The TileCal Energy Reconstruction for LHC Run2 and Future Perspectives

TL;DR

This paper presents an improved energy reconstruction algorithm for the TileCal detector in ATLAS during LHC Run 2, addressing pile-up effects with a refined optimal filtering method and discussing future deconvolution techniques.

Contribution

The work introduces a modified optimal filter algorithm that subtracts the pedestal estimated from calibration data, enhancing energy estimation accuracy during high luminosity conditions.

Findings

Improved energy estimation accuracy under high pile-up conditions.

Comparison shows the new method reduces bias and uncertainties.

Future deconvolution method under validation for offline reconstruction.

Abstract

The TileCal is the main hadronic calorimeter of ATLAS and it covers the central part of the detector ($|\eta|$ < 1.6). The energy deposited by the particles in TileCal is read out by approximately 10,000 channels. The signal provided by the readout electronics for each channel is digitized at 40 MHz and its amplitude is estimated by an optimal filtering algorithm. The increase of LHC luminosity leads to signal pile-up that deforms the signal of interest and compromises the amplitude estimation performance. This work presents the proposed algorithm for energy estimation during LHC Run 2. The method is based on the same approach used during LHC Run 1, namely the Optimal Filter. The only difference is that the signal baseline (pedestal) will be subtracted from the received digitized samples, while in Run 1 this quantity was estimated on an event-by-event basis. The pedestal value is estimated through special calibration runs and it is stored in a data base for online and offline usage. Additionally, the background covariance matrix will also be used for the computation of the Optimal Filter weights for high occupancy channels. The use of such information reduces the bias and uncertainties introduced by signal pile-up. The performance of the Optimal Filter version used in Run 1 and Run 2 is compared using Monte Carlo data. The efficiency achieved by the methods is shown in terms of error estimation, when different conditions of luminosity and occupancy are considered. Concerning future work, a new method based on linear signal deconvolution has been recently proposed and it is under validation. It could be used for Run 2 offline energy reconstruction and future upgrades.