The precision of future neutrino experiments, such as the Deep Underground Neutrino Experiment (DUNE) and Hyper-Kamiokande (Hyper-K), critically depends on estimating neutrino energy with an uncertainty of a few MeV. A central challenge in achieving this precision is modeling the re-interactions of hadrons produced in neutrino scattering with atomic nuclei, known as final-state interactions (FSI). These FSIs modify the energy and momentum of detected particles, complicating the reconstruction of the initial neutrino energy.

A recent study employed state-of-the-art neutrino interaction event generators to evaluate the impact of FSI modeling on the kinematic and calorimetric energy estimators used by Hyper-K and DUNE, respectively. Both semiclassical intranuclear cascades (INC), which dominate current simulations, and a microscopic treatment based on relativistic mean-field calculations were considered. The results indicate that plausible variations in the FSI model introduce uncertainties in neutrino energy estimation that equal or exceed the precision required for the projected neutrino oscillation sensitivities in both experiments. This underscores the need for careful FSI modeling to obtain robust constraints from near detectors.

The study also reveals that DUNE and Hyper-K are sensitive to different aspects of FSI models. Hyper-K's energy estimation is more affected by pion absorption and nuclear effects that go beyond the semiclassical paradigm. On the other hand, DUNE's energy estimation is more sensitive to how hadronic energy is shared between visible and invisible energy sources in the detector. These findings have significant implications for neutrino oscillation analyses and highlight the need for key experimental and theoretical developments to control uncertainties associated with FSI modeling.