Probabilistic Particle Flow Algorithm for High Occupancy Environment

TL;DR

This paper introduces a probabilistic particle flow algorithm that enhances energy resolution and particle hypothesis estimation in high occupancy collider environments, especially under conditions of high luminosity and coarse detector segmentation.

Contribution

The paper presents a novel probabilistic technique for resolving overlapping calorimeter energy deposits, improving energy resolution and providing uncertainty estimates for particle reconstruction.

Findings

Improved jet energy resolution in high occupancy environments.

Reliable performance in tau lepton decay jet reconstruction.

Enhanced sensitivity for physics analyses through uncertainty estimation.

Abstract

Algorithms based on the particle flow approach are becoming increasingly utilized in collider experiments due to their superior jet energy and missing energy resolution compared to the traditional calorimeter-based measurements. Such methods have been shown to work well in environments with low occupancy of particles per unit of calorimeter granularity. However, at higher instantaneous luminosity or in detectors with coarse calorimeter segmentation, the overlaps of calorimeter energy deposits from charged and neutral particles significantly complicate particle energy reconstruction, reducing the overall energy resolution of the method. We present a technique designed to resolve overlapping energy depositions of spatially close particles using a statistically consistent probabilistic procedure. The technique is nearly free of ad-hoc corrections, improves energy resolution, and provides new important handles that can improve the sensitivity of physics analyses: the uncertainty of the jet energy on an event-by-event basis and the estimate of the probability of a given particle hypothesis for a given detector response. When applied to the reconstruction of hadronic jets produced in the decays of tau leptons using the CDF-II detector at Fermilab, the method has demonstrated reliable and robust performance.