Speaker
Description
The Deep Underground Neutrino Experiment (DUNE) is a next-generation long-baseline neutrino experiment utilizing large liquid argon time-projection chambers (LArTPCs). It is highly sensitive to the electron neutrino burst from a galactic core-collapse supernova. This neutrino burst can be used to point back to the supernova. The primary directional information comes from elastic scattering on electrons; however the dominant interaction is the charged-current absorption of $\nu_e$ on $^{40}\text{Ar}$.
The cross section for this interaction has two distinct nuclear transitions: Fermi and Gamow-Teller, each with different angular distributions for the outgoing electron. Because the resulting de-excitation gamma cascades differ between the two, distinguishing them is an opportunity for achieving precise supernova pointing as a complement to elastic scattering events. We describe machine-learning-based improvements to DUNE’s supernova pointing algorithms, including exploration of the use of subtle topological differences within a highly sparse, low-energy environment.
