LHCb - HLTs & Alignment and Calibration at LHCb

The data volume generated from the LHCb detector reaches 5TB/s in Run 3 conditions. In order to record data to permanent storage, this volume needs to be reduced by a factor 400. To achieve this, the trigger system makes a full reconstruction of events and a selection of specific signals of interest in an approach called real-time analysis. After the first stage of the trigger (HLT1), the second stage (HLT2) performs an offline-quality reconstruction and selection of physics signatures. There are four main components in this reconstruction phase: charged particle pattern recognition, Kalman fit, calorimeter reconstruction and particle identification. After the reconstruction is performed, the selection is made using different selection algorithms tuned for a specific signal topology or analysis. This process is enabled thanks to the quasi-real-time alignment and calibration, which provides the most accurate parameters for reconstruction and selections to provide the best possible resolution of physics parameters and maximize the selection efficiency.