Systematic Uncertainties in the Analysis of the Reactor Neutrino Anomaly

Abstract

We examine uncertainties in the analysis of the reactor neutrino anomaly, wherein it is suggested that only about 94% of the emitted antineutrino flux was detected in short baseline experiments. We find that the form of the corrections that lead to the anomaly are very uncertain for the 30% of the flux that arises from forbidden decays. This uncertainty was estimated in four ways, is as large as the size of the anomaly, and is unlikely to be reduced without accurate direct measurements of the antineutrino flux. Given the present lack of detailed knowledge of the structure of the forbidden transitions, it is not possible to convert the measured aggregate fission beta spectra to antineutrino spectra to the accuracy needed to infer an anomaly. Neutrino physics conclusions based on the original anomaly need to be revisited, as do oscillation analyses that assumed that the antineutrino flux is known to better than approximately 4%.

Figures (3)

• Figure 1: The fit to the electron spectrum for $^{235}$U (left) for two different assumptions on how to treat forbidden transitions, and the ratio of the corresponding antineutrino spectra to that of Schreckenbach (right). The electron data are those of schreck2. The electron spectra are fit assuming (a) all allowed GT branches, or (b) up to 30% forbidden GT transitions. In both cases the WM and FS corrections are included. When folded over the neutrino detection cross section, the case for all allowed (25% forbidden) transitions results in a 2.2% (0.06%) increase in the number of detectable antineutrinos.
• Figure 2: The ratio of the function $k(E_e,E_\nu)$ for $^{235}$U using ENDFB/VII.1 relative to using $E_\nu=K_{\beta}$borovoi.
• Figure 3: Different treatments of the forbidden GT transitions contributing to the antineutrino spectrum summed over all actinides in the fission burn in mid-cycle kopeikin of a typical reactor. The left panel shows the ratio of these antineutrino spectra relative to that using the assumptions of Ref. schreck2. The right panel shows the spectra weighted by the detection cross section, where the additional curve in black uses the assumptions of Ref. schreck2. The spectra are strongly distorted by the forbidden operators, being lower below the peak and in some cases more than 20% larger above the peak than Ref. schreck2. The corresponding change in the number of detectable antineutrinos relative to schreck2 is -0.75%, 5.8% and 1.85% for the $0^-, 1^-$, and $2^-$ forbidden operators, respectively.