-quarters. Improved neutron tagging strengthens diffuse-background searches, aids supernova event classification, and provides additional information for atmospheric-neutrino and rare-event analyses."
    },
    {
      "heading": "Proton decay and other searches for new physics",
      "text": "Super-Kamiokande was designed not only to observe neutrinos but also to search for the instability of ordinary matter. Some grand unified theories place strong, weak, and electromagnetic interactions within a common framework and predict that the proton is not absolutely stable. Observation of proton decay would demonstrate baryon-number violation and reveal physics at energy scales far beyond the direct reach of existing accelerators.\n\nWater provides a large number of potential decaying protons. Its hydrogen nuclei contain free protons, while its oxygen nuclei contain both bound protons and neutrons. The standard 22,500-tonne fiducial volume contains approximately \\(7.5\\times10^{33}\\) protons. Monitoring such a target for many years makes it possible to test lifetimes vastly longer than the age of the Universe: the experiment does not wait for one proton to age, but watches an enormous population for rare decays.\n\nA benchmark channel is\n\n\\[p\\rightarrow e^++\\pi^0.\\]\n\nThe neutral pion rapidly decays into two gamma rays, so the event can produce three electron-like rings, or fewer if rings overlap or one photon is not separately reconstructed. For a free proton decaying at rest, the visible particles should have a total invariant mass near the proton mass and a total momentum near zero. Bound protons introduce broadening through nuclear motion and interactions of the pion inside oxygen.\n\nAnother important channel is\n\n\\[p\\rightarrow\\bar{\\nu}+K^+.\\]\n\nThe antineutrino is invisible, and the kaon is often below the Cherenkov threshold. Detection therefore relies on kaon-decay products, such as a muon and its subsequent decay electron, sometimes accompanied by a gamma ray from de-excitation of the residual nucleus. The contrasting signatures illustrate why each proposed decay channel needs a dedicated analysis rather than one universal proton-decay selection.\n\nAtmospheric-neutrino interactions constitute a principal background because they can accidentally imitate the visible products of nucleon decay. Reconstruction quality, total momentum, invariant mass, delayed particles, and neutron multiplicity help discriminate signal from background. Neutron tagging is particularly useful because atmospheric interactions often eject neutrons, although genuine decays of bound nucleons can also produce them through nuclear processes.\n\nNo proton decay has been established by Super-Kamiokande. For several leading channels, its published lower limits on partial lifetimes reach approximately \\(10^{34}\\) years or beyond. These are limits on lifetime divided by branching fraction, usually written \\(\\tau/B\\), and depend on the assumed decay mode. Additional searches address neutron decay, neutron–antineutron oscillations, unusual interactions, and possible neutrinos from dark matter annihilation. These analyses constrain particular models; an absence of signal does not by itself establish one unique alternative theory."
    },
    {
      "heading": "Operating phases, engineering changes, and experimental limits",
      "text": "Super-Kamiokande’s scientific record spans several detector configurations rather than one permanently unchanged instrument. The first phase, SK-I, ran from 1996 to 2001 with approximately 40 percent inner-detector photocathode coverage. A major interruption occurred on 12 November 2001, during refilling after maintenance, when the implosion of a photomultiplier initiated a pressure-wave chain reaction that destroyed more than half of the large inner-detector tubes.\n\nThe accident led to substantial engineering changes. Surviving and replacement tubes were fitted with protective assemblies designed to prevent an implosion from propagating through the detector. These combined transparent fronts with reinforced rear structures. Operation resumed in 2002 as SK-II with roughly half the original number of large tubes and approximately 19 percent photocathode coverage. The reduced light collection particularly affected low-energy measurements, but atmospheric-neutrino and other analyses continued.\n\nFull inner-detector coverage was restored for SK-III in 2006. SK-IV began in 2008 with upgraded electronics capable of recording photomultiplier signals more flexibly and supporting software-based triggering. This improved the handling of delayed activity, low-energy events, and closely spaced signals. Such changes matter physically because a delayed decay electron or neutron-capture signal can determine whether an event belongs to one interaction category or another.\n\nFurther refurbishment and water-system preparation preceded gadolinium loading. The essential engineering challenge was to preserve extremely high optical transparency while retaining the dissolved gadolinium and removing unwanted contaminants. Conventional purification methods can remove the desired solute along with impurities, so the SK-Gd system required selective treatment. Radiopurity was also important: adding a useful neutron-capture material would be counterproductive if it introduced excessive radioactive backgrounds.\n\nDespite these advances, the detector has fundamental limitations. Cherenkov thresholds hide slow particles, optical imaging does not provide complete particle tracking, and neutrino interactions on oxygen retain nuclear-model uncertainties. A finite target mass limits the statistics of rare processes. Distinguishing neutrino mass ordering, measuring the charge–parity phase, or detecting a diffuse supernova signal requires combining event samples and controlling systematic effects, not merely improving photomultiplier sensitivity.\n\nThe successor project, Hyper-Kamiokande, builds on the same water Cherenkov approach with a substantially larger target and newer photosensors. Super-Kamiokande supplies both a technical foundation and long-duration data sets that a new detector cannot reproduce immediately. Its accumulated atmospheric, solar, accelerator, and rare-event observations remain valuable because each probes different combinations of neutrino properties, backgrounds, and interaction mechanisms."
    }
  ],
  "links": [
    "neutrino",
    "japan",
    "water",
    "neutrino-oscillation",
    "standard-model",
    "particle-physics",
    "sun",
    "astrophysics",
    "supernova",
    "cosmic-ray",
    "photon",
    "cherenkov-radiation",
    "special-relativity",
    "speed-of-light",
    "refractive-index",
    "electric-charge",
    "weak-interaction",
    "muon",
    "electron",
    "calibration",
    "radioactivity",
    "measurement-uncertainty",
    "atmosphere-of-earth",
    "nobel-prize-in-physics",
    "nuclear-fusion",
    "neutron-star",
    "milky-way",
    "positron",
    "dark-matter"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}