TileCal Trigger Tower studies considering additional segmentation on the ATLAS upgrade for high luminosity at LHC

The Tile Calorimeter (TileCal) is the hadronic calorimeter covering the most central region of the ATLAS experiment at LHC. The TileCal readout consists of about 10000 channels and provides a compact information, called trigger towers (around 2000 signals), to the ATLAS first level online event selection system. The ATLAS upgrade program is divided in three phases: Phase 0 occurs during 2013- 2014 and prepares the LHC to reach peak luminosities of 10^34 cm2s-1; Phase 1, foreseen for 2018-1019, prepares the LHC for peak luminosity up to 2-3 x 10^34 cm2s-1, corresponding to 55 to 80 interactions per bunch-crossing with 25 ns bunch interval; and Phase 2 is foreseen for 2022-2023, whereafter the peak luminosity will reach 5-7 x 1034 cm2s-1 (HL-LHC). The ATLAS experiment is operating very well since 2009 providing large amount of data for physics analysis. The online event selection system (trigger system) was designed to reject the huge amount of background noise generated at LHC and is one of the main systems responsible for the quality of the acquired data in ATLAS. However, the LHC upgrade for high luminosity will increase the amount of interactions per bunch crossing, increasing the event pileup probability, complicating the task of the trigger system. Therefore, in order to cope with the new luminosity requirements, the ATLAS and its trigger system will upgrade its components and algorithms. The online trigger system in its first level will profit from the detectors’ full granularity after Phase 2 upgrade, but for Phase 1 TileCal could profit from an additional granularity in comparison with the current system. Thus, this work presents studies about the use of new algorithms considering the TileCal Trigger Tower additional longitudinal segmentation and improved resolution on jet identification for the ATLAS online trigger system. The longitudinal information from TileCal is used in order to improve the jet energy estimation, using a linear (Least Square) optimization technique. The energy resolution of the trigger tower information is varied from 1 GeV up to 256 MeV in order to evaluate its impact on jet tagging and, additionally, a bidimensional window is applied over the RoI envisaging the reduction of the pile up effect. The achieved results, through Monte Carlo simulations, shown that the TileCal longitudinal information reduces slightly the energy estimation error for jets at level one trigger but shows negligible effect on jet identification. The resolution improvement presented a noticeable impact on jet identification but noise issues from the TileCal trigger towers have to be further investigated. Finally, the use of the bidimensional window proved to reduce the pileup effect on jet identification at ATLAS level one trigger system.