UC Berkeley

A measurement of the smallest neutrino mixing angle, $\theta_{13}$, by observations of reactor $\bar{\nu}_e$ disappearance at the Daya Bay Reactor Neutrino Experiment over 1958 days, is described. Eight identically-designed antineutrino detectors (ADs) monitored the $\bar{\nu}_e$ flux produced by six 2.9 GWth nuclear reactors, with four ADs located close to the reactors (${\sim}350-600\text{m}$) monitoring the unoscillated flux, and four ADs located far from the reactors (${\sim}1500-1950\text{m}$) observing the oscillated flux. Inverse beta decay events were identified based on the coincidence of the prompt positron annihilation and the delayed neutron capture on hydrogen within the organic liquid scintillator target of each AD. A selection based on both the time delay and distance between the prompt and delayed events allowed for a strong suppression of the largest background, uncorrelated (accidental) coincidences of decays of radiocontaminants in the ADs. Differences in detection efficiency between ADs were constrained to within 0.66\%,dominated by the AD-uncorrelated uncertainty of the coincidence distance and time criteria. Comparison of the near AD and far AD observations, with appropriate adjustments for detector livetimes, AD-reactor baselines, and backgrounds, revealed a disappearance signal with minimal dependence on reactor modeling and detector response. A $\chi^2$ expression with nuisance parameters was constructed to model the impact of $\theta_{13}$ and systematic uncertainties on the predicted far AD observations given the near AD observations as input. A fit was performed using the exact three-flavor probability for $\bar{\nu}_e$ disappearance assuming the normal mass ordering, and found $\sin^2(2\theta_{13}) = 0.0731^{+0.0087}_{-0.0089}$.