Tau Reconstruction and Identification Performance in ATLAS

Using LHC collisions at a centre-of-mass energy of sqrt(s)=7 TeV recorded with the ATLAS detector, the performance of the reconstruction and identification algorithms for hadronic tau decays is studied. Although the dataset used here, corresponding to an integrated luminosity of 244 nb-1, contains a small number of real tau leptons, the background jets reconstructed as tau candidates can be used to assess performance aspects of these algorithms. Distributions of identification variables are compared in data and Monte Carlo samples, and the background efficiency of cut-based tau identification criteria is measured in a QCD dijet enriched data sample. Systematic effects on the background efficiency are also estimated. The rejection power of multi-variate tau identification discriminants such as boosted decision trees and projective likelihood methods is also investigated.