ANNIE Phase II Reconstruction Techniques

TL;DR

ANNIE Phase II introduces advanced reconstruction techniques using LAPPDs and machine learning to improve neutrino interaction measurements in a water Cherenkov detector at Fermilab.

Contribution

The paper presents novel reconstruction methods leveraging LAPPDs and machine learning for enhanced neutrino event analysis in the ANNIE experiment.

Findings

Improved vertex and direction resolution with LAPPDs

Effective neutrino and muon energy reconstruction using deep learning

Demonstrated the potential of fast-timing detectors in neutrino physics

Abstract

The Accelerator Neutrino Neutron Interaction Experiment (ANNIE) is a 26-ton Gd-doped water Cherenkov detector installed in the Booster Neutrino Beam at Fermilab. The experiment has two complementary goals: (1) perform the first measurement of the neutron yield from $\nu_{\mu}$ interactions as a function of Q$^2$ in order to constrain neutrino-nucleus interaction theoretical models, and (2) demonstrate the power of new fast-timing, position-sensitive detectors by making the first deployment of Large Area Picosecond PhotoDetectors (LAPPDs) in a physics experiment. In Phase I, ANNIE successfully performed neutron background measurements. To realise the Phase II measurements the ANNIE collaboration has developed several reconstruction techniques using the arrival time and position of the Cherenkov photons in the detector photomultipliers (PMTs) and LAPPDs. A maximum-likelihood fit is used to reconstruct the neutrino interaction vertex and direction. Machine and Deep Learning techniques are used for the muon and neutrino energy reconstruction. We present the results of ANNIE reconstruction techniques and the improvement we can get in resolution with the use of LAPPDs.