Neutrino oscillation

Back to mainpage

Laboratory experiments searching for neutrino oscillation

electron neutrino disappearance

Electron antineutrino disappearance experiments / Reactor neutrino experiments

These experiments are done close to powerful nuclear reactors. Each fission produces on average six electron antineutrinos, with a mean energy of a few MeV. A conventional uranium reactor produces a neutrino flux of 1.9 1020 /s /GW. Typically neutrinos are detected in liquid scintillators.

To measure the disappearance the neutrino flux must be known very well. In practice, to get reliable results one uses two or more detectors at different distances, one very close to the reactor. However, this kind of experiments are not sensitive to small mixing angles, despite a relatively large amount of events.

Experiment Location Baseline Observation Status Years
Gösgen Switzerland 37.9 m, 45.9 m, 64.7 m no osc finished 81-85
Bugey France 15 m, 40 m, 95 m no osc finished 1981-1994
Krasnoyarsk Russia 57 m, 57.6 m, 231.4 m no osc   19??
Chooz Ardennes, France 1 km O/E=0.98 ± 0.4 ± 0.4 finished 1997-1998
Palo Verde Arizona, U.S.A. 750 m O/E= 1.01 ± 0.024 ± 0.053 data taking ended 1998-2000 (July)
KamLAND Japan 180 km O/E= 0.611 ± 0.085 ± 0.041 taking data 2001-
San Onofre U.S.A. about km   rejected  

The results of Kamland after 766 ton-year exposure are consistent with the solar neutrino LMA-solution with delta m2 ≈ 8 10-5 eV2 and tan2t ≈ 0.40.

Kamland has also reported evidence of spectral distortion, which is a clear evidence for oscillation effects.

To see the reactor antineutrino background flux throughout the world calculated with the parameter values above click here.

muon neutrino to electron neutrino

Electron neutrino appearance experiments

Muon neutrinos are copiously produced in spallation reactions. Typically a proton with energy more than O(200 MeV) is shot to a fixed target, made of water or something else. The collision produces a large amount of pions, with all charges. For higher energies other mesons (e.g. K) are also produced.

A charged pion decays to a muon and a muon neutrino, with a lifetime 2.6 10-8 s. Positive pions typically decay at rest, since they are stopped in the matter. Negative pions, however, are quickly absorbed in the matter. Therefore, only decay in flight may be observable.

The muon subsequently decays to an electron, and two neutrinos, with a lifetime 2.2 10-6 s. The decay is practically isotropic.

The above decay scenario leads to muon neutrinos with energy spectrum of some tens of MeV, and electron neutrinos and muon antineutrinos with smaller mean energy. The appearance of high energy electron neutrinos is therefore a signal of neutrino oscillation.

Experiment Location Source Baseline Observation Status Years
BEBC Cern SPS   no osc finished - 1986?
CCFR Fermilab Tevatron 0.9 km to 1.4 km no osc taking data (?) 1990? -
E776 BNL AGS 1 km no osc finished 85-86
LSND Los Alamos LAMPF proton beam 30 m excess of electron antineutrinos:
40 ± 9
electron neutrinos: 18 ± 7
completed 1994-1998
Karmen Rutherford ISIS proton beam 18 m no osc data taking ended 1994-2001
Nomad Cern SPS 820 m no osc analyzing data 1995-1998
Minos Soudan mine, MS Main Injector at Fermilab 730 km   taking data 2003-
miniBooNE Fermilab Fermilab Booster 0.5 km/ 1 km   in progress 2003-
Icarus Gran Sasso Cern 732 km   under construction about 2005
Cosmos Fermilab Main Injector 1 km   rejected  

All experiments, except LSND, are consistent with no oscillation. The results of LSND can be interpreted as a signal of oscillation of muon neutrinos to electron neutrinos. Most of the parameter range explaining the LSND results are in disagreement with other experiments, particularly Karmen. However, there still seems to be a small area allowed by all experiments.

Muon neutrino disappearance

K2K experiment (KEK to Kamioka, 250 km) has reported results of neutrino oscillation. If there were no oscillation, they would expect 150.9 (+11.6, -10.0) muon events inside the fiducial volume of SuperKamiokande. However, they have only see 108 muon neutrino events. The result is consistent with the results from atmospheric neutrino measurements. K2K has also reported the first significant evidence for the energy dependence of the oscillation effect.

Experiment Location Source Baseline Observation Status Years
K2K Kamioka KEK beam 250 km   taking data 1999-2005

The K2K-experiment is running again after being stalled for about 2 years due to an accident at SuperK.
to tau neutrinos

Tau neutrino appearance experiments

These experiments use a high energy proton beam to produce muon neutrinos with high energy and high luminosity. As the neutrino beam does not contain any significant amount of tau neutrinos, an appearance of a charged tau lepton in a nearby detector is a signal of neutrino oscillation.

Taus cannot be detected in normal lepton detectors, because of their short lifetime. One uses two different techniques: Chorus has a very fine grained emulsion of AgBr, in which the charged particles leave a track. The tracks are identified by scanning the emulsion with a microscope, nowadays done automatically. Another way is to measure very accurately the tracks of other leptons, and conclude the existence of tau kinematically. This tehnique is used e.g. in Nomad.

Experiment Location Source Baseline Observation Status Years
E531 Fermilab accelerator 949 m no osc finished 86
CHORUS Cern SPS 820 m no osc scanning and analyzing data 1994-1997-
Nomad Cern SPS 820 m no osc analyzing data 1995-1998
OPERA Gran Sasso Cern 740 km   under construction 2005-
TOSCA Cern SPS     rejected -

No signals of oscillation have been seen, so far.


  • O = observed
  • E = expected
  • ev = events
  • no osc = no oscillation i.e. a signal consistent with no mixing
  • AGS = Alternating Gradient Synchrotron
  • SPS = Super Proton Synchrotron

Constraints for the mixing

Chooz: Chooz results
[Fogli et al 98]

Back to mainpage

Last modified 30.3.2005 (webmaster)