Performance of the ATLAS missing ET trigger with first sqrt(s) = 7 TeV data

We present an early study of the performance of the ATLAS trigger based on the vector (MET) and scalar (SumET) sums of the measured transverse energies, obtained with data collected during the first 2010 LHC runs, with proton-proton collisions at $sqrt{s}=7$ TeV. We demonstrate that because of the Level one (L1) configuration, calorimeter noise effects on top of minimum bias events don't significantly affect MET trigger rates. We also show that there are good correlations between Event Filter (EF) and reconstructed offline MET and SumET quantities, and that the trigger efficiency with respect to these offline quantities follow Monte Carlo expectations. From these studies, we conclude that MET and SumET HLT rejection can safely be enabled as soon as rates become too high for L1 rejection alone.