Performance Study of a New Cluster Splitting Algorithm for the Reconstruction of PANDA EMC Data

For high-energy π0 mesons, the angle between the two final-state photons decreases with the increase in the energy of the π0, which enhances the probability of overlapping electromagnetic showers. The performance of the cluster splitting algorithm in the EMC reconstruction is crucial for the mass resolution measurement of π0 with high energy. The cluster splitting algorithm is based on the theoretical lateral distribution of the electromagnetic showers. A simple implementation of the lateral distribution can be described as a (multi-)exponential function. In a realistic electromagnetic calorimeter, considering the granularity of the detector, the measured energy in a cell is actually the integral of the theoretical energy deposition, which deviates from the exponential function. Based on the simulation of the barrel EMC in the P¯ANDA experiment, a cluster splitting algorithm with a new lateral energy development function is developed. The energy resolution of overlapping showers with high energy has been improved.