Search for a Light Sterile Neutrino at Daya Bay

A search for light sterile neutrino mixing was performed with the first 217 days of data from the Daya Bay Reactor Antineutrino Experiment. The experiment’s unique configuration of multiple baselines from six 2.9 ${\mathrm{GW}}_{\mathrm{th}}$ nuclear reactors to six antineutrino detectors deployed in two near (effective baselines 512 m and 561 m) and one far (1579 m) underground experimental halls makes it possible to test for oscillations to a fourth (sterile) neutrino in the ${10}^{-3}{\mathrm{eV}}^2<|\mathrm{\Delta }{m}_{41}^2|<0.3{\mathrm{eV}}^2$ range. The relative spectral distortion due to the disappearance of electron antineutrinos was found to be consistent with that of the three-flavor oscillation model. The derived limits on ${\sin }^{2}2{\theta }_{14}$ cover the ${10}^{-3}{\mathrm{eV}}^2\lesssim |\mathrm{\Delta }{m}_{41}^2|\lesssim 0.1{\mathrm{eV}}^2$ region, which was largely unexplored.

Figure 1

Prompt energy spectra observed at EH2 (top) and EH3 (bottom), divided by the prediction from the EH1 spectrum with the three-neutrino best-fit oscillation parameters from the previous Daya Bay analysis [45]. The gray band represents the uncertainty of the three-neutrino oscillation prediction, which includes the statistical uncertainty of the EH1 data and all the systematic uncertainties. Predictions with ${\sin }^{2}2{\theta }_{14}=0.1$ and two representative $|\mathrm{\Delta }{m}_{41}^2|$ values are also shown as the dotted and dashed curves.

Figure 2

Comparison of the 95% exclusion limit sensitivities based on the confidence levels ${\mathrm{C}\mathrm{L}}_{\mathrm{s}}$ method for various combinations of EH’s data (see text for details). The sensitivities were estimated from an Asimov Monte Carlo data set that was generated without statistical or systematic variations. All the Daya Bay sensitivity curves were calculated assuming 5% rate uncertainty in the reactor flux except the dot-dashed one, which corresponds to a comparison of spectra only. Normal mass hierarchy was assumed for both $\mathrm{\Delta }{m}_{31}^2$ and $\mathrm{\Delta }{m}_{41}^2$. The dip structure at $|\mathrm{\Delta }{m}_{41}^2|\approx 2.4\times {10}^{-3}{\mathrm{eV}}^2$ was caused by the degeneracy between ${\sin }^{2}2{\theta }_{14}$ and ${\sin }^{2}2{\theta }_{13}$. The green dashed line represents Bugey’s [32] 90% confidence level (C.L.) limit on ${\overline{\nu }}_e$ disappearance and the magenta double-dot-single-dashed line represents the combined KARMEN and LSND 95% C.L. limit on ${\nu }_e$ disappearance from ${\nu }_e$-carbon cross section measurements [33].

Figure 3

Exclusion contours for the neutrino oscillation parameters ${\sin }^{2}2{\theta }_{14}$ and $|\mathrm{\Delta }{m}_{41}^2|$. Normal mass hierarchy is assumed for both $\mathrm{\Delta }{m}_{31}^2$ and $\mathrm{\Delta }{m}_{41}^2$. The red long-dashed curve represents the 95% C.L. exclusion contour with Feldman-Cousins method [54]. The black solid curve represents the 95% ${\mathrm{C}\mathrm{L}}_{\mathrm{s}}$ exclusion contour [55]. The parameter space to the right side of the contours is excluded. For comparison, Bugey’s [32] 90% C.L. limit on ${\overline{\nu }}_e$ disappearance is also shown as the green dashed curve.