Externally produced events  Atmospheric neutrinos  The atmospheric neutrino beam

Internally produced events

The events due to neutrino interactions inside the detector can be fully contained: the secondary particles are contained inside the fiducial volume of the detector and a good discrimination between events with muons and those with electrons is possible. In the case of Kamiokande, Superkamiokande, IMB and Soudan2, the Table  lists only the events satisfying the definition of quasi elastic. This means single-ring events in the case of the water Cerenkov detector. The shape of the ring is a signature for a muon or an electron. The main results from the inspection of Table  are: 1) The water Cerenkov experiments measure a significant deficit of muon neutrinos compared to electron neutrinos. Since the error in the prediction of R is small (5%), this result is statistically significant. Because the absolute flux of electron neutrinos is in rough agreement with the predictions (assuming the BGS flux calculation), the preferred hypothesis is that the deficit is due to $\nu_\mu \rightarrow \nu_\tau$ oscillations (and not to $\nu_\mu \rightarrow \nu_e$). The preferred values of the oscillation parameters are: $\delta m^{2}$ of the order of 0.01 eV² and sin $^2(\theta)$ near 1. Preliminary results (April 1997) of the Superkamiokande experiment are in agreement with the past measurement. 2) This effect is not seen in the fine grained calorimeters Frejus and NUSEX. Even if the errors are large they are in apparent contradiction to the water Cerenkov events. The Soudan2 detector gives intermediate results. The most remarkable feature of the Kamiokande partially contained (or multi GeV) data is the dependence of the flavour ratio on the measured zenith angle (related to the direction of the incident neutrinos). This angular dependence agrees with a model in which the path length of the downward $\nu_\mu$ is too short to oscillate and the upward $\nu_\mu$, travelling for a much longer distance, are attenuated by oscillations.  

Table: Internally produced events (contained and partially contained)

 

Experiment (contained events)	$R= \frac{(\mu/e)_{data}}{(\mu/e)_{MC}}$	Exposure (Kton year)
Kamiokande	0.61 $\pm 0.07_{stat} \pm 0.05_{syst}$	7.7
SuperKamiokande 1997	0.64$\pm$0.04$\pm$0.06	12.8
IMB	0.55 $\pm$ 0.05$\pm$0.12	7.7
Frejus 1995	1.0$\pm$0.15$\pm$0.08	2.0
Soudan2 1996	0.75$\pm$0.16$\pm$0.10	1.5
Nusex	0.99$\pm$0.29$\pm$?	0.74
partially contained events
Kamiokande (MultiGeV)	0.59$\pm$0.08	8.2
Frejus 1995	0.99$\pm$0.13$\pm$0.08	2.0