Full text
Selection of fully contained 𝜈μCC events in the ICARUS T600 detector at Fermilab Enrico Vedovato, on behalf of the ICARUS collaboration INFN - Università di Padova, email: [email protected] Neutrino Telescopes - September 29, 2025 to October 3, 2025 ICARUS T600 ·600 m from the beam source ·476 tons of LAr as active volume ·Two identical (3.6 x 3.9 x 19.6 m³) modules, and each module houses two TPCs separated by a common cathode ·Three wire planes that read ionization charge in each TPC Selection based on a series of cuts on variables associated to the event by Pandora: - SPILL - ·No CRT signal in 1.6μs beam spill window - SLICE - ·vertex in Fiducial Volume ·charge and light barycenters within 1m ·containment of all tracks (>5cm from edge) - PARTICLE - ·muon identification based on dE/dx vs residual range along the track ·μ is primary particle and longest in slice ·length μ >50cm Short-Baseline Neutrino Basic Selection Particle Interaction Hierarchy Selection Results Boosted Decision Tree (BDT) Selection Event Reconstruction BDT threshold to maximize eff * pur Selection procedure: ·Pre-selection to identify potential 𝜈μCC candidates (similar to basic selection but with looser cuts) ·Use BDT generated weights to discriminate between signal and background Signal Definition ICARUS work in progress ICARUS work in progress ICARUS work in progress ICARUS LAr-TPC 600m baseline 470t active volume SBND LAr-TPC 110 m baseline 112t active volume Booster Neutrino Beam (BNB) arxiv:1503.01520 60off axis from Main Injector (NuMI) ·The SBN Program includes two detectors exposed to the Booster Neutrino Beam: SBND and ICARUS, and they are all based on very similar Liquid Argon Time Projection Chamber (LArTPC) technology. Its main goal is to investigate the possible existence of 1 eV mass scale sterile neutrinos. ·The ICARUS T600 detector is the first large scale LArTPC ever realized and actively taking data. It allows to perform a three-dimensional and calorimetric reconstruction of particle trajectories in ultra-pure liquid Argon active volume. ICARUS data taking for physics within SBN started on June 9th 2022. DATA COLLECTION ·Decoding ·Deconvolution ·Region of Interest ·3D Reconstruction ·Pandora ·Interactions identification →“slices” ANALYSIS Track-score Shower Particle 1 Particle j Track Particle 2 ... (μ,π±,p) (e±,γ,π0) Slice 2 Slice 1 Slice n ... Spill ·360 PMTs for timing and trigger ·CRT systems to identify cosmic particles ·~3m concrete overburden Selection based on machine learning methods. BDT training on Monte Carlo (MC) samples to discriminate signal from background using many variables, to name a few: ·muon track-score ·dist. vertex-μstart ·muon length ·energy deposited in the last 5-10-15-20 cm ·vertex Y coord ·muon end Y coord ·muon dir. along Y&Z Blue: signal,Red: background ICARUS work in progress ICARUS work in progress ICARUS work in progress The basic selection finds more events with short muons but this comes at a significant purity cost. ·BDT bonus: provide a stronger reduction (almost a factor of 2) of background associated to cosmics and to neutrino events with the primary vertex outside the active volume. ·Visual studies on data samples confirm the results from the MC. · Basic BDT Neutrino + cosmics Efficiency = 75.5% Purity = 83.3% Efficiency = 76.6% Purity = 86.1% Data 5241 5050 Offbeam 151 81 Dirt events 95 65 ICARUS work in progress Muon length in the events selected by the two procedures considering both MC events and a reduced data statistics. ICARUS work in progress As progressively more conditions are applied to the selection, the > 50 cm requirement on the muon has the biggest impact on the efficiency. However, this cut is essential for better background rejection. The BDT preserves a nearly flat efficiency curve around the neutrino energy peak. What is next? ·Estimation of reconstructed neutrino energy ·Generalization by removing containment condition ICARUS work in progress ICARUS work in progress BDT Basic ICARUS work in progress The basic selection is more efficient than the BDT at selecting short muons. However... Goal: identification of 𝜈μCC fully contained interactions for 𝜈μdisappearance analysis ·𝜈μCC interaction with vertex in Fiducial Volume ·1 muon >50cm at the vertex ·fully contained
