Performance of shower deconstruction in ATLAS

This note describes the performance of the shower deconstruction algorithm, a jet tagging algorithm, using proton-proton collisions at a centre of mass energy of $\sqrt{s}= 8$ $\mathrm{\ Te\kern -0.1em V}$. Data recorded with the ATLAS detector corresponding to an integrated luminosity of $14.2$ fb$^{-1}$ are considered. Using small-radius jets to probe the substructure of a large-radius jet, shower deconstruction assigns a probability based on an approximate parton shower model, that the jet originated from a massive particle. In this note, the shower deconstruction algorithm is investigated to identify jets produced from boosted, hadronically decaying top quarks. The performance is evaluated using events enriched in top-quark pairs in the lepton plus jets final state and a sample of dijet events. The distribution of the shower deconstruction observable, the likelihood ratio $\chi_{\mathrm{SD}}$, is compared between data and simulation and the performance of shower deconstruction is compared to that of other tagging algorithms.