Revisiting leptonic nonunitarity

TL;DR

This paper investigates the current bounds on leptonic mixing matrix nonunitarity caused by potential high-scale neutrino states, showing that existing experiments limit nonunitarity to about 0.1% but cannot fully exclude it.

Contribution

It clarifies the constraints on leptonic nonunitarity without relying on neutrino oscillation data and corrects misconceptions about the sensitivity of oscillation experiments.

Findings

Experimental constraints bound nonunitarity to ~10^{-3}

Oscillation experiments cannot determine the overall normalization of the LMM

Future experiments like DUNE have limited power to exclude nonunitarity

Abstract

In the presence of extra neutrino states at high scales, the low-energy effective $3\times 3$ leptonic mixing matrix (LMM) is in general nonunitary. We revisit the question of what is our current knowledge of individual LMM matrix elements without assuming unitarity. We first demonstrate that a minimal set of experimental constraints suffices in bounding LMM nonunitarity parameters to the level of ${\cal O}(10^{-3})$, without the use of neutrino oscillation data. We then revisit oscillation results as a complementary cross-check, using different physics and different experimental techniques to probe a similar parameter space. We correct some common misconceptions in the neutrino nonunitarity literature resulting from an incautious treatment of input parameters. We find that neutrino oscillation experiments can constrain LMM nonclosure, but, contrary to claims in the literature, are completely insensitive to the overall normalization of the LMM. Thus we conclude that oscillation experiments, including the future DUNE experiment, have no power in excluding nonunitarity altogether.