The ATLAS transverse-momentum trigger performance at the LHC in 2011

The transverse momentum triggers of the ATLAS experiment at the CERN Large Hadron Collider are designed to select events with non-interacting particles passing through the detector and events with a large amount of outgoing momentum transverse to the beam axis. These triggers use global sums over the full calorimeter so are sensitive to measurement fluctuations and systematic changes anywhere in the detector. During the 2011 data-taking period, the LHC beam conditions for proton-proton collisions went through considerable evolution, starting with an average number of interactions per bunch crossing in a run, <mu>, of about 3, increasing to typical values of 7 to 15, and even including one run with <mu> of about 30. These changes were accompanied by changes in the bunch structure, including the number of filled bunches, how these were spaced, and the intensity of individual bunches. An increase in mu results in an increase of both the average energy deposit in the calorimeter and the energy-measurement fluctuations. Changes in beam conditions in turn necessitated changes in the calorimeter noise-suppression schemes used at various trigger levels. Transverse momentum distributions and trigger rates were impacted by all of these changes. This note contains a description of the transverse momentum triggers, the challenges faced in the 2011 data-taking period, the strategies used to deal with changes in the beam and detector, and characterization of the trigger performance in 2011. Even under these conditions, the trigger behavior was close to what was expected and allowed robust collection of data used for physics studies.