Strategy to Search for Single-Top Events using early Data of the ATLAS Detector at the LHC

This page contains approved plots and results as they appear in the ATL-PHYS-PUB-2010-003 note. Only the note contains all the relevant information and should thus be consulted if one of the plots is used.

Figure 1: Leading diagrams corresponding to the three production mechanisms of single-top events: (a) t-channel production, (b) Wt-channel associated production, (c) s-channel.

Figure 2: Figure 2: Stacked jet multiplicity distributions for preselected events, without b-tagging requirement (left plot) and for events with at least one b-tagged jet (right plot).

Figure 3: Reconstructed leptonic top-quark mass for events with exactly two jets that pass the event preselection.

Figure 4: Examples of input variables to the NN in the 3-jets Pretag data set. The right most bin indicates the overfow bin. Left side: HT distribution. Right side: m(j1j2) distribution. See text for details.

Figure 5: Examples of input variables to the NN in the 3-jets Pretag data set. The distributions are normalized to an integrated luminosity of 200 pb-1. Left side: HT . Right side: m(j1j2). The right most bin in each distribution contains the overfow bin.

Figure 6: The distributions of the NN output. Left side: Template distributions normalized to unit area. Right side: Expected distribution for an integrated luminosity of 200 pb-1.

Figure 7: Shape uncertainty for a JES shift of +/5% for the two most important processes, W+jets (left) and ttbar (right). The lower plots show the ratios of the template histograms and thus gives the relative change in each bin.

Figure 8: Effect of JES uncertainty on rates fitted in pseudo-experiments normalized to the expectation (Beta). Left side: distribution for W+jets. Right side: distribution for ttbar. The black histograms centered at one show the outcome of pseudo-experiments with unaltered JES, the default setting. The median of this distribution is taken as reference. To estimate the impact of an altered JES we perform ensemble tests in which the JES is altered by +5% or by -5%. The resulting distributions of Beta are shown: +5% (blue) and -5% (red). The median of these distributions is shifted by an amount Delta which indicates that we expect to measure a higher (or lower) rate of events in the case that the JES is differing by +/5%, compared to the rate we expect in case of an unaltered JES. We define the shift as a measure of the systematic uncertainty on our rate measurement.

Figure 9: Distribution of likelihood ratios discriminants in 200 pb-1 in the 2-jets tagged sample. Two discriminants are constructed, one for the rejection of ttbar events, Lttbar, (left) and another for the rejection of W+jets process, LW+jets, (right). The cross section determination is performed with events in which both likelihood ratios exceed 0.9.