Thorsten Lux

For more than a decade, IFAE has been contributing to several key experiments in the field of neutrino physics. Currently, it participates in the long-baseline neutrino oscillation experiments and joined in 2021 the next generation experiment Hyper-Kamiokande which will start data taking in 2027.

Activities in 2021

In 2021, the IFAE Neutrino group focussed its activities on the work for the upgrade project of the near detector, ND280, of the T2K experiment in which the principal investigator holds the important manager position of Co-Project-Leader. The ND280 Upgrade is part of the T2K-II phase which in addition to the ND280 Upgrade includes the increase of the beam power from 500 kW to 1.1 MW in 2026. For the upgrade, three new subdetectors will be installed in the existing ND280 detector:

  1. a novel scintillator tracker (SuperFGD) based on 2 million cubes (each of 1x1x1 cm3) which are read out via photo sensors coupled to wavelength shifting fibers traversing each cube along the 3 axis. This concept allows a truly 3D event reconstruction.
  2. two new high angle time projection chambers (HA-TPC) which will be placed below and above the SuperFGD. Important parts of these HA-TPCs are produced by NEXUS Projects SL (Martorell).
  3. 6 time-of-flight (TOF) panels, each consisting of 20 scintillator bars read out on both sides. The TOF system will provide time resolutions of about 150 ps sufficient to efficiently suppress background from neutrino interactions in the iron yoke of the ND280 magnet, previously used by the UA1 experiment at CERN.
Figure 1: Installation tests of one of the TOF panels.
A significant fraction of the assembly and commissioning work of the new subdetectors is carried out at CERN. For this several group members had long stays at CERN in 2021 participating in the preparation of this work, testbeams and commissioning of the final detectors.

Graph neural network for 3D classification

Due to the novelty of the SuperFGD which has very technology specific properties, the 3D event reconstruction based on three orthogonal 2D views requires the development of new event reconstruction algorithms. The three-dimensional reconstruction of particle tracks produced in neutrino interactions can be subject to ambiguities due to high multiplicity signatures in the detector or leakage of signal between neighbouring active detector volumes. Members groups had key contributions to this classification paper using Machine Learning techniques which was published in Phys.Rev.D 103 (2021) 3, 032005.
Figure 2: (left) The three 2D views from the SuperFGD. (right) the raw reconstructed 3D event

Figure 3: classification of the reconstructed cubes. One can clearly see that a short track out ghosts cubes would have been taken for real without classification