Three-dimensional fitting reconstruction for high-granularity calorimeters

Purpose:This paper presents a 3D fit method for reconstructing electromagnetic showers in high-granularity 3D calorimeters, aiming to improve the energy and angular reconstruction performance, to maximize the scientific potential of the detector.
Methods:Electromagnetic shower distribution is parameterized to calculate the expected energy deposition in the calorimeter for perpendicularly incident electromagnetic particles. A likelihood fit is then performed to extract precisely the energy and incidence direction of the particle.
Results:The proposed methodology achieves an energy resolution of approximately 1.4% and an angular resolution of ~0.45° for perpendicularly incident electrons at 200 GeV, compared to 1.5% energy resolution and over 1.0° angular resolution for the Convolutional Neural Network.
Conclusion:The results validate the method’s ability to accurately reconstruct the energy and angular information of perpendicularly incident electrons and significantly improve the angular resolution, which is crucial for high-energy gamma measurements.