No abstract

This final article about the CHOOZ experiment presents a complete description of the νe source and detector, the calibration methods and stability checks, the event reconstruction procedures and the Monte Carlo simulation. The data analysis, systematic effects and the methods used to reach our conclusions are fully discussed. Some new remarks are presented on the deduction of the confidence limits and on the correct treatment of systematic errors.

Initial results are presented from CHOOZ 1 , a long-baseline reactor-neutrino vacuum-oscillation experiment. The data reported here were taken during the period March to October 1997, when the two reactors ran at combined power levels varying from zero to values approaching their full rated power of 8.5 GW (thermal). Electron antineutrinos from the reactors were detected by a liquid scintillation calorimeter located at a distance of about 1 km. The detector was constructed in a tunnel protected from cosmic rays by a 300 MWE rock overburden. This massive shielding strongly reduced potentially troublesome backgrounds due to cosmic-ray muons, leading to a background rate of about one event per day, more than an order of magnitude smaller than the observed neutrino signal. From the statistical agreement between detected and expected neutrino event rates, we find (at 90% confidence level) no evidence for neutrino oscillations in the ν e disappearance mode for the parameter region given approximately by ∆m 2 > 0.9 10 −3 eV 2 for maximum mixing and sin 2 2θ > 0.18 for large ∆m 2 .1 The CHOOZ experiment is named after the new nuclear power station operated byÉlectricité de France (EdF) near the village of Chooz in the Ardennes region of France.