Physics Goals

The original physical goals of the experiment, as set out in the 2006 T2K proposal, were

  1. the discovery of νμ → νe (i.e., the confirmation that θ13 > 0);
  2. precision measurements of oscillation parameters in νμ disappearance;
  3. a search for sterile components in νμ disappearance by observation of neutral-current events (as neutral-current events are produced by all flavours of active neutrinos, a deficit would indicate an oscillation into sterile neutrinos).

As will be apparent from the page on “Neutrino Physics Today”, the first and third of these goals address the highest-priority questions in neutrino oscillations today: is the third angle non-zero, and do sterile neutrinos exist? The second goal is instrumental in achieving good measurements of θ13, and is also “automatic”, in the sense that an experiment designed to look for electron-neutrino appearance in θ13 mixing is necessarily in the right L/E regime to study θ23 mixing, so failing to design the experiment to do so would be a wasted opportunity.

Although designed as a stand-alone experiment for solar and atmospheric neutrinos. Super-Kamiokande has many advantages as a far detector for an oscillation experiment: it is very large and already well-understood, and it has excellent electron-muon separation, good energy resolution (2.5% at 1 GeV), and good control of backgrounds, which have already been extensively studied as part of its atmospheric neutrino oscillation analyses. As a water Cherenkov, it is most suited to measuring muons, electrons and photons in relatively low-multiplicity events; also, as a non-magnetic detector it cannot measure the charge of the produced lepton, and therefore cannot directly distinguish between neutrinos and antineutrinos.

The T2K near detector, ND280, was purpose-built for the experiment. Its main goals are:

  • measurement of the flux and spectrum of the νμ beam – necessary for accurate interpretation of disappearance data;
  • measurement of the νe content of the beam and its spectrum – necessary for appearance analyses;
  • accurate measurement of various neutrino interactions which either contribute to the backgrounds in the appearance and disappearance studies or are needed for the sterile-neutrino analysis.

The first two goals are standard for oscillation experiment near detectors and are fairly self-explanatory. An example of the last category is the measurement of neutral-current π0 production: ν + p/n → ν + p/n + π0. The π0 decays immediately into two photons, which will be detected in Super-K as electron-like (fuzzy) Cherenkov rings. The desirable consequence of this is that, if the rate of π0 production is well understood, this sample can be used to count neutral-current events, and therefore provide information on whether any muon-neutrinos are oscillating into sterile neutrinos instead of ντ or νe. The undesirable consequence is that if one photon is not detected, either because it has extremely low energy and does not produce a reconstructable ring, or because its ring overlaps so much with the other photon’s that the two are not separated, the event can be mistaken for νe appearance because the single electron-like ring looks like an electron. On both grounds, therefore, it is important to understand this reaction.

Although the March 2011 earthquake has caused an unscheduled delay in the experiment, the data collected before the earthquake have already yielded tentative indications of electron-neutrino appearance (see the KEK press release or, for technical details, the scientific paper). This is not, as yet, a definitive discovery: the probability that the apparent signal (six electron-like events on an estimated background of 1.5) represents a chance occurrence and not a real signal is only 0.7%, but physicists conduct hundreds of experiments every year, so the likelihood that one of those experiments sees something that has a 1% probability of occurring is in fact quite high. For this reason, physicists generally set the threshold for a definitive discovery at “5σ”, corresponding to a chance probability of about 1 in 1.7 million. Nevertheless, the fact that this result has been obtained with only a small fraction of the intended final data sample of the experiment is extremely encouraging, and suggests that the goal of measuring the elusive third mixing angle is finally within reach.