Bayesian evidence for non-zero theta_13 and CP-violation in neutrino oscillations

TL;DR

This paper applies Bayesian model selection to neutrino oscillation data, providing strong evidence for a non-zero theta_13 angle with current experiments, but no conclusive evidence for CP-violation.

Contribution

It introduces a Bayesian method for evaluating evidence of theta_13 and CP-violation, demonstrating its effectiveness with current neutrino data and future applicability.

Findings

Weak evidence for non-zero theta_13 without recent data

Overwhelming evidence for theta_13 > 0 with reactor experiments

No significant evidence for CP-violation or CP-conservation

Abstract

We present the Bayesian method for evaluating the evidence for a non-zero value of the leptonic mixing angle theta_13 and CP-violation in neutrino oscillation experiments. This is an application of the well-established method of Bayesian model selection, of which we give a concise and pedagogical overview. When comparing the hypothesis theta_13 = 0 with hypotheses where theta_13 > 0 using global data but excluding the recent reactor measurements, we obtain only a weak preference for a non-zero theta_13, even though the significance is over 3 sigma. We then add the reactor measurements one by one and show how the evidence for theta_13 > 0 quickly increases. When including the Double Chooz, Daya Bay, and RENO data, the evidence becomes overwhelming with a posterior probability of the hypothesis theta_13 = 0 below 10^-11 . Owing to the small amount of information on the CP-phase delta, very similar evidences are obtained for the CP-conserving and CP-violating hypotheses. Hence, there is, not unexpectedly, neither evidence for nor against leptonic CP-violation. However, when future experiments aiming to search for CP-violation have started taking data, this question will be of great importance and the method described here can be used as an important complement to standard analyses.