Refined Global Fit of KamLAND Data and the Daya Bay Antineutrino Energy Spectrum

Abstract

Recently, the JUNO experiment published its measurement of the solar neutrino oscillation parameters $Δm^2_{21}$ and $\sin^2θ_{12}$ based on 59 days of data, with central values differing by $0.2σ$ from those released by the KamLAND experiment in 2013. Meanwhile, short-baseline reactor neutrino oscillation experiments such as Daya Bay, RENO, and Double Chooz have observed significant deviations between the measured antineutrino spectrum and the Huber-Müller model around 5~MeV. To further investigate the impact of these deviations on the measurement of reactor neutrino oscillation parameters, a global analysis framework is constructed that is weakly dependent on the reactor antineutrino flux model. This framework is based on the independently measured $^{235}\mathrm{U}$ and $^{239}\mathrm{Pu}$ fission antineutrino spectra from the Daya Bay experiment, combined with the public data from KamLAND. First, using the Huber-Müller model, the KamLAND 2013 results are successfully reproduced to within $0.1σ$. Then, replacing the Huber-Müller model with the Daya Bay measured antineutrino spectra in a combined analysis, it is found that the best-fit value of the mass-squared difference $Δm^2_{21}$ decreases from $7.53^{+0.17}_{-0.16}\times10^{-5}\,\mathrm{eV^2}$ to $7.50^{+0.19}_{-0.18}\times10^{-5}\,\mathrm{eV^2}$, while the best-fit value of the mixing angle $\tan^2θ_{12}$ also shows a decreasing trend. This result is in better agreement with the latest JUNO measurement, suggesting that differences in the predicted reactor antineutrino spectra may be an important cause of the tension between the two experiments.