Joint Fit of Long-Baseline Accelerator Neutrino Experiments in GNA Software

Abstract

Long-baseline accelerator neutrino experiments are developing their neutrino data analysis in order to define the charge-parity phase ${{\delta }_{{{\text{CP}}}}}$ and the neutrino mass ordering which are important in describing neutrino oscillations and other phenomena of elementary particle physics. Since neutrinos interact very weakly it takes a long time to collect significant event statistics. Currently working experiments, NOvA and T2K, are not able to reach $5\sigma$ significance level separately to find out the values of unknown oscillation parameters. Future long-baseline accelerator neutrino experiment DUNE will also measure these oscillation parameters. These experiments have the similar composition for neutrino observation. A unified framework in the GNA software is developed to predict single and joint sensitivities of experiments of this type. It is expected that single NOvA and T2K sensitivities to the neutrino mass ordering and the charge-parity phase are much smaller than the expected DUNE sensitivities. From the analysis of joint sensitivities the hypothesis of the charge-parity conservation (${{\delta }_{{{\text{CP}}}}} = 0,\; \pm \pi$) is disfavored at $> 5\sigma$ significance level for 60% of possible ${{\delta }_{{{\text{CP}}}}}$ values for event rates accumulated by $36 \times {{10}^{{20}}}$ POT exposure in each regime in NOvA, $200 \times {{10}^{{20}}}$ POT exposure in T2K and after 7 yr of DUNE running time. DUNE sensitivities to the neutrino mass ordering after 2 yr are comparable to joint NOvA and T2K sensitivities.